couchdb-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Noah Slater <nsla...@tumbolia.org>
Subject Re: [VOTE] Apache CouchDB 1.2.0 -- Windows Binaries, round 3
Date Wed, 28 Mar 2012 17:27:43 GMT
Hello,

This is an important note from the PMC.

We were in error to call this thread a vote, and this is my fault.

To get things clear:

   - The PMC does not vote on binary packages. The PMC only votes on
   release artefacts, which are always source packages.
   - The Windows binaries are being provided by Dave Cottlehuber for
   convenience only, and do not constitute a release by the PMC.
   - We will host the Windows binaries in our project distribution
   directory, but we will add a note to inform users that any binary packages
   are not official releases and are provided for convenience only.
   - Binary packages which are referenced by the project (from the website,
   for instance) will always be built from a corresponding source release.
   - Dave Cottlehuber has done astounding work getting us to this point,
   and I am very pleased that so many people have provided feedback. In
   future, Dave will be asking for testers, but we will not be referring to
   this as a vote.
   - With my PMC hat on, and my Release Team hat on, I can tell you that we
   will try to run the release in parallel with any binary package testing. If
   there are significant problems in either, we will abort both. We will also
   try to announce the availability of binary packages along with the source
   release announcement.

Apologies that this has come so late in the thread. Be assured that it does
not change anything, and we are still moving forward as planned. But it is
important, as a PMC, that we get our terminology right, and set
expectations appropriately.

Thank you,

N

On Mon, Mar 26, 2012 at 11:16 PM, Randall Leeds <randall.leeds@gmail.com>wrote:

> On Mon, Mar 26, 2012 at 12:29, Nick North <north.n@gmail.com> wrote:
> > Not sure whether to be annoyed or not: when I run the replicator_db test
> > with Firebug installed it passes every time. I'm guessing that it is some
> > timing issue and the debugger slows things down enough to prevent it
> > happening.
> >
> > Similarly the Win8 view_compaction test works in my Firebugged Firefox.
> So
> > now I'm a more positive +1, though just a little peeved that it's not
> > possible to track down the problems.
>
> Well that is helpful, thanks. I'm compiling a private bullet list of
> test-related issues to bring to the group for attacking after 1.2.
> I'll take a look at this with an eye towards timing sensitive
> behaviour.
>
> -Randall
>
> >
> > Nick
> > On 25 March 2012 22:35, Dave Cottlehuber <dave@muse.net.nz> wrote:
> >
> >> On 25 March 2012 21:28, Nick North <north.n@gmail.com> wrote:
> >> > +1
> >> > Windows 7 Ultimate x64 SP1
> >> > Windows 8 Consumer Preview x64 ( on VMWare Player)
> >> > Firefox 11.0
> >> > Signature: is there a missing .asc file? I didn't see the signature
> file,
> >> > but am probably missing something obvious.
> >>
> >> I noticed that I didn't upload them at same time as the shas etc,
> >> they're there now. Sorry!
> >>
> >> 1.2.0_otp_R15B.exe.asc although they were missing up until
> >> > Md5 & sha OK.
> >> > No malware detected.
> >> > End-user verification OK.
> >> > Futon tests passed on Win7 with the usual exception of our friend
> >> > replicator_db, with the usual error:
> >> > Assertion failed: expected 'null', got
> >> >
> >>
> '{"_id":"foo666","_rev":"1-8f008c4354eb07d5fbfc399a84bc88a1","value":666}'
> >>
> >> We shouldn't be seeing that anymore. Do you feel like running this in a
> >> debugger
> >> and seeing what breaks?
> >>
> >> > Futon tests passed on Win8 with the exception of view_compaction,
> which
> >> > says:
> >> > Assertion failed: resp.view_index.disk_size < disk_size_before_compact
> >>
> >> If you gently re-run that a few times does it come right? I had varying
> >> results,
> >> still digging on what the issue is in this one.
> >>
> >> > I vote +1 on the rounds that the replicator_db error seems well-known
> and
> >> > not a genuine problem, and issues with Win8 can't really be addressed
> >> until
> >> > the final release of the OS. I was not prompted for a restart after
> >> > installation, by the way.
> >> > On 25 March 2012 18:49, Dave Cottlehuber <dave@muse.net.nz> wrote:
> >> >
> >> >> Hello,
> >> >>
> >> >> I would like to call a vote on the following Windows binaries based
> on
> >> >> the concurrent Apache CouchDB 1.2.0 release artefact from git:
> >> >>
> >> >> 654768d32bae3639756c18b200582093aa51f349
> >> >>
> >> >> https://people.apache.org/~dch/dist/1.2.0/
> >> >>
> >> >> These have been built using Microsoft SDK 7.1, and includes the
> >> >> following 3rd party code:
> >> >>
> >> >> Erlang R14B04 or R15B
> >> >> OpenSSL 0.9.8r
> >> >> wxWidgets 2.8.12
> >> >> Mozilla JavaScript 1.8.5
> >> >> cURL 7.23.1
> >> >> ICU 4.6.1
> >> >> Inno setup 5.4.3
> >> >>
> >> >> Please follow the test procedure before voting:
> >> >>
> >> >> https://wiki.apache.org/couchdb/Windows_Binary_Releases
> >> >>
> >> >> We encourage the whole community to download and test these
> >> >> release artifacts so that any critical issues can be resolved before
> the
> >> >> release is made. Everyone is free to vote on this release, so get
> stuck
> >> in!
> >> >>
> >> >> A+
> >> >> Dave
> >> >>
> >>
>

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