About tagging... It looks super ackward to me to tag EVERY commit with "build information". Building does not mean releasing... I prefer the other way around: we tag builds with commit information. Like that we know how to reproduce the build.
I'm ok with tagging build artifacts by explicitly saying "this is BUILD #X", which is ~= from "this is RELEASE #X". Moreover, before we needed to store ALL versions of the image because that was the way we versionned the system. Now, every commit in the #development branch should be buildable (delta some bugs of course ;)). This means that we can rebuild all images by just iterating the git history and building from scratch. No need to store all images. On Sun, Aug 6, 2017 at 9:40 AM, Guillermo Polito <guillermopol...@gmail.com> wrote: > I made a PR > > https://github.com/pharo-project/pharo/pull/185 > > this PR adds a script that will rename built archives accordingly. I > propose the following file names for the zip: > > Pharo${IMAGE_KIND}-7.0.0-arch.32bit.sha.${HASH}.build.${BUILD_NUMBER}.zip > > Where: > > IMAGE_KIND is the built product (core image, image with monticello, image > with metacello, full image...) > HASH is the commit hash > BUILD_NUMBER is the build number set by jenkins, or "nobuildnumber" if not > available > > > Moreover, this only makes part of a "prepare for upload" script and not > the main build process. That is because of simplicity right now. However, > if we put the build number by default in the main archives, there may be > several drawbacks: > > - PR's builds will have build numbers conflicting with the main builds > (which may be confusing) > - when building locally, we do not necessarily have a build number > available so it does not make much sense > > Once the PR is accepted/integrated, we should change the build process to > use this script. > > Guille > > > On Sun, Aug 6, 2017 at 9:05 AM, Guillermo Polito < > guillermopol...@gmail.com> wrote: > >> Hi, I created an issue: >> >> https://pharo.fogbugz.com/f/cases/20290/Add-build-number-to- >> uploaded-files >> >> I propose to keep both the sha and add the build number. Also, to follow >> semantic version conventions, we should use $- instead of $/. Something >> like: >> >> {Major}.{Minor}.{Patch}-sha.{sha}.build.{buildnumber} >> >> What do you think? I'll propose a pull request in some minutes. >> >> On Sat, Aug 5, 2017 at 6:17 PM, Stephane Ducasse <stepharo.s...@gmail.com >> > wrote: >> >>> Hi torsten >>> >>> Yes it sounds good to have MAJOR.MINOR.PATCH/BUILDNUMBER scheme >>> >>> On Fri, Aug 4, 2017 at 9:22 AM, Torsten Bergmann <asta...@gmx.de> wrote: >>> > Hi Ben, >>> > >>> > reason is that for Pharo 7 currently an sha git hash is used in the >>> file >>> > name >>> > instead of a (more clear) build number. >>> > >>> > See http://files.pharo.org/image/70/ >>> > >>> > This problem (which has more side effects on different sides, not only >>> the >>> > Launcher now) was discussed >>> > already yesterday on Discord #iceberg channel with Esteban and Pavel. >>> > >>> > The current sha based image file name scheme is not only confusing but >>> has >>> > some downsides. >>> > One can not easily remember the SHA or see which image is the latest, >>> or >>> > sort from recent >>> > images to older in a folder. >>> > >>> > If I understood correctly the reason to (initially) choose sha's in the >>> > image name has something >>> > to do with Travis and a discussion between Pavel, Esteban and Guille. >>> > >>> > I would vote for using Build numbers again. >>> > >>> > We would have several BENEFITS when keeping/returning to build numbers >>> for >>> > Pharo 7: >>> > - we do not change image file names, about box behavior, ... >>> compared to >>> > previous Pharo version < 7 >>> > (as we used image build number already in the past) >>> > - we tag each release as before and see it in Git (we can easily >>> reproduce) >>> > - the build number easily tells you which image is more recent (as >>> before) >>> > - we can easily sort when we have several images in a directory >>> > - a build number is more readable and recognizable by a human >>> (compared to >>> > the shas) >>> > - Pharo is not an "aliens" compared to the rest of the software world >>> as >>> > often software >>> > follows a MAJOR.MINOR.PATCH/BUILDNUMBER scheme (see semver.org) >>> > - we do not change the order in Launcher (higher numbers at the top to >>> > download more recent) >>> > >>> > According to the discussion with Esteban and Pavel it is technically >>> > possible to have build numbers again - >>> > it means to tag each commit again with a build number (we already did >>> this >>> > for Pharo 6, >>> > see https://github.com/pharo-project/pharo/releases) >>> > >>> > The outcome from yesterday was that Pavel will discuss again with >>> Guile on >>> > that topic. It would be good if others >>> > could comment on that topic too. Maybe we can return to the known build >>> > number scheme >>> > or (if there are problems with that) at least know the arguments why >>> we need >>> > to be exotic/different on >>> > this corner in the future. >>> > >>> > Thanks >>> > T. >>> > >>> > Gesendet: Donnerstag, 03. August 2017 um 15:41 Uhr >>> > Von: "Ben Coman" <b...@openinworld.com> >>> > An: "Pharo Development List" <pharo-dev@lists.pharo.org> >>> > Betreff: [Pharo-dev] PharoLauncher - uninformative Pharo7 template >>> names >>> > >>> > Attached is what I see for Pharo 7 images in PharoLauncher. >>> > I presume the top one is the latest, but Its a bit hard to tell :P >>> > Anyone else seeing this? >>> > >>> > Loading ConfigurationOfPharoLauncher-ChristopheDemarey.53 (latest) >>> > and doing "ConfigurationOfPharoLauncher loadDevelopment" >>> > which loads PharoLauncher-Core-ChristopheDemarey.116 (latest) >>> > >>> > This is with Pharo builds 60486 and 60510, and same VM for both... >>> > Win32 built on May 31 2017 03:09:04 GMT Compiler: 5.4.0 VMMaker >>> > versionString VM: 201705310241 >>> > CoInterpreter VMMaker.oscog-eem.2231 uuid: >>> > de62947a-7f40-4977-a232-e06a3a80c939 May 31 2017 >>> > >>> > (but btw, does that look strange? The 60510 image was lauched from the >>> > original 60486-PharoLauncher which said it was downloading the >>> matching VM, >>> > so I kind of expect each image to have a different VM ?? ) >>> > >>> > cheers -ben >>> >>> >> >> >> -- >> >> >> >> Guille Polito >> >> >> Research Engineer >> >> French National Center for Scientific Research - *http://www.cnrs.fr* >> <http://www.cnrs.fr> >> >> >> >> *Web:* *http://guillep.github.io* <http://guillep.github.io> >> >> *Phone: *+33 06 52 70 66 13 <+33%206%2052%2070%2066%2013> >> > > > > -- > > > > Guille Polito > > > Research Engineer > > French National Center for Scientific Research - *http://www.cnrs.fr* > <http://www.cnrs.fr> > > > > *Web:* *http://guillep.github.io* <http://guillep.github.io> > > *Phone: *+33 06 52 70 66 13 <+33%206%2052%2070%2066%2013> > -- Guille Polito Research Engineer French National Center for Scientific Research - *http://www.cnrs.fr* <http://www.cnrs.fr> *Web:* *http://guillep.github.io* <http://guillep.github.io> *Phone: *+33 06 52 70 66 13
