| Submitter | Philippe De Swert |
|---|---|
| Date | July 21, 2011, 10:31 p.m. |
| Message ID | <1311287464-13001-1-git-send-email-philippedeswert@gmail.com> |
| Download | mbox | patch |
| Permalink | /patch/8257/ |
| State | New, archived |
| Headers | show |
Comments
Dear Philippe, thank you for your patch. Am Freitag, den 22.07.2011, 01:31 +0300 schrieb Philippe De Swert: > * SCREV points to a broken revision which crashes due to badly initialized g_threads > See http://bugs.openembedded.org/show_bug.cgi?id=5409 > > Signed-off-by: Philippe De Swert <philippedeswert@gmail.com> > --- > recipes/gpe-mini-browser/gpe-mini-browser2_svn.bb | 2 -- > 1 files changed, 0 insertions(+), 2 deletions(-) > > diff --git a/recipes/gpe-mini-browser/gpe-mini-browser2_svn.bb b/recipes/gpe-mini-browser/gpe-mini-browser2_svn.bb > index 6bc89c8..2aa55b4 100644 > --- a/recipes/gpe-mini-browser/gpe-mini-browser2_svn.bb > +++ b/recipes/gpe-mini-browser/gpe-mini-browser2_svn.bb > @@ -2,8 +2,6 @@ DEPENDS = "gtk+ glib-2.0 libgpewidget webkit-gtk" > PV = "0.0.1+svnr${SRCPV}" > PR = "r0" > > -SRCREV = "9900" > - Please do not remove the `SRCREV` but replace it by a valid and working revision. Otherwise the builds would not be reproducible. > inherit autotools > > SRC_URI = "${GPE_EXTRA_SVN}" Thanks, Paul
Hi Paul, On 22/07/2011, Paul Menzel <paulepanter@users.sourceforge.net> wrote: > thank you for your patch. Thank you for the review. > Please do not remove the `SRCREV` but replace it by a valid and working > revision. Otherwise the builds would not be reproducible. > I already wondered who added one (as I originally added the svn recipe). Anyway I thought the whole idea of _svn etc were to have bleeding edge builds. If we wanted stable there would be a versioned recipe. Of course I can add a new SRCREV but it defeats the whole purpose of the recipe... Cheers, Philippe
On Fri, Jul 22, 2011 at 03:18:37PM +0300, Philippe De Swert wrote: > Hi Paul, > > On 22/07/2011, Paul Menzel <paulepanter@users.sourceforge.net> wrote: > > thank you for your patch. > > Thank you for the review. > > > Please do not remove the `SRCREV` but replace it by a valid and working > > revision. Otherwise the builds would not be reproducible. > > > > I already wondered who added one (as I originally added the svn > recipe). Anyway I thought the whole idea of _svn etc were to have > bleeding edge builds. If we wanted stable there would be a versioned > recipe. If you want it bleeding then you or your distribution will set SRCREV_pn-gpe-mini-browser2 = "${AUTOREV}" in local.conf or some distro .inc file like there is conf/distro/include/shr-autorev.inc conf/distro/include/fso-autorev.inc etc. but for rest of world is better to say which SRCREV was tested and should work for everybody. > Of course I can add a new SRCREV but it defeats the whole purpose of > the recipe... Main purpose of _svn recipes is to allow build project without proper release or old release etc.. picking random bleeding edge revision and expect to get building image with it everytime I need to build newer images is not purpose but insanity :). Regards,
Hi, Thanks for clarifying Martin. On 22/07/2011, Martin Jansa <martin.jansa@gmail.com> wrote: > On Fri, Jul 22, 2011 at 03:18:37PM +0300, Philippe De Swert wrote: >> I already wondered who added one (as I originally added the svn >> recipe). Anyway I thought the whole idea of _svn etc were to have >> bleeding edge builds. If we wanted stable there would be a versioned >> recipe. > > If you want it bleeding then you or your distribution will set > > SRCREV_pn-gpe-mini-browser2 = "${AUTOREV}" > > in local.conf or some distro .inc file > like there is > conf/distro/include/shr-autorev.inc > conf/distro/include/fso-autorev.inc > etc. > > but for rest of world is better to say which SRCREV was tested and > should work for everybody. Ok so things have changed since I made the recipe. Still catching up on all the new things in OE. I'll make a new patch with a fixed up SRCREV then. Discard this one. Cheers, Philippe
On Fri, Jul 22, 2011 at 04:06:43PM +0300, Philippe De Swert wrote: > Hi, > > Thanks for clarifying Martin. > > On 22/07/2011, Martin Jansa <martin.jansa@gmail.com> wrote: > > On Fri, Jul 22, 2011 at 03:18:37PM +0300, Philippe De Swert wrote: > >> I already wondered who added one (as I originally added the svn > >> recipe). Anyway I thought the whole idea of _svn etc were to have > >> bleeding edge builds. If we wanted stable there would be a versioned > >> recipe. > > > > If you want it bleeding then you or your distribution will set > > > > SRCREV_pn-gpe-mini-browser2 = "${AUTOREV}" > > > > in local.conf or some distro .inc file > > like there is > > conf/distro/include/shr-autorev.inc > > conf/distro/include/fso-autorev.inc > > etc. > > > > but for rest of world is better to say which SRCREV was tested and > > should work for everybody. > > Ok so things have changed since I made the recipe. Still catching up > on all the new things in OE. BTW: let's go to catch some oe-core :) Looking on that commit it was me changing SRCREV in recipe gpe-mini-browser2_svn -SRCREV = "9819" +SRCREV = "9900" because there was -SRCREV_pn-gpe-mini-browser2 ?= "9900" in conf/distro/include/sane-srcrevs.inc which was included by default and removed later after moving all SRCREVs to corresponding recipes. and before that it was converted from SRCDATE to SRCREV (+ in recipe SRCREV was added) http://git.openembedded.org/cgit.cgi/openembedded/commit/?id=e7814af646745ab8c6884c817d00475be3cd94e1 later SRCREV upgraded to 9900 by http://git.openembedded.org/cgit.cgi/openembedded/commit/?id=e7814af646745ab8c6884c817d00475be3cd94e1
Patch
diff --git a/recipes/gpe-mini-browser/gpe-mini-browser2_svn.bb b/recipes/gpe-mini-browser/gpe-mini-browser2_svn.bb index 6bc89c8..2aa55b4 100644 --- a/recipes/gpe-mini-browser/gpe-mini-browser2_svn.bb +++ b/recipes/gpe-mini-browser/gpe-mini-browser2_svn.bb @@ -2,8 +2,6 @@ DEPENDS = "gtk+ glib-2.0 libgpewidget webkit-gtk" PV = "0.0.1+svnr${SRCPV}" PR = "r0" -SRCREV = "9900" - inherit autotools SRC_URI = "${GPE_EXTRA_SVN}"
* SCREV points to a broken revision which crashes due to badly initialized g_threads See http://bugs.openembedded.org/show_bug.cgi?id=5409 Signed-off-by: Philippe De Swert <philippedeswert@gmail.com> --- recipes/gpe-mini-browser/gpe-mini-browser2_svn.bb | 2 -- 1 files changed, 0 insertions(+), 2 deletions(-)