Steve wrote:
> 
> It only appeared on the mirror here yesterday evening.

But I bet you have a place (Linux Weekly News or such)
where the announcements are made, and everybody knows in
a matter of seconds? Of course, downloading is another
issue (see the Slackware FTP...)

Didn't intend to offense you, by the way. I just thought
that it was a bit too much Linux specific and not POV
enough. And besides, this Linux program, err, I think
there are lots of places where you can talk about it,
right?  :-)

-- 
Adrien Beau - adr### [at] freefr - http://adrien.beau.free.fr
 Mes propos n'engagent que moi et en aucun cas mes employeurs

Post a reply to this message

Thorsten Froehlich wrote:
> 
> If it isn't stable and released than it is called alpha or beta, not final.
> GCC 3.0 is final, so it is stable.  If it is not, then it is from M$ ;-)
> Alternatively, it is just incompetence of the authors - their own list for
> testing included the Linux kernel...

Let's clarify some points.

IMNSHO...

Alphas are (should be) for testing the bare functionnalities
of a program, testing some new ideas, have the first feedback.
They are very buggy, and testers usually need to have means
to contact authors personnaly in order to solve problems
(that means a small team of testers)

Betas are (should be) for having most of the bugs squished,
and to have more feedback in terms of usability of the program.

But when no more bugs are found by the beta testers, and
everything seems stable and finished (as planned), it's time
for release.

And when the program is released, much more people use it,
and people use it for real-life needs, indeed, and so new,
hard-to-catch bugs are found. System incompatibilities, too,
because you can't test your program in every possible
configuration.

Most programs that I use have a .1 or .01 or whatever,
because that's the version the authors distribute. The .0
release is generally replaced by the same, bugless (we hope)
version.

Complex pieces of code like gcc, or the Linux kernel,
which use a lot of elaborate mechanisms, which deploys
for the first time new whole pieces of program (this is
a major realease guys!) have a 101% chance of containing
bugs.

I'm impressed to see lots of people rushing to use ReiserFS
to store their data, while there still are known bugs into
it! Using an instable program to do some calculus, or some
games is fine for me, but storing data unreliably... Wow.

-- 
Adrien Beau - adr### [at] freefr - http://adrien.beau.free.fr
 Mes propos n'engagent que moi et en aucun cas mes employeurs

Post a reply to this message

On 5 Jul 2001 18:39:51 -0400, ste### [at] zeroppsuklinuxnet (Steve) wrote:

>This means a kernel that will build with GCC 3.0

I know it compiled with egcs 1.1.2 fine as 2.4.6 is what I am running
now (and slowly replacing ipchains with iptables :) ). To try to
compile it with gcc 3.0 ... I'd rather e-mail Santa.


Peter Popov ICQ : 15002700
Personal e-mail : pet### [at] vipbg
TAG      e-mail : pet### [at] tagpovrayorg

Post a reply to this message

Adrien Beau wrote:
> 
> I'm impressed to see lots of people rushing to use ReiserFS
> to store their data, while there still are known bugs into
> it! Using an instable program to do some calculus, or some
> games is fine for me, but storing data unreliably... Wow.
> 

Yes, that is over the line for me - even though I otherwise have little
qualms over using beta software, including kernels. If the filesystem is
screwed, I'm screwed (what is that "backup" thing people keep talking
about?)
I did try building kernel 2.4.6 with gcc 3.0 BTW, but the result was
decidedly not stable. Looks like I'll have to switch back to 2.96 for
now; no biggie really, frankly I haven't seen *that* huge performance
leaps from using 3.0

-- 
Margus Ramst

Personal e-mail: mar### [at] peakeduee
TAG (Team Assistance Group) e-mail: mar### [at] tagpovrayorg
Home page http://www.hot.ee/margusrt

Post a reply to this message

On Mon, 09 Jul 2001 11:49:23 +0200, Adrien Beau
<adr### [at] sycomorefr> wrote:

>Of course, downloading is another issue (see the 
>Slackware FTP...)

Slack did something utterly stupid. They released Slackware 8.0 to the
public before the mirror sites got it. I got my hands on it less than
15 minutes after the isos were uploaded and did quite well, but in a
matter of hours the connection dropped to 2k (over a 5Mb airlink).
None of the mirrors had it yet, some had parts. I left it alone and
waited for the mirrors. It took sunet.se and funet.fi two days to get
it, and we all know how fast they are. Now, if only Slack had waited
for the mirrors... I would have had 8.0 up and running now.


Peter Popov ICQ : 15002700
Personal e-mail : pet### [at] vipbg
TAG      e-mail : pet### [at] tagpovrayorg

Post a reply to this message

Peter Popov wrote:
> 
> Now, if only Slack had waited
> for the mirrors... I would have had 8.0 up and running now.
> 

Then again, it probably would've been slashdotted anyway...

-- 
Margus Ramst

Personal e-mail: mar### [at] peakeduee
TAG (Team Assistance Group) e-mail: mar### [at] tagpovrayorg
Home page http://www.hot.ee/margusrt

Post a reply to this message

On Tue, 10 Jul 2001 01:54:02 +0300, Peter Popov wrote:
>On 5 Jul 2001 18:39:51 -0400, ste### [at] zeroppsuklinuxnet (Steve) wrote:
>
>>This means a kernel that will build with GCC 3.0
>
>I know it compiled with egcs 1.1.2 fine as 2.4.6 is what I am running
>now (and slowly replacing ipchains with iptables :) ). To try to
>compile it with gcc 3.0 ... I'd rather e-mail Santa.

It comples fine, no problems there, it's choosing the right sections
to include at configure time that's caused me the problems so far. 

--
Cheers
Steve              email mailto:ste### [at] zeroppsuklinuxnet

%HAV-A-NICEDAY Error not enough coffee  0 pps. 

web http://www.zeropps.uklinux.net/

or  http://start.at/zero-pps

  2:13am  up  5:04,  2 users,  load average: 1.02, 1.10, 1.13

Post a reply to this message

On Mon, 09 Jul 2001 22:56:44 +0300, Margus Ramst wrote:


>I did try building kernel 2.4.6 with gcc 3.0 BTW, but the result was
>decidedly not stable.

[snip]

I've found the performance of 2.4.6 to be much better tna 2.2.14 which
I have at the moment, the multitasking is so much quicker, and programs
open so much faster (as I've seen when booting with 2.4.6). 

--
Cheers
Steve              email mailto:ste### [at] zeroppsuklinuxnet

%HAV-A-NICEDAY Error not enough coffee  0 pps. 

web http://www.zeropps.uklinux.net/

or  http://start.at/zero-pps

  2:18am  up  5:08,  2 users,  load average: 1.00, 1.03, 1.08

Post a reply to this message

Peter Popov wrote:
> 
> Slack did something utterly stupid. They released Slackware 8.0 to the
> public before the mirror sites got it.

As Margus noted, there's always a Slashdotter hanging around
every FTP sites and that notifies the world as soon as something
happen.

The correct thing to do would have been to build the ISO on
another computer (Pat's ADSL box?) and give it to some mirrors
from that location, or, more easily done, put up a separate
FTP holding just the ISO, without anonymous access, so that
only authorized mirrors could get it.

But I think Pat was too tired to do anything else than
"upload the ISOs - post announcement - go to bed".

-- 
Adrien Beau - adr### [at] freefr - http://adrien.beau.free.fr
 Mes propos n'engagent que moi et en aucun cas mes employeurs

Post a reply to this message

Steve wrote:
> 
> I've found the performance of 2.4.6 to be much better tna 2.2.14 which
> I have at the moment, the multitasking is so much quicker, and programs
> open so much faster (as I've seen when booting with 2.4.6).
> 

Well, I'm running 2.4.5 right now; it's certainly better than 2.2
kernels with most stuff, but I've heard it has some serious bugs in its
VM code so I guess I really should do the upgrade.

-- 
Margus Ramst

Personal e-mail: mar### [at] peakeduee
TAG (Team Assistance Group) e-mail: mar### [at] tagpovrayorg
Home page http://www.hot.ee/margusrt

Post a reply to this message