Questions about Symbolics lisp machines
Nelson Rodriguez
I'm thinking on getting a Symbolics lisp machine for me and my friend to
play with it at home. I have searched the newsgropus and the internet but
I didn't find many precise information on these machines. Therefore I
want to ask the experts here.
We think about the XL1200, a 3620 and the 3630 machines but I will like
to gather all the answers and generate a little FAQ on the topic. You are
encourage to comment also on all the other Symbolics machines that you
know about.
The questions are:
- How loud and hot are they really? We only have a little "computer
study" and don't want to make it a sauna or go deaf :-)
- Would they run on 220V/50Hz electricity or do they need a transformer?
- What about ordinary IDE or SCSI HDs, CD-ROM and tapes? Would these work
on the bare machines or do they need extra hardware to work?
- How good are the machines networked? Do they understand TCP/IP, FTP,
NFS, ...?
- How much do they weigh? How tall, wide, depth are they?
- How much memory, HDs, ... support each machine?
- Which other cards would you recommend to install on the machines (like
FrameThrower, ...)? What will these cards do?
- Would the machines only work with an original Symbolics console,
keyboard, mouse? (I think the original consoles are pretty sensitive and
get broken fast?).
- Another questions?
I think another FAQ can/should be about the first installation
respectively maintenance of Genera on the machines. I found a page with a
little report on a MacIvory installation and I think for us first users
suche advices/hints are always very welcome.
At least: if you have any Genera (8.3), MacIvory or Symbolics
documentation that you would give away without too much pain, please send
me an email. If you know of a free or cheap machine, preferably in
Germany, you can write about it too :-)
Please feel free to write me on german, spanish or english [and if
necessary on french: more traduction work for a friend of mine :-)]
Thanks,
Nelson
Nelson Rodriguez
> <!doctype html public "-//w3c//dtd html 4.0 transitional//en">
> <html>
> Hi,
> <br>that's not an answer to your questions but what do you plan to do.
> As far as i know the symbolics is quite slow (i know a company that
> replaced them by sun's with
> <br>Golden CL beacause it was faster & cheaper) ?
> <br>Best
> <br>AHz
Hi Andy,
please if possible no html postings. It is hard to read with a "plain"
newsreader.
I know that the Symbolics machines are old and slow but you have Genera
on it. Since many years I have heard of Genera as THE development
platform and now I want to try it. I'm also fascinated with the idea of a
machine (hardware) build on Lisp. I love also such older "personal"
hardware more than todays "soulless" PCs :-)
I'm a system administrator and Lisp programming is a hobby of mine. There
is no need for speed for me at home :-)
Chao,
Nelson
Andy
BTW: i hope you can read that better. Netscape had HTML edit as default
;-(. Sorry.
Best
AHz
Scott McKay
It might look archaic, but nothing has yet come close.
Getting an old Symbolics box to play with Genera is a
fine thing to do. The Lisp environments sold by other
vendors are a total joke, in my humble opinion; only MCL
even makes a credible try, and its scope is a fraction of
what is on the Lisp machine.
An XL1200 ran fine in my office at Symbolics off of a normal
wall socket. The monitors are indeed fragile, but a burned-in
one will stay working indefinitely. They speak TCP/IP. The
more hardware you plug in, the more likely you are to find
something that breaks.
> "Andy" <a...@smi.de> wrote in message news:3C9A12FC...@smi.de...
> that's not an answer to your questions but what do you plan to do. As far
as i know the symbolics is quite slow (i know a company that
> replaced them by sun's with
> Golden CL beacause it was faster & cheaper) ?
> Best
> AHz
Carl Shapiro
> An XL1200 ran fine in my office at Symbolics off of a normal
> wall socket.
In my experiance, an XL1200 with a framethrower and an external SCSI
disk will pull close 8 amps, which should be just fine for home use.
However, it is a real challenge to find a UPS that will not get
overloaded as soon as you turn the machine on!
Ralf Kleberhoff
Nelson Rodriguez schrieb:
>
> [...]
>
> We think about the XL1200, a 3620 and the 3630 machines but I will like
> to gather all the answers and generate a little FAQ on the topic. You are
> encourage to comment also on all the other Symbolics machines that you
> know about.
I used a 3640 some years ago and a MacIvory.
> The questions are:
> - How loud and hot are they really? We only have a little "computer
> study" and don't want to make it a sauna or go deaf :-)
That's a problem with the 3600 series. They are hot.
You won't like them in the same room where you're sitting (in summer).
They are a bit loud, but that's not too bad.
The MacIvory is much better.
> - Would they run on 220V/50Hz electricity or do they need a transformer?
The 3640 runs on 220V/50Hz consuming ~1 kW (and more during startup).
Don't overload your power line.
The MacIvory is like your PC.
> - What about ordinary IDE or SCSI HDs, CD-ROM and tapes? Would these work
> on the bare machines or do they need extra hardware to work?
36xx - special hardware
MacIvory - Mac hardware (SCSI, CD-ROM etc.)
> - How good are the machines networked? Do they understand TCP/IP, FTP,
> NFS, ...?
Good networking support.
> - How much do they weigh? How tall, wide, depth are they?
3640: Twice the width and the depth of a PC big tower.
MacIvory: desktop or tower Macintosh.
> [...]
>
> Thanks,
>
> Nelson
Regards,
--- Ralf
Michael Parker
Mine doesn't have the framethrower, but the APC Back-UPS Pro 1400 seems
to work ok (knock on wood).
Michael Parker
> Hello,
>
> I'm thinking on getting a Symbolics lisp machine for me and my friend to
> play with it at home. I have searched the newsgropus and the internet but
> I didn't find many precise information on these machines. Therefore I
> want to ask the experts here.
>
> We think about the XL1200, a 3620 and the 3630 machines but I will like
> to gather all the answers and generate a little FAQ on the topic. You are
> encourage to comment also on all the other Symbolics machines that you
> know about.
36xx's will be cheaper, but a lot slower. And they aren't the snappiest
machines around nowadays anyway. But the environment is impressive, and
Dynamic Windows is amazingly cool.
> The questions are:
> - How loud and hot are they really? We only have a little "computer
> study" and don't want to make it a sauna or go deaf :-)
The XL1200's aren't particularly quiet. They aren't heaters, though.
They do prefer ambient temps to be on the cool side, though (say mid-60's).
> - Would they run on 220V/50Hz electricity or do they need a transformer?
Don't know. Mine runs off a UPS plugged into a wall socket, but that's
US power specs. Presumably a machine that's been floating around in
Europe would handle the euro-spec power ok.
> - What about ordinary IDE or SCSI HDs, CD-ROM and tapes? Would these work
> on the bare machines or do they need extra hardware to work?
SCSI works, but the drives need to be able to handle odd-sized blocks.
> - How good are the machines networked? Do they understand TCP/IP, FTP,
> NFS, ...?
yes.
> - How much do they weigh? How tall, wide, depth are they?
An XL1200 will fit nicely to the side of a desk. It'll be a few
inches shorter, and a few inches less deep. But it's sizeable.
> - How much memory, HDs, ... support each machine?
Depends on what you want to do. They do have really good VM.
> - Which other cards would you recommend to install on the machines (like
> FrameThrower, ...)? What will these cards do?
Framethrowers are video capture/output devices.
> - Would the machines only work with an original Symbolics console,
> keyboard, mouse? (I think the original consoles are pretty sensitive and
> get broken fast?).
Yes. If the consoles are well broken-in they should be ok, though. Dave
Schmidt fixed mine the last time it went out.
> - Another questions?
>
> I think another FAQ can/should be about the first installation
> respectively maintenance of Genera on the machines. I found a page with a
> little report on a MacIvory installation and I think for us first users
> suche advices/hints are always very welcome.
If you've never maintained one of these machine before, it can
be overwhelming -- they just don't do anything remotely like any
other machine on the planet. I used one in college, but never
had to do the admin-type stuff, so when I got mine awhile back, I
had an "interesting" few weeks. The documentation is very good
and complete, except when it isn't. The slug mailing list is a
good source for help.
Paolo Amoroso
http://www.abstractscience.freeserve.co.uk/symbolics/
Paolo
--
EncyCMUCLopedia * Extensive collection of CMU Common Lisp documentation
http://www.paoloamoroso.it/ency/README
[http://cvs2.cons.org:8000/cmucl/doc/EncyCMUCLopedia/]
Michael Parker
> Genera's programming environment is still incomparable.
> It might look archaic, but nothing has yet come close.
> Getting an old Symbolics box to play with Genera is a
> fine thing to do. The Lisp environments sold by other
> vendors are a total joke, in my humble opinion; only MCL
> even makes a credible try, and its scope is a fraction of
> what is on the Lisp machine.
This statement piqued my curiosity. I've not used MCL, but
after playing on my wife's new iMac, I have recently found
myself considering switching camps. What sorts of things
does MCL have that Lispworks and ACL are missing?
Dr. Edmund Weitz
> Genera's programming environment is still incomparable. It might
> look archaic, but nothing has yet come close. Getting an old
> Symbolics box to play with Genera is a fine thing to do. The Lisp
> environments sold by other vendors are a total joke, in my humble
> opinion;
This is not the first time I heard that opinion, and I'd like to know
what made Lisp Machines so different from the current Lisp
environments - unfortunately, I don't have access to one. Could you
provide a few examples? Also, why do you think the current
implementations are so much behind? Is it that certain things simply
aren't feasible without an OS that supports them? Is the Open Genera
environment that ran[*] on Alpha machines comparable to the LispM
environment or is it also lacking something?
> only MCL even makes a credible try, and its scope is a
> fraction of what is on the Lisp machine.
I've installed the trial version of MCL on my wife's PowerBook and I
couldn't see any significant advantage over, say, LispWorks. But I
only played with it for about an hour so I probably missed a lot of
things. Could you elaborate how MCL differs from the
Windows/Linux/Unix implementations currently available?
Thanks in advance for satifying my curiosity,
Edi.
[*] Or does it still run, meaning it's still available and works with
current Alpha machines?
--
Dr. Edmund Weitz
Hamburg
Germany
The Common Lisp Cookbook
<http://cl-cookbook.sourceforge.net/>
Kent M Pitman
> "Scott McKay" <s...@attbi.com> writes:
>
> > Genera's programming environment is still incomparable. It might
> > look archaic, but nothing has yet come close. Getting an old
> > Symbolics box to play with Genera is a fine thing to do. The Lisp
> > environments sold by other vendors are a total joke, in my humble
> > opinion;
>
> This is not the first time I heard that opinion, and I'd like to know
> what made Lisp Machines so different from the current Lisp
> environments - unfortunately, I don't have access to one. Could you
> provide a few examples?
Try:
http://groups.google.com/groups?hl=en&selm=sfwpu5b5224.fsf%40shell01.TheWorld.com
http://groups.google.com/groups?hl=en&selm=sfwyalstyq7.fsf%40world.std.com
[The references in this:
http://groups.google.com/groups?hl=en&selm=joswig-1312981610100001%40194.163.195.67
seem to have moved. It would be nice if Rainer Joswig could post a new
set of homes.]
> Also, why do you think the current
> implementations are so much behind?
Budgets were a lot more free back then, and Lisp Machines had the luxury
of programming for the future instead of for the current month's budget.
They could afford to be visionary.
> Is it that certain things simply
> aren't feasible without an OS that supports them?
Technically? No.
In the current marketplace? Hard to tell. The Lisp Machine got
tremendous advantage from having a coordinated set of products instead
of a heterogeneous market of uncoordinated products. Then again, I'm pretty
sure that this is not a fatal issue--a number of software environments
(gnu in the free software domain, java in the commercial software domain)
have done semi-coordinated activities. The issue is "direction".
> Is the Open Genera
> environment that ran[*] on Alpha machines comparable to the LispM
> environment or is it also lacking something?
>
> > only MCL even makes a credible try, and its scope is a
> > fraction of what is on the Lisp machine.
>
> I've installed the trial version of MCL on my wife's PowerBook and I
> couldn't see any significant advantage over, say, LispWorks.
I suspect the issue SWM meant is "environmental interconnect", not
"raw power". That's probably hard to see at first glance.
> But I
> only played with it for about an hour so I probably missed a lot of
> things. Could you elaborate how MCL differs from the
> Windows/Linux/Unix implementations currently available?
>
> Thanks in advance for satifying my curiosity,
> Edi.
>
> [*] Or does it still run, meaning it's still available and works with
> current Alpha machines?
"current" Alpha machines is an odd term, but yes, it does work.
Dr. Edmund Weitz
> Try:
> http://groups.google.com/groups?hl=en&selm=sfwpu5b5224.fsf%40shell01.TheWorld.com
>
> http://groups.google.com/groups?hl=en&selm=sfwyalstyq7.fsf%40world.std.com
Thanks for the pointers, that sounds very interesting and indeed
one-of-a-kind. The first one I should have read myself because I was
already a c.l.l regular at that time but I must have somehow missed it
in a pre-Christmas hurry... :)
> "current" Alpha machines is an odd term, but yes, it does work.
I suppose you're referring to the fact that the whole Alpha platform
looks doomed, but last time I looked you could still buy Alpha
workstations from Compaq. (I've never used them, though. Maybe these
are the same machines they sold five years ago and there is no active
development anymore.)
Thanks again,
Edi.
Rahul Jain
> I suppose you're referring to the fact that the whole Alpha platform
> looks doomed, but last time I looked you could still buy Alpha
> workstations from Compaq. (I've never used them, though. Maybe these
> are the same machines they sold five years ago and there is no active
> development anymore.)
Compaq has killed the EV8, but the EV7 will still be released, or at
least that's their latest promise.
--
-> -/ - Rahul Jain - \- <-
-> -\ http://linux.rice.edu/~rahul -=- mailto:rj...@techie.com /- <-
-> -/ "Structure is nothing if it is all you got. Skeletons spook \- <-
-> -\ people if [they] try to walk around on their own. I really /- <-
-> -/ wonder why XML does not." -- Erik Naggum, comp.lang.lisp \- <-
|--|--------|--------------|----|-------------|------|---------|-----|-|
(c)1996-2002, All rights reserved. Disclaimer available upon request.
Bill Clementson
> Kent M Pitman <pit...@world.std.com> writes:
>
> > Try:
> > http://groups.google.com/groups?hl=en&selm=sfwpu5b5224.fsf%40shell01.TheWorld.com
> >
> > http://groups.google.com/groups?hl=en&selm=sfwyalstyq7.fsf%40world.std.com
>
> Thanks for the pointers, that sounds very interesting and indeed
> one-of-a-kind. The first one I should have read myself because I was
> already a c.l.l regular at that time but I must have somehow missed it
> in a pre-Christmas hurry... :)
I've also been intrigued by the comments on c.l.l about LispMs and have
tried to find out how they differ from current lisp implementations. Unfortunately,
there isn't a lot of generally available information. Some internet sites that I
have found useful are:
http://fare.tunes.org/LispM.html
http://www.abstractscience.freeserve.co.uk/symbolics/
http://www.cbbrowne.com/info/lisposes.html
http://kogs-www.informatik.uni-hamburg.de/~moeller/symbolics-info/symbolics.html
Also, there is a (rather pricey) summary of Symbolics documentation available called
_Lisp Lore : A Guide to Programming the Lisp Machine_
It is available from Amazon but you might be able to pick up a used copy as well:
http://www.amazon.com/exec/obidos/ASIN/0898382289/qid=1016922531/sr=1-2/ref=sr_1_2/103-2126716-9830203
To date however, Kent's comments on this list have shed more light on why the LispMs
were unique (and better in some ways from current lisp implementations) than the other
sources that I have looked at.
--
Bill Clementson
Scott McKay
seems to make a lot of difference.
Disclaimer: I am not an, nor ever have been, an MCL developer.
MCL is now open-source, I believe.
"Michael Parker" <mpark...@hotmail.com> wrote in message
news:9f023346.02032...@posting.google.com...
Eric Moss
>looks doomed, but last time I looked you could still buy Alpha
FWIW, I chatted with the OpenGenera folks, and they are taking votes on
what platform to port it to. PowerPC/OSX is one, and I'm sure there are
others--what those might be is anyone's guess. ;)
Eric
Paolo Amoroso
> Genera's programming environment is still incomparable.
[...]
> fine thing to do. The Lisp environments sold by other
> vendors are a total joke, in my humble opinion; only MCL
> even makes a credible try, and its scope is a fraction of
How much of Genera's functionality can be implemented on contemporary, non
Lisp based machines and operating systems? I'd like to know whether the
lack of such functionality in current Lisp environments is "merely" due to
limited resources and commercial/marketing decisions, or to intrinsic
technical limitations of contemporary hardware/software.
Kent M Pitman
> On Fri, 22 Mar 2002 13:38:53 GMT, "Scott McKay" <s...@attbi.com> wrote:
>
> > Genera's programming environment is still incomparable.
> [...]
> > fine thing to do. The Lisp environments sold by other
> > vendors are a total joke, in my humble opinion; only MCL
> > even makes a credible try, and its scope is a fraction of
>
> How much of Genera's functionality can be implemented on contemporary, non
> Lisp based machines and operating systems? I'd like to know whether the
> lack of such functionality in current Lisp environments is "merely" due to
> limited resources and commercial/marketing decisions, or to intrinsic
> technical limitations of contemporary hardware/software.
Usually people use hardware/software as a kind of "I dont mean to quibble
over the detail here" but I would like to note my belief that it's nontrivial
that "/architecture/OS" perhaps belongs here.
By architecture, I mean it didn't hurt that the individual components were
written all to plug into a common architecture (which happened to be
Lisp-based, but where it was more important that it was of a common base).
[My understanding, though it's something I personally used only a little,
is that MIT Multics got similar power in some regards by being organized
around a common set of PL/1-based paradigms for control, memory management,
etc. I think this meant, for example, that cross-application error handlers
could run using a common protocol, not dissimilar to the one that CL uses
now. These ideas flowed into the LispM's "New Error System" (NES) and
from there to CL.]
By OS, I obviously mean that it didn't hurt that it was a Lisp-based OS.
The way many people learned about the Lisp Machine was literally to do this:
1. Find some application that did what you want, whether it was a
user application or even a text editor or a window system pop-up
dialog, and then press Control-Meta-Suspend, which forced entry to
the debugger. At this point, they could see the whole stack down
to the code that launched the executing process from the operating
system. (In all such stack frames was Lisp data that was subject
to the Lisp function INSPECT, by the way. That was helpful as
well.)
2. Find a stack frame that looked interesting and press Control-E
to edit the source code for the current stack frame. This
automatically knew where the source code was without you telling
it anything. You would then just read the source code to learn
how things worked. If any function or variable or class/flavor or
package in the source was unknown to the person reading the code,
they could just type Meta-. which would prompt for a function name
and then either type or click on the name in question.
or
1. You could also just go to a listener (shell) and type
:Show Herald to see a list of what's loaded; that output is
clickable so you can click a particular system name to see its
components, though you could always :Show System Components and
type a system name to see the compile script for the system if you
want to read its sources rooting that way instead of from source.
You could get to the editor with Meta-click on a filename in the
output of this command. If you didn't know this, the always-present
"mouse documentation line" would tell you this fact.
The reasons modern OS/architectures are already hampered are not due to
hardware/software per se, but due to lack of buy-in to these common
paradigms from each and every application, contributing to a coherent
sense of whole:
- The lack of a mouse-sensitive typeout area. It's not quite the same
to put it this way, and you'd have to see Dynamic Windows to understand,
and even then it does DW a massive disservice to sum it up this way,
but it's as if all interactions happened in HTML. That is, there's more
to what you see than what is typed on the screen. In HTML, what backs up
a link is a URL. (In DW, what backs up mouse-sensitivity is an internal
understanding of a hierarchy of reasons why various output occurred,
different parts of which become clickable depending on input syntax.)
- The lack of a keyboard with Control, Meta, Super, Hyper, Symbol, and Shift
[or minor variations thereof: in some architectures the Symbol key is
replaced by one or more shift keys named Top, Greek, Front] which can be
chorded with keys and mouse clicks to get an enormous number of actions
on a single click. For example, Super-Meta-mouse-Right might do something
quite different than mouse-Right.
- The lack of a dedicated "mouse doc line" to explain the enormously powerful
set of options for the mouse at any given moment. As you pressed shift
keys up and down, you'd see the key commands for that shift combination.
Super- commands tended to be meta-operations for inspecting the state of
the window system, Meta- were often commands for editing, etc. But the
mouse doc line helped you learn. It would say things like:
L: Select M: Inspect R: Menu
These were not fixed strings but were composed from smaller pieces of data
contributed in an object-oriented way by the input context of the process
with control of the keyboard.
- The lack of a dedicated "wholine" explaining the state of the machine.
From this line, one got not only the time but the input package, whether
the current process was executing or in a wait state (and what that state
was), whether a GC was happening and which of several GC phases was going
on, whether disk paging was occurring, whether a file was open or servers
were accessing the machine, and what percent done a file access or other
progress-monitored-actions were.
- That all executing software should carry around a full symbol table
that is cross-indexed with _correct_ source. This goes WELL BEYOND
the present and I-personally-find-utterly-useless claim that software
is "open source". 99.99% of the time, it might as well not be
"open source" to me because I have no idea if (a) the source is present,
(b) where the source is, (c) whether something that I find that looks
like the source _is_ the source, and (d) if it is the right general source,
whether it's the right version. If on the Lisp Machine the source
had been patched, Meta-. would tell you that there were multiple
sources in play and what the various source files were. The list
would end up in an editor buffer in Zmacs so you could see that a
certain source file was patched by later code loaded from certain
patch files.
- On Unix/Linux, sources vary in their home, find is syntactically
weird and hard to use for newbies as a "getting started" tool (and
is impeded by symbolic links), grep is hard to use across directory
levels, patches are mysterious, file conventions are not
primitively understood by the editor, commands do not exist to ask
what software is loaded and to browse source trees and patch files,
etc. And certainly none of such stuff is clickable beginning from a
user interface available to an utter newbie. In the Lisp Machine,
the initial herald (sort of its /etc/motd) is clickable to see all
this info, so a person with a curious mind and no experience would be
able to find all of this information with absolutely no knowledge
of command structure or file system organization -- not even the need
for file system syntax is presupposed.
- That the debugger should offer symbolically named stack frames to users
[and that the names of these stack frames should be meaningful; the
difference between EDITOR:MAKE-BUFFER and OPERATING-SYSTEM:MAKE-BUFFER
should be visually presented, no matter what people who dislike packages
think, since often only the "EDITOR:" or "OPERATING-SYSTEM:" is the point
of interest when one is getting started...]
- That restarts should be provided by any application on the stack and
visible across applications.
- The lack of common address space making the process of reaching into the
internals of another program and seeing its data a mere matter of
understanding a few relatively simple data structures in a way that
you're already used to managing other high-level data rather than a trip
to a bookstore, a licensing of a set of expensive foundation classes, and
a three month course for several thousand dollars taught either by
Microsoft or by a starving Free Software person who has to make his money
somehow since he can't make it on selling program.
I'm sure there are other elements. But maybe you can see by some of these
I've enumerated that the problem is more than is within power of a single
application writer to fix. It's about a whole community mindset and a
willingness of that entire community to contribute to the illusion.
The following two recursively arranged stories may help to see the problem
that confronts a community with a desire to do this; and, incidentally,
this story also probably makes it clear why it may not be an accident that
Scott McKay and others point to MCL as the closest contender to LispM-like
culture...
Someone once grumbled to me about a certain application that when it was
going to offer a submenu under a menu, the menu only said "Foo" and not
"Foo..." like on the Mac. Why don't you make it add "..." in the case of
a submenu. I explained that it couldn't know because it was just going to
call a function and it didn't know what the function was going to do.
"But on the Mac--" the person said, trailing off. I shook my head and
explained it this way: The Mac does not know either. Rather, an illusion
is maintained by the individual force of will of every menu writer for the
Mac, obeying a convention that they must wire the string "..." to the end
of menu items in certain cases just so people will _think_ that the Mac
actually does this. I then proceeded to tell the following story:
We often expect people to grow up with things like "common sense" and
"manners" and "ethics". The point has been made by some, though, that
these things don't come out of nowhere. They are the individual
product of parents everywhere, laboriously programming each and every
instance of "human being" with tons of stuff that we want to believe
is just "automatically there". But there is no mechanism for making
the download, recognition, and use of such stuff automatic. We are
just lucky when parents buy into this level of responsibility, and
such fragile illusions will survive only as long as parents, in the
aggregate, care. Society is the sum of these individual, unsung heroic
acts.
Paolo Amoroso
> MCL is now open-source, I believe.
Only the kernel/compiler, not the environment:
OpenMCL
http://openmcl.clozure.com
Michael Parker
> On Fri, 22 Mar 2002 13:38:53 GMT, "Scott McKay" <s...@attbi.com> wrote:
>
> > Genera's programming environment is still incomparable.
> [...]
> > fine thing to do. The Lisp environments sold by other
> > vendors are a total joke, in my humble opinion; only MCL
> > even makes a credible try, and its scope is a fraction of
>
> How much of Genera's functionality can be implemented on contemporary, non
> Lisp based machines and operating systems? I'd like to know whether the
> lack of such functionality in current Lisp environments is "merely" due to
> limited resources and commercial/marketing decisions, or to intrinsic
> technical limitations of contemporary hardware/software.
Well, you're not likely to write an OS in CL on modern hardware
although I'm not sure that's really even desirable in today's
heterogeneous environment.
Keep in mind though, that as nice as Genera was, it didn't
interface to Oracle, didn't support CORBA or COM or a lot
of things that are frequently more critical in todays world,
and to which the non-Lispm vendors have devoted substantial
resources. Rightly so IMO, since they're still around and
the otherwise superior Lispms are a fading memory. The number
of applications for which a closed-world solution is sufficient
is pretty small nowadays.
Julian Stecklina
> I suggest that you check this site:
>
> http://www.abstractscience.freeserve.co.uk/symbolics/
Does somebody know of a in detail description of a processor designed
to run Lisp software? Especially the instruction set would be
interesting.
Regards,
Julian
--
Meine Hompage: http://julian.re6.de
Um meinen oeffentlichen Schluessel zu erhalten:
To get my public key:
http://math-www.uni-paderborn.de/pgp/
Joe Marshall
> Does somebody know of a in detail description of a processor designed
> to run Lisp software? Especially the instruction set would be
> interesting.
http://www.delphion.com/details?pn=US04887235__
http://fare.tunes.org/tmp/emergent/memo444.htm
http://www.unlambda.com/lispm/memo528.html
ftp://publications.ai.mit.edu/ai-publications/pdf/AIM-514.pdf
http://www.swiss.ai.mit.edu/~mhwu/scheme86/sbt.ps
http://fare.tunes.org/tmp/emergent/kmachine.htm
~jrm
Bruce Miller
... A _gem_ of an article. Thanks Kent!!
It seems the main thing missing today is the culture (or
even the will) to construct, & then use, the appropriate
infractructure and paradigms.
I know you have certain reservations about the open source/free
software/whatever idea, and I'm not disagreeing with a lot of your
points. But it seems to me that in the current climate, that's
the only place there would be a chance (minimal though it may
be) of recreating the necessary infrastructure.
Both in terms of getting enough programming mindshare to implement
it, and enough user mindshare for it to matter.
One could imagine leveraging some sort of next generation
gcc/glibc/gtk (and/or kde stuff), package system, and (x)emacs
to achieve a lot of what the LispM had.
The technical details would be overwhelming enough, but the
politics would probably do it in.
Alas, the worst problem is that most people don't know what
they're missing!
bruce
Paolo Amoroso
wrote:
> The way many people learned about the Lisp Machine was literally to do this:
[...]
I posted my request before seeing other articles in this thread, so I was
about to follow up with a "never mind, I've seen the replies" post. But I
really appreciate the additional information you have provided.
> The reasons modern OS/architectures are already hampered are not due to
> hardware/software per se, but due to lack of buy-in to these common
> paradigms from each and every application, contributing to a coherent
> sense of whole:
[...]
> I'm sure there are other elements. But maybe you can see by some of these
> I've enumerated that the problem is more than is within power of a single
> application writer to fix. It's about a whole community mindset and a
> willingness of that entire community to contribute to the illusion.
This is why I think that Lisp is a civilization, and the Lisp Machines are
among its most advanced cultural centers.
> this story also probably makes it clear why it may not be an accident that
> Scott McKay and others point to MCL as the closest contender to LispM-like
> culture...
I am aware that Scott knows "a couple" of things about mouse sensitive
typeout-areas :)
Christopher C. Stacy
Michael> Keep in mind though, that as nice as Genera was, it didn't
Michael> interface to Oracle, didn't support CORBA or COM or a lot
Michael> of things that are frequently more critical in todays world,
Michael> and to which the non-Lispm vendors have devoted substantial
Michael> resources. Rightly so IMO, since they're still around and
Michael> the otherwise superior Lispms are a fading memory. The number
Michael> of applications for which a closed-world solution is sufficient
Michael> is pretty small nowadays.
Those things weren't significant (or extant) for most of the life of
Genera, and could have been very easily interfaced to. I whipped
up a database interface (like the Oracle network interface) very
quickly, at least once, and I am sure many others did that, also.
Genera was the paragon of network-based interoperability, until
Symbolics declined to the point where it no longer had resources
to expend on that sort of thing (sometime around the second time
that it laid off most of the staff). In that context, Lisp went
from being the most connected programming language to one of the
least well-connected.
Kent M Pitman
> >>>>> On 24 Mar 2002 07:23:59 -0800, Michael Parker ("Michael") writes:
> Michael> Keep in mind though, that as nice as Genera was, it didn't
> Michael> interface to Oracle, didn't support CORBA or COM or a lot
> Michael> of things that are frequently more critical in todays world,
> Michael> and to which the non-Lispm vendors have devoted substantial
> Michael> resources. Rightly so IMO, since they're still around and
> Michael> the otherwise superior Lispms are a fading memory. The number
> Michael> of applications for which a closed-world solution is sufficient
> Michael> is pretty small nowadays.
>
> Those things weren't significant (or extant) for most of the life of
> Genera, and could have been very easily interfaced to. I whipped
> up a database interface (like the Oracle network interface) very
> quickly, at least once, and I am sure many others did that, also.
Indeed, moreover it did connect to a great many things that were the
interfaces of choice from when the company was a going concern. To the
point that a number of companies bought Lisp Machines as, essentially,
"connectivity servers".
Friedrich Dominicus
> "Scott McKay" <s...@attbi.com> writes:
>
> > Genera's programming environment is still incomparable. It might
> > look archaic, but nothing has yet come close. Getting an old
> > Symbolics box to play with Genera is a fine thing to do. The Lisp
> > environments sold by other vendors are a total joke, in my humble
> > opinion;
>
> This is not the first time I heard that opinion, and I'd like to know
> what made Lisp Machines so different from the current Lisp
> environments - unfortunately, I don't have access to one. Could you
> provide a few examples?
Now what do you think about a desktop where everything can be
interactive? Where everthing can bear some information to browse?
An environment where date are not plain character files but structured
objects? Where really all tools are related and can be used from
anywhere?
> Also, why do you think the current
> implementations are so much behind? Is it that certain things simply
> aren't feasible without an OS that supports them? Is the Open Genera
> environment that ran[*] on Alpha machines comparable to the LispM
> environment or is it also lacking something?
According to the docs they are just missing the "low-level" stuff like
accessing SCSI deviced directly and the like. I would think OpenGenera
is still a LispOS.
> Dr. Edmund Weitz
> Hamburg
If it ever happened that you'll visit the south (somewhere around
Karlsruhe) just call me and come on a visit. You than can get your
hands on an LispOS system.
Regards
Friedrich
Friedrich Dominicus
> On Fri, 22 Mar 2002 13:38:53 GMT, "Scott McKay" <s...@attbi.com> wrote:
>
> > Genera's programming environment is still incomparable.
> [...]
> > fine thing to do. The Lisp environments sold by other
> > vendors are a total joke, in my humble opinion; only MCL
> > even makes a credible try, and its scope is a fraction of
>
> How much of Genera's functionality can be implemented on contemporary, non
> Lisp based machines and operating systems? I'd like to know whether the
> lack of such functionality in current Lisp environments is "merely" due to
> limited resources and commercial/marketing decisions, or to intrinsic
> technical limitations of contemporary hardware/software.
Well technically I can't see why it should not work anywhere
else. It may need a shift towards another way of thinking. E.g the
Unix folks have decided that everthing will be treated as flat
character streams the Lisp people have decided that structured objects
are the way to go. All the tools are adopted to suit this wish...
Regards
Friedrich
Jon Allen Boone
I'm trying really hard to get a handle around this LISPm environment
issue. I really appreciate the time that folks take to educate people
like me. I probably briefly touched a LISPm once or twice (I did
spend 7 years at CMU), but was so totally ignorant of LISP at that
point that I can only evoke faint echos of possible "past lives"
involving such interaction by looking at the pictures of LISPms put up
on the web...
I've followed this topic as much as time allows [including threads
from years ago].
cst...@theworld.com (Christopher C. Stacy) writes:
>>>>>> On 24 Mar 2002 07:23:59 -0800, Michael Parker ("Michael") writes:
> Michael> Keep in mind though, that as nice as Genera was, ...
>
> Genera was the paragon of network-based interoperability, ...
From what I gather to this point, the consensus would seem to be the
following:
* LISPm environments were much better than anything available today
* Genera was the best of the LISPm environments
* It is not worthwhile to contemplate recreating a pure LISPm
environment. Instead LISP environments on current host operating
systems [Linux/MacOS/Solaris/Windows/&tc] should be extended...
My conclusion follows this brief interlude that isn't totally
off-topic:
When I first went to CMU in 1988, we often found it inconvenient to
use the Andrew Window Manager wm (or Computer Club version, cwm) due
to what we'd now call performance limitations of the machines in use
there [Sun 3s, MicroVax, IBM RTs, eventually DECStation 3100s]. So,
we'd work right on the console, as much as we could. Since some
consoles [at least DS 3100] didn't directly support terminal emulation
[such as vt100], a friend of mine wrote a terminal emulator called
niftyterm, which even supported different font sizes, &tc. Once we
had niftyterm, we just had to load up Gnu Emacs and we could do pretty
much anything [non-graphical] we wanted from within this environment.
All of which leads me to the conclusion that the "most worthwhile"
modern Lisp environment that would come close to Genera would be
somehow integrating the best ideas from the Franz and Xanalys
environments with something like {GNU|X}Emacs.
* {GNU|X}Emacs alone doesn't cut it because of the poor CL support
and the lack of things like class browsers, &tc.
* I'm experimenting with the no-$ version of ACL, using XEmacs as
the editor with the Franz-supplied Lisp-interaction modules. So,
perhaps this combination is the best..?
* I like the no-$ version of LispWorks Personal [running under
Linux], but, of course, I'd need to have the $900 version of
LispWorks Professional to avoid the 4-hour environment meltdown
issue, as well as the fact that their Editor [at least in the
Personal edition] isn't a total plug-in-replacement for
{GNU|X}Emacs.
So, it would seem that if I have understood correctly, the best
thing would be to:
* start with either ACL or LispWorks
* implement a better editor from within CL [likely ignoring ELisp
compatibility]
* provide equivalent functionality to major ELisp systems such as
Gnus, W3, &tc. but in Common Lisp rather than ELisp.
Would this be an environment worth putting together? Should I just
beg Symbolics Technologies to port OpenGenera to an Intel chipset?
Is there anyone anywhere near Philadelphia with a LISPm that
wouldn't mind having me drop by for a visit? I'd really love to spend
some small amount of time [maybe up to an hour] observing someone
interacting with a LISPm [even non-Symbolics systems...]
-jon
--
------------------
Jon Allen Boone
ipmo...@delamancha.org
Thomas Bushnell, BSG
> Well technically I can't see why it should not work anywhere
> else. It may need a shift towards another way of thinking. E.g the
> Unix folks have decided that everthing will be treated as flat
> character streams the Lisp people have decided that structured objects
> are the way to go. All the tools are adopted to suit this wish...
A more serious problem is that in Unix flat character streams are
interpreted *everywhere* by code and logic duplicated a jillion times
over.
Consider the hair and pain involved in making Unicode work on
GNU/Linux systems with UTF-8. This is the easiest way to go, and even
so it's lots of work converting a jillion applications to work right.
And this is because "character stream" is *not* a well defined
concept; Unix historically only has ASCII character streams, and from
this comes a giant problem.
Unix is *filled* with such issues, where instead of an abstract type a
specific encoding is used--not just character streams, but that's
what's on top of my mind at present.
This problem is solvable in many different languages (not just Lisp
variants); Smalltalk is also able to do a perfectly good job. What it
takes is understanding the great importance of *abstracting* away
issues of representation.
And, of course, an attitude of *not* reimplementing everything a
jillion times just because things are in separate processes. Care to
guess how many times different argument parser routines there are in
an average GNU/Linux system? I pick that one because argument parsing
is actually, a *total waste*, forced by the use of a "shell"--another
wasted concept unnecessary in a Real System (like the various lispms
had, and like Sky [should it ever happen] will have).
Thomas
Kent M Pitman
Plus, Dick Gabriel's "worse is better" phenomenon is in play.
Erik Naggum
| Consider the hair and pain involved in making Unicode work on GNU/Linux
| systems with UTF-8. This is the easiest way to go, and even so it's lots
| of work converting a jillion applications to work right. And this is
| because "character stream" is *not* a well defined concept; Unix
| historically only has ASCII character streams, and from this comes a
| giant problem.
No, this is an important mistake. Unix has a well-defined concept of an
"octet stream". This is _never_ what you really want. On top of this
"octet stream" Unix has, via C's lack of a real character type, given its
users the notion that a _character_ is just the same as a small integer
that happens to fit in an octet. All of this is unfortunately wrong.
A character and its encoding are different concepts. An encoding and its
(external) representation are different concepts. An external
representation and the numeric values of whatever unit it is made up of
are different concepts. By conflating all four concepts into one, Unix
has held text processing and computing in general back several decades.
This is fairly ironic, since Unix started out as a text processing
vehicle.
One result of this character = small number = octet confusion is that
"variable length" encodings are seriously frightening to Unixoid coders
(partly because of the observation that you make that all the code that
deals with octet-stream -> anything-else interpretation and has caused
such a problem with stable and well-defined standards such as ISO 2022
that the IETF was utterly unable to use any existing standards for the
representation of multi--character-set "documents" and "streams", and so
had to invent both MIME (extremely crude structured objects in mail) and
a charset property at an extremely high level, such that mixing charsets
became extremely verbose and difficult. This is also why some people
think Unicode sucks because it may force programmers to deal with
characters differently than "just assume 16 bits" instead of the old
"just assume 8 bits".
| Care to guess how many times different argument parser routines there are
| in an average GNU/Linux system? I pick that one because argument parsing
| is actually, a *total waste*, forced by the use of a "shell"--another
| wasted concept unnecessary in a Real System (like the various lispms had,
| and like Sky [should it ever happen] will have).
I think you overreact now. The biggest problem here is that _everything_
in Unix is an octet stream, even strings, and program arguments are just
strings. (The fact that you need to parse the "string" from beginning to
end to find the in-band terminator (which cannot even be escpaed) makes
it a stream, and the "pointer" you have into a stream to read the current
position is just like the position in a stream.)
Unix is in fact so streams-based that it is nearly _impossible_ to work
with structured objects. Everywhere an object wants to go, it has to be
marshalled into and out of an octet-stream--based external format, both
in arguments and in pipelines. It is as you had to call a function foo
like (eval (format nil "(foo~@{ ~A~})" <arguments>)). Hey, I just
reinvented Tcl.
Of course, every object must have an external representation of _some_
sort to communicate it with external programs, but marshalling to and
from octet stream should preserve the object-ness. Lisp and things like
ASN.1 enable the preservation of objectness in marshalling, and many
other attempts have been made. But treating everything like a string
without futher adornment or syntax (which Unix shells and, ironically,
SGML and XML do) is just plain wrong.
On the other hand, there _are_ times when you want to just copy a file or
ship across a network bit by bit, in which case the octet stream might
seem the only alternative. This is not really a situation that the user
or even (application) programmer needs to be exposed to.
///
--
In a fight against something, the fight has value, victory has none.
In a fight for something, the fight is a loss, victory merely relief.
Robert Strandh
> So, it would seem that if I have understood correctly, the best
> thing would be to:
>
> * start with either ACL or LispWorks
>
> * implement a better editor from within CL [likely ignoring ELisp
> compatibility]
>
> * provide equivalent functionality to major ELisp systems such as
> Gnus, W3, &tc. but in Common Lisp rather than ELisp.
Another possibility would be to help us write McCLIM and use it with
CMUCL and/or SBCL to create such an environment. In the McCLIM CVS
tree there is an embryo of an implementation of Emacs in CL.
McCLIM could very well become the basis of the user interaction part
of an environment similar to Genera. It would then be interesting to
see whether a flat address space on a 64-bit machine could be used for
a multi-user Lisp environment.
As far as the rest of your list is concerned, Gilbert Baumann has a
(I think working) web browser in CL, which (if I remember correctly)
he is adapting to McCLIM. The Eclipse window manager is written
entirely in Common Lisp (with CLX) so there is another piece of the
puzzle.
> Would this be an environment worth putting together?
It would be a HUGE task, so it may not be "worth" it to you. But
certainly, if you do it, it would be a great service to the
community.
--
Robert Strandh
---------------------------------------------------------------------
Greenspun's Tenth Rule of Programming: any sufficiently complicated C
or Fortran program contains an ad hoc informally-specified bug-ridden
slow implementation of half of Common Lisp.
---------------------------------------------------------------------
Barry Margolin
Bruce Miller <doc...@md.prestige.net> wrote:
>On Sun, 24 Mar 2002 07:10:34 -0500, Kent M Pitman wrote:
>... A _gem_ of an article. Thanks Kent!!
>
>It seems the main thing missing today is the culture (or
>even the will) to construct, & then use, the appropriate
>infractructure and paradigms.
Right. It takes quite a bit of work to ensure that everything is
integrated. It's also much easier when there's a single organization doing
all the development.
Currently, the closest anyone sees to the kind of stuff I remember from my
LispM days is "web-enabled" applications, which allow you to click on links
in the application (e.g. mail/news readers, Acrobat reader, RealPlayer) and
automatically be transferred to the web browser for related information.
But this is just a small taste of how the LispM integrated things.
Practically everything on the screen was "live data", and they all had
associated context-based menus.
Take what you may be used to with plug-ins and web-enabled applications,
and multiply the convenience and integration by 100, and that's what Genera
is like.
Unfortunately, I think it may be close to infeasible to implement something
like this in conventional systems. LispM's have the benefit of everything
being in a single, shared address space, so it's easy to develop relatively
ad hoc mechanisms for accessing related applications and data. In a more
conventional OS, it's necessary to design interprocess communications
protocols to effect all of this. It's hard to ensure that these protocols
provide all the flexibility to support features that you haven't even
thought of yet. So basically, all we have are a handful of basic requests,
like "tell the browser to go to this URL", and "tell the plug-in to display
this data in the browser window".
--
Barry Margolin, bar...@genuity.net
Genuity, Woburn, MA
*** DON'T SEND TECHNICAL QUESTIONS DIRECTLY TO ME, post them to newsgroups.
Please DON'T copy followups to me -- I'll assume it wasn't posted to the group.
Frode Vatvedt Fjeld
inspiring, and I'm curious to actually see/try it for myself. I don't
know of any Symbolics or LMI machines in the area, unfortunately. But
I do know my department is supposed to keep, in some near-mythological
storage area, an old Xerox lisp machine. My question is: Are they as
interesting, and worthwile to spend a few days or weeks to get
running?
--
Frode Vatvedt Fjeld
Kent M Pitman
> Take what you may be used to with plug-ins and web-enabled applications,
> and multiply the convenience and integration by 100, and that's what Genera
> is like.
>
> Unfortunately, I think it may be close to infeasible to implement something
> like this in conventional systems. LispM's have the benefit of everything
> being in a single, shared address space, so it's easy to develop relatively
> ad hoc mechanisms for accessing related applications and data. In a more
> conventional OS, it's necessary to design interprocess communications
> protocols to effect all of this. It's hard to ensure that these protocols
> provide all the flexibility to support features that you haven't even
> thought of yet. So basically, all we have are a handful of basic requests,
> like "tell the browser to go to this URL", and "tell the plug-in to display
> this data in the browser window".
Well, actually, what I find a more indicative pointer is that when COM
and CORBA and such thing came along, it seemed like there would
finally be an opportunity for people to publish protocols for
customization of applications by plugin. While a few limited places
did create "markets" (photoshop filters come to mind, or different
graphics editors for editing a picture seen in a text editor), there
wasn't a cultural excitement about doing this in a place where money
was not to be made. Users don't do this kind of thing in one-liners,
nor do they have personal choice to easily override the defaults in
most such applications, _especially_ not in a system-wide way and
_especially_ not in a source-code form that can be trivially shared
with a friend by saying "put this line of text in your init file".
I would have expected that I could have taken system dialogs and
notepad-like input windows and specified, in some general way, "I
prefer to use Emacs everywhere on my system that text editing is done"
and had makers of such tools suddenly be aware. That was what I
understood to be the dream of COM and CORBA. But instead what I got
as "If you have really a lot of money (substitute "time" and
"knowledge", equally scarce commodities, on a "free software system),
you can maybe compile up an interface that one or two applications
will grudgingly acknowledge.
I don't know if this was a failure of leadership or a cynical
realization on the part of commercial leadership that this much power
would make it easy to lock out a particular vendor at the touch of a
button. It certainly is something that couldn't have taken off if use
of the underlying connectivity substrate was something that had to be
licensed, and it certainly is something that wasn't easy to do if you
couldn't enforce that everyone used the _same_ connectivity substrate.
In the Lisp Machine, everyone didn't have to use the same
connectivity, but common substrates were there for people who wanted
to use them, and the leadership showed how use of those common
substrates would make everyone happier.
I have often called early Lisp the "socialist state", by which I mean
the time when mommy/daddy (vendors) took care of us children (users).
What has enabled the modern free market is the loss of
synchronization, which held a whole community back in order to make
sure we were all on the same page. The cost of being able to move as
fast as the modern world does is that we are no longer all on the same
page, and so any attempt to solve a problem that requires global
synchronization is a lost cause. Some would say, and I'd agree, that
it was inevitable that the socialist state would fall, at least on the
first round, to the capitalistic one that doesn't protect everyone
equally. And I think it's inevitable that we'd lose things like what
we had.
If you want something to hope for [and be careful what you do hope for]
it's probably that the capitalist market will so quickly outrun our
needs that it will drown us in things we have no desire or use for. As
it does, companies will go out of business because a great deal of their
base is left behind in a world where such products make no sense or where
they can't be afforded. In the end, this meltdown is likely to produce
monopolies (social or economic; linux is of the former kind and is not
exempt from this criticism), and monopolies are a lot like socialist
states. So the "we know better than you" effect, whether from Microsoft
(as the "single mind knowing better than the world") or from Linux
(as the "multiple minds making something so big that even though anyone
could in principle do it again, no one is willing to try"), is likely to
come again. And as we sit there in a computational glut that is not
controllable in either case (Microsoft because it doesn't need to care
what customers think and Linux because its advocates, in the aggregate,
don't believe in being told what to do, and don't have a way to be told
even if they wanted to be) that everything else will fall and there will
be again only one way. And in such a world, all other competition having
been killed, maybe there will be a chance for LispMs again. If anyone
cares to bother, or even remembers the dream, both doubtful.
Barry Margolin
>connectivity, but common substrates were there for people who wanted
>to use them, and the leadership showed how use of those common
>substrates would make everyone happier.
The articles I've read about Tim Berners-Lee's ideas for the "Semantic Web"
suggest that he thinks XML and HTTP can be used as the new substrate for
this. I'll believe it when I see it.
And that is still only the data storage substrate. It's still up to
individual applications to integrate with each other.
Kent M Pitman
> In article <sfwk7s0...@shell01.TheWorld.com>,
> Kent M Pitman <pit...@world.std.com> wrote:
> >In the Lisp Machine, everyone didn't have to use the same
> >connectivity, but common substrates were there for people who wanted
> >to use them, and the leadership showed how use of those common
> >substrates would make everyone happier.
>
> The articles I've read about Tim Berners-Lee's ideas for the "Semantic Web"
> suggest that he thinks XML and HTTP can be used as the new substrate for
> this. I'll believe it when I see it.
I agree this is a stretch.
In the Lisp Machine, things like function calls and plists were intrinsic
to the representation. The notation offered direct access to power already
present.
XML and HTTP are 'marshalling' (to use the CORBA term) of functionality
and representations not necessarily present in the underlying representation.
This allows data hiding and is not without virtue, but in a sense, part of
the power of the Lisp Machine is about the inability to hide data. It would
not surprise me, btw, if the power of the brain works likewise; certainly it
would be an interesting factoid to know.
> And that is still only the data storage substrate. It's still up to
> individual applications to integrate with each other.
I'm not sure I understood where you were going with this remark.
I'm pretty sure I agree with the second sentence but am not sure why
the first is introducing it.
Barry Margolin
>> And that is still only the data storage substrate. It's still up to
>> individual applications to integrate with each other.
>
>I'm not sure I understood where you were going with this remark.
>I'm pretty sure I agree with the second sentence but am not sure why
>the first is introducing it.
Integrated applications require two things: a way to share data in a common
format (the single address space and CLOS/Flavors data representation on
the LispM, HTTP and XML in the Semantic Web), and ways for the applications
to tell each other to do things. On the LispM, presentation methods
provide the latter, but there isn't an analogous protocol for interaction
among applications on conventional systems.
Eric Moss
>software/whatever idea, and I'm not disagreeing with a lot of your
>points. But it seems to me that in the current climate, that's
>the only place there would be a chance (minimal though it may
>be) of recreating the necessary infrastructure.
>Both in terms of getting enough programming mindshare to implement
>it, and enough user mindshare for it to matter.
Wouldn't it be nice if when companies "gave up" on a product for whatever
reason, they opened the source? Maybe we should make that a rule of
incorporation--as long as you keep trying, the work is yours, but don't
hoard it if you can't or won't let it benefit the world. Or something like
that.
As it stands, we have to rely on the public conscience of owners, who often
lack the public conscience of the original authors. For example, I have
seen more than a few nice pieces of software get bought up by some company
who does nothing with them. The current users can't get support and their
investment is lost, even though they are willing to pay. This happens
because either the company bought the code to stifle it, or they got
sidetracked or couldn't sell enough to make it attractive for them, and
couldn't or wouldn't sell to someone to whom it was attractive.
I know I'd be frustrated to see my code not sell but be popular in the
open-source world, but it's got to be better for everyone to not have it
cease being.
Eric
Thomas Bushnell, BSG
> As it stands, we have to rely on the public conscience of owners, who often
> lack the public conscience of the original authors.
This bleeds back into the public conscience of the original authors.
When they sell to the highest bidder, in disregard of what will happen
after they leave, they have weakened or abandoned their public
conscience too.
Kent M Pitman
> Wouldn't it be nice if when companies "gave up" on a product for
> whatever reason, they opened the source?
This would mean that the original author did not get the benefit.
> Maybe we should make that a rule of incorporation--as long as you
> keep trying, the work is yours, but don't hoard it if you can't or
> won't let it benefit the world. Or something like that.
It is overly simplistic to try to boil down all uses of intellectual
property to hoarding. I find the mere use of this term without
qualification as to what is going on just as offensive as one would
think if one substituted the word "murder" for all uses of "kill", or
the word "destroy" for all uses "edit".
Capitalism and intellectual property rights are not about benefiting
the world in the micro. It is about the idea that the sum of the
actions of a number of people people motivated by private gain will,
collectively, benefit the world more than people who have nothing to
gain.
You may have missed it, but an experiment was tried in the Soviet
Union in the last century where people didn't have anything personal
to gain and while it may be true that the things that got made
benefited people evenly, most analysts seem to find the evidence clear
that the sum total of things that anyone was motivated to make was
much smaller than would have been collectively made if there had been
an individual stake in it for the people making them.
> As it stands, we have to rely on the public conscience of owners,
> who often lack the public conscience of the original authors.
It isn't clear that this is bad. Even if a company stops doing
business, someone still owns those assets and is probably trying to
figure out how to make something out of it. Their situation will not
be improved by having the government involuntarily yank from them the
assets they have left. I hate seeing this picture painted as if the
rich were keeping things from the poor, and then having the solution
be to take from the very poor and give to those who don't plan to make
a profit anyway... What a weird kind of neo-Robin-Hood effect that is.
We don't take from the rich, but we notice that when people are laying
in a ditch with a broken leg that the gold in their wallet is not getting
used to benefit the economy, so we take it and have it shipped to a far
away land that uses gold not for economic power but for "Monopoly money"
(pardon the ironic name of this last metaphor--I mean for gaming).
Also, even if it was right to rob someone when they were economically down
for the good of others, what makes you think the original authors would
or should open source it? My second choice would be to have ownership
revert to the people who did the coding, still for their own personal gain.
If they want to give away their work, that's great. But they should not
be obliged to. They already put time into making this stuff. Why shouldn't
they get more benefit than the average person?
But neither of these solutions is good. Read on.
> For example, I have seen more than a few nice pieces of software get
> bought up by some company who does nothing with them.
Me, too. But maybe the loss of that software is the price of our collective
education in writing better contracts with employers. Nothing forbids you
from asking an employer to make such an agreement part of your contract.
> The current
> users can't get support and their investment is lost, even though
> they are willing to pay. This happens because either the company
> bought the code to stifle it, or they got sidetracked or couldn't
> sell enough to make it attractive for them, and couldn't or wouldn't
> sell to someone to whom it was attractive.
>
> I know I'd be frustrated to see my code not sell but be popular in the
> open-source world, but it's got to be better for everyone to not have it
> cease being.
I don't agree.
Suppose I do a big multi-year project that embodies a great deal of my
personal knowledge of how to do something. I can always get another job
with someone else to do the same thing over if there is no one out there
competing with it. But if someone open-sources what I did, then I am
competing with my own self. You may think I deserve this, but my point is
that it is not unambiguously better for me to be competing against two or
five or ten years of my own work. You might also think it allows me that
much of a head start on a new project, but that's irrelevant. Once it's
open-sourced, my competitors have the same head start. Open source does
nothing except raise the base against which progress is measured. It adds
to the world but not to the commercial world. For people who want to make
money on the game, it just makes them run very fast and denies the value of
all the running they have ever done in the past because they receive no
ongoing compensation for that. It's a good recipe to get sick and die and
have someone say to you "tough luck. we don't owe you for past work. we
paid you for that already. you have no entitlement to get sick, to go
on vacation, nor to retire".
I don't by any means lament that the work I put into Symbolics is privately
held. I hope someone figures out how to make some money from it.
Their figuring out how to make money on it will benefit me far more than
a bunch of hobbyists figuring out how to distribute it for free and further
driving down the amount I can charge for other things I make for pay.
Free software may be differently motivated, but it has the same commercial
effect as dumping. It puts legitimate investors at a commercial disadvantage
because they have to compete with people who didn't pay money to get the
value they are offering. What built things like Lisp Machines was the
extra capital that came from briefly succeeding wildly enough that it was
possible for a while to invest in the future. But the Lisp Machine also
disproves the central thesis of most free software rants--that capitalism
leads to people just having free money forever at the expense of the little
guy. The fact is that if you charge too much for something, you don't stay
in business. So the market already corrects itself against people who
overcharge... it didn't need free software to correct that. Free software
is having a completely different effect than it intends, which is to make
the computer business a non-business. There may still be computing if it
succeeds, but it won't be a place to make money. It will be more like
farming ... where anyone can buy the seeds needed to plant a garden, but no
one but a few big companies can find a way to make any money off of it.
The economy of scale won't be right.
Thomas Bushnell, BSG
> [....] Open source does nothing except raise the base against which
> progress is measured. It adds to the world but not to the
> commercial world. For people who want to make money on the game, it
> just makes them run very fast and denies the value of all the
> running they have ever done in the past because they receive no
> ongoing compensation for that.
Why should software be exempt from the same rules as the rest of the
world? General Motors, for example, makes cars. Each car they
produce is (more or less) that much more value added to the world.
And their profit is (more or less) proportional to the number of cars
they sell.
But they get absolutely no ongoing profit for all the great cars they
sold in the 50s. They don't get to say "hey, you should keep paying
us for that great car you drove back in 1953!" Instead, they must
*keep running*, and do so *very fast*, because, macro-economically,
that makes for better/faster/cheaper cars for everyone.
> It's a good recipe to get sick and die and have someone say to you
> "tough luck. we don't owe you for past work. we paid you for that
> already. you have no entitlement to get sick, to go on vacation,
> nor to retire".
Which is *exactly* what *everyone* in *every* industry expects aside
from computer programmers. Steel workers have to bargain with their
employers to get things like sick leave, vacation, and retirement
benefits. If you want such things, in a capitalist economy, you need
to arrange for them from the profits you make. You are not owed for
past work, and if you want to be paid for past work, you have to
bargain for it *at the time* as a condition of doing the work.
> What built things like Lisp Machines was the
> extra capital that came from briefly succeeding wildly enough that it was
> possible for a while to invest in the future.
Naw, what built things like Lisp Machines was the free software
efforts of people in the MIT AI lab, until the day came that Symbolics
decided the owned all that collective work.
> Free software is having a completely different effect than it
> intends, which is to make the computer business a non-business.
> There may still be computing if it succeeds, but it won't be a place
> to make money.
Hogwash. It won't be a place to make billions, but why should
computer programmers make more than steel workers or building
contractors or the rest of the world that doesn't get to do the job
once and expect to keep getting paid for it?
You know, people told RMS way back when "this will kill computer
programmin as a profession". He said "hogwash". Well, Unipress went
under. But you know what? Unipress Emacs *sucked*. Meanwhile, there
are all kinds of programmers making happy comfortable livings doing
exclusively free software.
Your claim that "it won't be a place to make money" might have been
plausible fifteen years ago, but it turns out that you're just wrong.
Thomas
Erik Naggum
| Why should software be exempt from the same rules as the rest of the
| world? General Motors, for example, makes cars.
No, they spend billions of dollars designing cars in a shroud of secrecy
that should frighten people, and then they sell instantiations of their
design to recover their design costs. Then they make a profit on the
sale of such instantiations after their cost has been recovered. Not all
of their designs pay off this way, and "profit" is simply a way to make
sure that you can afford mistakes, which is what innovation is all about.
If they had to give away the design to other competitors, new cars would
not be made, because there is no way anybody would spend 1 billion
dollars if they had to give away the means to produce 5 billion dollars
to people who had not spent that 1 billion to get those means.
| Each car they produce is (more or less) that much more value added to the
| world. And their profit is (more or less) proportional to the number of
| cars they sell.
What utter nonsense. Their costs are fairly fixed and there is not a
shadow of any profit until a really large number of cars have been sold.
After that, it is all profit. Pricing your products and selling enough
to break even with your development cost is damn hard.
| But they get absolutely no ongoing profit for all the great cars they
| sold in the 50s.
The cars are unimportant. The design is what counts. This is the same
as for the development of drugs. An astonishingly huge amount of money
goes into testing drugs, but if they could not protect their investments
with patents and secrecy, anybody could make more of the drug at
virtually no cost, which is precisely what happens when the patents
expire or some rinky-dink country with a lot of criminals in power and no
legal protection against theft of intellectual property reverse engineer
the drugs and make cheap knock-offs. It is not that the original drug
company does not make money on producing cheaper pills, but they do not
recover their research costs with cheaper pills, and thus cannot make
_more_ drugs. Requiring the drug companies to lower their profits to get
closer to "mass production cost" is the same as "give us your research".
| They don't get to say "hey, you should keep paying us for that great car
| you drove back in 1953!"
What do you think spare parts are? Why do you think original spare parts
cost so much more than cheap knock-offs? Why do you think that a car
manufacturer loses all interest in protecting you from harm if you do not
use original spare parts? They have invested in making things that they
know are safe to put together. They have no clue how safe it is to run a
car with unoriginal parts -- other things might break because that part
had different resonance characteristics, for instance.
> It's a good recipe to get sick and die and have someone say to you "tough
> luck. we don't owe you for past work. we paid you for that already. you
> have no entitlement to get sick, to go on vacation, nor to retire".
| Which is *exactly* what *everyone* in *every* industry expects aside from
| computer programmers.
That is because every other industry has _not_ conflated design and
production. I mean, an author gets paid for each book sold even though
he wrote it only once, but the guy who printed the book gets paid for
each book. Airing rights for TV shows pay the company that made it back
for its successful series so they can innovate and make things that do
not succeed. If all series or shows had to be succeesses on their own
and only recover their costs with no option of getting paid more after
the work had succeeded, we would have even less intelligent series and
shows on TV.
| You are not owed for past work, and if you want to be paid for past work,
| you have to bargain for it *at the time* as a condition of doing the
| work.
You know, in most industries, the designers of things are actually paid
for past work. It is called royalties. Industrial designers are paid in
large part by royalties. All kinds of inventors protect their inventions
so that they can, indeed, get paid for the use of that invention. This
is what the whole patent system is for. Copyright is about paying people
for past work. I marvel at the strangeness of the selective ignorance
that has managed to avoid learing about royalties in _all_ industries.
However, in most other industries, those who produce objects according to
a design are much more numerous than the designers. Mass production is
_such_ a wonderful thing, as it allows that many times more money to go
into research and development than hand-made individual objects where the
design work is even more miniscule compared to the production labor cost.
| Naw, what built things like Lisp Machines was the free software efforts
| of people in the MIT AI lab, until the day came that Symbolics decided
| the owned all that collective work.
What free software efforts? Were they "free" as in the billion-dollar
budgets of MIT, of building the environment that made it possible? As
far as I can tell, MIT has always been _extraordinarily_ good at making
money off of _their_ past work. In fact, they are so good at it that
they have remained a central and powerful research institution for a
_really_ long time, much longer than most of the commercial operations
that have spawned from their research. Why do you think this is?
| Hogwash. It won't be a place to make billions, but why should computer
| programmers make more than steel workers or building contractors or the
| rest of the world that doesn't get to do the job once and expect to keep
| getting paid for it?
Because they can use their intellect to get work that they keep getting
paid for. On other hand, the current crop of programmers are basically
steel workers and deserve even less money than unskilled labor because
they are, in fact, unskilled labor in an industry where people pay very
little for their production tools. That does not mean that everybody in
this industry is an unskilled worker.
| You know, people told RMS way back when "this will kill computer
| programmin as a profession". He said "hogwash". Well, Unipress went
| under. But you know what? Unipress Emacs *sucked*. Meanwhile, there
| are all kinds of programmers making happy comfortable livings doing
| exclusively free software.
But will they regret it later on? And what will they do when they get to
that point? Maybe some of us old farts already regret giving away our
work, or actively regret having to compete with people who have access to
our past intellectual investment. Not all that goes into research and
development end up visible in the products. E.g., what _not_ to use or
do never gets out there. Reverse engineering is generally not allowed
because it uncovers information not required to _use_ the object, but
instead is required to compete with the original designer on unfair
terms. The whole industrial world is chock full of these things.
| Your claim that "it won't be a place to make money" might have been
| plausible fifteen years ago, but it turns out that you're just wrong.
Kent is not wrong just because you are also right in some respects.
Acquire more bits.
Thomas Bushnell, BSG
I'm not particularly interested in a lengthy flame war; I pretty much
said what I meant to say. Of course I disagree with some of Erik's
criticisms, but it's pretty off topic to continue further. Just a few
brief observations.
Erik Naggum <er...@naggum.net> writes:
> If all series or shows had to be succeesses on their own
> and only recover their costs with no option of getting paid more after
> the work had succeeded, we would have even less intelligent series and
> shows on TV.
It's very interesting that you say this. Once there was essentially
nothing like a copyright system in Europe, which had no shortage of
brilliant works of art being produced. Indeed, the "true artist", who
produces the best art, art for the sake of art, is likely to produce
great art under any system that gives him enough money to live on,
which the patronage/performance system did quite nicely for musicians,
and the commission system did for playwrights.
Indeed, even today, the finest television programming I get comes in
from PBS. (Well, the best comedy is on C-SPAN.)
So it's not so clear what you say. It seems to me that the copyright
motive doesn't so much encourage *brilliance*, per se.
> | Naw, what built things like Lisp Machines was the free software efforts
> | of people in the MIT AI lab, until the day came that Symbolics decided
> | the owned all that collective work.
>
> What free software efforts? Were they "free" as in the billion-dollar
> budgets of MIT, of building the environment that made it possible?
So, when I say "free software", I mean libre, not gratis, that is,
free as in "free speech", not as in "free beer". I'm surprised you're
apparently unfamiliar with that usage. I'm sorry for creating the
confusion. In case other people were similarly misled by my
terminology, you can read up on what I meant at
http://www.gnu.org/philosophy.
> | Your claim that "it won't be a place to make money" might have been
> | plausible fifteen years ago, but it turns out that you're just wrong.
>
> Kent is not wrong just because you are also right in some respects.
> Acquire more bits.
I meant only narrowly that he was wrong in saying that making bucks is
impossible or unlikely in a free software (again, libre, not gratis),
not that I had demonstrated that everything he said was incorrect.
Thomas
Michael Parker
My point is not that the lispms couldn't be hooked into the world, but
that
a vendor can spend it's resources on lispm-like coherency, or it can
sacrifice that in order to more closely track the rest of the world.
It can't
really do both (not even with wads of resources), because the rest of
the world is changing so quickly.
I'm not saying that the Lispms couldn't track the rest of the world,
but
that it was manifestly more difficult to do so than it is for a GP
lisp
implementation. Franz and Harlequin didn't have to write their own
window manager -- they used the existing ones. They didn't have to
"whip up" a database interface -- they just had to wrap the existing
interfaces up with FFI. They didn't have to reimplement COM from
scratch. On the Lispms you would have had to -- for every non-statice
database you wanted to talk to, and every time MS came up with its
newest great thing (.NET, SOAP, non-Kerberos, etc).
The gp vendors gave up the lispm's coherency, since they were having
to live on top of other companies' (frequently lousy) software rather
than
reimplement it in a manner consistent with their existing paradigm.
But staying current (or at least no more than a generation behind) did
more
for their survival than consistency would have.
Frank A. Adrian
> The articles I've read about Tim Berners-Lee's ideas for the "Semantic
> Web" suggest that he thinks XML and HTTP can be used as the new substrate
> for
> this. I'll believe it when I see it.
As well you should. At its heart, XML's DTD's provide only syntactic
commonality. Nothing in the standard talks about a shared ontology or
semantics. And, as much as Berners-Lee would like to find something to
provide an ontological specificaton (i.e., his "Semantic Web"), it seems
unlikely that this set of ontological specifications will work that much
better than KIF, CYC's meta-levels, or any of the dozens of similar
approaches that have been tried previously. I fear that disappointment
will be the likely outcome.
But we can always hope that this goal will once again focus researchers on
the "interesting" problems in knowledge representation and manipulation.
This might actually push symbolic computation forward to a state where Lisp
might become viable again (remember that XML is just S-exprs in disguise
and it's a lot simpler to manipulate S-exprs than XML strings). But with
our luck, we'll be stuck with manipulating XML strings in Perl and/or Java
and/or PHP and/or any of the other web monstrosities created lately. To
say I am less than hopeful would be overstating the case.
faa
Kent M Pitman
> I'm not particularly interested in a lengthy flame war; I pretty much
> said what I meant to say. Of course I disagree with some of Erik's
> criticisms, but it's pretty off topic to continue further. Just a few
> brief observations.
Would that it were off topic. I don't think anything could be more central
to the discussion of Lisp than how to make a buck off of it. I would bet
that if you asked people what one choice more than any other motivates their
use of a language you wouldn't get "where they put their parens" but rather
"how quickly I can make enough money to pay my mortgage".
> Erik Naggum <er...@naggum.net> writes:
>
> > If all series or shows had to be succeesses on their own
> > and only recover their costs with no option of getting paid more after
> > the work had succeeded, we would have even less intelligent series and
> > shows on TV.
>
> It's very interesting that you say this. Once there was essentially
> nothing like a copyright system in Europe, which had no shortage of
> brilliant works of art being produced. Indeed, the "true artist", who
> produces the best art, art for the sake of art, is likely to produce
> great art under any system that gives him enough money to live on,
> which the patronage/performance system did quite nicely for musicians,
> and the commission system did for playwrights.
This statement is nonsense, by the way. It took forever for the snobs at
the Oscars to acknowledge that the enormous box office draw of many of
Spielberg's movies warranted any kind of award. There's a desire on the
part of some people to claim that "true art" cannot be "popular art". This
is rubbish. I'm fine with the pieces people call "true art" just so long
as they don't try to claim the _name_ "true art". Art is an individual
thing. And much of the art that I and others would trade our hard-earned
dollars for is not what you're calling "true art". That may be an unpleasant
truth, but it is a truth nonetheless. And that art, true or not, would not
be getting made if there were not lots of $$ available for having made it.
In many cases, only that large number of $$ is what has made it possible to
produce the cinematic techniques and to produce things Star Wars or Shrek
or (which I saw last night and enjoyed) Ice Age.
That doesn't mean other art wouldn't be made if you eliminated copyright.
It just means that a lot of things we're used to now would not get made.
> So, when I say "free software", I mean libre, not gratis,
Libre as in "it's free for everyone except the people who made it to decide
how it's used". That's not "free", that's "stolen". Like as in the term
that was used by some friends of mine when they used to go shoplifting:
"I think I'll 'liberate' this candy bar." I'm sure the candy bar felt good.
The people who ate it probably enjoyed its freedom. But the person who
made it wouldn't have bothered if they thought this was how it would end up.
> that is,
> free as in "free speech", not as in "free beer". I'm surprised you're
> apparently unfamiliar with that usage.
We just don't believe in this usage.
> I'm sorry for creating the
> confusion. In case other people were similarly misled by my
> terminology, you can read up on what I meant at
> http://www.gnu.org/philosophy.
It's been amply discussed.
> > | Your claim that "it won't be a place to make money" might have been
> > | plausible fifteen years ago, but it turns out that you're just wrong.
> >
> > Kent is not wrong just because you are also right in some respects.
> > Acquire more bits.
>
> I meant only narrowly that he was wrong in saying that making bucks is
> impossible or unlikely in a free software (again, libre, not gratis),
> not that I had demonstrated that everything he said was incorrect.
I did not say it was impossible to make bucks by doing OTHER things than
writing software.
I said that for those who want to make their business writing
software, free software injures their market and risks putting them
out of business. This seems to be a cost you're prepared to accept.
But then, you aren't in that kind of business. What's curious about
this kind of reasoning, in my experience, is that the usual motivation
for people going into free software is that they perceive some other
uncaring soul is out there making decisions they perceive to be
callously injuring the things _they_ care about. But I guess the
tactic of injuring without caring is ok when it's done for a noble
reason.
Kent M Pitman
> My point is not that the lispms couldn't be hooked into the world,
> but that a vendor can spend it's resources on lispm-like coherency,
> or it can sacrifice that in order to more closely track the rest of
> the world. It can't really do both (not even with wads of
> resources), because the rest of the world is changing so quickly.
But, importantly, the lispm was not just about editing lisp code. It
was a whole way of life for computation. It was about text editing,
mail reading, program development, commercial quality graphics creation,
expert system shells, etc. It was a whole operating system and a whole
marketplace for products... just not a very big market because of how
expensive they were.
Lisp vendors of today can certainly make their lisp development programs
that are integrated among the lisp development tools, but they can't make
other non-lisp applications follow the same integration. So there is a
limit to what they can seek to do.
Erik Naggum
| It's very interesting that you say this. Once there was essentially
| nothing like a copyright system in Europe, which had no shortage of
| brilliant works of art being produced. Indeed, the "true artist", who
| produces the best art, art for the sake of art, is likely to produce
| great art under any system that gives him enough money to live on, which
| the patronage/performance system did quite nicely for musicians, and the
| commission system did for playwrights.
Where do you think the people who could pay for these artists got their
money? I am not a big fan of theft and robbery and feudalism just to be
able to fund "real artists". They money has to come from _somehwere_,
the expression "to make money" to the contrary notwithstandning.
| Indeed, even today, the finest television programming I get comes in from
| PBS.
Which, to my utter dismay when I found out, is funded with tax money, the
supreme form of coercion in modern society, not much different from
"protection money" paid to criminal gangs.
| So it's not so clear what you say. It seems to me that the copyright
| motive doesn't so much encourage *brilliance*, per se.
It provides a means to get enough money to pay for the real artists.
Please note that the proceeds on successful books pays for all the books
the same poublishing company has published at a sometimes great loss. I
am at a loss for how "*brilliance*" is a function of any political
system. Some individuals do amazing stuff regardless of which political
system they live under, so obviously the brilliance of an individual is
not an argument for the political system. The question is rather at what
cost their brilliance comes, and where a brilliant person chooses to
direct his brilliance.
| So, when I say "free software", I mean libre, not gratis, that is, free
| as in "free speech", not as in "free beer".
*sigh* This really has become a "dimwith alert". People are free, and
some lunches are free, but that does not mean that some people are
lunches any more than it means that some lunches have freedom. Same with
software. However, this abuse of the word is well known, albeit counter-
productive, as some people actually imagine that the two do not overlap,
that one _cannot_ argue against freedom with costs, but all freedom has
very significant costs. Free is not free. (See how confusing that gets
and how capitalization does _not_ work to distinguis them? :)
| I'm surprised you're apparently unfamiliar with that usage.
Will you please stop insulting me with your own stupidity? This is so
goddman annoying. You _know_ I have worked with Richard Stallman both
personally and on GNU Emacs. Damnit, countering such moronic nonsense is
_really_ taxing on my patience with some people.
What would happen if you just assumed I knew both usages? MIT has some
of the most severe licensing agreements I have ever seen, and actually
taking your _own_ stuff out of MIT is no walk in the park. People pay
huge amounts of money for this right in many cases, so MIT can keep doing
world-class research. They also get paid enormous amounts of money to do
research on demand, but still keep rights to the research. MIT is the
_one_ place in the world I would choose to point out that organizations
do make money on past work if I were allowed only one example.
| I meant only narrowly that he was wrong in saying that making bucks is
| impossible or unlikely in a free software (again, libre, not gratis), not
| that I had demonstrated that everything he said was incorrect.
How do you know this? I mean, people argued that Pol Pot was not a
serious threat to anyone, either, because "lookit all them survivors!
must be millions and millions". For some people, it has indeed become
impossible to make money they wanted to make in a line of work they
appreciated, and they have gone on to green pastures. I mean, all the
Linux distribution companies seem to struggle. I have some inside
information that will probably become public in about four weeks unless
they manage to get new investors that yet another Linux distribution
company is about to fold. Many software "developers" are simple steel
workers and just code from 9 to 5 and think not at all, but these work in
the same kind of industries where the designs have been drawn up by other
people. Such software may well be free, because the designs have already
paid for itself in the sold products. Where the software is the probuct,
things look very different.
I regard Open Source and Free Software as building infrastructure, so
that people can profit on using it rather than having to fight the lack
of it. I consider this operation akin to the development of railroads in
the United States. Whatever private successes there were, it was
eventually nationalized and is now running at a loss, failing to compete
with the road system, which has been paid for by another public money
pool, although privately owned roads are still an option in many places,
and users pay for bridges and roads all over the place with tolls. This
is not quite how we can make software work, however, so the problem is
that people who want software to be "free" fail to grasp that in order to
build infrastructure, you cannot require that all usage be free, too.
You cannot give people a road "for free" and demand that cars that use it
must be given away by the car company, that goods shipped on it must be
"free" with the notion that "traffic should be free".
Tim Moore
>tb+u...@becket.net (Thomas Bushnell, BSG) writes:
>> So, when I say "free software", I mean libre, not gratis,
>
>Libre as in "it's free for everyone except the people who made it to decide
>how it's used". That's not "free", that's "stolen". Like as in the term
Do you have an example of this?
>that was used by some friends of mine when they used to go shoplifting:
>"I think I'll 'liberate' this candy bar." I'm sure the candy bar felt good.
>The people who ate it probably enjoyed its freedom. But the person who
>made it wouldn't have bothered if they thought this was how it would end up.
>
>> that is,
>> free as in "free speech", not as in "free beer". I'm surprised you're
>> apparently unfamiliar with that usage.
>
>We just don't believe in this usage.
"We?"
Tim
Thomas Bushnell, BSG
> Would that it were off topic. I don't think anything could be more central
> to the discussion of Lisp than how to make a buck off of it.
If Symbolics' code were free software, people would still be making
bucks off of it now. As it is, it seems likely to be permanently
dead, and make nobody any more bucks.
> This statement is nonsense, by the way. It took forever for the snobs at
> the Oscars to acknowledge that the enormous box office draw of many of
> Spielberg's movies warranted any kind of award. There's a desire on the
> part of some people to claim that "true art" cannot be "popular art". This
> is rubbish. I'm fine with the pieces people call "true art" just so long
> as they don't try to claim the _name_ "true art".
The reason I put "true art" in square quotes is precisely because it's
not a term I place any stock in. I'm not holding to some highbrow
definition. My point is that many many people produce really good art
for entirely non-commercial reasons, and more to the point, that the
commercial motives provided by the commission, patronage, and
performance systems are pretty darn good motives.
> Libre as in "it's free for everyone except the people who made it to decide
> how it's used". That's not "free", that's "stolen".
There are two confusions in this sentence.
First, copyright does *NOT* give authors control over "how it's
used". Authors have, in general, no rights to determine how it's
used. They can control copying and public performance, and other than
that, it can be used however the owner of the copy chooses.
Second, libre software is libre *by the decision of the author*. It's
not "stolen". It's not "liberated". It's a gift of the author to the
world at large.
None of the free software out there is somehow "stolen", but you make
up some slander to make it look like somehow GCC (or, say, GNU Common
Lisp) is "stolen". Hardly.
> We just don't believe in this usage.
1) Who is this "we"?
2) You don't believe that's my usage? What is it to "believe in a
usage"?
> I did not say it was impossible to make bucks by doing OTHER things than
> writing software.
Um, it's possible to make money *BY* writing free software. Really.
I know all kind of people doing it.
> I said that for those who want to make their business writing
> software, free software injures their market and risks putting them
> out of business. This seems to be a cost you're prepared to accept.
Why should I particularly *care* if they go out of business? This is
like saying it's unfair for McDonalds to produce cheaper hamburgers
because it runs the risk of putting Burger King out of business.
Or that Henry Ford was unfairly putting lots of buggy-whip makers out
of business.
> But then, you aren't in that kind of business. What's curious about
> this kind of reasoning, in my experience, is that the usual motivation
> for people going into free software is that they perceive some other
> uncaring soul is out there making decisions they perceive to be
> callously injuring the things _they_ care about. But I guess the
> tactic of injuring without caring is ok when it's done for a noble
> reason.
Wait, I thought you were the advocate of free market capitalism here.
Suddenly competition is a bad thing?
But, really, what is it that you want? What I want to do is
encourage people to write free software, because in that way they are
helping their neighbors more than if they wrote non-free software.
I'm not trying to overturn the copyright system.
What is it that you want to do?
Thomas Bushnell, BSG
> Where do you think the people who could pay for these artists got their
> money? I am not a big fan of theft and robbery and feudalism just to be
> able to fund "real artists". They money has to come from _somehwere_,
> the expression "to make money" to the contrary notwithstandning.
Um, there are plenty of really rich people around even today.
> Will you please stop insulting me with your own stupidity? This is so
> goddman annoying. You _know_ I have worked with Richard Stallman both
> personally and on GNU Emacs. Damnit, countering such moronic nonsense is
> _really_ taxing on my patience with some people.
Then why did you play dumb?
> What would happen if you just assumed I knew both usages? MIT has some
> of the most severe licensing agreements I have ever seen, and actually
> taking your _own_ stuff out of MIT is no walk in the park.
Yes, and I never said the contrary. What I said was that that
particular code was free software originally. MIT produces huge
amounts of free software, and it also produces things under severe
licensing agreements. I was speaking of lispm code that was decidedly
in the former category (until Slimbolics M-DEL Symbolics decided they
owned it). There's an example of theft!
Thomas
Kent M Pitman
> On Tue, 26 Mar 2002 05:02:55 GMT, Kent M Pitman <pit...@world.std.com> wrote:
> >tb+u...@becket.net (Thomas Bushnell, BSG) writes:
>
> >> So, when I say "free software", I mean libre, not gratis,
> >
> >Libre as in "it's free for everyone except the people who made it
> >to decide how it's used". That's not "free", that's "stolen".
> >Like as in the term
>
> Do you have an example of this?
There are several issues here.
First, I was speaking about the hypothetical (and apparently
aspired-to) world in which people have only free software as an
option. In that world, when a person creates a piece of software,
they cannot dictate its use but any other person can decide how to use
it. That seems odd to me since the one person who it would seem
should be favored in such a decision is the one who made it. The
intended function of copyright is to incentivize creation and sharing
by saying that some value goes to the creator for having graced the
world with their work, and further that some protection goes to the
creator in exchange for sharing.
But, presently, it is only the case that free software seeks to "take
down" other ventures that charge money. It does this by making it
expensive to compete because one has to recover startup costs that do
not need to be recovered in the free software market. But, on the
other hand, in order to be part of this "nice community of sharing", I
have to buy into rules that I do not politically agree with. It's a
kind of self-destructive extortion.
The following is not the only reason for my belief on this issue, but
it is a concrete example of how I've seen this play out in micro; it
parallels my belief about how these things work in society at large: I
lived for a while in a sort of communal home with some people where
there was a rule that said "if you put something down in a public
space, anyone may unconditionally use it". This didn't result in lots
of my things getting shared. It resulted in lots of things getting
locked in my room to avoid what I perceived as abuse. I was then
accused of being "stingy" and "antisocial", which also annoyed me. I
didn't mind sharing but wanted to share MY things on MY terms, not terms
dictated by others. To me, "antisocial" is thinking you can use my
property, those things that are the products of my labors, without
regard to how I want those things used. To me, "social" is about
caring about individuals. To me, it is "social" when picking up an
object to learn whose object that is and how they want it used. To me
it is "antisocial" to think you can go around thinking that it's your
divine right to use anything that's made in the world without regard
to who made it and what their needs are. It's amazing to me, but it
appears to be true, that there are people who have these words/concepts
in their head (socialness and antisocialness) but with the polarity
of the meaning reversed...
> >that was used by some friends of mine when they used to go shoplifting:
> >"I think I'll 'liberate' this candy bar." I'm sure the candy bar felt good.
> >The people who ate it probably enjoyed its freedom. But the person who
> >made it wouldn't have bothered if they thought this was how it would end up.
> >
> >> that is,
> >> free as in "free speech", not as in "free beer". I'm surprised you're
> >> apparently unfamiliar with that usage.
> >
> >We just don't believe in this usage.
> "We?"
Sorry. Not really a bid to be declared royalty. I was going out on
what I hoped was a fairly sturdy limb and believing at least Erik, who
was already involved in the discussion, was on the same side as me on
this issue. Perhaps even others who haven't spoken... But the use of
"We" doesn't require an army.
Erik Naggum
| It's amazing to me, but it appears to be true, that there are people who
| have these words/concepts in their head (socialness and antisocialness)
| but with the polarity of the meaning reversed...
My take on this is that it depends on whether you consider yourself an
(aspiring) member of the haves or the have-nots. If you think you will
never get a break and will always live on borrowed time, you want to make
the most of it. If you think you will you get the breaks you work hard
for and live on earned time, you want to make the most of it. The former
consider it rude not to share something someone has with you on your
terms. The latter considers it rude for someone not to share something
you have with them on your terms. Specifically, if you are an author,
you want to get the most out of publishing your work. But as a reader,
you want to get the most ouf ot reading that work, too. As an author,
you will not be thrilled if someone copies your work so you get no money
from it, but as a reader, you will not be thrilled if you are barred from
copying a work and giving it to a friend to "share" it. In other words,
if you consider yourself a producer, you have respect for producers and
consider their rights, but if you consider yourself a consumer, obtaining
that respect appears to be quite hard, probably because it means _not_
getting some stuff you want, while one who considers himself a producer
knows that he can get the stuff he wants only if he lets other producers
get their stuff, too.
Only those who are more likely to offer than to need, will understand
that one can not place any demand on other people to give them anything.
Those who only need, will not generally understand what it feels like to
offer something for others and have them thank you with more demands.
Money is an equalizer here that simply does not apply to the kind of
good-will that forms the currency of the source-sharing community.
| Sorry. Not really a bid to be declared royalty. I was going out on what
| I hoped was a fairly sturdy limb and believing at least Erik, who was
| already involved in the discussion, was on the same side as me on this
| issue. Perhaps even others who haven't spoken... But the use of "We"
| doesn't require an army.
Yup, count me in that "we". However, please note that I used to be quite
a fan of free software, but my goal was different from what I saw become
the prevalent goal. I have wanted to obtain and grant access to source
to those who wanted to learn from it, whether it would be helpful to the
development of that source code or not, just like good authoers have to
read a lot of good code to learn what you can simply not learn without
reading other people's works. I had no political agenda with this at
all, and I came to think it was a good idea on very different premises
than were touted. Also, the Open Source movement seems to be an anti-
movement, not a pro-movement, and specifically anti-Microsoft, which I
am, too, but Open Source is unfortunately able to destroy much more than
Microsoft if it succeeds. I have become highly critical of the focus and
the political agenda, not to mention the fact that selling anything in a
market where you compete with Open Source is so much harder, but this
change has been gradual. I consider free software to be a bad way to
lure innocent students and hobbyists to give away their work and time,
which is only a good idea if they consider it learning and marketing for
their future carreers, but that is not how these things usually work.
Erik Naggum
| Then why did you play dumb?
Thomas Bushenll, THINK! Is there another possibility that you have
overlooked because _you_ are so hung up in your abusage of the word
"free" that you cannot see anything else? As I said: Freedom has costs.
You seem to want to ignore that completely, much like almost every other
"Free" Software proponent.
Wade Humeniuk
news:sfwzo0v...@shell01.TheWorld.com...
> tmo...@sea-tmoore-l.dotcast.com (Tim Moore) writes:
> But, presently, it is only the case that free software seeks to "take
> down" other ventures that charge money. It does this by making it
> expensive to compete because one has to recover startup costs that do
> not need to be recovered in the free software market. But, on the
> other hand, in order to be part of this "nice community of sharing", I
> have to buy into rules that I do not politically agree with. It's a
> kind of self-destructive extortion.
>
Many of the free software users, though they say that there are other ways
to make money (support, distibutions, custom programming) when it comes down
to it will not themselves pay for those services. Many Linux Distros go
under because the free software community will not pay ANYTHING to people.
Consider a distro company like Caldera, say the have 30 people working full
time for them, ~$50,000 per person per year or $1,500,000 a year just in
salaries. Where does that money come from? Distro sales of course, (at $30
a CD, that is many many sales) but hardly any Linux people I know would pay
(on some kind of principle) for any Linux (maybe just $5.00 for someone else
burning the distro, or dowloading it instead). They take pride in not
spending anything. Meanwhile they burn their own supporters, like Caldera,
at the stake. The same attitude goes for CD ripping and MP3 music swapping.
Self destructive behaviour and an implicit belief that they should not have
to support anyone else.
> The following is not the only reason for my belief on this issue, but
> it is a concrete example of how I've seen this play out in micro; it
> parallels my belief about how these things work in society at large: I
> lived for a while in a sort of communal home with some people where
> there was a rule that said "if you put something down in a public
> space, anyone may unconditionally use it". This didn't result in lots
> of my things getting shared. It resulted in lots of things getting
> locked in my room to avoid what I perceived as abuse. I was then
> accused of being "stingy" and "antisocial", which also annoyed me. I
> didn't mind sharing but wanted to share MY things on MY terms, not terms
> dictated by others. To me, "antisocial" is thinking you can use my
> property, those things that are the products of my labors, without
> regard to how I want those things used. To me, "social" is about
> caring about individuals. To me, it is "social" when picking up an
> object to learn whose object that is and how they want it used. To me
> it is "antisocial" to think you can go around thinking that it's your
> divine right to use anything that's made in the world without regard
> to who made it and what their needs are. It's amazing to me, but it
> appears to be true, that there are people who have these words/concepts
> in their head (socialness and antisocialness) but with the polarity
> of the meaning reversed...
I have been doing yoga for 10 years and there is also a microcosm of the
real world in the limited setting of a class. Usually people have their own
mats in class. When the class is moving around there seems to be two types
of people, those who will walk and stand on other people's mats and those
that will walk around and avoid other people's mats. To step, walk, tample
other's people's mats is considered rude and inconsiderate. But there are
still people who do it. When you point out what they are doing, they do not
even know they are doing it, they are UNAWARE of what they are doing. The
even stranger thing is that keep with this type of behaviour until (usually)
they get kicked out of class or leave or replace their behaviour with what
is based on "following the letter" of the rules.
To not walk on other people's mats is acknowledging other people and that
they have a right to exist in the class (world). It is a sign of respect.
The mat walkers are self-absorbed.
Wade
Barry Margolin
Thomas Bushnell, BSG <tb+u...@becket.net> wrote:
>
>I'm not particularly interested in a lengthy flame war; I pretty much
>said what I meant to say.
Unfortunately, once it gets started, it's hard to stop. Now I'm going to
have to put this thread in my killfile, since its signal-to-noise ratio has
become unbearable.
Craig Brozefsky
> Many of the free software users, though they say that there are
> other ways to make money (support, distibutions, custom programming)
> when it comes down to it will not themselves pay for those services.
> Many Linux Distros go under because the free software community will
> not pay ANYTHING to people.
This seems a bit poorly reasoned to me. There was a glut of Linux
distros in the last few years and the vast majority of them were
hardly worth the amount of plastic used in burning their CDs. Also,
they were very often poorly planned, under-funded, and added very
little value above and beyond what was available from cheaper and free
sources. To blame their failure on their customer base is misguided.
--
Craig Brozefsky <cr...@red-bean.com>
http://www.red-bean.com/~craig
Ask me about Common Lisp Enterprise Eggplants at Red Bean!
Thomas Bushnell, BSG
> But, presently, it is only the case that free software seeks to "take
> down" other ventures that charge money. It does this by making it
> expensive to compete because one has to recover startup costs that do
> not need to be recovered in the free software market.
How is it that those costs are not equally present in the free
software market?
Thomas Bushnell, BSG
> * Thomas Bushnell, BSG
> | Then why did you play dumb?
>
> Thomas Bushenll, THINK! Is there another possibility that you have
> overlooked because _you_ are so hung up in your abusage of the word
> "free" that you cannot see anything else? As I said: Freedom has costs.
> You seem to want to ignore that completely, much like almost every other
> "Free" Software proponent.
Of course it has costs. Nobody would say the contrary. Free software
is just as costly to produce as non-free software. (That is, when you
exclude the non-value-added parts of non-free commercial software like
packaging, advertising, and so forth.) The actual programmer labor is
pretty much the same. How could it be otherwise?
Erik Naggum
| Free software is just as costly to produce as non-nfree software.
*sigh* That is so not the issue. _Freedom_ is not costless.
Kent M Pitman
Well now, that's hitting the nail on the head now isn't it? :-)
And again, it shows the adjectives turned on their heads in a way I
would generally agree with. Because it's often the producers, the ones
who actively do the enriching of society, who get called the "stingy" and
"hoarder" and "self-serving". While the takers [I don't think "consumers"
is the right term, because there are "consumers" who are producers, and they
are not "takers"] are never examined for their public conscience. The
mere "buying into the Good Thought" is enough to label them "of good
conscience".
In Capitalism, the way you tell the consumers who are takers from the
consumers who are not is that they ones who are takers have no money,
because they never produced. But in Free-Software-Socialism, it's
more like the certain religions who tell you it's a sin(!) to practice
birth control and that having kids is always good ... and that the
only thing that matters is not how much money those people make or who
they help, but "do they believe". And I have to cynically wonder if
this is a recipe for a good society or a recipe for more votes at the
voting booth. I'm sure there are many people of very good intent who
subscribe to those religions, because they aren't into designing their
own and are just out there shopping for something pre-packaged. But
for the people who designed the religion, I have to wonder why they
didn't say "yes, some people die before they have a chance to
contribute so we'll give them a free pass to heaven because they
didn't have a chance, and some people are born with a legitimate
handicap and we'll give them a free pass too, if they just do the best
they can with what they have, but from the rest of you we need good
works, not just belief, to get you beyond the pearly gates". But for
some reason, it was more important to describe the goals of the group
as merely "being of the party faithful" and not actually "doing
something".
But I can hear the Free Software people saying, "oh, but that's what
we want". We want those people who are capable of creating us
wonderful things to be chained to a chair and to have to keep making
us more and more wondrous things because they are capable more than
us, and in our socialist society of "from each according to his
means", they owe us. No, they don't. You can see by the dollars
flowing into a company that the contributions of some are valued more
than the contributions of others. And that' where the _real_
political divide is.
Curious that the rest of the world has such
disdain for certain things of the US -- like the meltdown of its
school system -- and yet wants to import ideas -- like free software
-- that are driven by the same faulty idea. When I was in 5th grade
and first heard of "grading on the curve", which was new to me, it
seemed like a bad idea. "If everyone does badly, some of us will
still pass?" I asked. "Won't people eventually notice?" I was just
in 5th grade and figured adults must see some different perspective
that I would eventually grow into. But I think we (or, at least, I)
now know that from such little problems grow big problems...
And while there are many good things that the US has done, one of the
good things we done that are not good is to confuse terms like "pass"
and "fail" (everyone who tries should get to pass, some say) or
concept like "give" and "take" (Democrats push for big government to
offer social programs but then label as "takers" the Republicans who
think those plans don't always work, and not the people who abuse
those plans and make the Republicans want to cancel the plans;
Republicans push for small government saying that people of good heart
will given privately anyway, but then do not police their own to say
that those who do not give privately are being bad; in the end, both
parties say that you're a "giver" if you recite their script, and a
"taker" if you don't).
Free software talks of leveling the playing field, but that's just
what it doesn't do. Because equal opportunity is not to be had when
value is taken from those who contribute and given freely to everyone.
Value, in the sense of marketplace value, is _not_ a birthright. If
everyone had a Picasso or a Monet on their wall, it would not be
value, it would be Mickey Mouse, and you couldn't sell one for $10 at
K-Mart. Value is about scarcity. People value what they cannot have.
And if you give them everything, as has been tried with parents who
have too much money, you don't produce an idyllic world full of people
who do great things for people, you produce a world of couch potatoes,
drug addicts, and machine guns in schools, becuase no one can figure
out what their purpose is any more. The things they produce will not
be "valued" but will be shrugged at as the "birthrights" of others who
had nothing to do with their creation.
> | Sorry. Not really a bid to be declared royalty. I was going out on what
> | I hoped was a fairly sturdy limb and believing at least Erik, who was
> | already involved in the discussion, was on the same side as me on this
> | issue. Perhaps even others who haven't spoken... But the use of "We"
> | doesn't require an army.
>
> Yup, count me in that "we". However, please note that I used to be quite
> a fan of free software, but my goal was different from what I saw become
> the prevalent goal. I have wanted to obtain and grant access to source
> to those who wanted to learn from it, whether it would be helpful to the
> development of that source code or not, just like good authoers have to
> read a lot of good code to learn what you can simply not learn without
> reading other people's works. I had no political agenda with this at
> all, and I came to think it was a good idea on very different premises
> than were touted. Also, the Open Source movement seems to be an anti-
> movement, not a pro-movement, and specifically anti-Microsoft, which I
> am, too, but Open Source is unfortunately able to destroy much more than
> Microsoft if it succeeds. I have become highly critical of the focus and
> the political agenda, not to mention the fact that selling anything in a
> market where you compete with Open Source is so much harder, but this
> change has been gradual. I consider free software to be a bad way to
> lure innocent students and hobbyists to give away their work and time,
> which is only a good idea if they consider it learning and marketing for
> their future carreers, but that is not how these things usually work.
Yes, I too started out very differently. And this is why it disturbs
me to see that this movement appears to appeals so much more to young
people than older ones. I think they think they have a "better way",
but I also think they don't understand that some of us had that same
"better way" back then. They look at us like children look at adults,
not seeing the continuum, and thinking that "adults" and "children"
are fundamentally different kinds of stuff, and thinking that adults
could never possibly have once been children. I don't mean to use the
terms adult/children pejoratively here. But it does have the timeline
aspect of the metaphor, and I can't avoid that. I've been on the
other side. And when I woke up, as children all do one day, I was
angry that I had been allowed to flounder for so long.
I came out of the same environment at MIT that Stallman was in, though
I was a few years later. We all did things for each other then, and
none of us thought of money then. We were all very happy but we
served no one, really. And we used very expensive computers we didn't
have to account for, and most of us were on grants or money from
parents so time didn't count either. Our whole notion of cost models
was skewed badly. In exactly the same way as McDonald's notion of the
"disposable container" is skewed because they don't pay for collecting
litter, for landfill space, for pollution effects. If they were taxed
on these values, the equations would balance better and the notion of
giving away little cardboard boxes as if they were low-cost might
change.
If students owed an onerous duty to parents or research organizations,
they would think twice before giving away the fruits of their labors
on the grounds that "it cost them nothing to make". They would see
that what they were giving away free was costing them personally by
taking valuable time they'd already spent and turning it to no value.
I claim that the reason they give this stuff away is that they don't
realize that there are possible universes in which they might sell it,
and the reason they don't live in a world where they might sell it is
that they are surrounded by people intent on lessening the value of
what they do. The value they put in, their time, is no different than
the value put in by a struggling art student. But the art student
probably tries to sell what they can get away with selling, and it
wouldn't surprise me if the typical art student has a keener and more
personal understanding of marketplaces than many computer
scientists... maybe not formalized, but practical understanding.
Anyway, later, after I finished my time as a student, I saw people get
rich on things I'd done years before. As just one of several
examples, in about 1980, I had a nice database of all the restaurants
in Cambridge, MA, for example, indexed by price, quality, ethnicity,
hour they were open, geographical distance, etc. When we were hungry,
it was easy to make a choice of food--we just logged in and consulted
the program. Since then, people have made good money on such things.
Am I mad? Not at the people who made money, not if I'm honest. Ideas
aren't copyrightable, and no one stole the actual specific programs
I'd written. What they did was very legit. The simple truth is that
I had done the work to turn that idea into a program _already_, I
just didn't have the sense to sell it. So no one ever saw it. We'd
start it up, grin at it, and say "cool hack. now let's go eat before
that one closes too..." No reason to do otherwise, or so I thought.
No millions seemed ready to be made and the only restaurant we could
afford was about to close. The hint of money wouldn't have ruined
this mix, it would have gotten this idea to the market years faster
than it otherwise did.
Money isn't a bad thing. It's a good thing. It translates one
person's value system to another. When you give me dollars you may be
saying "it's worth it to me to give up my ability to purchase a
houseplant in order to buy your product" while I might say "I'm glad
to have the money for a new toaster". You and I may value different
things, yet you can still incentivize me. And, most importantly, it's the
way society tells you that you're onto something it cares about. Some
movies make money and others don't. That drives the creation of more
movies that people like. Do some that should be made get lost in the
wash? Probably. But not as many as get lost if there's no monetary
incentive to make movies....
Money, by the way, is like information. It can be freely copied, too.
It's just little pieces of paper ... or sometimes even just bits in a
computer. It's no more expensive for a bank to hold a lot of money
in your account than a small amount. But, nevertheless, it would lose
its commercial value if you did allow copying. Why do people understand
this fact about money but not about software?
And when there's no more money to be made on computer programming,
there's gonna be no more grant money except from people with specific
needs. Sure, there will still be the occasional corporation or
government with a specific market need they want you to service, but
that will be the "peasants in the field" metaphor. The chance for
computer scientists to direct where the world can go will last for as
long as computer scientists have money. And if the Free Software
people get their way, that won't be much longer.
Bulent Murtezaoglu
[...]
KMP> ... I don't think anything could
KMP> be more central to the discussion of Lisp than how to make a
KMP> buck off of it.
Hmm, I am almost sure you are exaggerating some here. The tradition
you yourself have been a part of does not seem to have making money
off of ideas as its main motivation. I don't think McCarthy was
running around telling his students and staff "our numbers are bad,
implement this damn eval ASAP." Of course making a living is
important, but the "most central" is a bit pushing it.
KMP> I would bet that if you asked people what
KMP> one choice more than any other motivates their use of a
KMP> language you wouldn't get "where they put their parens" but
KMP> rather "how quickly I can make enough money to pay my
KMP> mortgage".
It depends on which people you ask at what part of their lives.
[...]
>> So, when I say "free software", I mean libre, not gratis,
KMP> Libre as in "it's free for everyone except the people who
KMP> made it to decide how it's used". That's not "free", that's
KMP> "stolen". [...]
I dislike this as much as I dislike Stallman's overloading of 'free.'
If the maker says 'take this, you can use it in whatever manner you
want except you have to abide by such and so rules that _I_ set' it is
hardly theft. I used to occasionally loan money to friends or outright
give them money as a gift, I'd say stuff like "do whatever you want
with it except spending it in a bar." Now by your stretched analogy,
this is no different than a non-drinker strealing from me. Not so, I
voluntarily give up my control over that money with some stipulations.
I read the thread from the beginning over again to find out what exactly
pushed you to write this, and I couldn't find anything. Would you care
to maybe write some more and tell us why you wish to play the same game
with 'theft' as the GNU zealots are playing with 'freedom' ?
[...]
KMP> I said that for those who want to make their business writing
KMP> software, free software injures their market and risks
KMP> putting them out of business.
This is clearly correct. Free (ie no money) software has been used
by not-so-benign companies to do just that. (If anyone cares to respond
please do not repeat libre != gratis arguments.)
KMP> This seems to be a cost you're
KMP> prepared to accept. But then, you aren't in that kind of
KMP> business. What's curious about this kind of reasoning, in my
KMP> experience, is that the usual motivation for people going
KMP> into free software is that they perceive some other uncaring
KMP> soul is out there making decisions they perceive to be
KMP> callously injuring the things _they_ care about. But I guess
KMP> the tactic of injuring without caring is ok when it's done
KMP> for a noble reason.
Motivations aside, the effect you described in the previous paragraph
seems inescapable unless there's some magical mechanism that sees to
it that monetary costs are automatically incurred whenever somebody
gets some use out of some software regardless of whether the creator
of that software wants compensation. As you might be well aware,
several companies are trying to make money off of reselling help/labor
that most of us give away for free on usenet. You yourself have
probably produced much elucidating prose on many issues for free that
nonetheless had enormous value and could probably have been saleable.
You set a great example, in my opinion, by your well thought out and
extremely well written prose. If more able people follow your example
a potential market will be killed. Are you aware of this? Does it
only become offensive when some cultish behaviour emerges and not
otherwise?
cheers,
BM
Lieven Marchand
> It's amazing to me, but it appears to be true, that there are people
> who have these words/concepts in their head (socialness and
> antisocialness) but with the polarity of the meaning reversed...
Look up Proudhon's theories on property and be prepared to be amazed ;-)
--
Lieven Marchand <m...@wyrd.be>
She says, "Honey, you're a Bastard of great proportion."
He says, "Darling, I plead guilty to that sin."
Cowboy Junkies -- A few simple words
Thomas Bushnell, BSG
> And again, it shows the adjectives turned on their heads in a way I
> would generally agree with. Because it's often the producers, the ones
> who actively do the enriching of society, who get called the "stingy" and
> "hoarder" and "self-serving".
But the producers of free software--who *ARE* producers--earn nothing
but scorn. Why is that?
Thomas Bushnell, BSG
> Money isn't a bad thing. It's a good thing. It translates one
> person's value system to another.
Money is a good thing. But my value system is simply not denominated
in dollars.
Thomas
Thomas Bushnell, BSG
> * Thomas Bushnell, BSG
> | Free software is just as costly to produce as non-nfree software.
>
> *sigh* That is so not the issue. _Freedom_ is not costless.
Nope, freedom is very dearly bought, and all the more precious.
Thomas Bushnell, BSG
> Free software talks of leveling the playing field, but that's just
> what it doesn't do. Because equal opportunity is not to be had when
> value is taken from those who contribute and given freely to everyone.
But it's not *taken* from anybody. It's freely *given* by those who
contribute.
Nils Goesche
It's not the producers per se, but rather the propagandists who
demand all software to be free. These groups have a non-empty
intersection but are not equal.
Regards,
--
Nils Goesche
"Don't ask for whom the <CTRL-G> tolls."
PGP key ID 0x42B32FC9
Craig Brozefsky
> Would that it were off topic. I don't think anything could be more central
> to the discussion of Lisp than how to make a buck off of it. I would bet
> that if you asked people what one choice more than any other motivates their
> use of a language you wouldn't get "where they put their parens" but rather
> "how quickly I can make enough money to pay my mortgage".
And if this was the question foremost on their mind it would seem
obvious to me that they would get over nostalgiac attempts at
reproducing an older market model based on the sales of "tools" and
"libraries" in CL by small or independent vendors, and move towards
solving problems that people with the appropriate financial resources
are willing to fund. Laments, no matter how well put, are not going
to change the trend towards the commodification of tools that has been
going on since the start of industrialization.
This shift in perspective might have, and *HAS HAD*, some useful
results, not limited to paying several people's mortgages and
producing tools and libraries that the community can, and is, using.
Those tools can then help other people use CL in their work, widening
the pool of capital available to the CL community.
An inability to deal with this change of financial context has
resulted in lots of verbage about how Free Software is destroying the
community's chance of furthering its "collective" interest and
backhanded suggestions that some of the people who are writing CL code
now are sociopaths because they give it away.
I know we've been thru this before but I'm not about to stand for an
argument, no matter how verbose, eloquently spoken or rationalized,
that is going to call people actually contributing working code and
helping others use CL in their business or hobby or entertainment,
destructive sociopaths. I know of several cases, just dealing with
the FS CL tools I've contributed too, where people have been able to
get their company using CL, or have started to build up a company
around CL. To a certain extent I can understand the personal stress
this change has caused people, but when it turns into unproductive
attacks (disguised as attempts at balancing the scales of rhetoric or
otherwise) on other Lispers working to expand the community and
empower others in it, I lose patience.
The capital for expanding CL's capabilities is not coming thru direct
investment in CL tool vendors, it is being funneled thru people
writing solutions to other problems using CL and then sharing the
tools used to solve those problems when possible. There are no fringe
languages really capable of attracting the capital investment for pure
tool vendors anymore, since the ROI in such cases is dwarfed by the
size of the tool market in mainstream languages.
I say this as an owner of a CL shop, as a programmer in a CL shop, as
a contributor to and major author of several CL tool packages
distributed as Free Software, and as someone who also has co-workers
"paying their mortgages" with licenses for a vertical applications
written entirely in CL.
In the interest of heading off yet another giant thread dominated by a
few loquascious participants, I submit this post as my final
contribution. I think that the code that I and the other Free
Software Sociopaths(tm) have produced, and which the community is
presently using, speak louder than any psuedo-civilized whining that
appears in this newsgroup.
--
Craig Brozefsky <cr...@red-bean.com>
Free Software Sociopath(tm) http://www.red-bean.com/~craig
Kent M Pitman
> ... As you might be well aware,
> several companies are trying to make money off of reselling help/labor
> that most of us give away for free on usenet.
(Indeed on some days I've thought of entering this business myself.)
> You yourself have
> probably produced much elucidating prose on many issues for free that
> nonetheless had enormous value and could probably have been saleable.
Possibly so. There's also a difference between a "loss leader" and
"giving away the store". But the line is subtle and subjective.
And this place isn't just advice. It's also comradery and coalition
building and other things. There's not a bright line here between
chit chat and advice. The value those other companies sell may be
"slicing through the chit chat" so in that sense they may be in a
different area, by design.
> You set a great example, in my opinion, by your well thought out and
> extremely well written prose. If more able people follow your example
> a potential market will be killed. Are you aware of this? Does it
> only become offensive when some cultish behaviour emerges and not
> otherwise?
If someone offered to pay me for what I am doing now, I would not insist
on doing it free. :-)
I would want it to be at a reasonable price, but not just out of the
goodness of my heart--rather because I think most of the market would
be missed if I didn't.
I do think that, along with the good it does, there is some degree
harm done both by the fact that newsgroups don't cost and don't pay.
I don't have a good theory of how to fix the problem of competing with
legitimate advice-giving services at this time. I would not engage in
a political fight with someone who did come up with a theory of how to
offer a more equitable system than the present one.
Thomas Bushnell, BSG
> In article <873cync...@becket.becket.net>, Thomas Bushnell, BSG wrote:
> > Kent M Pitman <pit...@world.std.com> writes:
> >
> >> And again, it shows the adjectives turned on their heads in a way I
> >> would generally agree with. Because it's often the producers, the ones
> >> who actively do the enriching of society, who get called the "stingy" and
> >> "hoarder" and "self-serving".
> >
> > But the producers of free software--who *ARE* producers--earn nothing
> > but scorn. Why is that?
>
> It's not the producers per se, but rather the propagandists who
> demand all software to be free. These groups have a non-empty
> intersection but are not equal.
I'm happy to let free software quietly take over. We don't need to
change the laws.
But I will certainly be a propagandist if that means encouraging
people to write free software instead of non-free software.
Thomas
Kent M Pitman
I can't see this statement and not hear "I'm just annoyed that the things
other people value enough to part with money over aren't the things I would."
Let's just say I'd be very surprised if in a world where people were
paying dollars for the things you valued, you'd have the same
"principled" stance.
Money is the tangible acceptance of the notion that others don't value
things the same as we do. It is the acceptance of a multicultural world,
and all that that implies.
It's funny, too, that the people who I find most likely to claim money
is evil are the ones who are most likely to preach cultural tolerance.
It's almost like they both want to be tolerant and they also want to
control the set of things they are tolerant of...
Kent M Pitman
Because they offer nothing of value. Where "value" is defined by "something
above the baseline of what you can get for free".
Kent M Pitman
Those people don't "freely" give it. They give it subject to a complicated
contract that restricts uses.
"Freely given" is possible to do by putting something in the Public Domain.
Thomas Bushnell, BSG
> tb+u...@becket.net (Thomas Bushnell, BSG) writes:
>
> > Kent M Pitman <pit...@world.std.com> writes:
> >
> > > Money isn't a bad thing. It's a good thing. It translates one
> > > person's value system to another.
> >
> > Money is a good thing. But my value system is simply not denominated
> > in dollars.
>
> I can't see this statement and not hear "I'm just annoyed that the things
> other people value enough to part with money over aren't the things I would."
The reason I don't part with money for commercial software (in
general) is not because I'm stingy; it's because I want certain rights
that they authors of that software are not willing to give me.
Since I try to be an honorable person who keeps my commitments, I
would keep that commitment if I made it; as a result, I abstain from
making it. (But, there are exceptions; there have been cases where I
*do* willingly part with the freedoms that I greatly value. However,
what accounts for the exceptions is *not* the utility of the software
in question but rather other, secondary, considerations.)
I'm not annoyed that other people have different interests and are
willing to accept restrictions that I do not accept. However, I think
they may well be foolish in accepting those restrictions, and I
suggest that people not accept them. It's not a major deal for me;
it's not like I go around pestering all my Windoze-loving friends and
trying to convince them they should abandon non-free software.
> Let's just say I'd be very surprised if in a world where people were
> paying dollars for the things you valued, you'd have the same
> "principled" stance.
Huh? I'm happy to accept dollars for the things I value. But the
presence of dollars does not account for the value. I value them
whether or not they come with money attached.
By contrast, when it comes to handing out my money (or time), I
certainly do pay close attention to handing it out only to the things
I value--either on strictly utiliatarian grounds, such as when I give
Starbucks money to keep pouring coffee down my throat, or on
political, moral, or religious grounds, such as when I give money to
political campaigns, write free software, play bowling with my
friends, and give money to religious or charitable institutions.
If I received dollars for bowling with my friends, I'd accept it
happily. It wouldn't "taint" the pleasure of bowling.
> It's funny, too, that the people who I find most likely to claim money
> is evil are the ones who are most likely to preach cultural tolerance.
> It's almost like they both want to be tolerant and they also want to
> control the set of things they are tolerant of...
Nobody here has claimed that money is evil. Money is a good thing,
just not the only thing.
And, anyway, the free software movement is not about removing them
money from software.
Thomas
Lars Magne Ingebrigtsen
> Because they offer nothing of value. Where "value" is defined by
> "something above the baseline of what you can get for free".
Well, that's an, uh, interesting definition of "value".
I assume that you can get air for free around where you live. Does
air have no value for you, then?
--
(domestic pets only, the antidote for overdose, milk.)
la...@gnus.org * Lars Magne Ingebrigtsen
Jon Allen Boone
Kent, you wrote:
> Free software may be differently motivated, but it has the same
> commercial effect as dumping. It puts legitimate investors at a
> commercial disadvantage because they have to compete with people who
> didn't pay money to get the value they are offering.
I'm not paid to do software development full time, although I do it
incidentally as part of my job. I also recognize that I could go into
business for myself and make software development my full-time job.
It turns out, however, that at this point I value the benefits [pun
intended] that come along with having someone else assume most of the
risk [not to mention the paperwork and other administrative issues]
involved in running a company. In other words, I prefer to be an
employee.
Suppose, for the sake of argument, that I wanted to switch to being
a full-time software developer. The market puts a certain valuation
on my software development skills, influenced by things like the fact
that I don't use most of the commercial tools that professional
software developers use, etc. So, to switch from my current
profession into software development would likely involve taking a
significant pay cut.
It seems to me that the rational economic behavior for me is to
continue to draw the larger salary for my "day job" and moon-light as
a software developer [provided it is allowed by the contract I have
with my employer]. I would expect that if I were to do this enough, I
could build up my market value as a software developer. Is there a
market for part-time software developers other than the open-source
community?
It seems that even most contract work is set up with the
preconceived notion that you'll work "full-time" for a limited
duration rather than "part-time" for a longer duration...
-jon
--
------------------
Jon Allen Boone
ipmo...@delamancha.org
Lars Magne Ingebrigtsen
> tb+u...@becket.net (Thomas Bushnell, BSG) writes:
>
>> Kent M Pitman <pit...@world.std.com> writes:
>>
>> > Free software talks of leveling the playing field, but that's just
>> > what it doesn't do. Because equal opportunity is not to be had when
>> > value is taken from those who contribute and given freely to everyone.
>>
>> But it's not *taken* from anybody. It's freely *given* by those who
>> contribute.
>
> Those people don't "freely" give it. They give it subject to a complicated
> contract that restricts uses.
First you say that value is "taken" (by some unnamed agency) and
"given freely" to the world. And then you say that "freely" doesn't
apply at all.
Perhaps you should pursue some other rhetorical line.
Thomas Bushnell, BSG
> Because they offer nothing of value. Where "value" is defined by "something
> above the baseline of what you can get for free".
In other words, the food bank I work at offers nothing of value?
Funny that, because the people we help seem awfully grateful.
Kent M Pitman
> Kent M Pitman <pit...@world.std.com> writes:
>
> > Because they offer nothing of value. Where "value" is defined by
> > "something above the baseline of what you can get for free".
>
> Well, that's an, uh, interesting definition of "value".
>
> I assume that you can get air for free around where you live. Does
> air have no value for you, then?
What makes you think air is free?
I think a lot of air pollution comes because people don't think of it
as having value, exactly because they can get it for free.
Most people come to value what costs them.
That's why they don't value teachers.
[And probably why you don't value my remark. If I charged you more,
studies on cognitive dissonance say that, rhetoric claiming "it would
never happen to me" notwithstanding, you'd believe me more. Heh...]
Lars Magne Ingebrigtsen
> That's why they don't value teachers.
You're playing semantic games. "Has no value" is something different
than "places no value on".
> [And probably why you don't value my remark. If I charged you more,
> studies on cognitive dissonance say that, rhetoric claiming "it would
> never happen to me" notwithstanding, you'd believe me more. Heh...]
:-) Sure. People place more value on things that they've paid a lot
for. They hire in expensive consultants to tell them what they
already know. That's a pathology, though.
Kent M Pitman
> Is there a
> market for part-time software developers other than the open-source
> community?
I think so, but I don't know how big it is. I know someone who does
database programming on a per-case basis but basically none of the
sales and marketing. They have an agent who does all the promotion
and billing and just subcontracts the actual work to whoever can do it.
But I suspect situations like this are daily made more rare by two
factors: (a) the ability of free software to drive down the breakeven
price for individual providers to near zero, requiring a great deal of
economy of scale to make any money and (b) the desire of the companies
big enough to have that economy of scale to minimize their
administrative costs by getting a few full-time employees instead of
many part-time ones.
Personally, I consider this a very interesting area. I'd love for my
company, HyperMeta, to grow up into something that had a distributed
campus of people working at home, emphasized employee quality of life
in exchange for "value provided" rather than "hours spent", etc. But
getting from here to there involves first getting enough dollars in
that I can afford to experiment. And getting to that point involves
making money at all. And making money at all involves competing with
people giving away stuff for free. And that means not making very
much money, so working longer. So don't hold your breath.
> It seems that even most contract work is set up with the
> preconceived notion that you'll work "full-time" for a limited
> duration rather than "part-time" for a longer duration...
This is because the people who are in control of the purse strings
control the world. And the people who know a lot about computer science
are walking like lemmings toward a cliff that falls off into a world
where it's not possible to use their key talent to make any money.
So they don't get empowered, and their ideas never come into play.
But at least when they come home from their day job doing whatever it is
that people pay an hourly wage for, they can enjoy themselves making
things for free that they can't afford to buy because their day job
doesn't pay enough...
Thomas Bushnell, BSG
> > But it's not *taken* from anybody. It's freely *given* by those who
> > contribute.
>
> Those people don't "freely" give it. They give it subject to a complicated
> contract that restricts uses.
That all depends on which license in question. (They are not
generally, in fact, contracts.)
The GPL might be complicated; I don't know. I've understood it for so
long it's hard for me to judge. I don't think it restricts uses at
all, though it does restrict copying some. However, it restricts
copying far less than the average commercial license, which is what
you are comparing it to, right?
The new BSD license, or the X license, are not complicated at all, and
I don't see how one could ever say they restrict uses.
Thomas
Thomas Bushnell, BSG
If free software is such a harmful dangerous thing, Kent, then why do
you use it to read and post news?
Thomas
Lars Magne Ingebrigtsen
> If free software is such a harmful dangerous thing, Kent, then why do
> you use it to read and post news?
Isn't that a rather ad hominem line of discussion?
Bulent Murtezaoglu
ThomasB> If free software is such a harmful dangerous thing, Kent,
ThomasB> then why do you use it to read and post news?
I was wondering what pushed KMP to stretch his analogies. This wonderfully
clever top posting provides the answer. Thank you.
cheers,
BM
Kent M Pitman
It depends on whether the people are being grateful as in "wanting to
work to return the favor or otherwise provide compensating value to
the world" or just grateful like "they've learned to smile and say
thank you but keep coming back for more".
We all have our days when things don't quite balance and it helps to have
friends.
But the point of my remark was that friendship can get easily taken
advantage of if it is truly offered unconditionally as a right.
It isn't a comment about your goodwill, it is a comment about what kinds
of things pass for gratefulness.
Give me a fancy car and I'll be "grateful" too. Maybe I'll say "thank
you" several times and even bake you a cake. Then again, if the car
was free and it gets wrecked, I may not value it the same as if it was
a car I bought with my own money where I had to yield something else
of value in exchange for getting it...
Thomas Bushnell, BSG
> tb+u...@becket.net (Thomas Bushnell, BSG) writes:
>
> > If free software is such a harmful dangerous thing, Kent, then why do
> > you use it to read and post news?
>
> Isn't that a rather ad hominem line of discussion?
It would be ad hominem if it were offered to prove that Kent's ideas
are wrong.
Actually, I suspect Kent of (minor) hypocrisy, telling us that we
should all pay for software we use, that if we don't, it's valueless,
and the like. I don't think he really believes that.
Thomas
Nils Goesche
Why are they grateful? Because you offer them something they'd
have to pay for elsewhere, because it has value. But if you
would supply the whole region with free food, and everybody likes
your food, the price for food, and hence its value, would indeed
drop to zero there.
Regards,
--
Nils Goesche
Ask not for whom the <CONTROL-G> tolls.
PGP key ID 0xC66D6E6F
Thomas Bushnell, BSG
> > In other words, the food bank I work at offers nothing of value?
> > Funny that, because the people we help seem awfully grateful.
>
> It depends on whether the people are being grateful as in "wanting to
> work to return the favor or otherwise provide compensating value to
> the world" or just grateful like "they've learned to smile and say
> thank you but keep coming back for more".
Even if they are just mooching for no good reason, it's still clear to
me that the food has value, indeed, they even value it.
> It isn't a comment about your goodwill, it is a comment about what kinds
> of things pass for gratefulness.
You were talking about *value*, not about gratefulness. The fact that
they *want* the food, whether or not they pay, whether or not they
could, demonstrates that they do *value* it.
You are confusing value, cost, and price. I recommend any basic
economics textbook.
Thomas
Ingvar Mattsson
I suspect that Kent wants us to think before claiming that others
should produce Free (as in libre) Software or Free (as in beer)
Software and carefully consider the choice we make. Also that we
should not claim that anyone else should act according to what we
want.
This is, however, blatant extrapolation and I expect Kent to correct
me if I am badly wrong, say I am correct if I am or just ignore me as
he sees fit.
//Ingvar
--
Self-referencing
Five, seven, five syllables
This haiku contains
Thomas Bushnell, BSG
> Why are they grateful? Because you offer them something they'd
> have to pay for elsewhere, because it has value. But if you
> would supply the whole region with free food, and everybody likes
> your food, the price for food, and hence its value, would indeed
> drop to zero there.
And people would *still* value it, because without it they would die.
Nils Goesche
> You were talking about *value*, not about gratefulness. The fact that
> they *want* the food, whether or not they pay, whether or not they
> could, demonstrates that they do *value* it.
No. It's the fact that they /would/ pay for it if
(i) They had any money
(ii) You wouldn't give it away for free
> You are confusing value, cost, and price. I recommend any basic
> economics textbook.
I think it's you who needs to read something more about this.
But please note that both the bible and Karl Marx are a bit
dated, when you buy a book about it :-)
Nils Goesche
Yes, that's probably the reason they'd pay for it if it weren't
available for free. What is do hard to understand about this?
Lars Magne Ingebrigtsen
> Yes, that's probably the reason they'd pay for it if it weren't
> available for free. What is do hard to understand about this?
Your point.
Kent is saying that free stuff has no value. You're saying that free
stuff has no cost. One of these statements are true.
Nils Goesche
> Nils Goesche <n...@cartan.de> writes:
>
> > Yes, that's probably the reason they'd pay for it if it weren't
> > available for free. What is do hard to understand about this?
>
> Your point.
>
> Kent is saying that free stuff has no value. You're saying that free
> stuff has no cost. One of these statements are true.
Yes ;-)
It is indeed so, that free stuff has no cost (remember that often
there are hidden costs), that is the definition of ``free''.
Food isn't free, BTW. That free stuff has no value is a trivial
corollary.
Thomas Bushnell, BSG
> Yes, that's probably the reason they'd pay for it if it weren't
> available for free. What is do hard to understand about this?
Nothing at all. But by that measure, there's nothing about GNU Emacs
that is "valueless", since people did pay for the best they could get
before it was available. Heck, people *still* pay for GNU Emacs
today.
Kent said it had no value, and therefore was not a contribution to the
world. But that failed to take into account that it's the sort of
thing many people *would* pay for if they had to.
Thomas