bk-3.2.0 released

bk-3.2.0 released

[Larry McVoy]

BitKeeper Users,

BK/Pro 3.2.0 has been released and is in the BK download area,

    http://bitmover.com/download

We're departing from our previous tradition of letting the word out slowly,
people didn't like that.  Because of that, we suggest that cautious users
wait for a week or two to make sure that any lingering bugs have been caught.
The rest of you go bang on it and let us know if we missed anything please.

As always, please use "bk sendbug" to file any bug reports for this
release.

The next effort from us will be a commercial release we hope to have
out in mid-summer.

Thanks and enjoy,

The BitKeeper Team

===============================================================================
Release notes for BitKeeper version 3.2.0.   These release notes cover
the following topics:

. Upgrade Notes
. New Features
. Noteworthy bug fixes

Upgrade notes
----------------------------------------------------------------------------

  On Microsoft Windows, BitKeeper no longer requires (or uses) Cygwin.
  You may use Cygwin if you like but BitKeeper does not need it; BK
  is completely self contained.  

  Registry management, uninstall, and upgrades should work substantially
  better than they have in the past.

New Features
----------------------------------------------------------------------------

  A simple pager called 'bk more' is included and will be used if no
  suitable pager can be found.

  Under Windows the SCCS dirs are now created with the "hidden" attribute.

  HPUX 11.11 is now a supported platform.

  A new command called 'bk lease'.  This is primarily for openlogging users.
  Please see 'bk helptool lease' for more information.

Noteworthy Bug Fixes
----------------------------------------------------------------------------

  2004-01-19-002: Fix resolve bug where 'vr' (view remote) didn't work
  2004-01-27-001: Fix include/exclude processing in 'bk export'
  2004-02-06-002: bk export: fix error case for missing src directory
  2004-03-12-001: Don't delete a RESYNC directory we didn't create
  2004-04-19-001: Fix regression in bk send
  2004-04-23-001: Fix HTTP requests to match standards
  
  Remove an extra unneeded consistency check on the first commit
  after every clone.

  Fixed bug in citool that causes comments for pending files to not
  show up in the bottom window when the ChangeSet file is selected.

  Fix problem in BkWeb where the "N weeks" times didn't match
  reality.

  Improve documentation for which characters are legal in tags and add more
  error checks on tag names.
						   
  Prevent people from passing bad options to 'bk comments'.
  
  Always write keys to BitKeeper/etc/csets-in instead of revs or
  serial numbers.

  Fix a case where 'bk repogca' could return the wrong answer.

  Fix case where BitKeeper didn't work correctly with PGP 8.0 on
  Windows.

  Set time stamps correctly after running 'bk fix'.

-- 
---
Larry McVoy                lm at bitmover.com           http://www.bitkeeper.com

Re: bk-3.2.0 released

[Christoph Hellwig]

On Tue, May 18, 2004 at 04:32:38PM -0700, Larry McVoy wrote:
> BitKeeper Users,
> 
> BK/Pro 3.2.0 has been released and is in the BK download area,
> 
>     http://bitmover.com/download

And what makes this relevant for lkml?  I'm happy BK user, but this just
seems to be compltely offtopic here..

Re: bk-3.2.0 released

[Larry McVoy]

On Wed, May 19, 2004 at 07:51:28AM +0100, Christoph Hellwig wrote:
> On Tue, May 18, 2004 at 04:32:38PM -0700, Larry McVoy wrote:
> > BitKeeper Users,
> > 
> > BK/Pro 3.2.0 has been released and is in the BK download area,
> > 
> >     http://bitmover.com/download
> 
> And what makes this relevant for lkml?  I'm happy BK user, but this just
> seems to be compltely offtopic here..

A) Lots of BK users here.
B) We like people to stay up to date (we got burned by some user getting us
   to debug a problem that we had long since fixed).
C) I've been yelled at for not broadcasting about new releases (and now 
   I'm yelled at for doing so.  Sigh).

It would be nice if you suggested a better approach, I'm at a loss as to
how to keep everyone happy.
-- 
---
Larry McVoy                lm at bitmover.com           http://www.bitkeeper.com

Re: bk-3.2.0 released

[Jan-Benedict Glaw]

[-- Attachment #1: Type: text/plain, Size: 1012 bytes --]

On Wed, 2004-05-19 07:02:59 -0700, Larry McVoy <lm@bitmover.com>
wrote in message <20040519140259.GA18977@work.bitmover.com>:
> On Wed, May 19, 2004 at 07:51:28AM +0100, Christoph Hellwig wrote:
> > On Tue, May 18, 2004 at 04:32:38PM -0700, Larry McVoy wrote:

> > > BK/Pro 3.2.0 has been released and is in the BK download area,
> > And what makes this relevant for lkml?  I'm happy BK user, but this just
> > seems to be compltely offtopic here..
> 
> It would be nice if you suggested a better approach, I'm at a loss as to
> how to keep everyone happy.

Just create a bk-announce mailing list... News for those that want to
read them, and even an archive of bk's timeline.

MfG, JBG

-- 
   Jan-Benedict Glaw       jbglaw@lug-owl.de    . +49-172-7608481
   "Eine Freie Meinung in  einem Freien Kopf    | Gegen Zensur | Gegen Krieg
    fuer einen Freien Staat voll Freier Bürger" | im Internet! |   im Irak!
   ret = do_actions((curr | FREE_SPEECH) & ~(NEW_COPYRIGHT_LAW | DRM | TCPA));

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: bk-3.2.0 released

[Larry McVoy]

On Wed, May 19, 2004 at 04:11:15PM +0200, Jan-Benedict Glaw wrote:
> On Wed, 2004-05-19 07:02:59 -0700, Larry McVoy <lm@bitmover.com>
> wrote in message <20040519140259.GA18977@work.bitmover.com>:
> > On Wed, May 19, 2004 at 07:51:28AM +0100, Christoph Hellwig wrote:
> > > On Tue, May 18, 2004 at 04:32:38PM -0700, Larry McVoy wrote:
> 
> > > > BK/Pro 3.2.0 has been released and is in the BK download area,
> > > And what makes this relevant for lkml?  I'm happy BK user, but this just
> > > seems to be compltely offtopic here..
> > 
> > It would be nice if you suggested a better approach, I'm at a loss as to
> > how to keep everyone happy.
> 
> Just create a bk-announce mailing list... News for those that want to
> read them, and even an archive of bk's timeline.

There is one, it was copied on the mail, but lots of people who use BK
for the kernel aren't on it.

I don't see the problem.  What's the big deal about a post that says a
new version of a widely used tool is available?  If someone posted that
there is a new version of gcc available is that off topic?  
-- 
---
Larry McVoy                lm at bitmover.com           http://www.bitkeeper.com

Re: bk-3.2.0 released

[Kevin P. Fleming]

Larry McVoy wrote:

> I don't see the problem.  What's the big deal about a post that says a
> new version of a widely used tool is available?  If someone posted that
> there is a new version of gcc available is that off topic?  

No, this list gets monthly (if not more often) announcements of new 
versions of the "shortlog" tool and the other related tool, both of 
which are used pretty much exclusively by Linus and Marcelo. I haven't 
seen anyone complain about those, and their frequency is 10 times that 
of the BK update announcements.

Re: bk-3.2.0 released

[John Bradford]

Quote from "Kevin P. Fleming" <kpfleming@backtobasicsmgmt.com>:
> Larry McVoy wrote:
> 
> > I don't see the problem.  What's the big deal about a post that says a
> > new version of a widely used tool is available?  If someone posted that
> > there is a new version of gcc available is that off topic?  
> 
> No, this list gets monthly (if not more often) announcements of new 
> versions of the "shortlog" tool and the other related tool, both of 
> which are used pretty much exclusively by Linus and Marcelo.

What license is "shortlog" released under?

John.

Re: bk-3.2.0 released

[Tomas Szepe]

On May-19 2004, Wed, 15:36 +0100
John Bradford <john@grabjohn.com> wrote:

> Quote from "Kevin P. Fleming" <kpfleming@backtobasicsmgmt.com>:
> > Larry McVoy wrote:
> > 
> > > I don't see the problem.  What's the big deal about a post that says a
> > > new version of a widely used tool is available?  If someone posted that
> > > there is a new version of gcc available is that off topic?  
> > 
> > No, this list gets monthly (if not more often) announcements of new 
> > versions of the "shortlog" tool and the other related tool, both of 
> > which are used pretty much exclusively by Linus and Marcelo.
> 
> What license is "shortlog" released under?

I think it must be the "I'll annoy you with bullshit releases every
other day" license, but I didn't look.

-- 
Tomas Szepe <szepe@pinerecords.com>

Re: bk-3.2.0 released

[Jan-Benedict Glaw]

[-- Attachment #1: Type: text/plain, Size: 690 bytes --]

On Wed, 2004-05-19 07:16:48 -0700, Larry McVoy <lm@bitmover.com>
wrote in message <20040519141648.GB18977@work.bitmover.com>:

> new version of a widely used tool is available?  If someone posted that
> there is a new version of gcc available is that off topic?  

Yes, it was. Even miscompilation reports are mostly OT, since they
should go to GCC's bugzilla.

MfG, JBG

-- 
   Jan-Benedict Glaw       jbglaw@lug-owl.de    . +49-172-7608481
   "Eine Freie Meinung in  einem Freien Kopf    | Gegen Zensur | Gegen Krieg
    fuer einen Freien Staat voll Freier Bürger" | im Internet! |   im Irak!
   ret = do_actions((curr | FREE_SPEECH) & ~(NEW_COPYRIGHT_LAW | DRM | TCPA));

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: bk-3.2.0 released

[Larry McVoy]

On Wed, May 19, 2004 at 04:26:56PM +0200, Jan-Benedict Glaw wrote:
> On Wed, 2004-05-19 07:16:48 -0700, Larry McVoy <lm@bitmover.com>
> wrote in message <20040519141648.GB18977@work.bitmover.com>:
> 
> > new version of a widely used tool is available?  If someone posted that
> > there is a new version of gcc available is that off topic?  
> 
> Yes, it was. Even miscompilation reports are mostly OT, since they
> should go to GCC's bugzilla.

Sigh.  This strikes me as "aha, Larry's posted something, let's see if
we can get him hooked on some trollbait, that would be fun, it's been
a while!"  

I'll pass if you don't mind.  If you really have a problem then mention
it to DaveM who will educate me at the next Sharks hockey game.
-- 
---
Larry McVoy                lm at bitmover.com           http://www.bitkeeper.com

Re: bk-3.2.0 released

[Jan-Benedict Glaw]

[-- Attachment #1: Type: text/plain, Size: 1608 bytes --]

On Wed, 2004-05-19 07:48:52 -0700, Larry McVoy <lm@bitmover.com>
wrote in message <20040519144852.GC18977@work.bitmover.com>:
> On Wed, May 19, 2004 at 04:26:56PM +0200, Jan-Benedict Glaw wrote:
> > On Wed, 2004-05-19 07:16:48 -0700, Larry McVoy <lm@bitmover.com>
> > wrote in message <20040519141648.GB18977@work.bitmover.com>:
> > > new version of a widely used tool is available?  If someone posted that
> > > there is a new version of gcc available is that off topic?  
> > 
> > Yes, it was. Even miscompilation reports are mostly OT, since they
> > should go to GCC's bugzilla.
> 
> Sigh.  This strikes me as "aha, Larry's posted something, let's see if
> we can get him hooked on some trollbait, that would be fun, it's been
> a while!"  

Not at all! You *do* post useful/helpful/serious/whatnot emails (eg.
those about the corruption-at-page's-end topic). It's just that I like
to see noise on this list going down, which includes announcments eg. of
binutils/gcc, bitkeeper, but also the constant announcments for
yet-some-more-email-addresses-added patches for some rarely used script.

So, let's just try to keep noise low, I won't even send more messages to
this thread (accepting whatever you make out of it), but I'll just
continue hacking DZ11 ...

MfG, JBG

-- 
   Jan-Benedict Glaw       jbglaw@lug-owl.de    . +49-172-7608481
   "Eine Freie Meinung in  einem Freien Kopf    | Gegen Zensur | Gegen Krieg
    fuer einen Freien Staat voll Freier Bürger" | im Internet! |   im Irak!
   ret = do_actions((curr | FREE_SPEECH) & ~(NEW_COPYRIGHT_LAW | DRM | TCPA));

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: bk-3.2.0 released

[Larry McVoy]

On Wed, May 19, 2004 at 05:00:40PM +0200, Jan-Benedict Glaw wrote:
> On Wed, 2004-05-19 07:48:52 -0700, Larry McVoy <lm@bitmover.com>
> wrote in message <20040519144852.GC18977@work.bitmover.com>:
> > On Wed, May 19, 2004 at 04:26:56PM +0200, Jan-Benedict Glaw wrote:
> > > On Wed, 2004-05-19 07:16:48 -0700, Larry McVoy <lm@bitmover.com>
> > > wrote in message <20040519141648.GB18977@work.bitmover.com>:
> > > > new version of a widely used tool is available?  If someone posted that
> > > > there is a new version of gcc available is that off topic?  
> > > 
> > > Yes, it was. Even miscompilation reports are mostly OT, since they
> > > should go to GCC's bugzilla.
> > 
> > Sigh.  This strikes me as "aha, Larry's posted something, let's see if
> > we can get him hooked on some trollbait, that would be fun, it's been
> > a while!"  
> 
> Not at all! You *do* post useful/helpful/serious/whatnot emails (eg.
> those about the corruption-at-page's-end topic). It's just that I like
> to see noise on this list going down, which includes announcments eg. of
> binutils/gcc, bitkeeper, but also the constant announcments for
> yet-some-more-email-addresses-added patches for some rarely used script.

Some people complain that we don't announce BK releases and some people
complain that we do.  Given that we do a release about twice a year I
think complaining about it is over the top.  I'm quite happy to stop
announcing it here but the cost of that is I'll just turn off support
for the older releases on openlogging.org and bkbits.net.  That's not
very polite to the BK users (it's completely legal, you are supposed
to stay current under the terms of the BKL) but it would keep the list
free of these oh-so-annoying and oh-so-frequent posts.  See the tradeoff?
Whether you do or don't, I thought about this and felt it was better
to post than to not post.  If you had been in my position I suspect 
you'd do the same thing.  So how about we let this little tempest in
a teapot die now?  I'm sure we have better things to do with our time.
-- 
---
Larry McVoy                lm at bitmover.com           http://www.bitkeeper.com

Re: bk-3.2.0 released

[Nicolas Pitre]

On Wed, 19 May 2004, Larry McVoy wrote:

> Some people complain that we don't announce BK releases ...

Just tell them to subscribe to the BK announcement list.
Then no one will complain anymore.


Nicolas

Re: bk-3.2.0 released

[Valdis.Kletnieks]

[-- Attachment #1: Type: text/plain, Size: 924 bytes --]

On Wed, 19 May 2004 16:26:56 +0200, Jan-Benedict Glaw said:
> On Wed, 2004-05-19 07:16:48 -0700, Larry McVoy <lm@bitmover.com>
> wrote in message <20040519141648.GB18977@work.bitmover.com>:
> > new version of a widely used tool is available?  If someone posted that
> > there is a new version of gcc available is that off topic? =20
> 
> Yes, it was. Even miscompilation reports are mostly OT, since they
> should go to GCC's bugzilla.

Given the number of times (it's well past non-zero) we've changed around kernel
source in order to work around a bug in one or another widely-distributed
release of gcc, I think miscompile reports *are* on topic.

I seem to remember that gcc 3.4 did oddness with inlining strcpy that resulted
in a source tweak to lib/strings.c not too long ago, as just one example...

And there was the "read/write constraints and registers" warning recently...

And there was.... you get the idea...


[-- Attachment #2: Type: application/pgp-signature, Size: 226 bytes --]

Re: bk-3.2.0 released

[Stan Bubrouski]

Guys,

Honestly this crap is getting very silly even though you are all
serious.  People announce things on this list all the time that
are not really a part of the kernel but affect development, and
with the traffic on this list anyways, I don't see how Larry
announcing a new release of BitKeeper is offtopic or innappropriate.

Furthermore, it seems like any time Larry posts anything not in a thread
starting with [patch] or [oops] someone goes off on some political or
otherwise rant and its getting really really old.  The threads become
long and littered with opinions that contribute nothing but someone's
$0.2 in a case where it makes no difference at all except to the
bandwith the posts complaining about posts wastes.

Lighten up I guess is all I'm saying.

-sb


On Wed, 2004-05-19 at 16:06, Valdis.Kletnieks@vt.edu wrote:
> On Wed, 19 May 2004 16:26:56 +0200, Jan-Benedict Glaw said:
> > On Wed, 2004-05-19 07:16:48 -0700, Larry McVoy <lm@bitmover.com>
> > wrote in message <20040519141648.GB18977@work.bitmover.com>:
> > > new version of a widely used tool is available?  If someone posted that
> > > there is a new version of gcc available is that off topic? =20
> > 
> > Yes, it was. Even miscompilation reports are mostly OT, since they
> > should go to GCC's bugzilla.
> 
> Given the number of times (it's well past non-zero) we've changed around kernel
> source in order to work around a bug in one or another widely-distributed
> release of gcc, I think miscompile reports *are* on topic.
> 
> I seem to remember that gcc 3.4 did oddness with inlining strcpy that resulted
> in a source tweak to lib/strings.c not too long ago, as just one example...
> 
> And there was the "read/write constraints and registers" warning recently...
> 
> And there was.... you get the idea...
> 

Re: bk-3.2.0 released

[Christoph Hellwig]

On Wed, May 19, 2004 at 07:16:48AM -0700, Larry McVoy wrote:
> I don't see the problem.  What's the big deal about a post that says a
> new version of a widely used tool is available?  If someone posted that
> there is a new version of gcc available is that off topic?  

Yes.

Re: bk-3.2.0 released

[Christoph Hellwig]

On Wed, May 19, 2004 at 07:02:59AM -0700, Larry McVoy wrote:
> > And what makes this relevant for lkml?  I'm happy BK user, but this just
> > seems to be completely offtopic here..
> 
> A) Lots of BK users here.
> B) We like people to stay up to date (we got burned by some user getting us
>    to debug a problem that we had long since fixed).
> C) I've been yelled at for not broadcasting about new releases (and now 
>    I'm yelled at for doing so.  Sigh).
> 
> It would be nice if you suggested a better approach, I'm at a loss as to
> how to keep everyone happy.

in addition to the announce list you should probably put it prominently
on your websites (bitmover and probably bkbits).

Re: bk-3.2.0 released

[John Bradford]

Quote from Larry McVoy <lm@bitmover.com>:
> C) I've been yelled at for not broadcasting about new releases (and now 
>    I'm yelled at for doing so.  Sigh).
> 
> It would be nice if you suggested a better approach, I'm at a loss as to
> how to keep everyone happy.

Ask somebody else at Bitmover to make the announcement :-).

John.

Re: bk-3.2.0 released

[Bill Davidsen]

Christoph Hellwig wrote:
> On Tue, May 18, 2004 at 04:32:38PM -0700, Larry McVoy wrote:
> 
>>BitKeeper Users,
>>
>>BK/Pro 3.2.0 has been released and is in the BK download area,
>>
>>    http://bitmover.com/download
> 
> 
> And what makes this relevant for lkml?  I'm happy BK user, but this just
> seems to be compltely offtopic here..

Announcements regarding kernel tools seem relevant, as is any other tool 
change which might affect the kernel creation, maintenence, compilation 
or install. Considering the amount of stuff virtually everyone reads 
past in this list, reports on features not of interest or 2.4 kernels no 
longer used, why did you single out this short and informational message 
as worthy of comment?

I personally have a folder of release announcements, even for things I 
don't use, just so I can find them quickly if my interest or business 
needs change. I wish there was a single release announcements list, but 
there isn't, let's keep useful information HERE!

-- 
bill davidsen <davidsen@tmr.com>
   CTO TMR Associates, Inc
   Doing interesting things with small computers since 1979

Re: bk-3.2.0 released

[Jeff Sipek]

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

On Wednesday 19 May 2004 02:51, Christoph Hellwig wrote:
> On Tue, May 18, 2004 at 04:32:38PM -0700, Larry McVoy wrote:
> > BitKeeper Users,
> >
> > BK/Pro 3.2.0 has been released and is in the BK download area,

Thanks for the announcement.

Jeff.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFArCGVwFP0+seVj/4RAnNjAJ0X+iwHwWI5WVofmd9+twdYweOuygCfRJa5
fQynPZmVDWDPaFbNZgsgCTw=
=2dj+
-----END PGP SIGNATURE-----

Re: bk-3.2.0 released

[Vojtech Pavlik]

On Tue, May 18, 2004 at 04:32:38PM -0700, Larry McVoy wrote:
> BitKeeper Users,
> 
> BK/Pro 3.2.0 has been released and is in the BK download area,
> 
>     http://bitmover.com/download

Any chance of a native x86-64 version? 


-- 
Vojtech Pavlik
SuSE Labs, SuSE CR

Re: bk-3.2.0 released

[Larry McVoy]

On Sat, May 29, 2004 at 11:54:20AM +0200, Vojtech Pavlik wrote:
> On Tue, May 18, 2004 at 04:32:38PM -0700, Larry McVoy wrote:
> > BitKeeper Users,
> > 
> > BK/Pro 3.2.0 has been released and is in the BK download area,
> > 
> >     http://bitmover.com/download
> 
> Any chance of a native x86-64 version? 

We don't have any x86-64 machines but we could get one.  I asked about this
a while back and people told me that there was no point, the x86 one worked
perfectly.  Can you tell me what having a native one would gain?  If there
is any gain we'll do it.
-- 
---
Larry McVoy                lm at bitmover.com           http://www.bitkeeper.com

Re: bk-3.2.0 released

[Hugo Mills]

[-- Attachment #1: Type: text/plain, Size: 1121 bytes --]

On Sat, May 29, 2004 at 06:04:36AM -0700, Larry McVoy wrote:
> On Sat, May 29, 2004 at 11:54:20AM +0200, Vojtech Pavlik wrote:
> > On Tue, May 18, 2004 at 04:32:38PM -0700, Larry McVoy wrote:
> > > BK/Pro 3.2.0 has been released and is in the BK download area,
> > > 
> > >     http://bitmover.com/download
> > 
> > Any chance of a native x86-64 version? 
> 
> We don't have any x86-64 machines but we could get one.  I asked about this
> a while back and people told me that there was no point, the x86 one worked
> perfectly.  Can you tell me what having a native one would gain?  If there
> is any gain we'll do it.

   It'll allow BK to be run on machines which have a pure 64-bit
userspace (for example, Debian's current amd64 port), without having
to resort to a 32-bit chroot to run the 32-bit BK binary.

   Hugo.

-- 
=== Hugo Mills: hugo@... carfax.org.uk | darksatanic.net | lug.org.uk ===
  PGP key: 1C335860 from wwwkeys.eu.pgp.net or http://www.carfax.org.uk
         --- Nothing right in my left brain. Nothing left in ---         
                             my right brain.                             

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]

Re: bk-3.2.0 released

[Larry McVoy]

On Sat, May 29, 2004 at 02:15:10PM +0100, Hugo Mills wrote:
> On Sat, May 29, 2004 at 06:04:36AM -0700, Larry McVoy wrote:
> > On Sat, May 29, 2004 at 11:54:20AM +0200, Vojtech Pavlik wrote:
> > > On Tue, May 18, 2004 at 04:32:38PM -0700, Larry McVoy wrote:
> > > > BK/Pro 3.2.0 has been released and is in the BK download area,
> > > > 
> > > >     http://bitmover.com/download
> > > 
> > > Any chance of a native x86-64 version? 
> > 
> > We don't have any x86-64 machines but we could get one.  I asked about this
> > a while back and people told me that there was no point, the x86 one worked
> > perfectly.  Can you tell me what having a native one would gain?  If there
> > is any gain we'll do it.
> 
>    It'll allow BK to be run on machines which have a pure 64-bit
> userspace (for example, Debian's current amd64 port), without having
> to resort to a 32-bit chroot to run the 32-bit BK binary.

OK, that's more than enough of a reason.  We'll get a native one in the next
week or so.
-- 
---
Larry McVoy                lm at bitmover.com           http://www.bitkeeper.com

Re: bk-3.2.0 released

[Larry McVoy]

>    It'll allow BK to be run on machines which have a pure 64-bit
> userspace (for example, Debian's current amd64 port), without having
> to resort to a 32-bit chroot to run the 32-bit BK binary.

OK, I found a reasonable looking box at 

    http://www.bzboyz.com/store/product4170.html

ASUS K8V basic, AMD 64 2800+, $346.  Gotta love hardware prices.

Next question is what distro you'd prefer to be the build OS.
We typically use Red Hat, in this case that would be Fedora Core 1.
Since that was one of the first (?) distros for x86-64 there shouldn't
be any libc issues, right?  What we want is to use the _oldest_ possible
distro/libc that works going forward.  So is Fedora 1 OK with you?  Any
nay sayers?
-- 
---
Larry McVoy                lm at bitmover.com           http://www.bitkeeper.com

Re: bk-3.2.0 released

[Bryan O'Sullivan]

On Sat, 2004-05-29 at 08:47, Larry McVoy wrote:

> So is Fedora 1 OK with you?  Any
> nay sayers?

Fedora Core 1 and SuSE 9.0 use roughly the same version of glibc (2.3.2,
with some divergent patches), and the two seem fairly compatible in that
regard based on a little testing I just did.

If you use any libstdc++ stuff, you'll want to build on SuSE 9.0, which
has an older libstdc++.

	<b

Re: bk-3.2.0 released

[David Lang]

actually I think that Suse / UnitedLinux were first (for what little
difference it makes :-)

what libraries do you need? we could have those of us who have x86_64
systems could report the versions we have on our distros.

David Lang

P.S. you may be surprised by the perfomance improvements. I recently
benchmarked apache as almost doubling it's performance when going from
32bit code to 64 bit code on the same hardware/distro/config file and
things I have read on the internet show similar improvements for database
software

On Sat, 29 May 2004, Larry McVoy wrote:

> Date: Sat, 29 May 2004 08:47:14 -0700
> From: Larry McVoy <lm@bitmover.com>
> To: Hugo Mills <hugo-lkml@carfax.org.uk>, Vojtech Pavlik <vojtech@suse.cz>,
>      bitkeeper-announce@work.bitmover.com, linux-kernel@vger.kernel.org
> Subject: Re: bk-3.2.0 released
>
> >    It'll allow BK to be run on machines which have a pure 64-bit
> > userspace (for example, Debian's current amd64 port), without having
> > to resort to a 32-bit chroot to run the 32-bit BK binary.
>
> OK, I found a reasonable looking box at
>
>     http://www.bzboyz.com/store/product4170.html
>
> ASUS K8V basic, AMD 64 2800+, $346.  Gotta love hardware prices.
>
> Next question is what distro you'd prefer to be the build OS.
> We typically use Red Hat, in this case that would be Fedora Core 1.
> Since that was one of the first (?) distros for x86-64 there shouldn't
> be any libc issues, right?  What we want is to use the _oldest_ possible
> distro/libc that works going forward.  So is Fedora 1 OK with you?  Any
> nay sayers?
> --
> ---
> Larry McVoy                lm at bitmover.com           http://www.bitkeeper.com
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
>

-- 
"Debugging is twice as hard as writing the code in the first place.
Therefore, if you write the code as cleverly as possible, you are,
by definition, not smart enough to debug it." - Brian W. Kernighan

Re: bk-3.2.0 released

[Vojtech Pavlik]

On Sat, May 29, 2004 at 06:04:36AM -0700, Larry McVoy wrote:
> On Sat, May 29, 2004 at 11:54:20AM +0200, Vojtech Pavlik wrote:
> > On Tue, May 18, 2004 at 04:32:38PM -0700, Larry McVoy wrote:
> > > BitKeeper Users,
> > > 
> > > BK/Pro 3.2.0 has been released and is in the BK download area,
> > > 
> > >     http://bitmover.com/download
> > 
> > Any chance of a native x86-64 version? 
> 
> We don't have any x86-64 machines but we could get one.  I asked about this
> a while back and people told me that there was no point, the x86 one worked
> perfectly.  Can you tell me what having a native one would gain?  If there
> is any gain we'll do it.

Well, yes, of course, the x86 version works just fine, because x86-64 is
backward compatible.

The benefits won't be huge:

	a marginal speed increase due to more registers
	no need for 32-bit libs on the system and in memory
	enough address space for bk, should it ever need more
	a general feeling of doing things right ;)	

-- 
Vojtech Pavlik
SuSE Labs, SuSE CR