[Pharo-project] download scripts for the new infrastructure
Hi, we need to modify (or create new) http://pharo.gforge.inria.fr/ci/ciPharo20.sh and related scripts to use the new infrastructure. Has someone rights and time to do that? Cheers, -- Pavel
On 2012-11-14, at 14:32, Pavel Krivanek <pavel.krivanek@gmail.com> wrote:
Hi,
we need to modify (or create new) http://pharo.gforge.inria.fr/ci/ciPharo20.sh and related scripts to use the new infrastructure. Has someone rights and time to do that?
the ci*.sh are already self contained and only download stuff from the files statically served under http://pharo.gforge.inria.fr/ci/* so the scripts should be fine and contain no references to the build server. So the infrastructure (will go) goes like this (comments and improvements welcome!) http://ci.inria.fr/pharo - Pharo.image => push results to http://pharo.gforge.inria.fr/ci/image/ - CogVM / NBCogVM => push results http://pharo.gforge.inria.fr/ci/vm/ all other build scripts should be in the form: ---------------------------------------------------------------------------------- wget --quiet -qO - http://pharo.gforge.inria.fr/ci/ciPharo20Cog.sh | sh ./vm.sh Pharo.image save $JOB_NAME --delete-old REPO=http://smalltalkhub.com/mc/dh83/nabujito/main ./vm.sh $JOB_NAME.image config $REPO ConfigurationOfNabujito --install=development ./vm.sh $JOB_NAME.image test --junit-xml-output "Nabujito-.*" zip -r $JOB_NAME.zip $JOB_NAME.image $JOB_NAME.changes ---------------------------------------------------------------------------------- GOAL: - 1 single bash scripts with contents similar to the example above - No additional build.sh / install.st scripts - No more Jenkins dependencies (copy scripts, getting VM, getting Image) - Always use the latest VMs / Images provided through http://pharo.gforge.inria.fr/ci/ciPharo20Cog.sh and Co best cami
It would be fine to have an option in ConfigurationCommandLineHandler to print job [version] for CI build description. Pharo version and Configuration version. Options for selection of VM and Pharo versions would be fine too in such script. For example if you want run the job on StackVM on Pharo 1.4 or Pharo Kernel etc. -- Pavel On Thu, Nov 15, 2012 at 3:11 PM, Camillo Bruni <camillobruni@gmail.com> wrote:
On 2012-11-14, at 14:32, Pavel Krivanek <pavel.krivanek@gmail.com> wrote:
Hi,
we need to modify (or create new) http://pharo.gforge.inria.fr/ci/ciPharo20.sh and related scripts to use the new infrastructure. Has someone rights and time to do that?
the ci*.sh are already self contained and only download stuff from the files statically served under http://pharo.gforge.inria.fr/ci/*
so the scripts should be fine and contain no references to the build server.
So the infrastructure (will go) goes like this (comments and improvements welcome!)
http://ci.inria.fr/pharo - Pharo.image => push results to http://pharo.gforge.inria.fr/ci/image/ - CogVM / NBCogVM => push results http://pharo.gforge.inria.fr/ci/vm/
all other build scripts should be in the form: ---------------------------------------------------------------------------------- wget --quiet -qO - http://pharo.gforge.inria.fr/ci/ciPharo20Cog.sh | sh
./vm.sh Pharo.image save $JOB_NAME --delete-old
REPO=http://smalltalkhub.com/mc/dh83/nabujito/main ./vm.sh $JOB_NAME.image config $REPO ConfigurationOfNabujito --install=development ./vm.sh $JOB_NAME.image test --junit-xml-output "Nabujito-.*"
zip -r $JOB_NAME.zip $JOB_NAME.image $JOB_NAME.changes ----------------------------------------------------------------------------------
GOAL: - 1 single bash scripts with contents similar to the example above - No additional build.sh / install.st scripts - No more Jenkins dependencies (copy scripts, getting VM, getting Image) - Always use the latest VMs / Images provided through http://pharo.gforge.inria.fr/ci/ciPharo20Cog.sh and Co
best cami
On 15 Nov 2012 16:33, "Pavel Krivanek" <pavel.krivanek@gmail.com> wrote:
It would be fine to have an option in ConfigurationCommandLineHandler to print job [version] for CI build description. Pharo version and Configuration version.
Yes, that's definitely missing, the last step to create the numbered pharo zip files.
Options for selection of VM and Pharo versions would be fine too in such script. For example if you want run the job on StackVM on Pharo 1.4 or Pharo Kernel etc.
I will some more separate scripts to fetch vms by name. I just want to keep the scripts dead simple, because after 10 years of programming I still haver no clue about bash ;)
-- Pavel
On Thu, Nov 15, 2012 at 3:11 PM, Camillo Bruni <camillobruni@gmail.com> wrote:
On 2012-11-14, at 14:32, Pavel Krivanek <pavel.krivanek@gmail.com>
wrote:
Hi,
we need to modify (or create new) http://pharo.gforge.inria.fr/ci/ciPharo20.sh and related scripts to use the new infrastructure. Has someone rights and time to do that?
the ci*.sh are already self contained and only download stuff from the files statically served under http://pharo.gforge.inria.fr/ci/*
so the scripts should be fine and contain no references to the build
server.
So the infrastructure (will go) goes like this (comments and
improvements welcome!)
http://ci.inria.fr/pharo - Pharo.image => push results to
http://pharo.gforge.inria.fr/ci/image/
- CogVM / NBCogVM => push results http://pharo.gforge.inria.fr/ci/vm/
all other build scripts should be in the form:
wget --quiet -qO - http://pharo.gforge.inria.fr/ci/ciPharo20Cog.sh | sh
./vm.sh Pharo.image save $JOB_NAME --delete-old
REPO=http://smalltalkhub.com/mc/dh83/nabujito/main ./vm.sh $JOB_NAME.image config $REPO ConfigurationOfNabujito --install=development ./vm.sh $JOB_NAME.image test --junit-xml-output "Nabujito-.*"
zip -r $JOB_NAME.zip $JOB_NAME.image $JOB_NAME.changes
GOAL: - 1 single bash scripts with contents similar to the example above - No additional build.sh / install.st scripts - No more Jenkins dependencies (copy scripts, getting VM, getting Image) - Always use the latest VMs / Images provided through
http://pharo.gforge.inria.fr/ci/ciPharo20Cog.sh and Co
best cami
participants (3)
-
Bruni camillo -
Camillo Bruni
-
Pavel Krivanek