Nabble - Gnu - Arch

tla: file descriptor leak, causes panic when removing temporary directory

2009-03-31T09:31:39Z

Package: tla
Version: 1.3.5+dfsg-14
Tags: patch

When computing a changeset (e.g. for the "tla changes --diffs" command), tla
creates a temporary directory with a name that begins in ,,what-changed.
This directory is cleaned up at the end of the operation.

Symbolic links are compared by writing the target of the link into a
text file under that directory, then relying on usual "diff" behaviour.

Unfortunately, tla doesn't explicitly close this text file. As a result:
(1) it uses more file descriptors than necessary;
(2) on some filesystems (e.g., OpenAFS 1.4.8 and newer) tla dies with
an I/O panic as the rmdir() operation on the parent directory fails with
ENOTEMPTY, even though the unlink() of the file itself succeeded, because
the file is still open.

The attached patch has been tested and gets rid of symptom (2). I expect
it to help with (1) as well. The variables that hold the file descriptor
numbers go out of scope shortly afterwards.

#! /bin/sh /usr/share/dpatch/dpatch-run
## 07-changeset-fd-leak.dpatch by Sergio Gelato <Sergio.Gelato@...>
##
## All lines beginning with `## DP:' are a description of the patch.
## DP: Remember to close file descriptors before they go out of scope.

@DPATCH@
diff -urNad tla-1.3.5+dfsg-14~/src/tla/libarch/make-changeset.c tla-1.3.5+dfsg-14/src/tla/libarch/make-changeset.c
--- tla-1.3.5+dfsg-14~/src/tla/libarch/make-changeset.c 2006-07-20 08:34:40.000000000 +0200
+++ tla-1.3.5+dfsg-14/src/tla/libarch/make-changeset.c 2009-03-31 00:19:39.449388785 +0200
@@ -1485,6 +1485,9 @@
safe_printfmt (orig_out_fd, "%s\n", orig_target);
safe_printfmt (mod_out_fd, "%s\n", mod_target);

+ safe_close (orig_out_fd);
+ safe_close (mod_out_fd);
+
lim_free (0, patch_basename_path);
lim_free (0, orig_patch);
lim_free (0, mod_patch);

_______________________________________________
Gnu-arch-users mailing list
Gnu-arch-users@...
http://lists.gnu.org/mailman/listinfo/gnu-arch-users

GNU arch home page:
http://savannah.gnu.org/projects/gnu-arch/

From forum: Gnu - Arch - Users

Re: please look into gittorrent

2008-12-05T01:07:28Z

Hi,

Am Freitag, 5. Dezember 2008 schrieb Thomas Lord:
>
> Arch does better than git in taxonimizing versioned objects and in its
> management of coding history, branching, merging, etc.... but Arch
> shares with gittorrent this idea of distributed, decentralized revision
> control -- free software source code should be just sort of "floating"
> on a meta-net, on the Internet, over a P2P layer -- just so. This is a
> political goal because of the "decentralization" part.

I tried something like that back in the days of "larch".

http://www.unix-ag.uni-kl.de/~conrad/Archives/DSDiF/index.html

Basically, the idea was to host larch archives within freenet.
The proof-of-concept worked, but then tla was born, and somehow
the project fell asleep...

The only thing that's really needed for hosting arch archives
in freenet is an efficient way to mirror local archives into
freenet and vice versa.

Bye,
Peter
--
Peter Conrad Tel: +49 6102 / 80 99 072
[ t]ivano Software GmbH Fax: +49 6102 / 80 99 071
Bahnhofstr. 18 http://www.tivano.de/
63263 Neu-Isenburg

Germany

_______________________________________________
Gnu-arch-users mailing list
Gnu-arch-users@...
http://lists.gnu.org/mailman/listinfo/gnu-arch-users

GNU arch home page:
http://savannah.gnu.org/projects/gnu-arch/

From forum: Gnu - Arch - Users

Re: please look into gittorrent

2008-12-04T19:36:51Z

lkcl's overview and related:
GitTorrent, The Movie
http://www.advogato.org/article/994.html

On Thu, Dec 04, 2008 at 04:50:41PM -0800, Andy Tai wrote:

> http://code.google.com/p/gittorrent/
>
> On Thu, Dec 4, 2008 at 4:11 PM, Thomas Lord <lord@...> wrote:
>
> > Please look into gittorrent.
> >
>
>
> --
> Andy Tai, atai@...

> _______________________________________________
> Gnu-arch-users mailing list
> Gnu-arch-users@...
> http://lists.gnu.org/mailman/listinfo/gnu-arch-users
>
> GNU arch home page:
> http://savannah.gnu.org/projects/gnu-arch/

_______________________________________________
Gnu-arch-users mailing list
Gnu-arch-users@...
http://lists.gnu.org/mailman/listinfo/gnu-arch-users

GNU arch home page:
http://savannah.gnu.org/projects/gnu-arch/

From forum: Gnu - Arch - Users

Re: please look into gittorrent

2008-12-04T16:50:41Z

http://code.google.com/p/gittorrent/

On Thu, Dec 4, 2008 at 4:11 PM, Thomas Lord <lord@...> wrote:

Please look into gittorrent.

--
Andy Tai, atai@...

_______________________________________________
Gnu-arch-users mailing list
Gnu-arch-users@...
http://lists.gnu.org/mailman/listinfo/gnu-arch-users

GNU arch home page:
http://savannah.gnu.org/projects/gnu-arch/

From forum: Gnu - Arch - Users

please look into gittorrent

2008-12-04T16:11:44Z

Please look into gittorrent. While there are still important (at least
to my mind) gaps between git and Arch (things missing from the former
that are at least pointed out by the latter) nevertheless gittorrent
looks initially like an important development that helps narrow the gap.
I wanted to go someplace similar with GNU Arch.

Arch does better than git in taxonimizing versioned objects and in its
management of coding history, branching, merging, etc.... but Arch
shares with gittorrent this idea of distributed, decentralized revision
control -- free software source code should be just sort of "floating"
on a meta-net, on the Internet, over a P2P layer -- just so. This is a
political goal because of the "decentralization" part.

We also have a lot of work to do on the economics of the emerging
ecosystem of free software source code and because economics "wants" to
be "transactional" -- and because of the nature of the natural unit of a
"transaction" in software source code development and support -- the
economics of how free software can be a career and the formality imposed
by a global-scale, distributed, decentralized revision control system
are closely intertwined. Issuing a "commit" command should be an
economically significant act -- a "transaction" in both senses of the
word.

But that's a larger, future topic, for now. For now: please do look
into gittorrent and share your impressions. Have they actually made
progress on distributed, decentralized revision control? Are their
politics in order?

Thanks,
-t

_______________________________________________
Gnu-arch-users mailing list
Gnu-arch-users@...
http://lists.gnu.org/mailman/listinfo/gnu-arch-users

GNU arch home page:
http://savannah.gnu.org/projects/gnu-arch/

From forum: Gnu - Arch - Users

Re: Inclusion of Bzr into the GNU system

2008-11-26T00:14:45Z

On Wed, 2008-11-26 at 13:18 +0900, Stephen J. Turnbull wrote:

> So there's no longer room for proof-of-concept implementations, and
> that's all that Arch ever was to Tom Lord, at least that's what he
> said when he was busking for (financial) contributions. He even
> regrets many of the UI and feature concessions he made to the crew of
> developers who later became the nucleus of the Bazaar project. Making
> Arch into a contender again will require a genius (or, to be specific,
> Tom). But Tom, too, has moved on I think. While there are many ideas
> in Arch that haven't made it into other VCSes even today, the
> value-add to implementing them in Yet Another DVCS probably isn't that
> high.

I have not "moved on" I've been slain. I'm
left with no career or career prospects, no
savings, no resources to invest in much by way
of software development and a damaged reputation
as I enter my 44th year. 'Tis not so deep as a
well nor wide as a church door but 'tis enough;
'twill serve.

I spend a decent amount of time, these days,
observing a couple of trees and some birds
that live nearby. Oh, and, today I saw a cat
do something surprising. This appears to be about
what's left for me.

Were it otherwise I might say:

I don't care about the "GNU project" per se,
anymore, because I don't think that there is
any such project other than in name only. There
is no coherently expressed organizing set of goals.
There is no true strategy. There is no project
there, no matter what it's called. I'd argue
that there once was a GNU project and that it
was killed deliberately by Cygnus and Cygnus'
friends although I must also give due credit
to RMS for folding like a house of cards under
their pressure.

I do care about the progress of software in
society. I do think software freedom is important.
I do think there is a social policy need for
something worthy of the name "GNU project" but
as I say: no such thing exists. It got killed
and I would say it got killed to make way for
the Open Source Industrial Complex.

Were there a GNU project I think there is much
from the Arch project that would be worth
contemplating. For example, much in Arch is
applicable to the challenge of developing a
distributed, decentralized, transactional file
system and I would also argue that such a bit of
technology would help considerably to promote
software freedom. But there is no GNU project or
anything like it and so why go into such matters?

I think that one thing that was and remains
under-appreciated about Arch is that it was
an attack on the business models of the GNU/Linux
vendors and the "large, well funded, famous
projects". It was a technological attack
on the necessity of those firms in their present
form.

We, as a generally free-wheeling, crazy chaotic,
catch-as-catch can community of free software
developers *can* -- *without painful effort* --
displace the need for big, centralized, lock-in
GNU/Linux vendors and create stable distributions
and support that is *more reliable* than the current
vendors. We can do all that and capture their
revenue streams into a process that democratically
distributes the money among contributors. We
can do all of that in a decentralized way so that
the arising of those replacement products is
an emergent property of our community practices --
that is, we can "distribute control".

Arch was *by design* a first step in that direction
and so *naturally* it "had" to be rudely treated
by capital. Had Arch succeeded, Canonical, Red Hat,
Linus, Collabnet, et al. would all have had to radically
change business models sooner rather than later.

And so I am slain....

-t

_______________________________________________
Gnu-arch-users mailing list
Gnu-arch-users@...
http://lists.gnu.org/mailman/listinfo/gnu-arch-users

GNU arch home page:
http://savannah.gnu.org/projects/gnu-arch/

From forum: Gnu - Arch - Users

Inclusion of Bzr into the GNU system

2008-11-25T20:15:27Z

deadlyhead writes:

> But Python, while ostensibly free software,

Even RMS admits that it is strictly freer than GPL software, but based
on the long-term interests of the community he decided that GNU policy
should be oriented to *freedom-preserving* software, and not merely
free software. For that reason he defined "software freedom" in such
a way as to allow restrictive licenses like the GPL, and the GPL is
(at least to a pretty good approximation) the most restrictive license
that still qualifies as a free software license (even the GFDL does
not, although some instances of the GFDL do).

> has a rather weak license and in the past its community has shown a
> bit of hostility toward maintaining the Four Freedoms. This
> greatly troubles me for a project that has been accepted as part of
> the GNU system.

I guess you refuse to use X11, Perl, TeX, Apache, OpenSSH, and the BSD
shell utilities, then? The GNU System has a long history of
incorporating free software even though it it not copyleft.

The license is easy enough to fix. You can create a GNU Python
project which simply GPLs each release of Python as it comes out (I
told you Python's license was freer than the GPL, and this is an
important example of that freedom -- go ahead and try the reverse on
FSF-owned software if you don't believe me), and test whether Bazaar
works with the GPLed version. If not, you complain to the Bazaar
devs.

Also, it's a fundamental principle of copyleft theory that you
separate the platform from the licensed work. If people want to port
Emacs to Windows, this is not a problem for the Emacs project as long
as it doesn't provide features that make using Emacs on Windows more
attractive than using Emacs on free platforms. The rationale is that
Emacs becomes no less free, and all else equal Emacs users on Windows
will find it easier to migrate to free platforms because they don't
have to worry about learning a new editor.

For those reasons, I rather doubt that the discussion of admitting
Bazaar as a GNU project (which AFAIK is kept private and simply
consists of getting the developers to sign a document saying they
adhere to GNU goals) considered whether Python was GNUish or not.

It certainly did not come up at all in the discussion of a DVCS for
Emacs, which RMS summarily terminated by deciding on Bazaar because it
was an active GNU project. This, despite the presence of advocates of
GNU Arch---who actually have maintained a working Arch mirror of Emacs
for years. Even Tom Lord, who participated in that discussion, was
unable to effectively argue for Arch.

> This is just my perspective. I'd like to see what others have to
> say for the inclusion of Bazaar as a GNU project, how that sits
> with the GNU Arch community, its benefits and detriments, and what
> that means for the future of Arch.

Sad to say, I don't think Arch has a future beyond maintaining
existing projects that use Arch. Distributed revision control has
moved on. Git provides performance and the most intuitive database
schema, Mercurial gives almost as good performance and a convenient
UI, Darcs is patch-oriented, has some nice UI features, and
automatically computes optimal merge strategies for you, and Bazaar is
a GNU project (acceptable to RMS as long-term host for his first-born
child) for freedom-lovers.

So there's no longer room for proof-of-concept implementations, and
that's all that Arch ever was to Tom Lord, at least that's what he
said when he was busking for (financial) contributions. He even
regrets many of the UI and feature concessions he made to the crew of
developers who later became the nucleus of the Bazaar project. Making
Arch into a contender again will require a genius (or, to be specific,
Tom). But Tom, too, has moved on I think. While there are many ideas
in Arch that haven't made it into other VCSes even today, the
value-add to implementing them in Yet Another DVCS probably isn't that
high.

_______________________________________________
Gnu-arch-users mailing list
Gnu-arch-users@...
http://lists.gnu.org/mailman/listinfo/gnu-arch-users

GNU arch home page:
http://savannah.gnu.org/projects/gnu-arch/

From forum: Gnu - Arch - Users

Re: Inclusion of Bzr into the GNU system

2008-11-25T19:26:21Z

> This is just my perspective. I'd like to see what others have to say for
> the inclusion of Bazaar as a GNU project, how that sits with the GNU Arch

The fact that Bzr and Arch cover basically the same needs doesn't mean
they can't both be GNU projects. There is very little competition in
the "race to be accepted as GNU project".
So to a large extent, the two are mostly unrelated. If anything, Bzr's
acceptance as a GNU package might be remotly linked to Arch's lack of
development activity, but that's about it.

Stefan

_______________________________________________
Gnu-arch-users mailing list
Gnu-arch-users@...
http://lists.gnu.org/mailman/listinfo/gnu-arch-users

GNU arch home page:
http://savannah.gnu.org/projects/gnu-arch/

From forum: Gnu - Arch - Users

Re: Inclusion of Bzr into the GNU system

2008-11-25T15:50:35Z

Hi, just some general comments:

GNU Arch is not in active development, nor maintenance... (the latter my fault; the former, don't blame me).

Software's value depends on many factors; resources behind active development are one of them... in that regard bzr beats Arch hands down. There may be technical viewpoints favoring one or the other, but bzr also has one more thing superior to Arch: its (command line) user interface is much better due to the efforts spent on the design.

Python is free software; and how the Python inventor thinks of software freedom is not relevant because that is not connected to the developers of bzr. Not a reason to use or not to use Python. (I do not use Python myself)

Bzr is a GNU project; the issue of whether GNU should have accepted it is outside the scope of the GNU Arch community; frankly.

Arch is still a GNU project and contribution is welcome; and I welcome a better person to come along to take over the maintainership if such a hero exists.

Andy

On Tue, Nov 25, 2008 at 3:11 PM, deadlyhead <deadlyhead@...> wrote:

I've been sitting on this for a long time, but it's been bothering me a bit and I need to get some others' perspectives on it.

I'm a fan of GNU Arch. Seriously. I'm not a major software author at all, but I do have need of a revision control system, and of all that I've used, GNU Arch is the one that I really feel fits how I work. The fact that it is distributed, that it uses forward patching, that it has a sane, usable interface and more make it ideally suited to my needs. Most of all, it's a GNU project, and thus I am ensured that in using it I retain my freedom.

The fact that GNU Arch has a smaller user base than some revision systems has always been a bit of a downer to me, but has not deterred my use of it in my own projects. When Bazaar forked, then rewrote, GNU Arch, I was fairly unimpressed, because potential users of Arch were then pulled over to Bzr. I can't fault either the Bazaar team nor the users; this is a freedom protected by free software. I would have preferred to see those same efforts make there way into GNU Arch (though I have read of the reasons why they weren't), but what really galls me is the means of developing Bazaar and the subsequent product.

Bazaar is written in Python. I've used Python, and though I personally don't prefer it (I'm a Guile user), I see its merits and why so many programmers are drawn to it. But Python, while ostensibly free software, has a rather weak license and in the past its community has shown a bit of hostility toward maintaining the Four Freedoms. This greatly troubles me for a project that has been accepted as part of the GNU system. I would like to have seen the deliberations over accepting Bazaar as a GNU project and know whether this issue was brought up.

My personal wish would be to see the continued development of GNU Arch (1.x or 2, either works) as a GNU project. I feel that these programs are much more in line with GNU philosophy and thus a better choice for the system. I realize that the ``best'' way to go about this is to take an active role in the development of either piece of software, and I'd love to, though I feel barely confident in my abilities to do the software justice. (I'm a passable C programmer at best. Along with being an avid Guile user, my interests lie squarely with Ada programming and GNAT. Talk about marginalizing my own project acceptance!) I feel that it would be worthwhile to use all GNU-supported languages for GNU projects, and GNU Arch extended with Guile would be a worthy showpiece for GNU's extension language as well. Maybe I'm just a fanboy, but using GNU's tools when working on a GNU project seems _right_ in so many ways.

This is just my perspective. I'd like to see what others have to say for the inclusion of Bazaar as a GNU project, how that sits with the GNU Arch community, its benefits and detriments, and what that means for the future of Arch. I'm inclined to get involved with Arch if it's active in any way still (but don't think I could push start it if it's stalled), but due to my reservations with Bazaar, involvement in it, let alone use of it, is unlikely. (It doesn't even work on my Debian box! WTFpython!?)

Thanks for reading.

-- deadlyhead

_______________________________________________
Gnu-arch-users mailing list
Gnu-arch-users@...
http://lists.gnu.org/mailman/listinfo/gnu-arch-users

GNU arch home page:
http://savannah.gnu.org/projects/gnu-arch/

From forum: Gnu - Arch - Users

Inclusion of Bzr into the GNU system

2008-11-25T15:11:18Z

I've been sitting on this for a long time, but it's been bothering me a
bit and I need to get some others' perspectives on it.

I'm a fan of GNU Arch. Seriously. I'm not a major software author at
all, but I do have need of a revision control system, and of all that
I've used, GNU Arch is the one that I really feel fits how I work. The
fact that it is distributed, that it uses forward patching, that it has
a sane, usable interface and more make it ideally suited to my needs.
Most of all, it's a GNU project, and thus I am ensured that in using it
I retain my freedom.

The fact that GNU Arch has a smaller user base than some revision
systems has always been a bit of a downer to me, but has not deterred my
use of it in my own projects. When Bazaar forked, then rewrote, GNU
Arch, I was fairly unimpressed, because potential users of Arch were
then pulled over to Bzr. I can't fault either the Bazaar team nor the
users; this is a freedom protected by free software. I would have
preferred to see those same efforts make there way into GNU Arch (though
I have read of the reasons why they weren't), but what really galls me
is the means of developing Bazaar and the subsequent product.

Bazaar is written in Python. I've used Python, and though I personally
don't prefer it (I'm a Guile user), I see its merits and why so many
programmers are drawn to it. But Python, while ostensibly free
software, has a rather weak license and in the past its community has
shown a bit of hostility toward maintaining the Four Freedoms. This
greatly troubles me for a project that has been accepted as part of the
GNU system. I would like to have seen the deliberations over accepting
Bazaar as a GNU project and know whether this issue was brought up.

My personal wish would be to see the continued development of GNU Arch
(1.x or 2, either works) as a GNU project. I feel that these programs
are much more in line with GNU philosophy and thus a better choice for
the system. I realize that the ``best'' way to go about this is to take
an active role in the development of either piece of software, and I'd
love to, though I feel barely confident in my abilities to do the
software justice. (I'm a passable C programmer at best. Along with
being an avid Guile user, my interests lie squarely with Ada programming
and GNAT. Talk about marginalizing my own project acceptance!) I feel
that it would be worthwhile to use all GNU-supported languages for GNU
projects, and GNU Arch extended with Guile would be a worthy showpiece
for GNU's extension language as well. Maybe I'm just a fanboy, but
using GNU's tools when working on a GNU project seems _right_ in so many
ways.

This is just my perspective. I'd like to see what others have to say
for the inclusion of Bazaar as a GNU project, how that sits with the GNU
Arch community, its benefits and detriments, and what that means for the
future of Arch. I'm inclined to get involved with Arch if it's active
in any way still (but don't think I could push start it if it's
stalled), but due to my reservations with Bazaar, involvement in it,
let alone use of it, is unlikely. (It doesn't even work on my Debian
box! WTFpython!?)

Thanks for reading.

-- deadlyhead

_______________________________________________
Gnu-arch-users mailing list
Gnu-arch-users@...
http://lists.gnu.org/mailman/listinfo/gnu-arch-users

GNU arch home page:
http://savannah.gnu.org/projects/gnu-arch/

From forum: Gnu - Arch - Users

[bug #11964] Bug in tla undo with file ids in {arch} directory

2008-09-13T07:44:28Z

Follow-up Comment #1, bug #11964 (project gnu-arch):

All artists by Rock | MP3 Download | mp3ultima.com
http://mp3ultima.com/genres/genre332/Rock/

_______________________________________________________

Reply to this item at:

<http://savannah.gnu.org/bugs/?11964>

_______________________________________________
Message sent via/by Savannah
http://savannah.gnu.org/

_______________________________________________
Bug-gnu-arch mailing list
Bug-gnu-arch@...
http://lists.gnu.org/mailman/listinfo/bug-gnu-arch

Nabble - Gnu - Arch

tla: file descriptor leak, causes panic when removing temporary directory

Re: please look into gittorrent

Re: please look into gittorrent

Re: please look into gittorrent

please look into gittorrent

Re: Inclusion of Bzr into the GNU system

Inclusion of Bzr into the GNU system

Re: Inclusion of Bzr into the GNU system

Re: Inclusion of Bzr into the GNU system

Inclusion of Bzr into the GNU system

[bug #11964] Bug in tla undo with file ids in {arch} directory

Re: botched invariant pest

botched invariant pest

[bug #23931] Implicit pointer convention breaks binaries on many architectures

[bug #23931] Implicit pointer convention breaks binaries on many architectures

Re: GNU Arch intro in russian

GNU Arch intro in russian

[bug #17626] tla update deletes directories and files

[bug #7101] solaris build changes (package-framework)

[bug #12751] stdarg missuse breaks on sparc-sun-solaris

[bug #10145] Directory named core should not violate default naming conventions

[bug #12751] stdarg missuse breaks on sparc-sun-solaris

[bug #15915] MACHINE_ALIGNMENT miscalculated on PPC32

[bug #12751] stdarg missuse breaks on sparc-sun-solaris

[bug #7101] solaris build changes (package-framework)

[bug #14739] tla crashes with some options

[bug #7853] Test framework assumes 'tar' is GNU tar

Fwd: Bazaar becomes a GNU project

[bug #13261] double tag on FC3

[bug #7853] Test framework assumes 'tar' is GNU tar

[bug #14739] tla crashes with some options

[bug #12566] tla tree-lint report false "These explicit ids have no corresponding file"

[bug #5251] |tla undo -o mychanges| barfs badly

rev. ctl and file systems

Re: revc