[9fans] Maybe it is april fool's after all ...

[9fans] Maybe it is april fool's after all ...

[Ronald G Minnich]

when is a .so not a .so? when it's a ld script of course!

yup, /lib/libc.so on my fc4 box is a ..... script.

which is why l4 would not build ... /bin/ld from olden times does
not understand the GROUP command below.


[rminnich@q ~]$ file /usr/lib/libc.so
/usr/lib/libc.so: ASCII C program text
[rminnich@q ~]$ cat !$
cat /usr/lib/libc.so
/* GNU ld script
    Use the shared library, but some functions are only in
    the static library, so try that secondarily.  */
OUTPUT_FORMAT(elf32-i386)
GROUP ( /lib/libc.so.6 /usr/lib/libc_nonshared.a  AS_NEEDED ( 
/lib/ld-linux.so.2 ) )
[rminnich@q ~]$



ron

Re: [9fans] Maybe it is april fool's after all ...

[Charles Forsyth]

>>when is a .so not a .so? when it's a ld script of course!
>>yup, /lib/libc.so on my fc4 box is a ..... script.

curiously, i was just installing fc4 on a machine today.
following an i/o error on a few blocks of a hard drive containing an ext3,
that had emptied /usr/etc to the confusion of all, subsequent use of
redhat-something's fsck destroyed the whole file system rather instantly,
so i thought it might be time for an upgrade.
fc4's installation said there was an i/o error on installation,
that was ``a serious problem'', it could not continue, OK?
as usual my ``that's not OK, actually'' met with a blank face,
and OK rebooted the system.  i think it's all a hint to replace the drive.
clearly, from ron's message, i'm in for a good time!
we'll know computing has cracked it when that sort of thing
is a question on ``who wants to be a millionaire?''.
all helps to keep those synapses active.

Re: [9fans] Maybe it is april fool's after all ...

[Bruce Ellis]

i like it when chip companies ask "why would you want to do that"
and the obvious answer is "because the data sheet says i can
and i want to".  i've been thru this with a few vendors.

brucee

On 1/20/06, Charles Forsyth <forsyth@terzarima.net> wrote:
> >>when is a .so not a .so? when it's a ld script of course!
> >>yup, /lib/libc.so on my fc4 box is a ..... script.
>
> curiously, i was just installing fc4 on a machine today.
> following an i/o error on a few blocks of a hard drive containing an ext3,
> that had emptied /usr/etc to the confusion of all, subsequent use of
> redhat-something's fsck destroyed the whole file system rather instantly,
> so i thought it might be time for an upgrade.
> fc4's installation said there was an i/o error on installation,
> that was ``a serious problem'', it could not continue, OK?
> as usual my ``that's not OK, actually'' met with a blank face,
> and OK rebooted the system.  i think it's all a hint to replace the drive.
> clearly, from ron's message, i'm in for a good time!
> we'll know computing has cracked it when that sort of thing
> is a question on ``who wants to be a millionaire?''.
> all helps to keep those synapses active.
>
>

Re: [9fans] Maybe it is april fool's after all ...

[Devon H. O'Dell]

The specific reason many Linuxes do this is for i386 compat on AMD64
architectures. It's horribly broken, but is only slightly worse than
the way FreeBSD implements it, which is to have /usr/lib32

--Devon

Re: [9fans] Maybe it is april fool's after all ...

[jmk]

Compatibility mode? Why? Surely you just recompile the programme for
the target machine, you have the source, after all. A minute, tops.

--jim

On Fri Jan 20 06:59:15 EST 2006, devon.odell@gmail.com wrote:
> The specific reason many Linuxes do this is for i386 compat on AMD64
> architectures. It's horribly broken, but is only slightly worse than
> the way FreeBSD implements it, which is to have /usr/lib32
> 
> --Devon

Re: [9fans] Maybe it is april fool's after all ...

[andrey mirtchovski]

On 1/20/06, jmk@plan9.bell-labs.com <jmk@plan9.bell-labs.com> wrote:
> Compatibility mode? Why? Surely you just recompile the programme for
> the target machine, you have the source, after all. A minute, tops.
>
> --jim

for a while back then the globus toolkit didn't work in 64-bit mode on
the opterons. one needed to compile it 32-bit which meant installing a
second set of everything, including the X11 libraries :)

on every node.

Re: [9fans] Maybe it is april fool's after all ...

[Paweł Lasek]

On 1/20/06, jmk@plan9.bell-labs.com <jmk@plan9.bell-labs.com> wrote:
> Compatibility mode? Why? Surely you just recompile the programme for
> the target machine, you have the source, after all. A minute, tops.
>
> --jim

It comes from the fact that some of the code is not-so-portable -
OpenOffice 1x being popular example (don't know about 2.0), as it
didn't compile on amd64 (some bugs in the code, remember, it comes
back from StarDivision and OS/2 :D). There's a lot more of it,
thaknfully limited mostly to closed-source software.

M$ is having the same trouble, having to supply 386 IE for basically
every architecture NT is working on (From what I remember, 386 IE is
included in Alpha, IA-64 and AMD64 ones) :)



--
Paweł Lasek
"Once a hitokiri, always a hitokiri. This will never change" - Jine-Ei

Re: [9fans] Maybe it is april fool's after all ...

[jmk]

Excuses, excuses. I thought the mantra was "Given enough
eyeballs all bugs are shallow"? Maybe they should augment the
rooms full of penguins typing code with rooms full of spiders
giving it the eight-eyeball lookover. Also, what is behind
the eyeballs, the skills can be shallower than the bugs.

--jim

On Fri Jan 20 17:20:26 EST 2006, pawel.lasek@gmail.com wrote:
> On 1/20/06, jmk@plan9.bell-labs.com <jmk@plan9.bell-labs.com> wrote:
> > Compatibility mode? Why? Surely you just recompile the programme for
> > the target machine, you have the source, after all. A minute, tops.
> >
> > --jim
> 
> It comes from the fact that some of the code is not-so-portable -
> OpenOffice 1x being popular example (don't know about 2.0), as it
> didn't compile on amd64 (some bugs in the code, remember, it comes
> back from StarDivision and OS/2 :D). There's a lot more of it,
> thaknfully limited mostly to closed-source software.
> 
> M$ is having the same trouble, having to supply 386 IE for basically
> every architecture NT is working on (From what I remember, 386 IE is
> included in Alpha, IA-64 and AMD64 ones) :)
> 
> 
> 
> --
> Paweł Lasek
> "Once a hitokiri, always a hitokiri. This will never change" - Jine-Ei
On Fri Jan 20 17:20:26 EST 2006, pawel.lasek@gmail.com wrote:
> On 1/20/06, jmk@plan9.bell-labs.com <jmk@plan9.bell-labs.com> wrote:
> > Compatibility mode? Why? Surely you just recompile the programme for
> > the target machine, you have the source, after all. A minute, tops.
> >
> > --jim
> 
> It comes from the fact that some of the code is not-so-portable -
> OpenOffice 1x being popular example (don't know about 2.0), as it
> didn't compile on amd64 (some bugs in the code, remember, it comes
> back from StarDivision and OS/2 :D). There's a lot more of it,
> thaknfully limited mostly to closed-source software.
> 
> M$ is having the same trouble, having to supply 386 IE for basically
> every architecture NT is working on (From what I remember, 386 IE is
> included in Alpha, IA-64 and AMD64 ones) :)
> 
> 
> 
> --
> Paweł Lasek
> "Once a hitokiri, always a hitokiri. This will never change" - Jine-Ei

Re: [9fans] Maybe it is april fool's after all ...

[Bruce Ellis]

it would certainly be more fun if the a significant amount of
open source wasn't open crap.

brucee

On 1/21/06, jmk@plan9.bell-labs.com <jmk@plan9.bell-labs.com> wrote:
> Excuses, excuses. I thought the mantra was "Given enough
> eyeballs all bugs are shallow"? Maybe they should augment the
> rooms full of penguins typing code with rooms full of spiders
> giving it the eight-eyeball lookover. Also, what is behind
> the eyeballs, the skills can be shallower than the bugs.
>
> --jim
>
> On Fri Jan 20 17:20:26 EST 2006, pawel.lasek@gmail.com wrote:
> > On 1/20/06, jmk@plan9.bell-labs.com <jmk@plan9.bell-labs.com> wrote:
> > > Compatibility mode? Why? Surely you just recompile the programme for
> > > the target machine, you have the source, after all. A minute, tops.
> > >
> > > --jim
> >
> > It comes from the fact that some of the code is not-so-portable -
> > OpenOffice 1x being popular example (don't know about 2.0), as it
> > didn't compile on amd64 (some bugs in the code, remember, it comes
> > back from StarDivision and OS/2 :D). There's a lot more of it,
> > thaknfully limited mostly to closed-source software.
> >
> > M$ is having the same trouble, having to supply 386 IE for basically
> > every architecture NT is working on (From what I remember, 386 IE is
> > included in Alpha, IA-64 and AMD64 ones) :)
> >
> >
> >
> > --
> > Paweł Lasek
> > "Once a hitokiri, always a hitokiri. This will never change" - Jine-Ei
> On Fri Jan 20 17:20:26 EST 2006, pawel.lasek@gmail.com wrote:
> > On 1/20/06, jmk@plan9.bell-labs.com <jmk@plan9.bell-labs.com> wrote:
> > > Compatibility mode? Why? Surely you just recompile the programme for
> > > the target machine, you have the source, after all. A minute, tops.
> > >
> > > --jim
> >
> > It comes from the fact that some of the code is not-so-portable -
> > OpenOffice 1x being popular example (don't know about 2.0), as it
> > didn't compile on amd64 (some bugs in the code, remember, it comes
> > back from StarDivision and OS/2 :D). There's a lot more of it,
> > thaknfully limited mostly to closed-source software.
> >
> > M$ is having the same trouble, having to supply 386 IE for basically
> > every architecture NT is working on (From what I remember, 386 IE is
> > included in Alpha, IA-64 and AMD64 ones) :)
> >
> >
> >
> > --
> > Paweł Lasek
> > "Once a hitokiri, always a hitokiri. This will never change" - Jine-Ei
>

Re: [9fans] Maybe it is april fool's after all ...

[erik quanstrom]

just because you can't see the source, doesn't make it any better.

the worst code i ever worked with was icem/ddn, a cad program
that was a mixture of fortran, c and c++. complete with holorith
data, computed-gotos, overlays, and holorith data.

that was also the program, of all that i've worked on, that 
customers liked most.

- erik

Bruce Ellis <bruce.ellis@gmail.com> writes

| 
| it would certainly be more fun if the a significant amount of
| open source wasn't open crap.
| 
| brucee
| 
| On 1/21/06, jmk@plan9.bell-labs.com <jmk@plan9.bell-labs.com> wrote:
| > Excuses, excuses. I thought the mantra was "Given enough
| > eyeballs all bugs are shallow"? Maybe they should augment the
| > rooms full of penguins typing code with rooms full of spiders
| > giving it the eight-eyeball lookover. Also, what is behind
| > the eyeballs, the skills can be shallower than the bugs.
| >
| > --jim
| >

Re: [9fans] Maybe it is april fool's after all ...

[jmk]

On Fri Jan 20 23:28:05 EST 2006, quanstro@quanstro.net wrote:
> just because you can't see the source, doesn't make it any better.
> ...
> - erik
> ...

the original point was that it doesn't matter whether you can
see the source or not, the effect is the same - failure to deal
with complexity results in an inability to simply compile a
working programme on new hardware running the same O/S and have
it work. instead, hacks are done and baggage is carried around
to keep the old binary working.

--jim

Re: [9fans] Maybe it is april fool's after all ...

[uriel]

Subject: Re: BK, deltas, snapshots and fate of -pre...
From: Alexander Viro (viro@math.psu.edu)
To: LKML
Date: Sun Apr 21 2002 - 01:27:56 EST
Url: http://groups.google.com/group/fa.linux.kernel/msg/caede4c7fd703c4e

[...]
When it comes to software there are three systems of beliefs.
One of them:

    * Thou shalt know by your heart that all software sucks.
    * Beware of those who say that their software does not suck, for they
      are either fools or liars.
    * Beware of those who give you garments and do not allow to mend them,
      for sooner or later thou shalt find what needs mending.
    * But beware also of those who give you badly rotten garments and say
      "Thou shalt prefer that above everything, for thou art allowed to
      mend it".
    * Thou shalt not treat software as a living being, for it is not one.
    * Choose a license of thine liking for sofware thou writest and do not
      blame those who choose differently for software they write.
    * Know when to say "It can be mended, I shalt do that" and when to
      say "It is rotten beyond repair".
    * Choose free over non-free when it is better or when thou art willing
      to fix what is broken.
    * When shit happens, think how to fix it.

Another:

    * All software wants to be free
    * Thou shalt not use non-free software
    * Thou shalt not mention non-free software
    * Thou shalt make all thine software free
    * Thou shalt choose free above working, even if free one is broken
      beyond repair
    * When shit happens, add new features

and the last one:

    * Our 3133t! K3wl! Software! Does Not Suck!!!
    * Always choose our software above everything else
    * When shit happens, we add new features

If you happen to believe in second variant, you have my condolence as
long as you don't force your beliefs on everybody else. If you choose
to emulate door-to-door pests^H^H^H^Hreachers - don't expect to be
treated differently.

Re: [9fans] Maybe it is april fool's after all ...

[Skip Tavakkolian]

>     * Our 3133t! K3wl! Software! Does Not Suck!!!
>     * Always choose our software above everything else
>     * When shit happens, we add new features

some open source projects remind me of public swimming pools
in the middle of summer: noisy, full of kids and not hygienic; the kids
however, think it's k3wl.

Re: [9fans] Maybe it is april fool's after all ...

[Bruce Ellis]

i think ken said "open sewer" at one stage.

brucee

On 1/21/06, Skip Tavakkolian <9nut@9netics.com> wrote:
> >     * Our 3133t! K3wl! Software! Does Not Suck!!!
> >     * Always choose our software above everything else
> >     * When shit happens, we add new features
>
> some open source projects remind me of public swimming pools
> in the middle of summer: noisy, full of kids and not hygienic; the kids
> however, think it's k3wl.

Re: [9fans] Maybe it is april fool's after all ...

[Brantley Coile]

And a sufficient quantity of bad source is the same, or worse, than
no source at all.


> On Fri Jan 20 23:28:05 EST 2006, quanstro@quanstro.net wrote:
>> just because you can't see the source, doesn't make it any better.
>> ...
>> - erik
>> ...
> 
> the original point was that it doesn't matter whether you can
> see the source or not, the effect is the same - failure to deal
> with complexity results in an inability to simply compile a
> working programme on new hardware running the same O/S and have
> it work. instead, hacks are done and baggage is carried around
> to keep the old binary working.
> 
> --jim

Re: [9fans] Maybe it is april fool's after all ...

[Ronald G. Minnich]

> the original point was that it doesn't matter whether you can
> see the source or not, the effect is the same - failure to deal
> with complexity results in an inability to simply compile a
> working programme on new hardware running the same O/S and have
> it work. instead, hacks are done and baggage is carried around
> to keep the old binary working.
>

it's even worse than that, because people learn from the bad code. We're
raising a whole generation of people who think awful stuff is how it's
done ...

ron

Re: [9fans] Maybe it is april fool's after all ...

[Bruce Ellis]

anyone who has thought "i wonder how linux does this" will
have gone through all the layers of misguided inexperience.

brucee

On 1/22/06, Ronald G. Minnich <rminnich@lanl.gov> wrote:
>
> > the original point was that it doesn't matter whether you can
> > see the source or not, the effect is the same - failure to deal
> > with complexity results in an inability to simply compile a
> > working programme on new hardware running the same O/S and have
> > it work. instead, hacks are done and baggage is carried around
> > to keep the old binary working.
> >
>
> it's even worse than that, because people learn from the bad code. We're
> raising a whole generation of people who think awful stuff is how it's
> done ...
>
> ron
>

Re: [9fans] Maybe it is april fool's after all ...

[David Leimbach]

On 1/20/06, jmk@plan9.bell-labs.com <jmk@plan9.bell-labs.com> wrote:
> On Fri Jan 20 23:28:05 EST 2006, quanstro@quanstro.net wrote:
> > just because you can't see the source, doesn't make it any better.
> > ...
> > - erik
> > ...
>
> the original point was that it doesn't matter whether you can
> see the source or not, the effect is the same - failure to deal
> with complexity results in an inability to simply compile a
> working programme on new hardware running the same O/S and have
> it work. instead, hacks are done and baggage is carried around
> to keep the old binary working.

Yep, anyone who's ever tried to use FreeBSD ports on x86-64 platforms
as 64bit applications.

Totally not worth the aggravation...  At least not if you don't have
to, I was just curious one day.

Dave

Re: [9fans] Maybe it is april fool's after all ...

[Skip Tavakkolian]

>> the original point was that it doesn't matter whether you can
>> see the source or not, the effect is the same - failure to deal
>> with complexity results in an inability to simply compile a
>> working programme on new hardware running the same O/S and have
>> it work. instead, hacks are done and baggage is carried around
>> to keep the old binary working.
> 
> Yep, anyone who's ever tried to use FreeBSD ports on x86-64 platforms
> as 64bit applications.

wouldn't a VM be a viable option, since we're talking about one
processor family?  those cantankerous applications, that don't port
easily, can be run under qemu/xen/etc. with appropriate os.
everything old is new again.

Re: [9fans] Maybe it is april fool's after all ...

[Brantley Coile]

> that was also the program, of all that i've worked on, that 
> customers liked most.

I guess some sort of usefullness is why X windows still lives.

Re: [9fans] Maybe it is april fool's after all ...

[David Leimbach]

On 1/20/06, Devon H. O'Dell <devon.odell@gmail.com> wrote:
> The specific reason many Linuxes do this is for i386 compat on AMD64
> architectures. It's horribly broken, but is only slightly worse than
> the way FreeBSD implements it, which is to have /usr/lib32
>
> --Devon
>

Funny, the linux distros I've used have "/usr/lib64"  SUSE was one
that used this, but then again, I must admit, I don't use linux
anywhere near as much as I used to.

Dave

Re: [9fans] Maybe it is april fool's after all ...

[Ronald G Minnich]

David Leimbach wrote:
> On 1/20/06, Devon H. O'Dell <devon.odell@gmail.com> wrote:
> 
>>The specific reason many Linuxes do this is for i386 compat on AMD64
>>architectures. It's horribly broken, but is only slightly worse than
>>the way FreeBSD implements it, which is to have /usr/lib32
>>
>>--Devon
>>
> 
> 
> Funny, the linux distros I've used have "/usr/lib64"  SUSE was one
> that used this, but then again, I must admit, I don't use linux
> anywhere near as much as I used to.
>

/usr/lib64 &tc. is very, very common linuxices. I would prefer /usr/lib32.

recompilation is not a common option on linux, which has tons of 
something we don't see much on plan 9: third party software.

ron

Re: [9fans] Maybe it is april fool's after all ...

[jmk]

On Fri Jan 20 12:42:42 EST 2006, rminnich@lanl.gov wrote:
> 
> recompilation is not a common option on linux, which has tons of 
> something we don't see much on plan 9: third party software.
> 
> ron

That sounds a lot like Windows - "I got this code from somewhere else
and it's too difficult/awkward or impossible to move it to my new system
so I will carry the baggage around forever".

So there's no free freedom after all?

--jim

Re: [9fans] Maybe it is april fool's after all ...

[Dave Lukes]

 > So there's no free freedom after all?
 >
 > --jim

"The price of freedom is eternal memory upgrades".

An aside:
I also find it slightly ironic that,
out of all the free/open/whatever licenses,
the FSF's is probably the most restrictive,
due to the encumbrances it places on it's victims^h^h^h^h^h^h^husers.

What *I* want is a s/w licence which says something like:
"You can do whatever you want with this except tell lies about it".

DaveL

Re: [9fans] Maybe it is april fool's after all ...

[Lyndon Nerenberg]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


| What *I* want is a s/w licence which says something like:
| "You can do whatever you want with this except tell lies about it".

You just described the three-clause BSD license.

- --lyndon

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (FreeBSD)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFD0S4lSoWjgnL/KRcRAnFAAJ4pSKsaj4qLHZLAP9Y7C0unGMcOwwCfeShT
IOK91i3+qVGsI+x3k9+pfuQ=
=sALL
-----END PGP SIGNATURE-----

Re: [9fans] Maybe it is april fool's after all ...

[Marina Brown]

Dave Lukes wrote:

> > So there's no free freedom after all?
> >
> > --jim
>
> "The price of freedom is eternal memory upgrades".
>
> An aside:
> I also find it slightly ironic that,
> out of all the free/open/whatever licenses,
> the FSF's is probably the most restrictive,
> due to the encumbrances it places on it's victims^h^h^h^h^h^h^husers.
>
> What *I* want is a s/w licence which says something like:
> "You can do whatever you want with this except tell lies about it".
>
> DaveL
>
>
I've been a fan of the BSD licenses.  They are so simple and free.

-- Marina Brown

Re: [9fans] Maybe it is april fool's after all ...

[Ronald G Minnich]

jmk@plan9.bell-labs.com wrote:

> That sounds a lot like Windows - "I got this code from somewhere else
> and it's too difficult/awkward or impossible to move it to my new system
> so I will carry the baggage around forever".
> 
> So there's no free freedom after all?

yep. By 3rd party software I mean stuff like compilers, libraries, CAD 
tools, and so on. It's going to be a while before it's all 64-bit for 
linux, I would guess. And, seeing the stuff that they do in gnu* 
nowadays, that's hardly surprising.

I mean, ld scripts where a .so should be? holy mackeral.

ron