Tcl for Evil Geniuses
Sean Woods
their leaves, winter blankets the northern hemisphere, and another TCL
IS DYING thread gets posted to clt.
One issue that was brought up at the 14th Conference's Town Hall
meeting was that of publicity. While the concern at the time was
publicity for the Conference, I think it speaks to a larger issue: why
don't we have hordes and hordes of Tcl developers like other (you fill
in the blank) languages.
One item that is abundantly clear is that while more hands to reap the
harvest are nice, the quality of the new hands is more important than
the quantity. We don't need more programmers so much as more
programmers who are experienced and competant in both C and Tcl.
And competant people in C and or Tcl are members of an elite kind of
society rivaled only by the martial arts, or certain artisan trades.
So I propose that we take an elitist view. Tcl is a secret society. We
have our own handshakes. Our own customs. Members of the society are
highly placed in the engineering and technology community from around
the world. When I hear about what we all work on, I don't think it too
much of a stretch to add a Jacob's ladder or two, and a hunchback
doing the legwork. At our conference, all we really need is a large
chair with buttons, and someone stroking a cat.
What we need to do is exploit this image. Turn the "You aren't popular
na na na na na na" on it's ear and retort: "Yes, but we know what we
are doing." While we welcome all, to fit in requires a certain mindset
and devotion to the craft.
How do we exploit this image? T-shirts that play off of the cultural
'evil geniuses'. Bond villians. Mad (or Rocket) scientists. Military
strategists. The sorts of images that the public will readily relate
to "smart, but I don't understand what they are doing."
I want people who are looking to attend the proceedings to not get
this image of old men talking about plumbing. I want them to think
that the trilateral commission is getting together to decide how high
the hem lines on skirts should be this year.
And having attended a few conferences thus far, that latter is
actually closer to the truth.
--The Hypnotoad
Donal K. Fellows
> Members of the society are
> highly placed in the engineering and technology community from around
> the world.
Don't be ridiculous. Everyone knows there are no Secret Tcl
Illuminati.
Donal.
suchenwi
Sean Woods schrieb:
> So I propose that we take an elitist view. Tcl is a secret society. We
> have our own handshakes. Our own customs. Members of the society are
> highly placed in the engineering and technology community from around
> the world. When I hear about what we all work on, I don't think it too
> much of a stretch to add a Jacob's ladder or two, and a hunchback
> doing the legwork. At our conference, all we really need is a large
> chair with buttons, and someone stroking a cat.
>
> What we need to do is exploit this image. Turn the "You aren't popular
> na na na na na na" on it's ear and retort: "Yes, but we know what we
> are doing." While we welcome all, to fit in requires a certain mindset
> and devotion to the craft.
Amen, brother! What a good contrast to the usual FUD about Tcl's
future. We are the future.
And don't forget that as few as we are, we are pretty well connected -
not only at conferences, but with chat, Wiki, and even c.l.t :^)
Larry W. Virden
wrote:
> Everyone knows there are no Secret Tcl
> Illuminati.
I think _that_ is what I want on my tee shirt. Perhaps with an image
of a feather and a candle illuminating it...
Sean Woods
Ron Fox
> So I propose that we take an elitist view. Tcl is a secret society. We
> have our own handshakes. Our own customs. Members of the society are
> highly placed in the engineering and technology community from around
> the world. When I hear about what we all work on, I don't think it too
> much of a stretch to add a Jacob's ladder or two, and a hunchback
> doing the legwork. At our conference, all we really need is a large
> chair with buttons, and someone stroking a cat.
RF
Sean Woods
caterpillar.
sd
extent it deserves. I'm a small fish in a large international company
and yet I am the only one that uses tcl. Every time I create an
application to be used at any one of our facilities around the world,
I have to first explain to their IT manager and employees (some don't
speak English) what tcl is (even veteran IT personnel never heard of
it), and then justify why using tcl fits their needs and won't cause
regrets. I then to have to go through the same drill a while later.
Also, the tcl community being relatively small leads to long wait
times for problems in packages to be fixed. I had to abandon a couple
of important projects using tcl because I needed to use a package
which contains some limitations that could've easily been fixed,
IMHO. Explaining this to my boss was no fun, especially when I brag
about tcl like a proud parent.
I respect all the regulars on this board that help people like me in
need. Since I am not a naturally talented programmer who can
contribute to clt like most of you can, my humble way of contributing
comes in the form of buying tcl books and software licenses, and share
them with other individuals whom might show interest.
I honestly believe the tcl community needs to go the extra mile and
promote this excellent tool.
Just my 2 cents
/sd
suchenwi
> Also, the tcl community being relatively small leads to long wait
> times for problems in packages to be fixed. I had to abandon a couple
> of important projects using tcl because I needed to use a package
> which contains some limitations that could've easily been fixed,
> IMHO. Explaining this to my boss was no fun, especially when I brag
> about tcl like a proud parent.
Umm.. Tcl is open-source ("free") software which you can get for
$0.00. And the same amount refunded if disappointed...
If your big company needs changes to packages, there are consultants
available who can deliver (for >$0.00), e.g. http://activestate.com or
http://evolane.com
> I respect all the regulars on this board that help people like me in
> need. Since I am not a naturally talented programmer who can
> contribute to clt like most of you can, my humble way of contributing
> comes in the form of buying tcl books and software licenses, and share
> them with other individuals whom might show interest.
Buying books is good (thank you!), though the market is surely
declining. And licenses? If you buy licenses from ActiveState, your
bug reports will surely be taken more seriously there...
> I honestly believe the tcl community needs to go the extra mile and
> promote this excellent tool.
Oh well, the Tcl community altogether goes many miles, but what are we
to say against popular fashion? I think it's good to have a choice of
many languages, as long as we still can choose Tcl for where it works
best. (In other cases, I prefer awk). But the hype's not on Tcl's
side... so what, we code away, and deliver simple, elegant tools that
just do the job.
sd
bauersa...@siemens.com> wrote:
> sd schrieb:
>
> > Also, the tcl community being relatively small leads to long wait
> > times for problems in packages to be fixed. I had to abandon a couple
> > of important projects using tcl because I needed to use a package
> > which contains some limitations that could've easily been fixed,
> > IMHO. Explaining this to my boss was no fun, especially when I brag
> > about tcl like a proud parent.
>
> Umm.. Tcl is open-source ("free") software which you can get for
> $0.00. And the same amount refunded if disappointed...
Correct. My intention is to provide business to a company that
maintains/develops and packages tcl as a whole
> If your big company needs changes to packages, there are consultants
> available who can deliver (for >$0.00), e.g.http://activestate.comorhttp://evolane.com
Good point. I will contact activestate regarding this issue, even
though they did not author that package, nor do they distribute it.
> Oh well, the Tcl community altogether goes many miles, but what are we
> to say against popular fashion? I think it's good to have a choice of
> many languages, as long as we still can choose Tcl for where it works
> best. (In other cases, I prefer awk). But the hype's not on Tcl's
> side... so what, we code away, and deliver simple, elegant tools that
> just do the job.
I'm all for having many sensible languages floating around. But the
fact that tcl is so obscure works to my disadvantage as a builder of
modest applications. In the IT world, I found it hard to convince
people to use a product built from something they never heard of.
Regards,
/sd
John Kelly
> On Oct 3, 12:03 pm, suchenwi wrote:
>> the Tcl community altogether goes many miles, but what are we
>> to say against popular fashion? ... so what, we code away
If I ruled the world, programmers would still use keypunch machines
and IBM System/360's. IMO, it never got any better than that.
> the fact that tcl is so obscure works to my disadvantage as a
> builder of modest applications. In the IT world, I found it
> hard to convince people to use a product built from something
> they never heard of.
So by definition, Tcl developers must therefore be, illuminati.
--
Internet service
http://www.isp2dial.com/
David Gravereaux
I'll buy that on a shirt.
--
David Gravereaux <davy...@pobox.com>
[species:human; planet:earth,milkyway(western spiral arm),alpha sector]
Sean Woods
perceptible form, I can concentrate my forces while the enemy is
fragmented. The pinnacle of military deployment approaches the
formless: if it is formless, then even the deepest spy cannot discern
it nor the wise make plans against it."
--Sun Tzu, The Art of War
Jawn
>
> - Show quoted text -
I would buy the t-shirt too. ;-)
John
nobodywhoareyou
>
> I'm all for having many sensible languages floating around. But the
> fact that tcl is so obscure works to my disadvantage as a builder of
> modest applications. In the IT world, I found it hard to convince
> people to use a product built from something they never heard of.
>
For me, so much comes down to the name. I really, truly hate the
"tickle" thing. Tools are cool, "control languages" imply power but
"tickle" - ugh. It diminishes the language before anyone can even see
the potential.
I learned coding on the midrange platform and I have no trouble seeing
and thinking t-c-l, since I spent many years writing CL scripts but if
I say it out loud, people correct me with the "tickle" pronunciation
and all I can do is cringe. (I'm a middle-aged female, so if I can't
say it, no wonder the guys won't touch it.) I appreciate that back in
the day, it might have seemed cute and user-friendly but I bet usage
would double if the community could reformulate the name.
Why wasn't "tackle" used initially? Tackle boxes are cool. It even
fits with the feather icon. :-)
Terry
John Kelly
<terr...@gmail.com> wrote:
>I say it out loud, people correct me with the "tickle" pronunciation
>and all I can do is cringe. (I'm a middle-aged female, so if I can't
>say it, no wonder the guys won't touch it.) I appreciate that back in
>the day, it might have seemed cute and user-friendly but I bet usage
>would double if the community could reformulate the name.
Yeah, who wants to go on a date with Tom Raper. If I had a bad name,
I would change it.
Cameron Laird
sd <sha...@hotmail.com> wrote:
.
.
.
>I'm all for having many sensible languages floating around. But the
>fact that tcl is so obscure works to my disadvantage as a builder of
>modest applications. In the IT world, I found it hard to convince
>people to use a product built from something they never heard of.
.
.
Another approach several of us have employed--often
successfully!--is to say, "I built this project with
a high-quality C library."
I'm entirely serious.
Sean Woods
suchenwi
John Kelly schrieb:
> Yeah, who wants to go on a date with Tom Raper. If I had a bad name,
> I would change it.
I have no problems with "tickle", but everyone who has can always call
the language "Tool Control", which may sound stronger :^)
Bryan Oakley
> Another approach several of us have employed--often
> successfully!--is to say, "I built this project with
> a high-quality C library."
>
> I'm entirely serious.
So... TCL stands for:
The high quality
C
Library
That works :-)
--
Bryan Oakley
http://www.tclscripting.com
Ron Fox
>The< C Library
I sense a logo opportunity here.
RF
Jochem Huhmann
> Another approach several of us have employed--often
> successfully!--is to say, "I built this project with
> a high-quality C library."
"It was *that* high-quality I didn't even need to write a single line of
C to make my application work." ;-)
Jochem
--
"A designer knows he has arrived at perfection not when there is no
longer anything to add, but when there is no longer anything to take away."
- Antoine de Saint-Exupery
John Kelly
>John Kelly schrieb:
In the case of Tcl, I don't really think a new name will improve its
popularity.
There are too many technologies and too many tools. They can't all
stand out among the crowd. Some become popular, others don't. Who
knows why. That's just the way the wind blows, something man cannot
change.
Donal K. Fellows
> cla...@lairds.us (Cameron Laird) writes:
>
>> Another approach several of us have employed--often
>> successfully!--is to say, "I built this project with
>> a high-quality C library."
>
> "It was *that* high-quality I didn't even need to write a single line of
> C to make my application work." ;-)
That was just a few runtime configuration settings. Not programming at
all. After all, programming requires tens of thousands of Indian wage
slaves and investments of hundreds of millions of dollars. Everyone
knows that!
Donal.
sleb...@gmail.com
<donal.k.fell...@manchester.ac.uk> wrote:
> Jochem Huhmann wrote:
> > cla...@lairds.us (Cameron Laird) writes:
>
> >> Another approach several of us have employed--often
> >> successfully!--is to say, "I built this project with
> >> a high-quality C library."
>
> > "It was *that* high-quality I didn't even need to write a single line of
> > C to make my application work." ;-)
>
> That was just a few runtime configuration settings. Not programming at
> all.
GOLD!
Boss: Which bit of code did you modify to fix that bug?
Tcler: None. I just modified the config files!
saves me from having to write code change documentation ;-)
Gerald W. Lester
> ...
>
>> If your big company needs changes to packages, there are consultants
>> available who can deliver (for >$0.00), e.g.http://activestate.comorhttp://evolane.com
>
> Good point. I will contact activestate regarding this issue, even
> though they did not author that package, nor do they distribute it.
There are several independent consultants that you can also contact.
Also please understand that ActiveState is not the only company that "that
maintains/develops and packages tcl as a whole".
--
+--------------------------------+---------------------------------------+
| Gerald W. Lester |
|"The man who fights for his ideals is the man who is alive." - Cervantes|
+------------------------------------------------------------------------+
Mark Stucky
> Cameron Laird wrote:
>> Another approach several of us have employed--often
>> successfully!--is to say, "I built this project with
>> a high-quality C library."
>>
>> I'm entirely serious.
>
> So... TCL stands for:
>
> The high quality
> C
> Library
>
> That works :-)
>
>
I was thinking
TCL : Top-quality C Library
:)
--Mark
stevel
I've heard of people doing the following
#!/usr/bin/env tclsh
# Java inside
...
"Yes boss, it's got Java inside"
Sean Woods
Enterprise
Vertical
Integration
Language
On Oct 3, 2:57 pm, nobodywhoareyou <terryo...@gmail.com> wrote:
>
> For me, so much comes down to the name. I really, truly hate the
> "tickle" thing. Tools are cool, "control languages" imply power but
> "tickle" - ugh. It diminishes the language before anyone can even see
> the potential.
>
> Terry
Cameron Laird
So how many of our secrets shall we surrender today?
Cesar Rabak
non initiated...
miguel
You now owe me a new keyboard - mine has Java inside!
Should know better than reading clt while sipping my coffee :P
davidn...@gmail.com
> na na na na na na" on it's ear and retort: "Yes, but we know what we
> are doing." While we welcome all, to fit in requires a certain mindset
> and devotion to the craft.
That sort of works for languages like Lisp and Haskell, but
realistically, they are pretty "fancy" (academically speaking)
languages, so the snoot comes natural.
In the long run, though, I think for widespread popularity, the ivory
tower is not a winning strategy.
davidn...@gmail.com
bauersa...@siemens.com> wrote:
> sd schrieb:
>
> > Also, the tcl community being relatively small leads to long wait
> > times for problems in packages to be fixed. I had to abandon a couple
> > of important projects using tcl because I needed to use a package
> > which contains some limitations that could've easily been fixed,
> > IMHO. Explaining this to my boss was no fun, especially when I brag
> > about tcl like a proud parent.
>
> Umm.. Tcl is open-source ("free") software which you can get for
> $0.00. And the same amount refunded if disappointed...
>
> If your big company needs changes to packages, there are consultants
> available who can deliver (for >$0.00), e.g.http://activestate.comorhttp://evolane.com
Sure, but the cost benefit analysis that some people are going to do
is "well, the Python/Javascript/Ruby/Blub package do this works just
fine".
That's not true in all cases, Ruby isn't very mature in a lot of its
packages either, but that sort of herd mentality isn't entirely
stupid, either, unless you're really getting an order of magnitude
more something by using a particular language.
davidn...@gmail.com
> On Wed, 03 Oct 2007 13:06:50 -0700, suchenwi
>
> >John Kelly schrieb:
> >> Yeah, who wants to go on a date with Tom Raper. If I had a bad name,
> >> I would change it.
> >I have no problems with "tickle", but everyone who has can always call
> >the language "Tool Control", which may sound stronger :^)
>
> In the case of Tcl, I don't really think a new name will improve its
> popularity.
>
> There are too many technologies and too many tools. They can't all
> stand out among the crowd. Some become popular, others don't. Who
> knows why. That's just the way the wind blows, something man cannot
> change.
Luck is an element, but there's skill, too. It's called "marketing",
and not in the evil dilbert sense of trying to sell junk, or
advertising, but in figuring out what people want and delivering it to
them in such a way as to make them happy. Look at the success of
Ruby on Rails... *that* is a masterful bit of hype. It's built on
something high quality though - it's not just about pushing junk.
It's been a while since I've dredged it up here, but this is a little
piece I wrote a while ago:
http://www.welton.it/articles/programming_language_economics.html
It's hardly the last word, but I owe my interest in this stuff
partially to Tcl's flagging popularity, which got me to thinking about
why, deciding it certainly wasn't a quality problem, and asking what
it could be...
Donal K. Fellows
> It doesn't matter... our hermetic language is incomprehensible by the
> non initiated...
That inspires me to think that we need a T-shirt with:
"I'm not a secret illuminatus, honest!"
on the front, and a Tcl feather and website URL on the back.
Donal.
Uwe Klein
> "If I determine the enemy's disposition of forces while I have no
> perceptible form, I can concentrate my forces while the enemy is
> fragmented. The pinnacle of military deployment approaches the
> formless: if it is formless, then even the deepest spy cannot discern
> it nor the wise make plans against it."
>
> --Sun Tzu, The Art of War
World of NULL-A
uwe
Uwe Klein
> Maybe we should change the name to
>
> Enterprise
> Vertical
> Integration
> Language
A white angora cat?
uwe
Colin Macleod
> Maybe we should change the name to
>
> Enterprise
> Vertical
> Integration
> Language
>
Can't help thinking of Jumba in "Lilo and Stitch" whose
"Evil Genius" experiments rarely turned out as planned :-)
(http://en.wikipedia.org/wiki/Jumba_Jookiba)
Colin.
Sean Woods
I can see the "EVIL in a Nutshell" book now. Complete with a dark hand
stroking the cat.
skuh...@web.de
I really can't see, why people complain about the name and
pronunciation of Tcl. "Tickle" is not the problem. I work in Germany
and many people here don't even know, what the english word "tickle"
means (very sad in IT-environments...), but nevertheless they have
wrong preconceptions about Tcl. It's not the name to blame...
OTOH we have for example GNU. Come on, a whole software/OS/compiler
environment named after a stupid looking nordic animal (ever took a
closer look at the logo?)... Or Python, a language named after a
(sadly) long time ago split british commedy group...! Look at the
Python-Logo, how stupid does that poor snake look! Talk about Java,
named after a kind of coffee, which is obviously a perfect mataphor
for doing nothing (have a break and a cup of coffee). And so on, you
can find dozens of ridiculous names for languages or other projects
which are nevertheless taken very seriously by people.
Marketing *is* important, I totally agree with that, but changing the
name for marketing-reasons of a technical product is not the solution.
The trick is not to make up a catchy name. The goal must be to connect
the name with the positive attributes it deserves. These connections
must be anchored in peoples minds. In the world of programmers this
means also to make some good and effective marketing noise, but also
to show the audience, how powerfull and how vivid, agile
and up to date the language, it's toolboxes (and supporting libs) and
it's development is. In the long term, IMHO, this is the only way to
bring programmers (and managers) attention back to Tcl, and this is,
where at least partly the Tcl-community lacks some effective efforts.
Regards
Stephan
ZB
> It's hardly the last word, but I owe my interest in this stuff
> partially to Tcl's flagging popularity, which got me to thinking about
> why, deciding it certainly wasn't a quality problem, and asking what
> it could be...
I think, it could be just because of the difference in the syntax. Tcl's
syntax is unusual; much different of many other popular languages - and not
everyone is able to dispose of his habits (or just doesn't want to). Say,
why some guys have invented "L" ("C-like TCL")?
One example: in 8-bit era, although many Forth-implementations were
available for every (I think so) home-computer (remember that term? ;),
Forth never became popular, because it was much too different from BASIC.
Which was first computer language mastered (if any) by 8-bit microcomputer
owner.
There was a LOGO (with its "turtle graphics") much hype was about it - where
is LOGO today? It was too different either. And it hasn't any support for
polymorphism, etc. ;)
--
ZB
Cameron Laird
<skuh...@web.de> wrote:
.
.
.
>OTOH we have for example GNU. Come on, a whole software/OS/compiler
>environment named after a stupid looking nordic animal (ever took a
>closer look at the logo?)... Or Python, a language named after a
.
.
There's a breakdown here. I read "nordic" as "Scandinavian",
generally; are you writing it in the sense of "be-horned,
particularly when the pair is relatively short and upturned"?
Malmström has a lot to answer for ...
suchenwi
> There was a LOGO (with its "turtle graphics") much hype was about it - where
> is LOGO today? It was too different either. And it hasn't any support for
> polymorphism, etc. ;)
Logo was kind of a Lisp variant with much less parentheses (in its
early days planned to be taught in kindergartens), and in some aspects
comparable to Tcl. There is still a somehow active community, the
flagship release is seemingly MSW Logo (see http://www.softronix.com/logo.html
), they have their comp.lang.logo group...
BTW, MSW Logo is very cool for its 3D graphics
skuh...@web.de
Cameron Laird wrote:
> There's a breakdown here. I read "nordic" as "Scandinavian",
> generally; are you writing it in the sense of "be-horned,
> particularly when the pair is relatively short and upturned"?
> Malmström has a lot to answer for ...
Hmm... I mean these "Wildebeest": http://en.wikipedia.org/wiki/Gnu ,
which of course doesn't look very stupid, especially if it stands in
front of you ;-) , but I was referring to GNU (www.gnu.org) and the
logo. - Maybe I used the wrong words...
Stephan
ZB
> Logo was kind of a Lisp variant with much less parentheses [..]
> BTW, MSW Logo is very cool for its 3D graphics
I know, what is LOGO. It was just another example of cool - as you noticed -
language, which seems to be underestimated.
Actually, my feelings are similar to Peter Newman's, mentioned on the page
http://wiki.tcl.tk/4138 ("Things holding Tcl back")
--
ZB
Uwe Klein
these as "nordic" , right?
Reminds me of a norwegian ranting about flighty southerners.
Took me some time to understand that he was referring to a
danish coworker.
uwe
Donal K. Fellows
> Reminds me of a norwegian ranting about flighty southerners.
>
> Took me some time to understand that he was referring to a
> danish coworker.
Well shucks! I'd have assumed he was referring to someone from Trondheim...
Donal.
John Kelly
wrote:
> my feelings are similar to Peter Newman's, mentioned on the
> page http://wiki.tcl.tk/4138 ("Things holding Tcl back")
I see it says:
> there are some artifacts that trip people up ... Once these
> folks (all electrical engineers I should add) are bitten, they
> avoid the language
Anything that clears the engineers out of the room is OK with me.
:-D
--
Internet service
http://www.isp2dial.com/
davidn...@gmail.com
> Marketing *is* important, I totally agree with that, but changing the
> name for marketing-reasons of a technical product is not the solution.
> The trick is not to make up a catchy name. The goal must be to connect
> the name with the positive attributes it deserves. These connections
> must be anchored in peoples minds. In the world of programmers this
> means also to make some good and effective marketing noise, but also
> to show the audience, how powerfull and how vivid, agile
> and up to date the language, it's toolboxes (and supporting libs) and
> it's development is. In the long term, IMHO, this is the only way to
> bring programmers (and managers) attention back to Tcl, and this is,
> where at least partly the Tcl-community lacks some effective efforts.
I've never been able to write it quite that well. You are 100%
correct.
Uwe Klein
> On 04 Oct 2007 14:09:02 GMT, ZB <zbREMOVE_THIS@AND_THISispid.com.pl>
> wrote:
>
>
>>my feelings are similar to Peter Newman's, mentioned on the
>>page http://wiki.tcl.tk/4138 ("Things holding Tcl back")
>
>
> I see it says:
>
>
>>there are some artifacts that trip people up ... Once these
>>folks (all electrical engineers I should add) are bitten, they
>>avoid the language
>
>
> Anything that clears the engineers out of the room is OK with me.
you won't move me a teeny bit.
None of my <censored> pounds ;-)
Whats so bad about my co-engineers?
>
> :-D
uwe
>
>
nobodywhoareyou
> Sean Woods wrote:
> > Maybe we should change the name to
>
> > Enterprise
> > Vertical
> > Integration
> > Language
>
> pronunciation of Tcl. "Tickle" is not the problem. I work in Germany
> and many people here don't even know, what the english word "tickle"
> means (very sad in IT-environments...), but nevertheless they have
> wrong preconceptions about Tcl. It's not the name to blame...
>
"tickle" is a personal thing for me; there's no logic there and so no
way to really defend it. Truthfully, most of the resistance I see at
work is usually to the old Tk interface, which is not attractive on
WinXP. We do have some old home-grown tools developed with Perl/Tk
that are being phased out in favor of Java - instead of getting a face
lift with a newer version of Tk. Sometimes the old tools get
rewritten just so the latest maintainer can use their language of
choice but I've never seen a manager object to dumping a working
utility that had a TK gui component - which doesn't seem good for the
Tcl/Tk community.
The guys who work mainly on the Unix platform still use tclsh a lot,
though.
Terry
John Kelly
<uwe_klein_...@t-online.de> wrote:
>> Anything that clears the engineers out of the room is OK with me.
>you won't move me a teeny bit.
>None of my <censored> pounds ;-)
>Whats so bad about my co-engineers?
I don't dislike people who happen to be engineers.
Though nowdays it seems CS belongs to the college of engineering, it
wasn't always so. And now the grads think they're "engineers," even
though they don't have any bridges to build.
To me, programming is more architecture than engineering, so I would
never call myself an engineer. If had a time machine, I would start
over and study architecture, instead of CS. Engineers are just too
dull. ;-)
nobodywhoareyou
> Maybe we should change the name to
>
> Enterprise
> Vertical
> Integration
> Language
I'm thinking with the EVIL O'Reilly book on my shelf, people would
think twice before interrupting. And that works for me... ;-)
Terry
Uwe Klein
There is a reason architects do the flashy outside of buildings
and engineers have to bend over backwards to keep that fancy stuff
from crumbling down or being generally useless ;-)
In olden times these jobs used to be in one hand, and lots of stuff
from that time is a very acceptable combination of esthetics and sound
engineering. They look good _and_ they still stand soundly.
>
On a sidenote: a bit more of the engineering style
"have we covered all angles" approach
and less
"exceptionally bright ideas" that fail with an "exceptionally bright flash"
would be a boon for a signifcant number of (commercial) software projects.
>
uwe
Wojciech Kocjan
> You know that another season has come and gone, when the trees shed
> their leaves, winter blankets the northern hemisphere, and another TCL
> IS DYING thread gets posted to clt.
> [...]
Hi,
The thread inspired me to write a couple of things in my mind on a
Tcl-related blog I'm trying to create. Here's the link:
http://tclmentor.kocjan.org/2007/10/is-tcl-alive-and-well-or-dying.html
I hope you don't consider this spamming.
--
Wojciech Kocjan
Donal K. Fellows
> In olden times these jobs used to be in one hand, and lots of stuff
> from that time is a very acceptable combination of esthetics and sound
> engineering. They look good _and_ they still stand soundly.
The rest have mostly fallen down already. (Medieval cathedrals had a
real tendency to not stay standing properly, and it took a long time for
people to properly understand why...)
Donal.
Donal K. Fellows
> I'm thinking with the EVIL O'Reilly book on my shelf, people would
> think twice before interrupting. And that works for me... ;-)
Trouble is, there'd also be _EVIL for Dummies_ there. But the O'Reilly
book could have a picture of Dick Cheney on the cover.
Donal.
nobodywhoareyou
I'm lucky I wasn't drinking coffee or it would be sprayed over the
keyboard. I'm going to have to go home now and figure out a way to
get that picture out of my mind. (There isn't enough alcohol this
side of the Mississippi...)
Terry
Alan Anderson
sd <sha...@hotmail.com> wrote:
> In the IT world, I found it hard to convince
> people to use a product built from something they never heard of.
I have it worse -- someone in the IT group where I work *has* heard of
Tcl, considers it "shareware", and it is thus labeled as unsuitable for
production work.
skuh...@web.de
Uwe Klein wrote:
> > Hmm... I mean these "Wildebeest": http://en.wikipedia.org/wiki/Gnu ,
> > which of course doesn't look very stupid, especially if it stands in
> > front of you ;-) , but I was referring to GNU (www.gnu.org) and the
> > logo. - Maybe I used the wrong words...
> >
> > Stephan
> >
> You would have to live down under to be able to designate
> these as "nordic" , right?
Damn, this should have been my response! ;-) No, I'm in Germany and
was simply wrong, because I had an elk in mind, when I was writing
gnu...
Stephan
Uwe Klein
I spoke of engineers, not religious fanatics.
Churches used to be bombastic buildings sitting on the
foundations of roman castelli.
Roman infrastructure works are what i was thinking of.
>
> Donal.
uwe
Cameron Laird
>was simply wrong, because I had an elk in mind, when I was writing
>gnu...
.
.
Note for North Americans: European "elk" sometimes
most closely corresponds to what we call "moose".
Sean Woods
> Hi,
>
> The thread inspired me to write a couple of things in my mind on a
> Tcl-related blog I'm trying to create. Here's the link:
>
> http://tclmentor.kocjan.org/2007/10/is-tcl-alive-and-well-or-dying.html
>
> I hope you don't consider this spamming.
That's a very nice summation of where all of these threads end up.
I'll have you know that some of the best and brightest minds in my
head are actually working on the problem of getting different object
systems to talk to one another.
As for growth, one must remember that simply growing fast is not
always the best approach. Bamboo grows extremely fast, and while it is
reasonably strong, it's only useful if you need long cylinders. Pine
trees grow fast, but they rot fast too. One dry summer and you see
hundreds of square miles of dead pines. Oak trees are hardy, but they
take years to grow. You have the poplar, that is really a network of
root systems that can live for thousands of years. All we see are the
shoots above the surface. If we were to cut/burn/mow down an entire
stand of poplars, they would actually grow back. Then there is the
redwood and the sequoia. They do not grow very fast at all. The key to
their survival is that their wood does not rot.
Ruby from what I've seen is Bamboo. It's fast, but god help you if you
need something in a different shape.
Perl and PHP are Pine. It's fast. It's flexible. But if you don't
varnish it, it'll rot out in a season.
Java is the mighty oak.
Fortran and Cobol are definitely the redwoods of the Language forest.
Code written for them just sits quietly and runs.
C and Tcl are poplars. They exist on a different level.
Implementations shoot to the surface, exist, and die, but the source
is unchanged.
Donal K. Fellows
> I'll have you know that some of the best and brightest minds in my
> head are actually working on the problem of getting different object
> systems to talk to one another.
Yeah, the best minds are working on it, and I'm having a go too!
Donal.
EKB
> On Wed, 03 Oct 2007 13:06:50 -0700, suchenwi
>
> <richard.suchenwirth-bauersa...@siemens.com> wrote:
> >John Kelly schrieb:
> >> Yeah, who wants to go on a date with Tom Raper. If I had a bad name,
> >> I would change it.
> >I have no problems with "tickle", but everyone who has can always call
> >the language "Tool Control", which may sound stronger :^)
>
> In the case of Tcl, I don't really think a new name will improve its
> popularity.
>
> There are too many technologies and too many tools. They can't all
> stand out among the crowd. Some become popular, others don't. Who
> knows why. That's just the way the wind blows, something man cannot
> change.
>
> --
> Internet servicehttp://www.isp2dial.com/
I feel somewhat the same, in that there's a certain inevitability to
it, but I do think that "tickle" gets in the way. Here are the things
I see stopping my colleagues from adopting it, even after they see how
productive it can be:
1. They've never heard of it before
2. "Tickle?" heh heh
3. The syntax is unfamiliar, and when it comes to math...
4. They ask, "I really have to write [expr {$x + $y}] just to add two
numbers?"
5. It's too easy to develop an application quickly (yes, really)
6. They're already heavily invested in another language
#1 is the main problem, IMHO. They haven't heard of it, so they don't
want to adopt it, so it isn't adopted, so they haven't heard of it...
But the reason why people don't experiment with it is because of #s
1-6. #5 may sound odd, but I have begun to suspect that if there isn't
a big effort (or at least a big IDE) required to build applications,
as there is for Java, C#, and Delphi, then people don't think it's a
"serious" language. To be honest, #4 annoys me also, but the rest of
the syntax is so clean and powerful that I put that annoyance aside.
I don't see these objections being overcome by making any changes that
would leave the current Tcl user base happy. However, one possibility
did just occur to me: produce a book titled "Tool Command Language
(TCL) in the Enterprise," which would present a top-heavy but robust
approach to making applications using Tcl & Tile. It would avoid most
of the usual arguments for why to use Tcl, and just focus on those
that would be of interest for a corporate team. It would assume that
you will only create Tclkits (or equivalent), and will never assume
the user of the program will have a Tcl distribution installed. It
would include coding standards and frameworks and discuss
introspection heavily. It would list commercial sources for support,
and downplay the excellent free sources of support.
I couldn't/shouldn't be the author for such a book, and it might not
make it past the post. But it might just work.
John Kelly
> Here are the things I see stopping my colleagues from adopting it,
> even after they see how productive it can be:
>
> 1. They've never heard of it before
> #1 is the main problem, IMHO. They haven't heard of it, so they don't
> want to adopt it, so it isn't adopted, so they haven't heard of it...
I knew about Tcl, but never had time to learn it.
Then one day I looked at the UW Alpine webmail system. When I saw
they use Tcl for scripting, I wondered why. But I was interested in
their webmail system, so I became motivated to learn more about Tcl.
After coding an experimental server using threads/events in Tcl, I
realize Tcl won't do what I want. The interpreters consume too much
memory to scale to beyond a few hundred threads.
But at least now, I can read and write a little Tcl, and it's helped
me explore some design ideas I'm interested in.
Cameron Laird
.
.
Tcl's so clear and honest you can figure out its limits
in a few days. There's an important point there I wish
I knew how to articulate better; I know it's important,
though, in a world where Java teams bash their heads
against walls for months, investing what can be MILLIONS
of dollars, before figuring out they've made a ... sub-
optimal choice.
Donal K. Fellows
> After coding an experimental server using threads/events in Tcl, I
> realize Tcl won't do what I want. The interpreters consume too much
> memory to scale to beyond a few hundred threads.
Here's a small not-very-well-hidden secret. It's really rather hard to
do much better than Tcl in this area. Not impossible (Tcl's written in
C, so you can always go to C itself where the should be extra
optimization you can do) but definitely difficult, especially on a
limited budget. Of course our thread model can't run nearly as many
threads on a single processor machine as some system using green threads
(notably Python and Ruby) but we scale far better on parallel hardware
as we have a largely lock-free system. Then you've still got to use
event driven I/O on top of this, as events still scale better than
threads on a single processor as they're lighter weight. (The optimal
number of threads per processor and connections per thread? No idea! But
I have my usual advice here: Don't guess, measure!) The combination of
the two, with careful management of a few other things (getting to the
top takes effort) can give you really serious scalability power...
To start towards really high scalability, think carefully how much state
really needs to be per-user. It's often not that much other than the
bits in the database. At that point, you realize you're just mediating
between the user and the corresponding database entries.
Donal.
davidn...@gmail.com
<donal.k.fell...@manchester.ac.uk> wrote:
> John Kelly wrote:
> > After coding an experimental server using threads/events in Tcl, I
> > realize Tcl won't do what I want. The interpreters consume too much
> > memory to scale to beyond a few hundred threads.
>
> Here's a small not-very-well-hidden secret. It's really rather hard to
> do much better than Tcl in this area.
Erlang is a step above Tcl (and pretty much everything else) in that
department. The runtime itself has an event-based loop, like Tcl's,
however it also has a scheduler so that you can do processing on
events and not block. Furthermore, it has SMP support to divy up
"processes" (not OS processes, Erlang processes) to multiple cores.
Ok, so it's not so good in other departments, but I think that right
now, it's the leader in concurrency - at least in terms of practical
languages that you can use in a production environment.
Alan Anderson
> After coding an experimental server using threads/events in Tcl, I
> realize Tcl won't do what I want. The interpreters consume too much
> memory to scale to beyond a few hundred threads.
If "what you want" is to write an application that uses several hundred
threads, then Tcl probably isn't a good tool for you.
However, if "what you want" is to write a web interface to an IMAP
server, Tcl ought to be perfectly capable of doing it. It seems to me
in this case that threads are a design detail, not a requirement.
John Kelly
>Erlang is a step above Tcl (and pretty much everything else) in that
>department. The runtime itself has an event-based loop, like Tcl's,
>however it also has a scheduler so that you can do processing on
>events and not block. Furthermore, it has SMP support to divy up
>"processes" (not OS processes, Erlang processes) to multiple cores.
>Ok, so it's not so good in other departments, but I think that right
>now, it's the leader in concurrency - at least in terms of practical
>languages that you can use in a production environment.
I didn't know about it, thanks for the tip. Looks interesting.
nobodywhoareyou
> I don't see these objections being overcome by making any changes that
> would leave the current Tcl user base happy. However, one possibility
> did just occur to me: produce a book titled "Tool Command Language
> (TCL) in the Enterprise," which would present a top-heavy but robust
I'd buy this book in a heartbeat. And if you tossed "lean" into the
title somewhere, it wouldn't hurt the sales. ;-)
Terry (not sorry for griping about "tickle" now)
Sean Woods
> I don't see these objections being overcome by making any changes that
> would leave the current Tcl user base happy. However, one possibility
> did just occur to me: produce a book titled "Tool Command Language
> (TCL) in the Enterprise," which would present a top-heavy but robust
> approach to making applications using Tcl & Tile. It would avoid most
> of the usual arguments for why to use Tcl, and just focus on those
> that would be of interest for a corporate team. It would assume that
> you will only create Tclkits (or equivalent), and will never assume
> the user of the program will have a Tcl distribution installed. It
> would include coding standards and frameworks and discuss
> introspection heavily. It would list commercial sources for support,
> and downplay the excellent free sources of support.
I second that. So long as we call it EVIL. There are a lot of hard,
sometime arbitrary, and usually controversial decisions that would
have to be made in putting the package together. The whole idea of
stating emphatically that there is one good way of doing things flys
in the face of the Tcl way.
Thus, this work, while it would build on the tools and skills of the
Tcl Community, would have to take on a life of it's own as Tcl's EVIL
twin. Also, calling itself by a seperate name would give the
maintainer some leverage to say "Hey, you are writing in EVIL. EVIL is
as EVIL does we need to nail the jelly to the wall. If you want
flexibility, you want to code in Tcl, not EVIL."
EVIL will probably introduce some additional creature comforts that
may or may not be part of a standard Tcl release. Again, but starting
off as EVIL, we can always come to the light side later. But we don't
want to, for a minute, corrupt the purity of Tcl.
And having to design a concrete enterprise framework would require a
LOT of corrupting simplifications.
--The Hypnotoad
Donal K. Fellows
> EVIL will probably introduce some additional creature comforts that
> may or may not be part of a standard Tcl release. Again, but starting
> off as EVIL, we can always come to the light side later. But we don't
> want to, for a minute, corrupt the purity of Tcl.
That makes me wonder: can we get Google to use it?
Donal.
Mark Roseman
We wouldn't have been able to a few years ago, but nowadays I don't
think there are any obstacles on that front at all...
Mark
davidn...@gmail.com
Google have four programming languages for production code: Java,
Javascript, Python and C++ (presumably C as well), and seem very rigid
about it. You don't want to try and get behemoths using your language
(it's a tough fight), you want the guys who will be tomorrow's
behemoths to get started using your language.
Sean Woods
wrote:
Well I'm teaching it to High School kids now.