SVN fetcher behaviour

SVN fetcher behaviour

[Martyn Welch]

Hello,

I have just tried to build "task-base" for machine=x86, distro=generic.

It is failing to download gconf-dbus using "gconf-dbus_svn.bb". It seems
that the script does not specify a revision and it therefore trying to
download revision 1. This fails.

I have managed to get the script to continue buy hacking the script to
get it to download the current revision (608), by adding "rev=608" to
the SRC_URI.

I've looked at
http://www.openembedded.org/user-manual&dpage=src_uri_variable, which
(under "SVN") describes the behaviour of the svn fetcher. This doesn't
seem to specify what the default behaviour of the svn fetcher is when no
revision is specified, I'd assumed it would be the "head", though I am
now having some doubt.

Martyn

----
Martyn Welch
Principal Software Engineer
 
Radstone Digital Processing
Part of GE Fanuc Embedded Systems
Tove Valley Business Park
Towcester
Northants
England
NN12 6PF

Telephone: +44 (0) 1327 359444
Direct Dial: +44 (0) 1327 322748
Fax: +44 (0) 1327 322800
email: martyn.welch@radstone.co.uk
web: www.radstone.com

This email is private and confidential and for the addressee only. If
misdirected, please notify us by telephone and confirm that it has been
deleted from your system and any hard copies destroyed. You are strictly
prohibited from using, printing, distributing or disseminating it or any
information contained in it save to the intended recipient.
 


________________________________________________________________________
This e-mail has been scanned for all viruses by Star.The service is powered by MessageLabs. 
________________________________________________________________________

Re: SVN fetcher behaviour

[Koen Kooi]

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

Martyn Welch schreef:
> Hello,
> 
> I have just tried to build "task-base" for machine=x86, distro=generic.

use DISTRO="angstrom-2007.1", generic has been broken for a long time.

regards,

Koen
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (Darwin)

iD8DBQFHBM4gMkyGM64RGpERAlmfAKCFLdUSkG2DQz16eUdxxbH1T42wKACcDd4k
hsxqQCOcuZViQ+kvtq80be8=
=PILz
-----END PGP SIGNATURE-----

Re: SVN fetcher behaviour

[Philippe De Swert]

Hello all,

On Thu, 2007-10-04 at 11:57 +0100, Martyn Welch wrote:
> It is failing to download gconf-dbus using "gconf-dbus_svn.bb". It seems
> that the script does not specify a revision and it therefore trying to
> download revision 1. This fails.
> 
> I have managed to get the script to continue buy hacking the script to
> get it to download the current revision (608), by adding "rev=608" to
> the SRC_URI.

I ran into a similar problem recently and it seems you need to edit
../org.openembedded.dev/conf/distro/include/sane-srcdates.inc or make
sure it is included in your distro.

> I've looked at
> http://www.openembedded.org/user-manual&dpage=src_uri_variable, which
> (under "SVN") describes the behaviour of the svn fetcher. This doesn't
> seem to specify what the default behaviour of the svn fetcher is when no
> revision is specified, I'd assumed it would be the "head", though I am
> now having some doubt.

I think this should be the default behaviour, unfortunately it is not.
The expected behaviour is fetch the latest revision and build it but it
dies not.  I think ib general it is a great improvement for managing svn
revisions, but it sucks at building the latest and newest.

Cheers,

Philippe

Re: SVN fetcher behaviour

[Hans Henry von Tresckow]

On 10/5/07, Philippe De Swert <philippedeswert@scarlet.be> wrote:
>
> Hello all,
>
> On Thu, 2007-10-04 at 11:57 +0100, Martyn Welch wrote:
> > It is failing to download gconf-dbus using "gconf-dbus_svn.bb". It seems
> > that the script does not specify a revision and it therefore trying to
> > download revision 1. This fails.
> >
> > I have managed to get the script to continue buy hacking the script to
> > get it to download the current revision (608), by adding "rev=608" to
> > the SRC_URI.
>
> I ran into a similar problem recently and it seems you need to edit
> ../org.openembedded.dev/conf/distro/include/sane-srcdates.inc or make
> sure it is included in your distro.
>
> > I've looked at
> > http://www.openembedded.org/user-manual&dpage=src_uri_variable, which
> > (under "SVN") describes the behaviour of the svn fetcher. This doesn't
> > seem to specify what the default behaviour of the svn fetcher is when no
> > revision is specified, I'd assumed it would be the "head", though I am
> > now having some doubt.
>
> I think this should be the default behaviour, unfortunately it is not.
> The expected behaviour is fetch the latest revision and build it but it
> dies not.  I think ib general it is a great improvement for managing svn
> revisions, but it sucks at building the latest and newest.
>
> Cheers,
>
> Philippe
>
>
>
> _______________________________________________
> Openembedded-devel mailing list
> Openembedded-devel@lists.openembedded.org
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
>
<http://bugs.openembedded.org/show_bug.cgi?id=2897>

The default of "1" is set in conf/bitbake.conf. The reason for this is that
any svn recipe without an explicit SRCREV set will cause parsing to fail if
the build machine or the svn server is off line. If you can live with this
limitation, then you could comment that line out to get the default
behavior.



-- 
Henry von Tresckow (hvontres)

Re: SVN fetcher behaviour

[Leon Woestenberg]

Hello,

On 10/5/07, Hans Henry von Tresckow <hvontres@gmail.com> wrote:
> The default of "1" is set in conf/bitbake.conf. The reason for this is that
> any svn recipe without an explicit SRCREV set will cause parsing to fail if
> the build machine or the svn server is off line. If you can live with this
> limitation, then you could comment that line out to get the default
> behavior.
>
But I think what was asked if a default can be set that (a) does not
break parsing when servers are unreachable; and (b) does fetch HEAD
instead of revision 1.

Regards,
-- 
Leon

Re: SVN fetcher behaviour

[Hans Henry von Tresckow]

On 10/5/07, Leon Woestenberg <leon.woestenberg@gmail.com> wrote:
>
> Hello,
>
> On 10/5/07, Hans Henry von Tresckow <hvontres@gmail.com> wrote:
> > The default of "1" is set in conf/bitbake.conf. The reason for this is
> that
> > any svn recipe without an explicit SRCREV set will cause parsing to fail
> if
> > the build machine or the svn server is off line. If you can live with
> this
> > limitation, then you could comment that line out to get the default
> > behavior.
> >
> But I think what was asked if a default can be set that (a) does not
> break parsing when servers are unreachable; and (b) does fetch HEAD
> instead of revision 1.
>
> Regards,
> --
> Leon
>
> _______________________________________________
> Openembedded-devel mailing list
> Openembedded-devel@lists.openembedded.org
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel
>

I did see an Entry called "AUTOREV" in bitbake.conf but I am not sure if
that does what you want either, since using that in a recepie might still
break parsing when off-line..


-- 
Henry von Tresckow (hvontres)

Re: SVN fetcher behaviour

[Richard Purdie]

On Fri, 2007-10-05 at 21:56 +0200, Leon Woestenberg wrote:
> On 10/5/07, Hans Henry von Tresckow <hvontres@gmail.com> wrote:
> > The default of "1" is set in conf/bitbake.conf. The reason for this is that
> > any svn recipe without an explicit SRCREV set will cause parsing to fail if
> > the build machine or the svn server is off line. If you can live with this
> > limitation, then you could comment that line out to get the default
> > behavior.
> >
> But I think what was asked if a default can be set that (a) does not
> break parsing when servers are unreachable; and (b) does fetch HEAD
> instead of revision 1.

The answer to this is that its not possible. We need to be able to
connect to the server to work out which version we're going to build and
set a sane PV value. Nothing else inbetween will work *properly* given
the way OE/Bitbake works.

Cheers,

Richard