concerted-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jake Farrell <jfarr...@apache.org>
Subject Re: Website for Concerted
Date Fri, 23 Oct 2015 16:37:49 GMT
the process you described is close, but the release is not for binary
artifacts, but for a source release. we can choose to builds binary
artifacts to go along with the source release, but that must come first.

rough outline of the process is

- checkout a branch
- create source release (easiest thing to do here is use a git archive)
- sign file
- create md5 and sha sums
- push to svn release candidate dir
- draft email
- start release candidate vote on dev@
- if vote is successful then
- run release script which
- creates release tag off the given branch commit HEAD
- downloads release candidate from svn
- rename package by removing -rc#
- push to svn release dir
- send release email to dev@


our svn repo for this would be located under the incubator, paths would be
something like

release candidate -
https://dist.apache.org/repos/dist/dev/incubator/concerted/0.1.0
release -
https://dist.apache.org/repos/dist/release/incubator/concerted/0.1.0

-Jake

On Fri, Oct 23, 2015 at 12:22 PM, Gaurav Shukla <gshukla66@gmail.com> wrote:

> Hi Jake,
>
> I wanted to know what all things need to be done for a release.Script
> attached with CONCERTED-16
> <https://issues.apache.org/jira/browse/CONCERTED-16> was checking out a
> tag
> and building it from source.
>
> It created a tar ball,signed it with a key and then uploaded the same on
> svn repo
> but since we don't have a svn repo what do you want the script to do after
> checking out a tag and building it.
>
> So far I've edited the script to checkout a given tag
> build it from source
> create a tarball of files in build folder and create md5  checksum.
> and generate a email template.
>
> what should it do with tarball ?
>
> On Thu, Oct 22, 2015 at 1:04 AM, Jake Farrell <jfarrell@apache.org> wrote:
>
> > Highly recommend that this process for creating a release candidate and
> > then promoting it to a release be scripted and heavily documented to make
> > the process easier and as error free as possible. Here is a good example
> of
> > this
> >
> > release candidate -
> > https://github.com/apache/mesos/blob/master/support/vote.sh
> > release - https://github.com/apache/mesos/blob/master/support/release.sh
> >
> > -Jake
> >
> > On Wed, Oct 21, 2015 at 2:59 PM, Julian Hyde <jhyde@apache.org> wrote:
> >
> > > "Release" is not a git concept. It is a concept that github layers on
> > > top of git. The git concepts are branch, commit (with associated sha),
> > > tag.
> > >
> > > I recommend that you apply a git tag to each release CANDIDATE which
> > > will become the tag for the release if and when the vote passes.
> > >
> > > Each Calcite release has a corresponding news item with text and
> > > metadata. See the metadata at the top of
> > >
> > >
> >
> https://github.com/apache/incubator-calcite/blob/master/site/_posts/2015-09-02-release-1.4.0-incubating.md
> > > .
> > > This helps generate the "recent releases" list in
> > > http://calcite.incubator.apache.org/news/ and also the table in the
> > > downloads page http://calcite.incubator.apache.org/docs/downloads.html
> .
> > >
> > > Julian
> > >
> > >
> > > On Wed, Oct 21, 2015 at 10:09 AM, Gaurav Shukla <gshukla66@gmail.com>
> > > wrote:
> > > > Hi, Julian I was looking into a way to fetch releases from github
> what
> > I
> > > > found was most of the repos in apache don't have relases they have
> > tags.
> > > >
> > > > Is this a rule?
> > > >
> > > > Even calcite does not have releases
> > > >
> > > > https://api.github.com/repos/apache/incubator-calcite/releases
> > > >
> > > > https://api.github.com/repos/apache/incubator-calcite/tags
> > > >
> > > > I don't have problem with tags but to fetch date for those tags I
> have
> > to
> > > > find commit url and then find date for that commit.
> > > >
> > > > Is that how you displayed release date and user on calcite's website?
> > > >
> > > > On Wed, Oct 21, 2015 at 10:23 PM, Julian Hyde <jhyde@apache.org>
> > wrote:
> > > >
> > > >>
> > > >> > On Oct 20, 2015, at 12:39 AM, Atri Sharma <atri.jiit@gmail.com>
> > > wrote:
> > > >> >
> > > >> > I am fine with fetching contributors and committers from github
> > > (mentors,
> > > >> > please advice) as long it is clearly mentioned and understandable
> on
> > > the
> > > >> > website.
> > > >>
> > > >> We did that in Calcite and no one had a problem with it. Every
> > > contributor
> > > >> has a github account because pull requests are our preferred mode
of
> > > >> contributions.
> > > >>
> > > >> In fact this page is generated from that information, and the
> pictures
> > > >> come from github:
> > > >>
> > > >> http://calcite.incubator.apache.org/develop/
> > > >>
> > > >> The “organization” field was a little more contentious. Some people
> > feel
> > > >> that Apache contributors should leave their afiliations at the door
> > when
> > > >> working within Apache. So you should only list organizations if your
> > > >> community has broad consensus.
> > > >>
> > > >> Julian
> > > >>
> > > >>
> > > >> >
> > > >> > ASF Git repository is being set up right now.
> > > >> > On 20 Oct 2015 12:51, "Gaurav Shukla" <gshukla66@gmail.com>
> wrote:
> > > >> >
> > > >> >> By dead link I mean I get 404 project not found both on github
> > > mirror as
> > > >> >> well as git-wip.apache.org
> > > >> >> <https://git-wip-us.apache.org/repos/asf/incubator-concerted.git
> >
> > > >> >>
> > > >> >> I'll list PPMC members in html but current committers and
> > > contributors
> > > >> can
> > > >> >> be fetched directly from github repo so it will need not
be
> updated
> > > >> again
> > > >> >> and again.
> > > >> >>
> > > >> >> Also should I create a blog for this website using jekyll
or
> > should I
> > > >> use
> > > >> >> the one available in confluence ?
> > > >> >>
> > > >> >> On Tue, Oct 20, 2015 at 12:44 PM, Atri Sharma <atri@apache.org>
> > > wrote:
> > > >> >>
> > > >> >>> What do you mean by dead link?
> > > >> >>>
> > > >> >>> It works fine for me.
> > > >> >>>
> > > >> >>> I like the idea of activity feed. However we need a hard
list
> > > >> mentioning
> > > >> >>> current committers, PPMC members and contributors.
> > > >> >>>
> > > >> >>> Also a page for listing newbie JIRAS.
> > > >> >>>
> > > >> >>> Let us know if you have any questions.
> > > >> >>>
> > > >> >>> Regards,
> > > >> >>>
> > > >> >>> Atri
> > > >> >>>
> > > >> >>
> > > >> >>
> > > >> >>
> > > >> >> --
> > > >> >> Regards !
> > > >> >> Gaurav Shukla
> > > >> >> gauravshukla.xyz
> > > >> >> Phone: +91-837-584-7862
> > > >> >>
> > > >>
> > > >>
> > > >
> > > >
> > > > --
> > > > Regards !
> > > > Gaurav Shukla
> > > > gauravshukla.xyz
> > > > Phone: +91-837-584-7862
> > >
> >
>
>
>
> --
> Regards !
> Gaurav Shukla
> gauravshukla.xyz
> Phone: +91-837-584-7862
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message