FreeBSD vs. Linux
Andy
Could someone please give me a quick info about the differences between
Linux and FreeBSD, alternativly some good resources on the net where I can
learn about the differencies?
/Andy
Homer Simpson
Symbiotic Order
www.linux.org
--
- joe
http://www.mindspring.com/~joekiser/
"Andy" <news...@broadpark.no> wrote in message
news:avsi9i$j9mlm$1...@ID-48010.news.dfncis.de...
Rob MacGregor
I'd suggest you look at what you want to use the system for first, then look
at what fits that requirement. Starting off with assumptions about your
choice of OS might not produce the best result for you.
--
Rob MacGregor (MCSE)
The light at the end of the tunnel is an oncoming dragon.
JoeBlow
Simon Barner
You might want to have a look at the following thread in this news group:
Subject: FreeBSD versus Linux
Date: Mon, 30 Dec 2002 23:41:26 +0100
Msg-id: <auqi3v$ss4$1...@reader08.wxs.nl>
Simon
Donn Miller
The best thing to do is try a few Linux distros, and then try FreeBSD.
Use one or both, depending on your needs and what you like best.
Myself, I get turned off by some Linux distros, because they try to do
too much, i.e., they try to second-guess what the user's needs are.
FreeBSD only installs what you tell it to. For example, even Gentoo,
which is supposed to be so minimalist, installed enigmail, Microsoft
fonts, and configured Mozilla for anti-aliased fonts (which look
horrible on a laptop, BTW) when all I wanted was Mozilla. This seems to
be typical of quite a few Linux distros. Plus, FreeBSD and Linux more
or less do the same thing anyways, despite being totally different
internally. The documentation for FreeBSD (mainly man pages) is very
superbly kept up to date.
-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----== Over 80,000 Newsgroups - 16 Different Servers! =-----
charger
http://people.freebsd.org/~murray/bsd_flier.html
Choice is yours :)
--
www.chimba-team.tk
Croatian street racing
Peter Wu
> On Sun, 12 Jan 2003 21:13:34 +0100, Andy wrote:
>
>> Hi
>>
>> Could someone please give me a quick info about the differences
>> between Linux and FreeBSD, alternativly some good resources on the
>> net where I can learn about the differencies?
>>
>> /Andy
>
> http://people.freebsd.org/~murray/bsd_flier.html
I think this comparison is too old.. :(
"FreeBSD uses the UFS (Unix File System), which is a little more complex
than Linux's ext2."....
What about the performance benching between UFS and some new Linux File
Systems?
--
Cheers, Peter
ta...@lpthe.jussieu.fr
If one believes what is said about the new Reiserfs4, it should be much faster
than any existing filesystem, plus having some extremely interesting features
like virtual directories providing constructed views of the filesystem,
similar to views in sql, plugins allowing to do on the fly compression or cyphering,
search and indexation, etc.
: --
: Cheers, Peter
--
Michel TALON
Andy
Thanx, I'm checking it now, and it seems to give me the information I was
out after
/Andy
Jorge Godoy
> http://people.freebsd.org/~murray/bsd_flier.html
> Choice is yours :)
They should update that table. There are several Linux things that
aren't as said anymore. And, it should also read something like 'Red
Hat Linux', since it's RH's fault to turn on services by
default... Also, ext2 isn't reliable. Comparing FreeBSD with
SoftUpdates with an ext2 FS is like benchmarking a Ferrari with
popular cars... It should at least cite ext3. Besides, it's based on
kernel 2.2 (it cites the 'new 2.4'), whis isn't the mainstream anymore
and had several things that diminished performance. There are also
mistakes on the impossibility of releasing binary drivers for Linux
(see, e.g., NVidia's driver for X that runs perfectly on it or
winmodem devices that are kernel modules) and compatibility with
binaries for other OS's (Linux has an IBCS compatibility module).
On the development area, the notes on applications not being
interchangeable between distributions should be changed so that the
reader knows that it is due to different library versions; one can
always get the sorce code and required libraries and regenerate the
application on his or her favourite distribution of Linux.
I think that if it's going to be used *today*, this document should be
updated.
--
Godoy. <go...@metalab.unc.edu>
ta...@lpthe.jussieu.fr
: charger <chimba...@gmx.net> writes:
However, it is of course biased towards BSD, but gives rather honest
account of the situation. For example, he says that Linux performances are
rapidly converging towards FreeBSD performances, which is true. He says that
the FreeBSD ports system is amazing, which is true, but to be honest, the
Debian package system is also excellent and has even more packages. What is
blatantly false is the question of binary drivers, since such ugly things
exist for Linux since a long time (for example the nvidia driver you mention)
and are just now available for FreeBSD. Linux having a much bigger market
share has more attention from hardware developers, from the Java community,
etc. The library problem is a real plague in the Linux world, as everybody
knows. Of course one can recompile the apps, if they are provided in source
form (not always the case, think SAP for example), but it is not always
obvious to make them fit to the dependency system (rpm, etc.).
To turn back to the perfs problem, here is an interesting quote of a well
known Linux developer, I. Molnar who has developed the O(1) scheduling
algorithm which allows Linux to schedule thousands of threads efficiently:
-------------
A: What areas of the Linux kernel do you think still lags behind FreeBSD?
Ingo Molnar: there were two areas where i think we used to lag, the VM and the
block IO subsystem - both have been significantly reworked in 2.5. Whether the
VM got better than FreeBSD's remains to be seen (via actual use), but the
Linux VM already has features that FreeBSD does not have, eg. support for more
than 4 GB RAM on x86 (here i guess i'm biased, i wrote much of that code). But
FreeBSD's core VM logic itself, ie. the state machine that decides what to
throw out under memory pressure, how to swap and how to do IO, is top-notch. I
think with Andrew Morton's and Jens Axobe's latest VM and IO work we are
top-notch as well (with a few extras perhaps).
There's also an interesting VM project in the making, Arjan van de Ven's O(1)
VM code. [without doubt i do appear to have a sweet spot for O(1) code :-) ]
Rik van Riel has merged Arjan's code a couple of days ago. The code converts
every important VM algorithm (laundering, aging) to a O(1) algorithm while
still keeping the fundamentals - this is quite nontrivial for things like page
aging. It's in essence the VM overhead reduction work that Andrea Arcangeli
has started in 2.4.10, brought to the extreme. I have run Arjan's O(1) VM
under high memory pressure, and it's really impressive - kswapd (the central
VM housekeeping kernel thread), which used to eat up lots of CPU time under VM
load, has almost vanished from the CPU usage chart.
I do have the impression that the Linux VM is close to a conceptual
breakthrough - with all the dots connected we now have something that is the
next level of quality. The 2.5 VM has merged all the seemingly conflicting VM
branches that fought it out in 2.4, and the many complex subsystems involved
suddenly started playing in concert and produce something really nice.
---------
If now you add to that the fact that many exciting things occurred at the file
system level, for example xfs, and the new reiserfs4, one may wonder wether
FreeBSD is not becoming to lag behind. Of course softupdates+dirhash+dirpref
is something working very well, but the contenders have solid arguments.
There are certainly places where FreeBSD appears particularly obsolete, the
installer being one of these. The exciting new features in FreeBSD 5.0
come with a lot of delay, the release has been postponed for one year, and
the work is still not finished. Obviously this is bad. There is clearly more
momentum in the Linux camp.
: --
: Godoy. <go...@metalab.unc.edu>
--
Michel TALON
Jorge Godoy
> However, it is of course biased towards BSD, but gives rather honest
> account of the situation.
I haven't argued against that :-) I also thought it was biased, but
hey! He's "selling" FreeBSD.
I use both --- now I'm using Linux --- and I think both are excellent
systems.
> For example, he says that Linux performances are
> rapidly converging towards FreeBSD performances, which is true. He says that
> the FreeBSD ports system is amazing, which is true, but to be honest, the
> Debian package system is also excellent and has even more packages. What is
> blatantly false is the question of binary drivers, since such ugly things
> exist for Linux since a long time (for example the nvidia driver you mention)
> and are just now available for FreeBSD. Linux having a much bigger market
> share has more attention from hardware developers, from the Java community,
> etc. The library problem is a real plague in the Linux world, as everybody
> knows. Of course one can recompile the apps, if they are provided in source
> form (not always the case, think SAP for example), but it is not always
> obvious to make them fit to the dependency system (rpm, etc.).
Nowadays it isn't all that hard to create a package. One can 'borrow'
the source of another package and make the necessary modifications to
the rules that generate that package (that would be the <package>.spec
on rpm).
It will also take care of the dependencies for that specific
distribution version.
With regards to SAP, one should use one of their certified
distributions if he doesn't want any headaches or is an unexperienced
user.
> To turn back to the perfs problem, here is an interesting quote of a well
> known Linux developer, I. Molnar who has developed the O(1) scheduling
> algorithm which allows Linux to schedule thousands of threads efficiently:
>
> -------------
> A: What areas of the Linux kernel do you think still lags behind FreeBSD?
>
> Ingo Molnar: there were two areas where i think we used to lag, the VM and the
> block IO subsystem - both have been significantly reworked in 2.5. Whether the
> VM got better than FreeBSD's remains to be seen (via actual use), but the
> Linux VM already has features that FreeBSD does not have, eg. support for more
> than 4 GB RAM on x86 (here i guess i'm biased, i wrote much of that code). But
> FreeBSD's core VM logic itself, ie. the state machine that decides what to
> throw out under memory pressure, how to swap and how to do IO, is top-notch. I
> think with Andrew Morton's and Jens Axobe's latest VM and IO work we are
> top-notch as well (with a few extras perhaps).
Funny. I've been talking to Rik van Riel this morning at Conectiva,
here in .br, and we were discussing the Linux VM subsystem ans some
new things he has done with it.
> There's also an interesting VM project in the making, Arjan van de Ven's O(1)
> VM code. [without doubt i do appear to have a sweet spot for O(1) code :-) ]
> Rik van Riel has merged Arjan's code a couple of days ago. The code converts
> every important VM algorithm (laundering, aging) to a O(1) algorithm while
> still keeping the fundamentals - this is quite nontrivial for things like page
> aging. It's in essence the VM overhead reduction work that Andrea Arcangeli
> has started in 2.4.10, brought to the extreme. I have run Arjan's O(1) VM
> under high memory pressure, and it's really impressive - kswapd (the central
> VM housekeeping kernel thread), which used to eat up lots of CPU time under VM
> load, has almost vanished from the CPU usage chart.
Indeed, as it is now, the load on the CPU is a lot heavier while
compared to, e.g., FreeBSD. I feel that for some applications my
FreeBSD goes smoother than Linux. It also happens that some
applications run better with Linux than FreeBSD...
> I do have the impression that the Linux VM is close to a conceptual
> breakthrough - with all the dots connected we now have something that is the
> next level of quality. The 2.5 VM has merged all the seemingly conflicting VM
> branches that fought it out in 2.4, and the many complex subsystems involved
> suddenly started playing in concert and produce something really nice.
>
> ---------
>
> If now you add to that the fact that many exciting things occurred at the file
> system level, for example xfs, and the new reiserfs4, one may wonder wether
> FreeBSD is not becoming to lag behind. Of course softupdates+dirhash+dirpref
> is something working very well, but the contenders have solid arguments.
> There are certainly places where FreeBSD appears particularly obsolete, the
> installer being one of these. The exciting new features in FreeBSD 5.0
> come with a lot of delay, the release has been postponed for one year, and
> the work is still not finished. Obviously this is bad. There is clearly more
> momentum in the Linux camp.
I hope FreeBSD makes its installer a little more friendlier. And the
partition utility too. :-) I've installed it several times here and on
customers of my company, but it is still incredibly easier to install
a Linux system.
Particularly I like very much the ports system and how tied are things
in there, but with regards to binary distribution, Apt (and it also
runs with RPM files) is *extremely* efficient (I like it more than
pkg_add, but I also like it more than rpm by itself).
As I always say, there are good things in both OS. One has to know
both them and see which one will be more interesting to some
particular use.
--
Godoy. <go...@metalab.unc.edu>
Donn Miller
> If now you add to that the fact that many exciting things occurred at the file
> system level, for example xfs, and the new reiserfs4, one may wonder wether
> FreeBSD is not becoming to lag behind. Of course softupdates+dirhash+dirpref
> is something working very well, but the contenders have solid arguments.
Don't forget about snapshots and background fsck which is in 5.0.
Also, journalled filesystem sometimes require the occasional fsck
anyways, because the restore-from-journal feature isn't 100% perfect,
and can't correct major problems. Presumably, doing a fsck would be
better, and it can commence in the background. Also, isn't FreeBSD's
networking code still better? FreeBSD has had ipv6 for almost 2 years
now, but Linux's ipv6 code isn't mature yet. FreeBSD 5.0 will also
feature ACL's, so that aspect of it is up-to-date wrt Linux.
> There are certainly places where FreeBSD appears particularly obsolete, the
> installer being one of these. The exciting new features in FreeBSD 5.0
> come with a lot of delay, the release has been postponed for one year, and
> the work is still not finished. Obviously this is bad. There is clearly more
> momentum in the Linux camp.
Ah well, but who really cares about installers? How many times is a
person have to install an OS? Good OSes only need one install. And
FreeBSD's installer is still better than Slackware's, Gentoo, and many
other LFS-like distros. Also, FreeBSD is still more coherently and
sensibly laid out. And FreeBSD still doesn't have as many rabid
advocates as Linux, not even close. Plus, FreeBSD users tend to be
more mature.
If there's any one cause for Linux' momentum, it's because Linux tends
to be much more wild and crazy in the development side of things, but
FreeBSD is, and has always been, more conservative in its design and
other decisions. Linux has hype and all kinds of other frenzy about
it, which naturally attracts more developers and companies looking to
support the OS.
Jorge Godoy
> FreeBSD has had ipv6 for almost 2 years
> now, but Linux's ipv6 code isn't mature yet. FreeBSD 5.0 will also
> feature ACL's, so that aspect of it is up-to-date wrt Linux.
IPv6 in FreeBSD is incomparable to Linux. It works and is easy to
setup. Linux is still a child in the IPv6 arena.
> Ah well, but who really cares about installers? How many times is a
> person have to install an OS?
I care. While my customers won't care about it, we need to make
installations of it very often (thanks God!).
We are a consulting company and we have more than one customer ---
actually, we have new customers often (thanks God, again).
> Good OSes only need one install. And
> FreeBSD's installer is still better than Slackware's, Gentoo, and many
> other LFS-like distros.
This is a good reason for those distros be lagging behind easily
installable distributions.
> Also, FreeBSD is still more coherently and
> sensibly laid out. And FreeBSD still doesn't have as many rabid
> advocates as Linux, not even close. Plus, FreeBSD users tend to be
> more mature.
But they still need their job done quickly and efficiently. If we
spend too much time installing machines, we will be able to have less
customers than we should with something faster and easier.
--
Godoy. <go...@metalab.unc.edu>
ta...@lpthe.jussieu.fr
: I hope FreeBSD makes its installer a little more friendlier. And the
: partition utility too. :-) I've installed it several times here and on
: customers of my company, but it is still incredibly easier to install
: a Linux system.
Me too. I don't do FreeBSD installs frequently, fortunately, since i upgrade
with cvsup, but each time i have to do that, i end up hitting the wrong key,
doing something ugly, and having to reboot :-)
: Particularly I like very much the ports system and how tied are things
: in there, but with regards to binary distribution, Apt (and it also
: runs with RPM files) is *extremely* efficient (I like it more than
: pkg_add, but I also like it more than rpm by itself).
I think that the ports system and the ease of upgrade via cvsup are really
one of the strongest points of FreeBSD. Indeed the Debian system is really
very good, but with the recent addition of portupgrade, FreeBSD is on par.
Note that Debian concurs with FreeBSD for the ugliest installer around...
: As I always say, there are good things in both OS. One has to know
: both them and see which one will be more interesting to some
: particular use.
You are right. Basically both are very good and both have the same set of
applications. There are still some instances where one is stuck to using
only one, for example i have an ADSL USB modem which has only a Linux driver.
I fear that this situation becomes more and more common.
Eric Masson
Michel> Note that Debian concurs with FreeBSD for the ugliest installer
Michel> around...
The fact that the installer is ugly isn't so annoying imho. The real
point is that sometimes a wrong keystroke can lead to reboot cause
sysinstall won't work correctly after that (change wrong media setup or
other hiccups like that, sorry no PR atm, last time I used sysinstall, I
had too much work to take time to reproduce the case and fill and a PR)
I'd like to see sysinstall-ng for the ones interested and another
install method, with a shell and a few scripts.
Eric Masson
--
B: Quelqu'un a-t-il déjà fait fonctionner ce modem sous linux ????
G.P.: Je me demande si ça ne marcherait pas en faisant un proxy
avec un vieux PC sous Windows en réseau avec un PC sous Linux.
-+- in Guide de linuxien pervers : "le réseau, c'est simple comme un proxy"
ta...@lpthe.jussieu.fr
:> If now you add to that the fact that many exciting things occurred at the file
:> system level, for example xfs, and the new reiserfs4, one may wonder wether
:> FreeBSD is not becoming to lag behind. Of course softupdates+dirhash+dirpref
:> is something working very well, but the contenders have solid arguments.
: Don't forget about snapshots and background fsck which is in 5.0.
: Also, journalled filesystem sometimes require the occasional fsck
: anyways, because the restore-from-journal feature isn't 100% perfect,
: and can't correct major problems. Presumably, doing a fsck would be
: better, and it can commence in the background. Also, isn't FreeBSD's
: networking code still better? FreeBSD has had ipv6 for almost 2 years
: now, but Linux's ipv6 code isn't mature yet. FreeBSD 5.0 will also
: feature ACL's, so that aspect of it is up-to-date wrt Linux.
reiserfs4 has some features that, if they really work, are a conceptual
breakthrough in the domain.
For the networking code, i have posted recently here a reference to a
benchmark which is not very glorious for FreeBSD. Of course the problem seems
to be localized to netgraph.
http://derbian.org/pppoe/
: Ah well, but who really cares about installers?
Everybody in fact. This is the first thing you see when you discover an OS,
and a crappy installer gives a very bad idea of the system. There exist Linux
distros whose only strong point is the installer, like Mandrake, and a lot of
people use them because of that. There are excellent distros that few people
use, like Debian, because the installer is crappy. Exactly the same as when
you are writing a text, there must be an introduction, development and
conclusion, and the introduction has to be excellent. The introduction to the
OS is the installer. Note that J. Hubbard says similar things on his paper
which is referred to in FreeBSD web site.
: How many times is a
: person have to install an OS? Good OSes only need one install. And
: FreeBSD's installer is still better than Slackware's, Gentoo, and many
: other LFS-like distros. Also, FreeBSD is still more coherently and
: sensibly laid out.
Yes. Throwing everything under /usr like even good distros like Debian do
is nonsensical.
: And FreeBSD still doesn't have as many rabid
: advocates as Linux, not even close. Plus, FreeBSD users tend to be
: more mature.
This is not too much of an argument, i think. I could not care less about
the Linux advocates or the FreeBSD advocates. What is very nice with FreeBSD
is that you can easily find answers when you have problems.
: If there's any one cause for Linux' momentum, it's because Linux tends
: to be much more wild and crazy in the development side of things, but
: FreeBSD is, and has always been, more conservative in its design and
: other decisions. Linux has hype and all kinds of other frenzy about
: it, which naturally attracts more developers and companies looking to
: support the OS.
You are very right. And i am becoming to think that the Linux "model"
is showing great advantages compared to the FreeBSD one. When FreeBSD
has to develop the new KSE, only two or three developers do the job,
and it takes ages to be done. In the same time tons of people
contribute to Linux. Conservatism is the most direct way to ending
as a complete looser.
--
Michel TALON
David Magda
[...]
> I hope FreeBSD makes its installer a little more friendlier. And the
> partition utility too. :-) I've installed it several times here and on
> customers of my company, but it is still incredibly easier to install
> a Linux system.
OpenBSD's disklabel(8) has a wonderful feature that is activated with
the '-E' command-line option -- an interactive editor. Anyone know
how complex the code is?
Wheever I have to do editing of the filesystem/partition table (which
is rare), I always use /stand/sysinstall -- it makes things so much
easier.
--
David Magda <dmagda at ee.ryerson.ca>
Because the innovator has for enemies all those who have done well under
the old conditions, and lukewarm defenders in those who may do well
under the new. -- Niccolo Machiavelli, _The Prince_, Chapter VI
ta...@lpthe.jussieu.fr
: Jorge Godoy <go...@metalab.unc.edu> writes:
: [...]
:> I hope FreeBSD makes its installer a little more friendlier. And the
:> partition utility too. :-) I've installed it several times here and on
:> customers of my company, but it is still incredibly easier to install
:> a Linux system.
: [...]
: OpenBSD's disklabel(8) has a wonderful feature that is activated with
: the '-E' command-line option -- an interactive editor. Anyone know
: how complex the code is?
Perfectly true. By the way fdisk is better also. Once on the mailing lists
i have seen the question discussed of porting OpenBSD fdisk and disklabel
to FreeBSD. You may imagine the concert of lamentations that ensued ...
Our fdisk is better, less buggy, and so on. To make a long story short,
nothing was changed. From time to time Matt Dillon dares touching
something, and then all the old farts scream like old ladies. I have seen
quite insulting posts against M. Dillon who is one of the few
effectively introducing innovations.
--
Michel TALON
ta...@lpthe.jussieu.fr
: I'd like to see sysinstall-ng for the ones interested and another
: install method, with a shell and a few scripts.
: Eric Masson
I agree completely. One may have the poor man's method, like in OpenBSD with
a shell script on the console, or the luxury method like knoppix, that
is a shell script under a splendid desktop, and all the documentation and
utilities available. Both are preferable. Or a decent installer like the
Mandrake one, which incidentally is under GPL so could be reused easily.
I have read the paper by J. Hubbard, and while i agree with his critics of the
present system, i am extremely sceptic about the replacement he proposes.
I fear to see still a more ugly monster than the present sysinstall.
--
Michel TALON
David Magda
[...]
> something, and then all the old farts scream like old ladies. I have seen
The image that formed in my mind was quite hilarious.
I'd pay money to see that. Do you have any JPEGs? :>
JD
<ta...@lpthe.jussieu.fr> wrote in message news:b024nj$1kh6$2...@asmodee.lpthe.jussieu.fr...
> David Magda <dmagda...@ee.ryerson.ca> wrote:
>
> : Jorge Godoy <go...@metalab.unc.edu> writes:
> : [...]
> :> I hope FreeBSD makes its installer a little more friendlier. And the
> :> partition utility too. :-) I've installed it several times here and on
> :> customers of my company, but it is still incredibly easier to install
> :> a Linux system.
> : [...]
>
> : OpenBSD's disklabel(8) has a wonderful feature that is activated with
> : the '-E' command-line option -- an interactive editor. Anyone know
> : how complex the code is?
>
> Perfectly true. By the way fdisk is better also. Once on the mailing lists
> i have seen the question discussed of porting OpenBSD fdisk and disklabel
> to FreeBSD. You may imagine the concert of lamentations that ensued ...
> Our fdisk is better, less buggy, and so on. To make a long story short,
> nothing was changed. From time to time Matt Dillon dares touching
> something, and then all the old farts scream like old ladies.
>
opinions before making changes. He has hopefully moderated his behavior,
but the opinions of those who are responsible for the code can be
difficult to change.
Maybe Matt has gotten a 'bad rap', but it is difficult to erase the impressions
from earlier behavior.
John
Donn Miller
> Yes. Throwing everything under /usr like even good distros like Debian do
> is nonsensical.
Yes, but I think there's a document or standard out, FHS, or something
like that, which was drafted by members of the Linux community. It says
that everything should go under /usr and not /usr/local. Hopefully,
you'll know what I'm talking about.
The problem is that many Linux distros have a customizable or vague
notion of what constitutes a "base" system. In FreeBSD and the other
BSD's, the distinction is well-defined. For many Linux distros, Gentoo
being one of them, what constitutes a "base" system isn't fixed, but
varies by the individual. Hence, almost every package counts as being
part of "base". To that end, I think I've heard a blurb about Solaris
becoming much of the same way, at least they're doing away with /usr/local.
The major problem with keeping everything under /usr is it makes it hard
to blow away your external packages and re-install everything from
scratch, or to back up the base system and ignore the packages in
/usr/local.
I've heard all the unices were becoming the same way, because of Linux's
influence. No matter how superior or inferior Linux is to the BSD's, I
still think it would be wise for the Linux distributors to keep some
semblance of order by making it easy to distinguish between the base
system and external packages and reflecting this order in the directory
hierarchy. Take Gentoo, for example. There is a clear distinction of
what "BASE" is. Then it has a concept of what the "world". The world =
base system + external packages. So surely, Gentoo should put all
external packages into /usr/local, as it has a notion of what "base" is.
I think packages = world - base in Gentoo's case.
Donn Miller
> I agree completely. One may have the poor man's method, like in OpenBSD with
> a shell script on the console, or the luxury method like knoppix, that
> is a shell script under a splendid desktop, and all the documentation and
> utilities available. Both are preferable. Or a decent installer like the
> Mandrake one, which incidentally is under GPL so could be reused easily.
> I have read the paper by J. Hubbard, and while i agree with his critics of the
> present system, i am extremely sceptic about the replacement he proposes.
> I fear to see still a more ugly monster than the present sysinstall.
I think FreeBSD needs more newsgroups, and that having one dedicated to
the administration aspect of FreeBSD (which would include installer
design) would be a good idea. If I worked up an installer, I'd think
about a dual-license, one BSD and the other GPL, so it could be used for
both FreeBSD and some Linux distros.
u.r. faust
> I think FreeBSD needs more newsgroups, and that having one dedicated
> to the administration aspect of FreeBSD (which would include installer
> design) would be a good idea.
The newsgroups need to be under the comp.os.* hierarchy.
That is where one expects operating system discussions to be.
I only searched for this group since I was planning to use dummynet.
--
natsu-gusa ya / tsuwamono-domo-ga / yume no ato
summer grasses / strong ones / dreams site
Summer grasses,
All that remains
Of soldier's dreams
(Basho trans. Stryk)
Jorge Godoy
> I think FreeBSD needs more newsgroups, and that having one dedicated
> to the administration aspect of FreeBSD (which would include installer
> design) would be a good idea. If I worked up an installer, I'd think
> about a dual-license, one BSD and the other GPL, so it could be used
> for both FreeBSD and some Linux distros.
Linux distros can use BSDL for their software. You'd have to worry
about one license only...
--
Godoy. <go...@metalab.unc.edu>
Mark Weinem
> Could someone please give me a quick info about the differences between
> Linux and FreeBSD, alternativly some good resources on the net where I can
> learn about the differencies?
the best sites I know about:
1. http://sites.inka.de/mips/unix/bsdlinux.html
2. http://www.cons.org/cracauer/freebsd.html
both are from 2000 and compare FreeBSD with the old Linux 2.2 (running
the now outdated ext2 filesystem) - seems FreeBSD advocacy was easier
those days...
Even FreeBSD users recognize that the basic pkg tools are not very good
(and that's the reason why portupgrade exists).
Note that today, ports-like package management for third party software
is possible with a couple of Linux distributions. And there a plans to
enhance the Debian package management (http://telemetrybox.org/upm/).
Filesystems
Linux supports different Journaling Filesystems (JFS)
Easier with FreeBSD: only the UFS filesystem is available,
and the FreeBSD Kernel stays slim!
The ext3 filesystem for Linux is in productive state and
already avoids annoying filesystem checks after crashes.
Comparable functionality for FreeBSD is unstable but will be
soon available with FreeBSD 5.0
Perfomance und data security should be equal (at least from a
home user's point of view). Linux with a JFS may offer better
performance for some databases and mailservers.
Packet Filtering:
FreeBSD users have the choice between ipfw, ipf and User PPP
(or using them together) with readable filter rules. ipfw and
ipf are both very powerful packet filters.
Linux ships with iptables and some people claim that it has
better performance and needs fewer rules. Maybe that's right for
special purposes but convincing benchmarks are not available.
True is that the iptables syntax is very sick
USB Support:
FreeBSD supports only USB1 and there is not much documentation
about USB configuration available. More decent for USB2 devices are
Linux and NetBSD.
Ways to crash (aside from poor hardware):
A FreeBSD is easy to crash: Writing to floppy with write
protection, unloading Kernel modules while they are in use
No problems with Linux crashes here ;-)
Serious: aside from booting a Knoppix on PC with only 128MB RAM
or running the sucking Mandrake installer, I'm not experienced
in crashing Linux anymore.
Other things I like about FreeBSD
Configuring and installing a custom Kernel is easy: just editing a
few lines in a config file and typing five commands - ready.
The featureful and interactive user-ppp program for dialup.
Ciao, Mark
Raqueeb Hassan
hmm, I don't know, but if you talk about the rock solid stability,
then you may count on freeBSD. Well, Linux is doing really great this
days.
raqueeb hassan
augusta, ga
Zenin
:> I think FreeBSD needs more newsgroups, and that having one dedicated to
:> the administration aspect of FreeBSD (which would include installer
:> design) would be a good idea.
:
: The newsgroups need to be under the comp.os.* hierarchy.
:
: That is where one expects operating system discussions to be. I only
: searched for this group since I was planning to use dummynet.
But in the real world there are Unix systems, and other junk, thus
comp.unix.* vs comp.os.*. :-)
--
Z R /\ _ _ _ _
E H / \ | | |_ | _ | /\ |\ | / |_
N @ . O R G / \ |_ |_ |_ \_/ | / \ | \| \_ |_
I P "The Greatest Game You Never Played"
N S www.AllegianceHQ.org
Zenin
: ta...@lpthe.jussieu.fr wrote:
:> Yes. Throwing everything under /usr like even good distros like Debian do
:> is nonsensical.
:
: Yes, but I think there's a document or standard out, FHS, or something
: like that, which was drafted by members of the Linux community. It says
: that everything should go under /usr and not /usr/local. Hopefully,
: you'll know what I'm talking about.
Most "members of the Linux community" seem to believe in the kitchen
sink /usr mentality, so of course they would draft such into a
"Linux standard".
That doesn't mean it isn't wrong, and stupid.
: The problem is that many Linux distros have a customizable or vague notion
: of what constitutes a "base" system. In FreeBSD and the other BSD's, the
: distinction is well-defined. For many Linux distros, Gentoo being one of
: them, what constitutes a "base" system isn't fixed, but varies by the
: individual. Hence, almost every package counts as being part of "base".
: To that end, I think I've heard a blurb about Solaris becoming much of the
: same way, at least they're doing away with /usr/local.
Solaris has never used /usr/local like a BSD system. It's Sys5
mentality, thus /usr/local on Solaris is for "local" software that
isn't managed by the package management system. -GNU packaged
software loves to break this convention, but that's GNU's fault.
Package managed software under Solaris goes in /opt traditionally.
/opt however, is a mostly failed attempt at solving a tiny problem
(clashing names between packages) with a big hammer. Mostly because
it fails for small software packages, but does generally help for
larger ones. But a BSDish /usr/local can work just as well for
both, especially when linked into the ports system were conflicts
can be resolved by the community before users have to deal with it.
: The major problem with keeping everything under /usr is it makes it hard
: to blow away your external packages and re-install everything from
: scratch, or to back up the base system and ignore the packages in
: /usr/local.
The problems go much, much deeper then that. Linux has done more to
blow good SCM practices to hell then even Windows, which isn't easy.
But hey, only %0.0001 of the Linux world even knows what SCM is, let
alone what it's useful for. Hell, even Linus wasn't start enough to
start using a source control system until very recently, let alone
care about larger aspects of SCM. Like many (most?), he has no real
idea what SCM is all about or what good it does for anyone. The
Linux community, most often doing little more then echoing the Words
of their Prophet, doesn't understand or care either.
Thus with almost no exception, all Linux distros are a complete mess
and don't even know it.
: I've heard all the unices were becoming the same way, because of Linux's
: influence.
Oh goody. "Hey look at all the pathetically stupid ideas coming out
of the Linux camps! Lets adopt them all for the good of our
customers!".
Actually, what's sadly happening (albeit slowly) is that many of the
larger Unix venders are slowly phasing out their own brands in
(mistaken) favor of Linux. As hardware venders first there is solid
logic for them to move to *an* open Unix system, but the choice of
Linux for such a role has nothing to do with technical merits and
everything to do with hype. To be sure, MS is scared of Linux hype,
but not much of Linux tech.
While I'm an avid Unix user, I'll take Win2k/XP over Linux any day
of the week. Thankfully I can afford to use two machines, one for
my Unix work (FreeBSD) and one for everything else (W2k for office,
games). I'd love to converge my needs onto one system; the same
holy grail of the Linux movement. From all indications however,
while that won't happen on FreeBSD or Windows, it doesn't have a
chance in hell of happing on any Linux distro.
My next box will undoubtably be an Apple, replacing all my existing
workstation systems. And it will be the first Apple I've ever
owned. If XServer ever drops enough in price I might even consider
replacing my FreeBSD servers at some point with OS X.
Linux and FreeBSD aren't dead by any stretch, but the real enemy of
both in the long run isn't each other or even Redmond, but rather
the sleeping giant of Cupertino.
Donn Miller
> Actually, what's sadly happening (albeit slowly) is that many of the
> larger Unix venders are slowly phasing out their own brands in
> (mistaken) favor of Linux. As hardware venders first there is solid
> logic for them to move to *an* open Unix system, but the choice of
> Linux for such a role has nothing to do with technical merits and
> everything to do with hype. To be sure, MS is scared of Linux hype,
> but not much of Linux tech.
Exactly. And while it's nice that IBM is pumping billions of $$$ into
Linux, I really could care less about IBM. Digital was a far more
reputable company than IBM, and they're history. I guess they weren't
ruthless enough to survive. And companies sometimes need to be ruthless
in order to survive. I'm sure many other companies other than Microsoft
are such, but MS is the most often used example, and the most prominent,
so of course they get blamed for a lot of what's wrong with commercial
software. Also, don't forget about IBM's attempts to blow off the PC in
the very beginning. So I really don't care what IBM does or says.
Also, there's this thing going on where Linux is eating up all the unices.
But again, Microsoft is all about hype and marking too, and hence are
scared because the Linux camp is doing an awefully good job of marketing
their OS. What it boils down to is MS is afraid of the marketing and
mindshare competition of Linux, not necessarily the tech aspects of it,
as you've pointed out. If you want to find out what the better OS is,
just go to netcraft.com. But Linux advocates are full of excuses, like,
well, Linux had that jiffies bug which made the uptime counter roll over
ever XXX days. Or, well, Linux people had to reboot their systems when
the 2.2.X kernel was released, thus destroying their uptime. Hey, if
the jiffies counter rolls over every XXX days, doesn't this say
something about the quality of the OS? Why didn't they think of that?
> Linux and FreeBSD aren't dead by any stretch, but the real enemy of
> both in the long run isn't each other or even Redmond, but rather
> the sleeping giant of Cupertino.
So, you think Apple is the enemy? I'm not sure what you mean here. But
I'm not sure it is, as they are contributing to the open source
movement. Basically, I think FreeBSD has more order to it, and that
Linux, if it has a downside, is that it's so chaotic. But I'm sure a
lot of people will tell you that they like running and developing on
both FreeBSD and Linux. I'm not sure I know exactly what kind of a
person I am yet.
Steven White
---[snip]---
>
> I think FreeBSD needs more newsgroups, and that having one dedicated to
> the administration aspect of FreeBSD (which would include installer
> design) would be a good idea. If I worked up an installer, I'd think
> about a dual-license, one BSD and the other GPL, so it could be used for
> both FreeBSD and some Linux distros.
>
>
>
> -----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
> http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
> -----== Over 80,000 Newsgroups - 16 Different Servers! =-----
I certainly agree with the additional newsgroups. Reviewing more than
one hundred new posts each day is a little much. Having more groups
(assuming most posters stay on topic) thoughtfully divided amongst
newbies/support, admin, kernel/drivers, and applications might be useful.
Best Regards,
Steve
ta...@lpthe.jussieu.fr
: While I'm an avid Unix user, I'll take Win2k/XP over Linux any day
: of the week. Thankfully I can afford to use two machines, one for
: my Unix work (FreeBSD) and one for everything else (W2k for office,
: games). I'd love to converge my needs onto one system; the same
: holy grail of the Linux movement. From all indications however,
: while that won't happen on FreeBSD or Windows, it doesn't have a
: chance in hell of happing on any Linux distro.
While i agree with you that dropping everything under /usr is a bad practice,
i would not say that it is so dommageable as being a case for prefering
Windows over Linux. It is slightly annoying but not much more. After
all, once you have installed several hundred packages in FreeBSD you also
get a terrible mess under /usr/local. I agree that you can rm -rf /usr/local
and restart from a clean base, while you cannot do that in Linux except
reinstalling everything, but, first i never feel obliged to
rm -rf /usr/local, and second a complete Linux reinstall is still ways faster
than a FreeBSD install with a reasonable number of packages.
And a Linux box works much better than a Windows box (except for gaming, which
is in my opinion a stupid activity) and roughly the same as a FreeBSD box.
: My next box will undoubtably be an Apple, replacing all my existing
: workstation systems. And it will be the first Apple I've ever
: owned. If XServer ever drops enough in price I might even consider
: replacing my FreeBSD servers at some point with OS X.
My own experience with Mac OS X (i have several collegues who own Mac desktops
and laptops) is that the GUI is no more easy to use that any standard Open
Source GUI (KDE,...) no more faster, and that the underlying OS is far from
being as performant as Linux or FreeBSD. It used to be slow to the extreme,
now it is just reasonable. Nothing to be highly praised.
: Linux and FreeBSD aren't dead by any stretch, but the real enemy of
: both in the long run isn't each other or even Redmond, but rather
: the sleeping giant of Cupertino.
I am still waiting for the moment where Apple will displace Microsoft on
everybody's desktop. At least Windows behaves decently since Windows2000.
It will take some time for Apple to get similar performnce and functionality,
and by this time, Linux will have progressed a lot, and FreeBSD also, at least
i hope.
--
Michel TALON
Patrick Scheible
> Exactly. And while it's nice that IBM is pumping billions of $$$ into
> Linux, I really could care less about IBM. Digital was a far more
> reputable company than IBM, and they're history. I guess they weren't
> ruthless enough to survive.
No, DEC got arrogant and stopped caring about its customers. Dropping
the PDP-10 line. Ignoring bug reports. No matter what anyone's
problem is, the answer is a Vax running VMS... no more ability to see
your OS source code... Then when they finally introduced Unix it was
the worst implementation of Unix out there. Grossly overpriced
peripherals. DEC was great once upon a time, but its best days were
20 years ago.
-- Patrick
Bill Vermillion
Donn Miller <dmmi...@cvzoom.net> wrote:
>Zenin wrote:
>
>> Actually, what's sadly happening (albeit slowly) is that many of the
>> larger Unix venders are slowly phasing out their own brands in
>> (mistaken) favor of Linux. As hardware venders first there is solid
>> logic for them to move to *an* open Unix system, but the choice of
>> Linux for such a role has nothing to do with technical merits and
>> everything to do with hype. To be sure, MS is scared of Linux hype,
>> but not much of Linux tech.
>
>Exactly. And while it's nice that IBM is pumping billions of $$$ into
>Linux, I really could care less about IBM. Digital was a far more
>reputable company than IBM, and they're history. I guess they weren't
>ruthless enough to survive.
Some of their moves seemed to make them look like a company who
wanted to retain users by locking them into their product.
The Rainbow even required you buy pre-formatted floppies for awhile.
A couple of others tried that and also sold formatting utilities
too. Thing like that can harm an image and make it harder to
survive. And I will say their floppy drive - which may have been
a technological tour de force was pretty bizarre IMO. They did
some great things along with some stupid things.
>Also, don't forget about IBM's attempts to blow off the PC in
>the very beginning. So I really don't care what IBM does or says.
In which way. The current PC was their 3rd attempt at building a
small computer in that vein and the PC certainly should have had a
few more real engineers look at the design before it was finalized.
Anyone who had done any real work in cooling would have known you
do not blow air OUT of a system as you suck in dirt everywhere -
and across such important things as floppy dirve heads, and you
would suck air into the cabinet and in effect pressurize it.
Really nasty in the earliest ones when so many chips were in
sockets and trying to fill sockets laster where the air was sucked
through the sockets helping to oxidize them. Pretty crued IMO,
Bill
--
Bill Vermillion - bv @ wjv . com
Donn Miller
ta...@lpthe.jussieu.fr wrote:
> I am still waiting for the moment where Apple will displace Microsoft on
> everybody's desktop. At least Windows behaves decently since Windows2000.
> It will take some time for Apple to get similar performnce and functionality,
> and by this time, Linux will have progressed a lot, and FreeBSD also, at least
> i hope.
Indeed, Microsoft will rise to the task. They'll stop at nothing to
regain and/or maintain desktop dominance, all other unethical things
aside. I think your statement is true not only for Apple displacing
Microsoft, but Linux and FreeBSD as well. I think the latter will
manage to break into the desktop, but won't do any significant damage to
Windows dominance on the desktop, as Windows will continue to get more
reliable and easy to use. The only thing that will have any chance at
tripping up Windows is MS' licensing plans.
Donn Miller
Patrick Scheible wrote:
> No, DEC got arrogant and stopped caring about its customers. Dropping
> the PDP-10 line. Ignoring bug reports. No matter what anyone's
> problem is, the answer is a Vax running VMS... no more ability to see
> your OS source code... Then when they finally introduced Unix it was
> the worst implementation of Unix out there. Grossly overpriced
> peripherals. DEC was great once upon a time, but its best days were
> 20 years ago.
By "the worst implementation of Unix out there", do you mean Ultrix?
One person who ran BSD 4.X a long time ago said Ultrix was a
"bastardized version of BSD", and he didn't like it.
ta...@lpthe.jussieu.fr
: Donn Miller <dmmi...@cvzoom.net> writes:
From what i hear Digital True64 Unix was a very good Unix, and the Alpha
machines were very good also. Indeed in my experience DEC people were
extremely arrogant. You had to pray them to buy something and pay full price
when the Sun people were very ready to discounts.
: -- Patrick
--
Michel TALON
Patrick Scheible
>
>
>
> Patrick Scheible wrote:
>
> > No, DEC got arrogant and stopped caring about its customers. Dropping
> > the PDP-10 line. Ignoring bug reports. No matter what anyone's
> > problem is, the answer is a Vax running VMS... no more ability to see
> > your OS source code... Then when they finally introduced Unix it was
> > the worst implementation of Unix out there. Grossly overpriced
> > peripherals. DEC was great once upon a time, but its best days were
> > 20 years ago.
>
> By "the worst implementation of Unix out there", do you mean Ultrix?
> One person who ran BSD 4.X a long time ago said Ultrix was a
> "bastardized version of BSD", and he didn't like it.
Yes.
I could add completely mising the boat on personal computers and
workstations. Granted, it was very hard for any company used to
quality engineering to compete with commodity hardware, but DEC didn't
have any response at all, even in the high end of the workstation
market, for a decade or so.
-- Patrick
Patrick Scheible
True64 may have been a good Unix; I was thinking of the Ultrix days.
In the 80s and early 90s DEC seemed to be going after business
computing and no longer interested in the academic and scientific
users that were their core market earlier. But the business computing
users weren't that easy to woo away from IBM and other
business-oriented vendors... the academics could no longer afford
DEC's expense and arrogance about bug reports/feature requests.
Yeah, the Alpha machines were good hardware. And VMS wasn't all bad,
it just wasn't the answer to everything, especially in DECs
traditional academic and scientific markets. By the time there was a
choice of OSs, Alphas no longer had much of a performance lead over
Intel, and Intel was so much cheaper many users held their noses and
bought Intel.
-- Patrick
Bill Vermillion
<ta...@lpthe.jussieu.fr> wrote:
>David Magda <dmagda...@ee.ryerson.ca> wrote:
>
>: Jorge Godoy <go...@metalab.unc.edu> writes:
>: [...]
>:> I hope FreeBSD makes its installer a little more friendlier. And the
>:> partition utility too. :-) I've installed it several times here and on
>:> customers of my company, but it is still incredibly easier to install
>:> a Linux system.
>: [...]
>
>: OpenBSD's disklabel(8) has a wonderful feature that is activated with
>: the '-E' command-line option -- an interactive editor. Anyone know
>: how complex the code is?
>
>Perfectly true. By the way fdisk is better also. Once on the mailing lists
>i have seen the question discussed of porting OpenBSD fdisk and disklabel
>to FreeBSD. You may imagine the concert of lamentations that ensued ...
>Our fdisk is better, less buggy, and so on. To make a long story short,
>nothing was changed. From time to time Matt Dillon dares touching
_ _
/ \ __ _ _ __ __ _| |__
/ _ \ / _` | '__/ _` | '_ \
/ ___ \ (_| | | | (_| | | | |
/_/ \_\__,_|_| \__, |_| |_|
|___/
Bill Vermillion
<ta...@lpthe.jussieu.fr> wrote:
>Zenin <ze...@rhps.org> wrote:
>: While I'm an avid Unix user, I'll take Win2k/XP over Linux any day
>: of the week. Thankfully I can afford to use two machines, one for
>: my Unix work (FreeBSD) and one for everything else (W2k for office,
>: games). I'd love to converge my needs onto one system; the same
>: holy grail of the Linux movement. From all indications however,
>: while that won't happen on FreeBSD or Windows, it doesn't have a
>: chance in hell of happing on any Linux distro.
>While i agree with you that dropping everything under /usr is a
>bad practice, i would not say that it is so dommageable as being
>a case for prefering Windows over Linux. It is slightly annoying
>but not much more. After all, once you have installed several
>hundred packages in FreeBSD you also get a terrible mess under
>/usr/local.
Most packages/ports are pretty decent when it comes to removing
themselves when you find you don't need them. I'd prefer to have
more things in /usr/local than I needed instead of having something
I needed go away in the base system
For remote OS upgrades the /usr/local concept really makes it
pretty fail safe. I've become quite fond of /usr/local
> and second a complete Linux reinstall is still
>ways faster than a FreeBSD install with a reasonable number of
>packages.
I find a reinstall of the OS runs about 20-30 minutes on FreeBSD.
That of course is with a cvsup and build world and install kernel
done during daylight and the reboot and installworld done
late at night. {The colo is about 20 minutes away at 2AM much
longer during daylight hours. And working standing up between
rows of cabinets with a very high noise level from the four
Lieberts in that area is not a pleasant way to spend time}
Worst case is I go an boot locally with kernel.old about 2AM.
Haven't had to do that yet. I feel a whole lot better with
/usr/local. Maybe it's just a false sense of security but
I like it.
>And a Linux box works much better than a Windows box (except for
>gaming, which is in my opinion a stupid activity) and roughly
>the same as a FreeBSD box.
I saw some reviews on custom made MS gaming boxes and they cost
more than most servers. I think a game box with several polished
coats of autmotive finish is a bit much - but the gamers seem to be
part of the reasons we get faster machines all the time.
>: My next box will undoubtably be an Apple, replacing
>: all my existing workstation systems. And it will be
>: the first Apple I've ever owned. If XServer ever drops
>: enough in price I might even consider replacing my
>: FreeBSD servers at some point with OS X.
>My own experience with Mac OS X (i have several collegues who
>own Mac desktops and laptops) is that the GUI is no more easy to
>use that any standard Open Source GUI (KDE,...) no more faster,
>and that the underlying OS is far from being as performant as
>Linux or FreeBSD. It used to be slow to the extreme, now it is
>just reasonable. Nothing to be highly praised.
Performance isn't that bad on the 10.2. But as a BSD user I find
things lacking at the CLI level.
>: Linux and FreeBSD aren't dead by any stretch, but the
>: real enemy of both in the long run isn't each other
>: or even Redmond, but rather the sleeping giant of
>: Cupertino.
>I am still waiting for the moment where Apple will displace
>Microsoft on everybody's desktop.
That is going to be a long wait. :-)
Bill
ta...@lpthe.jussieu.fr
:>While i agree with you that dropping everything under /usr is a
:>bad practice, i would not say that it is so dommageable as being
:>a case for prefering Windows over Linux. It is slightly annoying
:>but not much more. After all, once you have installed several
:>hundred packages in FreeBSD you also get a terrible mess under
:>/usr/local.
: Most packages/ports are pretty decent when it comes to removing
: themselves when you find you don't need them. I'd prefer to have
: more things in /usr/local than I needed instead of having something
: I needed go away in the base system
This is exactly the argument that the Linux guys use to throw
everything under /usr: the package system easily handles removal
so there is no point ...
:> and second a complete Linux reinstall is still
:>ways faster than a FreeBSD install with a reasonable number of
:>packages.
: I find a reinstall of the OS runs about 20-30 minutes on FreeBSD.
: That of course is with a cvsup and build world and install kernel
: done during daylight and the reboot and installworld done
: late at night. {The colo is about 20 minutes away at 2AM much
: longer during daylight hours. And working standing up between
: rows of cabinets with a very high noise level from the four
: Lieberts in that area is not a pleasant way to spend time}
Well i was speaking of *complete* reinstall, that is base + ports.
Last time i did that on my laptop (because i shrinked the Dos
partition) it took me of the order of one day to get the laptop
to the previous point, with all ports configured as before and so on.
Still i am not sure i have not forgotten some detail which will
bother me one day. And yes i had kept backups of all config files
on the Dos partition. Of course reinstall of the base system is
very quick via cvsup, this is one of the strengths of FreeBSD.
--
Michel TALON
Donn Miller
ta...@lpthe.jussieu.fr wrote:
>
> Well i was speaking of *complete* reinstall, that is base + ports.
> Last time i did that on my laptop (because i shrinked the Dos
> partition) it took me of the order of one day to get the laptop
> to the previous point, with all ports configured as before and so on.
> Still i am not sure i have not forgotten some detail which will
> bother me one day. And yes i had kept backups of all config files
> on the Dos partition. Of course reinstall of the base system is
> very quick via cvsup, this is one of the strengths of FreeBSD.
I usually back up my FreeBSD install as follows to my Windows partition:
tar clpvf - / /var /usr | bzip2 -9v > freebsd.tbz
And this takes care of all my partitions. To re-install, all I do is
insert my FreeBSD install CD, do all my partitions how I want them, and
select a "minimal install". After it's installed, I boot into
single-user mode, mount all partitions, and then tar yxpvf freebsd.tbz
-C /. I don't have to screw around with re-installing anything (except
for the minimal system). The only problem is that I have to
recompile/reinstall certain things, such as tar, bzip2, init, sh, csh,
etc., as tar refuses to overwrite files that are in use.
ta...@lpthe.jussieu.fr
: And this takes care of all my partitions. To re-install, all I do is
: insert my FreeBSD install CD, do all my partitions how I want them, and
: select a "minimal install". After it's installed, I boot into
: single-user mode, mount all partitions, and then tar yxpvf freebsd.tbz
: -C /. I don't have to screw around with re-installing anything (except
: for the minimal system). The only problem is that I have to
: recompile/reinstall certain things, such as tar, bzip2, init, sh, csh,
: etc., as tar refuses to overwrite files that are in use.
Don, this is not what i call a complete reinstall. I wanted to get rid of all
cruft under /usr/local, so i started with an empty /usr/local and added
all packages (in the most recent version) that i had previously.
Then i went on configuring them, which may take time when the package is inn
or something equally abstruse.
--
Michel TALON
Jorge Godoy
> Don, this is not what i call a complete reinstall. I wanted to get
> rid of all cruft under /usr/local, so i started with an empty
> /usr/local and added all packages (in the most recent version) that
> i had previously. Then i went on configuring them, which may take
> time when the package is inn or something equally abstruse.
The same will happen for Linux. You'll also spend time configuring new
software, so I see no advantage on this for any of them. What may
happen is that you don't need to recompile everything, but this is
also true with FreeBSD...
A complete reinstall --- and update to the latest versions instead of
using what's on your CDs --- and reconfiguration takes a long time...
--
Godoy. <go...@metalab.unc.edu>
Bill Vermillion
<ta...@lpthe.jussieu.fr> wrote:
>Bill Vermillion <b...@wjv.comremove> wrote:
>:>While i agree with you that dropping everything under /usr is a
>:>bad practice, i would not say that it is so dommageable as being
>:>a case for prefering Windows over Linux. It is slightly annoying
>:>but not much more. After all, once you have installed several
>:>hundred packages in FreeBSD you also get a terrible mess under
>:>/usr/local.
>: Most packages/ports are pretty decent when it comes to removing
>: themselves when you find you don't need them. I'd prefer to have
>: more things in /usr/local than I needed instead of having something
>: I needed go away in the base system
>This is exactly the argument that the Linux guys use to throw
>everything under /usr: the package system easily handles removal
>so there is no point ...
But that's a shallow argument if you keep /usr/local on a separate
partition and then remake user. That's only thinking of add on
packages. /usr/local makes things so manageable. I suspect that
those who advocate everthing under /usr have not tried the
/usr/local concept.
>:> and second a complete Linux reinstall is still
>:>ways faster than a FreeBSD install with a reasonable number of
>:>packages.
>: I find a reinstall of the OS runs about 20-30 minutes on FreeBSD.
>: That of course is with a cvsup and build world and install kernel
>: done during daylight and the reboot and installworld done
>: late at night. {The colo is about 20 minutes away at 2AM much
>: longer during daylight hours. And working standing up between
>: rows of cabinets with a very high noise level from the four
>: Lieberts in that area is not a pleasant way to spend time}
>Well i was speaking of *complete* reinstall, that is base + ports.
That will take a bit of time. But that's the difference in the way
I think of things. Base is base and ports are ports. I think of
them as separate entities. So I guess it's all in your POV.
>Last time i did that on my laptop (because i shrinked the Dos
>partition) it took me of the order of one day to get the laptop
>to the previous point, with all ports configured as before and so on.
I don't have anthing with a DOS partition on it any more. And when
I used to put DOS partitions on UNIX systems it was about 30MB
for those stupid vendor utilities that you had to run under DOS
such as the older NIC card configs.
DOS looked good when I read the books. Then I installed it and
found it was so very slow that I went back to my other stuff and
then moved to *n*x about 6 months later for my prime work. That
was DOS 2.0. I also worked on a major project on that platform and
really got to dislike it's limits. But I made so much money
working 18+ hour days and the final day 25 hours on the clock and
we made the drop-dead deadling that I moved from audio engineering
into computers full-time. Gawd that was a long time ago.
Jens Schweikhardt
in <b07tts$mho$1...@asmodee.lpthe.jussieu.fr>:
# Bill Vermillion <b...@wjv.comremove> wrote:
...
# : I find a reinstall of the OS runs about 20-30 minutes on FreeBSD.
# : That of course is with a cvsup and build world and install kernel
# : done during daylight and the reboot and installworld done
# : late at night. {The colo is about 20 minutes away at 2AM much
# : longer during daylight hours. And working standing up between
# : rows of cabinets with a very high noise level from the four
# : Lieberts in that area is not a pleasant way to spend time}
#
#
# Well i was speaking of *complete* reinstall, that is base + ports.
# Last time i did that on my laptop (because i shrinked the Dos
# partition) it took me of the order of one day to get the laptop
# to the previous point, with all ports configured as before and so on.
# Still i am not sure i have not forgotten some detail which will
# bother me one day. And yes i had kept backups of all config files
# on the Dos partition. Of course reinstall of the base system is
# very quick via cvsup, this is one of the strengths of FreeBSD.
This sounds like you guys could be interested in FreeBSD From Scratch,
an article soon to appear in the FreeBSD article series. If you want to
review the tried-and-tested but still draft version and give some
feedback I'd be glad: http://www.schweikhardt.net/article.html
Regards,
Jens
--
Jens Schweikhardt http://www.schweikhardt.net/
SIGSIG -- signature too long (core dumped)
Jens Schweikhardt
in <kpsmvs3...@wintermute.casa>:
# ta...@lpthe.jussieu.fr writes:
#
#> Don, this is not what i call a complete reinstall. I wanted to get
#> rid of all cruft under /usr/local, so i started with an empty
#> /usr/local and added all packages (in the most recent version) that
#> i had previously. Then i went on configuring them, which may take
#> time when the package is inn or something equally abstruse.
#
# The same will happen for Linux. You'll also spend time configuring new
# software, so I see no advantage on this for any of them. What may
# happen is that you don't need to recompile everything, but this is
# also true with FreeBSD...
#
# A complete reinstall --- and update to the latest versions instead of
# using what's on your CDs --- and reconfiguration takes a long time...
Help is on the way! FreeBSD From Scratch can do this with two keystrokes.
Please review an article soon to appear in the FreeBSD article collection,
ta...@lpthe.jussieu.fr
: #
: # using what's on your CDs --- and reconfiguration takes a long time...
: Help is on the way! FreeBSD From Scratch can do this with two keystrokes.
: Please review an article soon to appear in the FreeBSD article collection,
: http://www.schweikhardt.net/article.html
I have read the article, nice article. I don't agree with everything
(i have been converted to the everything under / stuff :-) i don't like
mergemaster, etc.) but this is a step in a good direction, provided you
have a big disk.
:--
Michel TALON
jpd
On Thu, 16 Jan 2003 10:34:13 +0000 (UTC),
[snip]
> I am still waiting for the moment where Apple will displace Microsoft on
> everybody's desktop. At least Windows behaves decently since Windows2000.
> It will take some time for Apple to get similar performnce and functionality,
> and by this time, Linux will have progressed a lot, and FreeBSD also, at least
> i hope.
At the very least apple did not invent it's own conventions over perfectly
usable and long-established `real world' conventions just because they could.
And, something I think I admire most, they pulled off the great changeover
from mc68k to powerpc. It helps if you're controlling all the hardware, too,
ofcourse, but still. Of course micros~1 has !nt/alpha, sort-of, but there
does not exist any win9x/alpha because of too much assembly (and STILL it
doesn't run fast for any value of `fast'[1]).
[1] ObOpenGoal or `run'.
--
j p d (at) d s b (dot) t u d e l f t (dot) n l .
Bill Vermillion
Jens Schweikhardt <use...@schweikhardt.net> wrote:
>ta...@lpthe.jussieu.fr wrote
> in <b07tts$mho$1...@asmodee.lpthe.jussieu.fr>:
># Bill Vermillion <b...@wjv.comremove> wrote:
>...
># : I find a reinstall of the OS runs about 20-30 minutes on FreeBSD.
># : That of course is with a cvsup and build world and install kernel
># : done during daylight and the reboot and installworld done
># : late at night. {The colo is about 20 minutes away at 2AM much
># : longer during daylight hours. And working standing up between
># : rows of cabinets with a very high noise level from the four
># : Lieberts in that area is not a pleasant way to spend time}
># Well i was speaking of *complete* reinstall, that is base + ports.
># Last time i did that on my laptop (because i shrinked the Dos
># partition) it took me of the order of one day to get the laptop
># to the previous point, with all ports configured as before and so on.
># Still i am not sure i have not forgotten some detail which will
># bother me one day. And yes i had kept backups of all config files
># on the Dos partition. Of course reinstall of the base system is
># very quick via cvsup, this is one of the strengths of FreeBSD.
>This sounds like you guys could be interested in FreeBSD From Scratch,
>an article soon to appear in the FreeBSD article series. If you want to
>review the tried-and-tested but still draft version and give some
>feedback I'd be glad: http://www.schweikhardt.net/article.html
I looked at it and I have some problems with your your approach
and your reasons for this approach seem to imply you are doing
things in the incorrect order. eg install world workd but kernel
doesn't boot.
If the kernel doesn't boot you should go back to the original
kernel, and perhaps re-cvsup if needed, and fix things there before
you proceed. A broken installworld should not keep a kernel from
booting. Could you give a scenario where that would be true?
I've not seen this. But I've only run FreeBSD continually since
about 1996 and intermittently since 1994.
I never totally wipe partitions unless there is a compelling
reason. Maintaining remote servers does not permit that. Some
are only miles away but a couple are at least a day by car and a
half-day minimum by air. There is no need to wipe a partition. If
you are paranoid you can always make more slices and put a new
filesystem on the slice for the OS, but this is NOT the MS world
and things like that are NOT neccesary.
The last time I felt I had to wipe a partition was in a commercial
OS where there were some corrupt file deep in a hierarchy that
prevented a clean kernel re-link and with that I had a feeling
there were probably other parts that were corrup - so a new
filesystem was installed there. Keeping the 4 or 5 other file
systems on the other drives intact. Just the base OS filesystem
was remade.
You say that wiping a system means recompiling your ports and that
is true and is a good reason NOT to wipe them. I've had some
ports take hours to build what with DL'in files that are over 100MB
in length and the compile/install. For instance the apfsfilter
with th TeX portions consumed 300MB of disk space.
Your comments on installing second OSes on other partitions does
seem logical. It's just that your suppositions for stated
at first don't ring true for me. I've been self-employed
maintaining/adming Unix and quasi-Unix systems since about 1983 so
I think my comments have at least a little legitimacy.
I'll read the article further. I'm not saying your approach is
wrong for your needs - it's just wrong for the way I do things
and may be perfectly OK.
Don't take this as a flame but just a counter-view on what I see is
needed and what you see is needed.
Jens Schweikhardt
in <H90yy...@wjv.com>:
# In article <b0bt49$1r6a$1...@hal9000.schweikhardt.net>,
# Jens Schweikhardt <use...@schweikhardt.net> wrote:
...
#>This sounds like you guys could be interested in FreeBSD From Scratch,
#>an article soon to appear in the FreeBSD article series. If you want to
#>review the tried-and-tested but still draft version and give some
#>feedback I'd be glad: http://www.schweikhardt.net/article.html
#
# I looked at it and I have some problems with your your approach
# and your reasons for this approach seem to imply you are doing
# things in the incorrect order. eg install world workd but kernel
# doesn't boot.
#
# If the kernel doesn't boot you should go back to the original
# kernel, and perhaps re-cvsup if needed, and fix things there before
# you proceed. A broken installworld should not keep a kernel from
# booting.
Maybe I have been sloppy with my wording, booting as I meant it
also included starting init and process the rc files. There are
various scenarios how this can go wrong.
# Could you give a scenario where that would be true?
# I've not seen this. But I've only run FreeBSD continually since
# about 1996 and intermittently since 1994.
A few months ago there was a problem on 5-current with the 'test'
program/builtin used in many shell scripts. IIRC it was due to using a
non-implemented/removed syscall. The installworld got to the point where
it installed 'test'. All further test invocations resulted in SIGSYS.
Then I could not use various "install" targets any longer and the
system was hosed in many other places.
# I never totally wipe partitions unless there is a compelling
# reason. Maintaining remote servers does not permit that. Some
# are only miles away but a couple are at least a day by car and a
# half-day minimum by air. There is no need to wipe a partition. If
# you are paranoid you can always make more slices and put a new
# filesystem on the slice for the OS, but this is NOT the MS world
# and things like that are NOT neccesary.
True, FBSD from Scratch is not handy in some situations. However
wiping partitions has the advantage of removing old cruft that can
cause headaches for mysterious reasons. Witness the UPDATING entry,
20020831:
gcc has been upgraded to 3.2. It is not all binary compatible
with earlier versions of gcc for c++ programs. All c++
programs and libraries need to be recompiled.
Also, if you encounter g++ issues, rm /usr/include/g++/* before
doing an installworld to make sure that stale files are removed.
Many people have been bitten by this and "Why don't my C++ programs
compile anymore?" had almost become a FAQ.
# The last time I felt I had to wipe a partition was in a commercial
# OS where there were some corrupt file deep in a hierarchy that
# prevented a clean kernel re-link and with that I had a feeling
# there were probably other parts that were corrup - so a new
# filesystem was installed there. Keeping the 4 or 5 other file
# systems on the other drives intact. Just the base OS filesystem
# was remade.
Now that you mention it, possibly corrupt files from long ago that
you never noticed were corrupt are another reason to wipe.
# You say that wiping a system means recompiling your ports and that
# is true and is a good reason NOT to wipe them. I've had some
# ports take hours to build what with DL'in files that are over 100MB
# in length and the compile/install. For instance the apfsfilter
# with th TeX portions consumed 300MB of disk space.
Yes, the stage_2.sh in the article takes about 4hours on an AMD1800+,
the pigs being mozilla and X11. I recognize that such a long downtime is
not acceptable in many shops, but my objective was not to provide a
solution for all situations. I will add a paragraph on the "cons" of my
approach.
...
# Don't take this as a flame but just a counter-view on what I see is
# needed and what you see is needed.
Bill, I have read many sensible articles of yours, and this one is
certainly just one more in this line. I welcome your constructive
criticism.
Mike-UK
I've just finished reading this educational thread. I've learned quite a
bit more than I understood about unix systems. Thanks chaps!
Now for a few questions of my own if thats ok?
I could use a bit of input about using FreeBSD on the following...
Two pootas, Thin ethernet LAN, dialup external modem.
OffLine poota= (running W98 as the main OS)
Jetway (Aladin chipset) mainboard
AMD 500Mhz CPU (underdriven at 450Mhz)
64Mb SDRAM (single stick)
hda = 10Gb EIDE (Fujitsu)
hdc = 6Gb EIDE (Fujitsu)
hdb = CDWR unit (LG/EIDE)
hdd = 44x CDROM unit (?/EIDE)
Voodoo3 2000 16Mb graphics card
Kingston de5 NIC (Tulip?)
PS2 scroll-mouse (Trust)
16bit PNP 'Value' Soundblaster
OnLine poota
Intel chipset mainboard (?)
Intel P90 CPU
32Mb EDO RAM (2x2x4 8Mb sticks)
hda = 4Gb EIDE (Maxtor I think?)
hdc = Ancient 4x CDROM drive (EIDE)
4Mb Virge DX graphics card
Cheap pci ne2000 clone NIC (VIA chipset I think?)
Serial 3-button mouse
external USR 56k/V90 modem (?)
Q-1
Will FreeBSD have a problem installing a boot manager to boot a FreeBSD
installation on hdc3? (The Win98 partitions use up most of the drive.)
Q-2
As I've only got about 1Gb free on hdc, will FreeBSD fit? (I've got
VectorLinux on there at the moment taking up about 300Mb and about 128Mb
on hdc4 as swap.)
Q-3
Given that it may be obvious that Gnome will grind to a halt on the P90
poota, and as I like the IceWM that I've been using with VectorLinux, is
it practical for me to consider learning how to install/setup the IceWM
X-GUI on a minimum FreeBSD install on this machine? (I'll need Opera too
come to think of it as Phoenix takes too many tea-breaks on the P90!)
Q-4
I already have FreeBSD v4.5 on a CD I downloaded and played about with
some time ago. Would it be advisable to download the latest 4. series or
is updating the 4.5 install less of a task? (Dialup account!)
Q-5
Should I consider transfering from a Linux OS to FreeBSD "just a few new
things to get used to" or "Start again with the reading up and learning"?
The main reasons I'm considering FreeBSD are things like the ports
collection over the Linux "do this then that then the other, if it doesn't
work do this then....", the things I've read about the focus on system
security, this rather readable and educational NG, and the cute little
daemon. ;-)
One thing I will want to do is use the Midnight Commander. As I understand
things, this will mean using a "Linux compatibility" thingie? Is there a
FreeBSD equivalent (console file manager) available instead?
Your hints'n'tips, advised URLs, homepages and genial ribbings will all be
appreciated...
--
Signature line applied for/in the post...
Steve O'Hara-Smith
Mike-UK <snipthism...@blueyonder.co.uk> wrote:
MU> Q-1
MU> Will FreeBSD have a problem installing a boot manager to boot a
MU> FreeBSD installation on hdc3? (The Win98 partitions use up most of the
MU> drive.) Q-2
Should be no problem.
MU> As I've only got about 1Gb free on hdc, will FreeBSD fit? (I've got
MU> VectorLinux on there at the moment taking up about 300Mb and about
MU> 128Mb on hdc4 as swap.)
Tight but possible - no sources and probably no ports tree
at best a trimmed one and a lot of care with chucking out distfiles.
Less than ideal.
MU> Q-3
MU> Given that it may be obvious that Gnome will grind to a halt on the
MU> P90 poota, and as I like the IceWM that I've been using with
MU> VectorLinux, is it practical for me to consider learning how to
MU> install/setup the IceWM X-GUI on a minimum FreeBSD install on this
MU> machine? (I'll need Opera too come to think of it as Phoenix takes too
MU> many tea-breaks on the P90!) Q-4
IceWM and opera are both available - IceWM native - Opera is
the Linux one.
MU> I already have FreeBSD v4.5 on a CD I downloaded and played about with
MU> some time ago. Would it be advisable to download the latest 4. series
MU> or is updating the 4.5 install less of a task? (Dialup account!)
Given the glacial speed of buildworld on a P90 I would go
for an ftp install of 4.7. I wouldn't bother with the CD you'll just
be downloading things you don't need - perhaps the mini ISO.
MU> Q-5
MU> Should I consider transfering from a Linux OS to FreeBSD "just a few
MU> new things to get used to" or "Start again with the reading up and
MU> learning"?
They are both unix family and therefore pretty similar, there
is less tendency to plaster GUIs around in FreeBSD so there will be
reading and learning.
MU> One thing I will want to do is use the Midnight Commander. As I
MU> understand things, this will mean using a "Linux compatibility"
MU> thingie? Is there a FreeBSD equivalent (console file manager)
MU> available instead?
Midnight Commander - there's a native FreeBSD port. Opera will
drive you to needing the linux compatability stuff though.
--
C:>WIN | Directable Mirrors
The computer obeys and wins. |A Better Way To Focus The Sun
You lose and Bill collects. | licenses available - see:
| http://www.sohara.org/
Mike-UK
> MU> As I've only got about 1Gb free on hdc, will FreeBSD fit? (I've got
> MU> VectorLinux on there at the moment taking up about 300Mb and about
> MU> 128Mb on hdc4 as swap.)
>
> Tight but possible - no sources and probably no ports tree
> at best a trimmed one and a lot of care with chucking out distfiles.
> Less than ideal.
Fingers crossed then! :)
> MU> Q-3
> MU> Given that it may be obvious that Gnome will grind to a halt on the
> MU> P90 poota, and as I like the IceWM that I've been using with
> MU> VectorLinux, is it practical for me to consider learning how to
> MU> install/setup the IceWM X-GUI on a minimum FreeBSD install on this
> MU> machine? (I'll need Opera too come to think of it as Phoenix takes too
> MU> many tea-breaks on the P90!) Q-4
>
> IceWM and opera are both available - IceWM native - Opera is
> the Linux one.
Great! Thats one problem solved already!
> MU> I already have FreeBSD v4.5 on a CD I downloaded and played about with
> MU> some time ago. Would it be advisable to download the latest 4. series
> MU> or is updating the 4.5 install less of a task? (Dialup account!)
>
> Given the glacial speed of buildworld on a P90 I would go
> for an ftp install of 4.7. I wouldn't bother with the CD you'll just
> be downloading things you don't need - perhaps the mini ISO.
I'm bound to mess things up a few times, so I'll take a look at
downloading that mini.ISO and maybe the full 4.7.ISO over the following
week. I'm guessing that 4.7 is a significant upgrade from 4.5?
> MU> Q-5
> MU> Should I consider transfering from a Linux OS to FreeBSD "just a few
> MU> new things to get used to" or "Start again with the reading up and
> MU> learning"?
>
> They are both unix family and therefore pretty similar, there
> is less tendency to plaster GUIs around in FreeBSD so there will be
> reading and learning.
This would have made me shy away from FreeBSD, but I've been reading
though the newbie stuff available and I'm quite impressed with how
relevant it all is to the software, and how it seems to be all available
from a few well documented sources, rather than all over the place and in
zillions of different version. From what I've read so far, I get the
impression treating FreeBSD as a completely different approach to things
may be the best way to start?
> MU> One thing I will want to do is use the Midnight Commander. As I
> MU> understand things, this will mean using a "Linux compatibility"
> MU> thingie? Is there a FreeBSD equivalent (console file manager)
> MU> available instead?
>
> Midnight Commander - there's a native FreeBSD port. Opera will
> drive you to needing the linux compatability stuff though.
It just gets better! All I need now is for someone to tell me there is a
native browser as light on hardware resources as Opera, and I may
just have all I need in one go. Any suggestions?
Thanks for the info, most encouraging.
Mike
Darryl Okahata
> Midnight Commander - there's a native FreeBSD port. Opera will
> drive you to needing the linux compatability stuff though.
Opera exists natively for FreeBSD (don't know how well it works,
though). You don't need the Linux compatibility stuff.
--
Darryl Okahata
dar...@soco.agilent.com
DISCLAIMER: this message is the author's personal opinion and does not
constitute the support, opinion, or policy of Agilent Technologies, or
of the little green men that have been following him all day.
Steve O'Hara-Smith
Mike-UK <snipthism...@blueyonder.co.uk> wrote:
MU> place and in zillions of different version. From what I've read so
MU> far, I get the impression treating FreeBSD as a completely different
MU> approach to things may be the best way to start?
If you've got used to the GUI way then yes - once you get used
to the other way you may find that driving Linux configuration is
easier without the GUI too.
MU> It just gets better! All I need now is for someone to tell me there is
MU> a native browser as light on hardware resources as Opera, and I may
MU> just have all I need in one go. Any suggestions?
I gather I missed the native Opera - links is nice also (links
for text links -g for GUI).
ta...@lpthe.jussieu.fr
: Steve O'Hara-Smith <ste...@eircom.net> writes:
:> Midnight Commander - there's a native FreeBSD port. Opera will
:> drive you to needing the linux compatability stuff though.
: Opera exists natively for FreeBSD (don't know how well it works,
: though). You don't need the Linux compatibility stuff.
It works well, but i have the impression it is slower than the Linux version
i had before. printing is extremely slow and rather buggy. In fact Mozilla
is not much slower and has good printing, working Java, working Flash. I would
go to Mozilla (or a derivative, Phoenix,...) and keeping dillo at hand for
fast browsing.
--
Michel TALON
Dr. Richard E. Hawkins
Patrick Scheible <k...@itchy.serv.net> wrote:
>I could add completely mising the boat on personal computers and
>workstations. Granted, it was very hard for any company used to
>quality engineering to compete with commodity hardware, but DEC didn't
>have any response at all, even in the high end of the workstation
>market, for a decade or so.
Oh, to be able to black out memories of the Rainbow and Microvax :)
Although I believe that it eventually became possible to format your own
Rainbow diskettes, rather than buying them from DEC.
hawk
--
Richard E. Hawkins, Asst. Prof. of Economics /"\ ASCII ribbon campaign
doc...@psu.edu Smeal 178 (814) 375-4700 \ / against HTML mail
These opinions will not be those of X and postings.
Penn State until it pays my retainer. / \
Stud Ent
news:b061r5$h0$2...@asmodee.lpthe.jussieu.fr:
> My own experience with Mac OS X (i have several collegues who own Mac
> desktops and laptops) is that the GUI is no more easy to use that any
> standard Open Source GUI (KDE,...) no more faster, and that the
> underlying OS is far from being as performant as Linux or FreeBSD. It
> used to be slow to the extreme, now it is just reasonable. Nothing to
> be highly praised.
Yup...and for those who really used our Macs for living in the last
10 years OS X is simply unusable: selecting an object in Freehand takes ~
20-30 sec - very pathetic for an expensive boxy. Although, for a UNIX
junkie it's the most beautiful (although noisy) gadget. At this point,
Apple missing both class and mass, but I am watching with interest...
Stud Ent
@cvzoom.net:
> Indeed, Microsoft will rise to the task. They'll stop at nothing to
> regain and/or maintain desktop dominance, all other unethical things
> aside
Anybody who thinks that Apple is more ethical than MSFT needs to check in
detox volunteerly.
Zenin
: Zenin wrote:
>snip<
:> Linux and FreeBSD aren't dead by any stretch, but the real enemy of
:> both in the long run isn't each other or even Redmond, but rather
:> the sleeping giant of Cupertino.
: So, you think Apple is the enemy? I'm not sure what you mean here. But
: I'm not sure it is, as they are contributing to the open source movement.
They aren't the "enemy" as such, but they are competition. More
importantly at the rate and direction they are going they could very
likely make both Linux and FreeBSD irrelevant for a very large
number of uses, especially the Linux holy grail of the desktop
market. The only thing that's keeping OSX from possibly making both
completely irrelevant is the lack of x86 support.
: Basically, I think FreeBSD has more order to it, and that Linux, if it has
: a downside, is that it's so chaotic. But I'm sure a lot of people will
: tell you that they like running and developing on both FreeBSD and Linux.
: I'm not sure I know exactly what kind of a person I am yet.
I'm personally too old at my 29 years to waste my time with the
pointless chaos of Linux.
--
Z R /\ _ _ _ _
E H / \ | | |_ | _ | /\ |\ | / |_
N @ . O R G / \ |_ |_ |_ \_/ | / \ | \| \_ |_
I P "The Greatest Game You Never Played"
N S www.AllegianceHQ.org
Chuck Swiger
[ ... ]
> Anybody who thinks that Apple is more ethical than MSFT needs to check in
> detox volunteerly.
Has this helped you out any?
-Chuck
Chuck Swiger | ch...@codefab.com | All your packets are belong to us.
-------------+-------------------+-----------------------------------
"The human race's favorite method for being in control of the facts
is to ignore them." -Celia Green
Stud Ent
> They aren't the "enemy" as such, but they are competition. More
> importantly at the rate and direction they are going they could
> very likely make both Linux and FreeBSD irrelevant for a very
> large number of uses, especially the Linux holy grail of the
> desktop market. The only thing that's keeping OSX from possibly
> making both completely irrelevant is the lack of x86 support.
The rate of installed base is kinda assimptotic to zer0. The Cool Factor is
undeniable though.
ta...@lpthe.jussieu.fr
: Zenin <ze...@rhps.org> wrote in news:v3b038k...@news.supernews.com:
What do you mean by cool?
- the processor is cooler than a x86? True, doesn't produce as much heat.
- the operator is less stressed than under FreeBSD? True since he has all the
time to drink his coffee cup between the click to open a window and the
window ready to type in.
- the desktop is nicer than KDE3 with mosfet-liquid? Untrue, they are
basically the same.
- easier to config with netinfo? Untrue, this is a pile of crap.
--
Michel TALON
Chuck Swiger
> - easier to config with netinfo? Untrue, this is a pile of crap.
NetInfo worked better in 1990 under NEXTSTEP 2.x then it does now.
(But then, OS X 10.2 also broke NIS authentication.)
Ean Kingston
No, the cool factor is getting a commercially supported implementation of
FreeBSD (sort of) so you can thumb your nose at M$ then turn around and run
explorer on it.
--
Please remove the NOSPAM from my e-mail addres when replying directly.
I'm a good UNIX system administrator looking for work. If you happen to need
one, e-mail me.
Ted Spradley
Zenin <ze...@rhps.org> wrote:
> I'm personally too old at my 29 years to waste my time with the
> pointless chaos of Linux.
Oooh!! I *love* that!
--
Remember, more computing power was thrown away last week than existed in
the world in 1982. -- http://www.tom.womack.net/computing/prices.html
Donn Miller
Stud Ent wrote:
> Anybody who thinks that Apple is more ethical than MSFT needs to check in
> detox volunteerly.
Well, I'm running Windows .NET Server 2003 RC2, and it's been extremely
stable compared to Windows ME. But then again, what isn't more stable
than ME?
-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----== Over 80,000 Newsgroups - 16 Different Servers! =-----
Thomas Dickey
> On Mon, 27 Jan 2003 18:54:32 -0000
> Zenin <ze...@rhps.org> wrote:
>> I'm personally too old at my 29 years to waste my time with the
>> pointless chaos of Linux.
> Oooh!! I *love* that!
well, perhaps once he's had some experience, he'll have a different attitude.
--
Thomas E. Dickey <dic...@radix.net> <dic...@herndon4.his.com>
http://dickey.his.com
ftp://dickey.his.com
Dick Hoogendijk
> Ted Spradley <tsp...@spradley.org> wrote:
>>> I'm personally too old at my 29 years to waste my time with the
>>> pointless chaos of Linux.
>> Oooh!! I *love* that!
> well, perhaps once he's had some experience, he'll have a different
> attitude.
FreeBSD people were (are?) know for not being involved in these kind of
"flamy" discussions. I hope this'll continue.
--
dick -- http://www.nagual.st/ -- PGP/GnuPG key: F86289CE
++ Running FreeBSD 4.7 ++ Debian GNU/Linux (Woody)
Thomas Dickey
> begin quoting words of Thomas Dickey:
>> Ted Spradley <tsp...@spradley.org> wrote:
>>>> I'm personally too old at my 29 years to waste my time with the
>>>> pointless chaos of Linux.
>>> Oooh!! I *love* that!
>> well, perhaps once he's had some experience, he'll have a different
>> attitude.
>
> Can we please stop this annoying pro/anti linux threads.
> FreeBSD people were (are?) know for not being involved in these kind of
> "flamy" discussions. I hope this'll continue.
I'm not so much interested in the pro/anti thread, as the stupid comment
which was made.
Jay Dresser
>> drive you to needing the linux compatability stuff though.
>
> It just gets better! All I need now is for someone to tell me there is a
> native browser as light on hardware resources as Opera, and I may
> just have all I need in one go. Any suggestions?
Well I love Opera and can't imagine using anything else -- I'm hooked
on the features. Linux compatibility was not an issue for me, I
didn't install anything explicitly, it was just there. There are a
lot of things from Linux that run fine here.
I only use Konqueror (a kitchen-sink program that is a browser, ftp,
file manager, ...) as a file manager, but according to eWeek, it is
very lean and quick:
http://www.eweek.com/article2/0,3959,825902,00.asp
At last week's Macworld Expo, however, Apple Computer
Inc. launched a Web browsing salvo of its own, called Safari,
along with a reminder that there's still plenty of room for
innovation in this space.
eWeek Labs tested a beta version of Safari, which runs only on
Mac OS X 10.2 or later. Safari is based on the open-source
rendering engine KHTML. This is the engine that powers Konqueror,
the native Web browser of KDE (K Desktop Environment).
The choice of KHTML will likely end up benefiting Apple and KDE
alike: Apple has turned over to the KDE Project the bug fixes and
performance improvements it made to KHTML, and KDE developers
have expressed interest in moving forward with a common HTML
rendering back end for both Konqueror and Safari.
KHTML was selected as Safari's foundation because of its leanness
and speed, according to Apple, and these were the two attributes
that impressed us most in tests of the Apple browser. We've
always been disappointed with the speed of IE on the Mac, and
Safari delivers performance that's at least equivalent to that of
IE on Windows. What's more, Safari allowed us to block pop-up ads
in the same way that Mozilla, Konqueror and Opera can -- and IE
cannot.
...
--
Jay Dresser / use...@dresserfamily.org
Ted Spradley
Thomas Dickey <dic...@saltmine.radix.net> wrote:
> Dick Hoogendijk <di...@nagual.st> wrote:
> > begin quoting words of Thomas Dickey:
> >> Ted Spradley <tsp...@spradley.org> wrote:
> >>>> I'm personally too old at my 29 years to waste my time with the
> >>>> pointless chaos of Linux.
> >>> Oooh!! I *love* that!
> >> well, perhaps once he's had some experience, he'll have a different
> >> attitude.
> >
> > Can we please stop this annoying pro/anti linux threads.
> > FreeBSD people were (are?) know for not being involved in these kind
> > of"flamy" discussions. I hope this'll continue.
>
> I'm not so much interested in the pro/anti thread, as the stupid
> comment which was made.
Which "stupid comment which was made"? Zenin's, mine, or Dick
Hoogendijk's? And how is it interestig to you? Do try to express
yourself clearly.
I was tickled by Zenin's stupid comment about being too old. I have a
daughter who's 29. I *am* too old.
Jay Dresser
For me it just crashed, and it had no flash. I immediately went back
to 6.02 for Linux.
I was so *thrilled* when I saw Opera support FreeBSD. That was what I
was waiting for before I plunked down my money to support a great
browser. Sad it didn't work. If they ever get the native version to
work, they will get my money.
Thomas Dickey
> On 31 Jan 2003 16:16:35 GMT
> Thomas Dickey <dic...@saltmine.radix.net> wrote:
>> Dick Hoogendijk <di...@nagual.st> wrote:
>> > begin quoting words of Thomas Dickey:
>> >> Ted Spradley <tsp...@spradley.org> wrote:
>> >>>> I'm personally too old at my 29 years to waste my time with the
>> >>>> pointless chaos of Linux.
>> >>> Oooh!! I *love* that!
>> >> well, perhaps once he's had some experience, he'll have a different
>> >> attitude.
>> >
>> > Can we please stop this annoying pro/anti linux threads.
>> > FreeBSD people were (are?) know for not being involved in these kind
>> > of"flamy" discussions. I hope this'll continue.
>>
>> I'm not so much interested in the pro/anti thread, as the stupid
>> comment which was made.
> Which "stupid comment which was made"? Zenin's, mine, or Dick
> Hoogendijk's? And how is it interestig to you? Do try to express
> yourself clearly.
The first, of course.
> I was tickled by Zenin's stupid comment about being too old. I have a
> daughter who's 29. I *am* too old.
hmm - being too old implies that one has given up...
Ted Spradley
Thomas Dickey <dic...@saltmine.radix.net> wrote:
> hmm - being too old implies that one has given up...
Exactly! Given up on "the pointless chaos of Linux." That's what I
understood Zenin to be saying, that his experience had led him to give
up on it.
Of course, that's his opinion. But doesn't the "Linux Phenomenon" seem
a bit chaotic to you, compared to FreeBSD and FreeBSD people? Of
course, that's because there's so many more people involved, and the
average age is a bit younger. The more people, the more differing
opinions, confusion, and misinformation you have to expect.
Come on, can you honestly say that all those enthusiastic advocates
touting all those different "distributions" of basically the same stuff
doesn't feel just a little chaotic?
John Dyson
>
> On 31 Jan 2003 18:38:51 GMT
> Thomas Dickey <dic...@saltmine.radix.net> wrote:
>
> > hmm - being too old implies that one has given up...
>
> Exactly! Given up on "the pointless chaos of Linux." That's what I
> understood Zenin to be saying, that his experience had led him to give
> up on it.
>
> Of course, that's his opinion. But doesn't the "Linux Phenomenon" seem
> a bit chaotic to you, compared to FreeBSD and FreeBSD people? Of
> course, that's because there's so many more people involved, and the
> average age is a bit younger. The more people, the more differing
> opinions, confusion, and misinformation you have to expect.
>
> Come on, can you honestly say that all those enthusiastic advocates
> touting all those different "distributions" of basically the same stuff
> doesn't feel just a little chaotic?
>
30yrs or more really don't want to have to deal with the problems created
by the 'youngsters' who are reinventing the wheel over and over again. This
doesn't mean that there isn't some innovation coming from Linux, but it
is tiresome to have to deal with the 'flavor of the day.' This overhead
wouldn't be so bad, except MOST (not all) of the time, the 'new flavor' is
either slightly worse, slightly better, or about the same in 'quality',
but with differing tradeoffs. Seldom are the 'new flavors' incredibly
better.
Gratuitious incompatibility only wastes time for an end user. When
a developer spends a very slight effort to avoid gratuitious changes
that can mitigate a very large aggregate cost to the userbase.
Most people just DONT have time to waste. Sometimes we enjoy learning
something new, but the 'learning' process is often just necessary and
not the goal.
John
jpd
>
>
> Stud Ent wrote:
>
>
>> Anybody who thinks that Apple is more ethical than MSFT needs to check in
>> detox volunteerly.
>
> Well, I'm running Windows .NET Server 2003 RC2, and it's been extremely
> stable compared to Windows ME. But then again, what isn't more stable
> than ME?
<fx action="pushes donn so he falls over in the conveniently placed pool"/>
You aren't _that_ stable, apparently.
--
j p d (at) d s b (dot) t u d e l f t (dot) n l .
Thomas Dickey
> On 31 Jan 2003 18:38:51 GMT
> Thomas Dickey <dic...@saltmine.radix.net> wrote:
>> hmm - being too old implies that one has given up...
> Exactly! Given up on "the pointless chaos of Linux." That's what I
> understood Zenin to be saying, that his experience had led him to give
> up on it.
it's probably only going to get worse (a side-effect of improved technology)
> Of course, that's his opinion. But doesn't the "Linux Phenomenon" seem
> a bit chaotic to you, compared to FreeBSD and FreeBSD people? Of
> course, that's because there's so many more people involved, and the
> average age is a bit younger. The more people, the more differing
> opinions, confusion, and misinformation you have to expect.
not really - people tend to cluster around rather limited sets of ideas.
> Come on, can you honestly say that all those enthusiastic advocates
> touting all those different "distributions" of basically the same stuff
> doesn't feel just a little chaotic?
not here either - I tend to ignore the ones that don't put together a
product to the point where they're demonstrating some competance.
Zenin
>snip<
: I was tickled by Zenin's stupid comment about being too old. I have a
: daughter who's 29. I *am* too old.
My point was not being too old in general, but rather too old to
have time to pointlessly waste constantly dealing with yet another
Linuxism or enhancement handled badly. At just shy of 30 I've got
over 20 years of computing experience and nearly 15 of that solidly
with Unix systems. While I admit to not having nearly the
perspective of many, I likely have a better perspective then most.
To wit...
The more FreeBSD grows, the more it helps ME do MY work, giving ME
more time to spend on something else (like life).
The more Linux grows, the more time I waste dealing with ITS growing
pains and the LESS time I have left for MY work and thus the LESS
time I have to spend on MY life.
See the difference?
FreeBSD allows me to tinker, Linux forces me to.
Donn Miller
> FreeBSD allows me to tinker, Linux forces me to.
And what about Open and NetBSD? I've been wanting to try OpenBSD,
although it won't be as a server, but mainly a desktop.
Zenin
: Zenin wrote:
:> FreeBSD allows me to tinker, Linux forces me to.
:
: And what about Open and NetBSD? I've been wanting to try OpenBSD,
: although it won't be as a server, but mainly a desktop.
NetBSD has done great work on portability aspects, but lacks the
overall "polish" of FreeBSD. OpenBSD improves on the NetBSD code
base and *BSD in general with aggressive security work, but still
lacks the polish.
If you're on x86, 9 out of 10 times FreeBSD will be your best
*BSD choice. If you want BSD on a non-x86 system, NetBSD is a good
choice. If you need the added peace of mind wrt security, OpenBSD
can be a better choice...but not always. For instance, for my home
firewall I initially choose OpenBSD for obvious reasons, but since
99% of my time is on FreeBSD systems it was more then a chore to
keep myself upto date on OpenBSD and keep the system updated. For
me, because I have to maintain a strong background in FreeBSD
anyway, it was easier to keep a FreeBSD firewall upto date and thus
more secure then an OpenBSD system.
All things being equal OpenBSD is more secure; but all things are
never equal.
HTH
Donn Miller
> NetBSD has done great work on portability aspects, but lacks the
> overall "polish" of FreeBSD. OpenBSD improves on the NetBSD code
> base and *BSD in general with aggressive security work, but still
> lacks the polish.
One thing I noticed is that OpenBSD supports the Maestro-3 sound card,
and surprisingly, NetBSD doesn't. So maybe OpenBSD adds some other
improvements over NetBSD.
Thomas Dickey
> Ted Spradley <tsp...@spradley.org> wrote:
> >snip<
> : I was tickled by Zenin's stupid comment about being too old. I have a
> : daughter who's 29. I *am* too old.
> My point was not being too old in general, but rather too old to
> have time to pointlessly waste constantly dealing with yet another
> Linuxism or enhancement handled badly. At just shy of 30 I've got
> over 20 years of computing experience and nearly 15 of that solidly
people who cite figures like that seldom (never in my experience) know
anything worthwhile from it.
Zenin
: Zenin <ze...@rhps.org> wrote:
:> Ted Spradley <tsp...@spradley.org> wrote:
:> >snip<
:> : I was tickled by Zenin's stupid comment about being too old. I have a
:> : daughter who's 29. I *am* too old.
:
:> My point was not being too old in general, but rather too old to
:> have time to pointlessly waste constantly dealing with yet another
:> Linuxism or enhancement handled badly. At just shy of 30 I've got
:> over 20 years of computing experience and nearly 15 of that solidly
:
: people who cite figures like that seldom (never in my experience) know
: anything worthwhile from it.
And people who stop reading at the first opportunity to make a snide
and self-righteous remark seldom grasp the point being presented.
Had you actually read the thread, or indeed even simply the rest of
my message, you'd likely see this wasn't the typical case of
arguing, "I've got XYZ more experience and therefor I'm right", that
you would appear to be alluding to. But that would have implied
that you were looking to add constructive prose to the discussion,
which obviously is anything but the case given this and your earlier
comments in this thread.
Have a nice life.
*plonk*
--
-Zenin (ze...@rhps.org) From The Blue Camel we learn:
BSD: A psychoactive drug, popular in the 80s, probably developed at UC
Berkeley or thereabouts. Similar in many ways to the prescription-only
medication called "System V", but infinitely more useful. (Or, at least,
more fun.) The full chemical name is "Berkeley Standard Distribution".
Thomas Dickey
> Thomas Dickey <dic...@saltmine.radix.net> wrote:
> : Zenin <ze...@rhps.org> wrote:
> :> Ted Spradley <tsp...@spradley.org> wrote:
> :> >snip<
> :> : I was tickled by Zenin's stupid comment about being too old. I have a
> :> : daughter who's 29. I *am* too old.
> :
> :> My point was not being too old in general, but rather too old to
> :> have time to pointlessly waste constantly dealing with yet another
> :> Linuxism or enhancement handled badly. At just shy of 30 I've got
> :> over 20 years of computing experience and nearly 15 of that solidly
> :
> : people who cite figures like that seldom (never in my experience) know
> : anything worthwhile from it.
> And people who stop reading at the first opportunity to make a snide
> and self-righteous remark seldom grasp the point being presented.
tsk, tsk.
> Had you actually read the thread, or indeed even simply the rest of
> my message, you'd likely see this wasn't the typical case of
no, I got the point.
(you're turning your brain off, and bragging about it).
Bill Vermillion
John Dyson <dy...@iquest.net> wrote:
>Many of us who have worked with Unix and/or computing in general
>for appox 30yrs or more really don't want to have to deal with
>the problems created by the 'youngsters' who are reinventing the
>wheel over and over again.
It wouldn't be so bad if they kept re-inventing the wheel, but the
can't seem to make it round. A truly bumpy ride at times.
Bill
--
Bill Vermillion - bv @ wjv . com
John Dyson
with differing flavors.
FreeBSD (NetBSD, OpenBSD, etc) all have a lineage from the older, traditional
source code bases. The biggest problem with those OSen were the
decisions and implementations that would choose between a revolutionary
or evolutionary approach to change. In these cases, the choice exists.
For Linux, the choice is to re-implement, which is almost always an
ego-trip for newbie developers. Linux is essentially the same as *BSD
in external function, and has little true innovation relative to the
BSDs (and the BSDs have minor aspects that distinguish themselves.) In
some ways, one can overstate the differences, but in the scheme of
OS architecture, the differences are relatively small.
Unfortunately, alot of bright, intelligent kids are being seduced by
'yet another re-implementation' of the same old Unix story. There is
a SERIOUS need for our best and brightest to work towards networked
OSes (distributed OS designs), rather than hacking these OSes onto
a distributed OS architecture. (This is just one complaint that
constitutes my general lament.)
Much like the fact that TV (television) is good for sucking up time
that could be spent on more creative and skill building exercises, these
'yet another Unix clones' that express the 'I am as smart as McKusick,
Joy, or even Von Neumann' mentality simply waste the brainpower on the
same old tricks, with a futile quest showing a mee-too ethic, spending
time on the reinvention, over and over again.
For fun and interesting work, we can look at L4, Mach (even though it is
pathetically bloated, practically impractical :-)), Amoeba, Chorus, etc.
Maybe there is alot of work needed to create a 'real' advanced distributed
OS that can be used as a competent uniprocessor or SMP OS, but THE FUTURE
(not the past achievements of the truly innovative people) is where the
best, brightest brainpower should be spent.
I wouldn't be 'proud' to claim that I invented a 'newer, better cp command'
or even a 'newer, better c compiler', but would rather see more work spent
on things that haven't been done before. Even if a 'new' concept is
started, it won't bear fruit for 10yrs -- but as long as the brainpower
is sponged up with the almost passive entertainment mode of software
development, then the investment won't even take root - let alone ever
bear fruit.
John
Bill Vermillion
>> In article <3E3AE86F...@iquest.net>,
>> John Dyson <dy...@iquest.net> wrote:
>> It wouldn't be so bad if they kept re-inventing the wheel, but the
>> can't seem to make it round. A truly bumpy ride at times.
>I agree - the same mistakes (or slightly different ones) get re-created
>with differing flavors.
...
>Much like the fact that TV (television) is good for sucking up
>time that could be spent on more creative and skill building
>exercises, these 'yet another Unix clones' that express the 'I
>am as smart as McKusick, Joy, or even Von Neumann' mentality
>simply waste the brainpower on the same old tricks, with a
>futile quest showing a mee-too ethic, spending time on the
>reinvention, over and over again.
I've never had a formal computer education [it was in other fields
including engineering, broadcast and histroy]but so many of the books
I have were textbooks in Univeristy classes. So it was
self-education in that area - and not just sitting down and trying
to see what worked.
I also found the history of computing fascinating [partially
because of the history above] and have read many of the classic
in that area. The trouble with so many books like that which are
available now is that they are more recently written and not by
the people who were there then. It the old 'those who don't learn
from history ....'
I have Goldstein "Computers - from Pascal to Von Neuman" written in
1977 which give a more accurate view of the past. I found "The
Moore School Lectures" fascinating. That was a printing of the
transcripts of what was really the first class on computing given
back in 1946.
Von Neumann was on of the instructors and one of the 28 invited
students was Shannon who became famous for his informaiton theory.
The thrust of those lectures was to introduce the students to
electronic computing with tube devices showing their superiority
over the mechanical devices. When you look at schematics with
tubes cascaded to form storage registers you can appreciate what
we have today. Then you look at some of the stuff on quantum
computing of which I'm starting to get a vauge idea of how it
works and it's just amazing.
But all of this is progress - not just staying in one general area
and trying to find a different - and sometimes better way of doing
things.
One of the places where many parts of industry in the US has
fallen down was their reluctance to part with the past and insist
on trying to push the limits intead of exploring new ways.
In a field related to computing - the magnetic media - the US
companies seemed bound and determine to live with oxide types and
kept forumlating and doping things to get more improvements.
Others said forget oxide and go with metal particle and plated.
And they had problems. Sony's first particle tape had the
particles oxidize [rust] and was a failure, but they kept at it,
until eventually they got it right and almost all of the old
manufacturers in the US stopped making media because technology had
passed them by.
It was like 100 years ago when people felt the automobile would not
replace the horse, until the day they found themselve with a
warehouse full of buggy whips and no market.
A re-invention does is waste resources both momentarily and
temporally, while truly revoltionary idea move the world ahead, but
so often are not seen as worthwhile during their first shaky days
of discovery.
For me one of the more classic utterance in this area was back
in 1947 when John Shockey and his partner [whose name I forget]
invented a device that would act like a conductor or a resistor.
The named it after it's function and called it a tranfer-resistor.
The name was later shortened to transitor.
Shockely said [and this is an approximate quote] "It's interesting
but I can see no practical use for it". Of course now we see it
change the course of the planet. But what would have happened if
he had just been content on making a better tube?
One other problem of 'reinvention' is convincing others that your
way of doing something which has been done in other ways, is being
evolutionary enough to cause other to want to change to that
method. If you don't succeed we now have one more choice and
having a wealth of choices can lead to confusion as to which choice
to make. And to make an informed choice you have to learn more in
depth about each - and again that's a waste of time.
I tend to ramble a bit at times - but you get the idea I hope.