You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@couchdb.apache.org by Jan Lehnardt <ja...@apache.org> on 2014/11/10 15:01:03 UTC
The ApacheCon EU Announcements
Hey all,
it is ApacheCon EU next week and CouchDB has a full track of great talks, yay! \o/
http://events.linuxfoundation.org/events/apachecon-europe/program/schedule (Click on “Couch” in the side bar, to see all our talks :)
I’m scheduled to give a CouchDB State of the Union kind of talk on Monday morning and aside from where we are coming from and what the current state is, I’d like to make some forward looking statements about CouchDB 2.0.
The big news, of course, is the BigCouch merge and what this means for CouchDB. There are also various bits and pieces about things that we have done outside of that (the rcouch merge), as well as a few things that are part of the merge, but are worth pointing out on its own (like better compaction e.g.).
I’d like to do one better though. Bob Newson has already mentioned a few times that master is now in a state that it would be great if more people could get their hands on it.
To that end, I’d like to propose that we announce and release[1] a “Developer Preview” version of master as it is now (+ the _cluster_setup) work.
If you agree, what do you think we should be doing for this?
My short list is:
1. finish /_cluster_setup & Fauxton UI, to make cluster setup easier
2. write “how to get started” guide
3. figure out what a “release” could look like.
For 1. I can do on the backend, but I’d love some help from the Fauxton team for the UI bits.
For 2. I can do too, but any volunteers are welcome :D
For 3. I only have a vague understanding of how to do releases in a 2.0 world overall. Ideally, we have something that puts out nightly tarballs (easily done on our CI server). A next step and nice to have would be pre-built binaries for people to run, as well as a Docker/Vagrant/etc setup that people can use. But I don’t know how far away all of these are for us at this point. Can anyone shed some light on this?
* * *
Ideally, our Nov 17. Announcement says that everything is available right away.
Is there anything else that you think we absolutely need to do to make a “Developer Preview” of CouchDB 2.0? Do you see any significant issues with the things I outlined above?
Is there anything in particular that I should be mentioning in (one of) my ApacheCon EU talk(s)?
Thank you for your help!
Jan
--
[1]: This doesn’t have to be an ASF, capital R, Release,d putting up some tarballs somewhere would be fine.
Re: The ApacheCon EU Announcements
Posted by Andy Wenk <an...@nms.de>.
On Nov 10, 2014 9:07 PM, "Jan Lehnardt" <ja...@apache.org> wrote:
>
> marketing@ only:
>
> In addition to the technical discussion:
>
> Cloudant have offered to issue a press release on our behalf about the
Developer Preview.
>
> Is this something we’d like to take them up on?
>
> I’m +1. My only concern was to not use too many of these because we want
to make a big splash with the 2.0 release, but they said it is fine because
CouchDB hasn’t been the most vocal project in the past, so we can do this
one.
>
> Best
> Jan
> --
Definitely +1
I don't see any collision with the efforts for 2.0 . Quite the opposite is
the point imho. Because CouchDB never did too much press wise we should
start now slowly to get the attention from the press we need for the big
bang!
Thanjs Jan
>
>
> > On 10 Nov 2014, at 15:01 , Jan Lehnardt <ja...@apache.org> wrote:
> >
> > Hey all,
> >
> > it is ApacheCon EU next week and CouchDB has a full track of great
talks, yay! \o/
> >
> >
http://events.linuxfoundation.org/events/apachecon-europe/program/schedule
(Click on “Couch” in the side bar, to see all our talks :)
> >
> > I’m scheduled to give a CouchDB State of the Union kind of talk on
Monday morning and aside from where we are coming from and what the current
state is, I’d like to make some forward looking statements about CouchDB
2.0.
> >
> > The big news, of course, is the BigCouch merge and what this means for
CouchDB. There are also various bits and pieces about things that we have
done outside of that (the rcouch merge), as well as a few things that are
part of the merge, but are worth pointing out on its own (like better
compaction e.g.).
> >
> > I’d like to do one better though. Bob Newson has already mentioned a
few times that master is now in a state that it would be great if more
people could get their hands on it.
> >
> > To that end, I’d like to propose that we announce and release[1] a
“Developer Preview” version of master as it is now (+ the _cluster_setup)
work.
> >
> > If you agree, what do you think we should be doing for this?
> >
> > My short list is:
> > 1. finish /_cluster_setup & Fauxton UI, to make cluster setup easier
> > 2. write “how to get started” guide
> > 3. figure out what a “release” could look like.
> >
> >
> > For 1. I can do on the backend, but I’d love some help from the Fauxton
team for the UI bits.
> >
> > For 2. I can do too, but any volunteers are welcome :D
> >
> > For 3. I only have a vague understanding of how to do releases in a 2.0
world overall. Ideally, we have something that puts out nightly tarballs
(easily done on our CI server). A next step and nice to have would be
pre-built binaries for people to run, as well as a Docker/Vagrant/etc setup
that people can use. But I don’t know how far away all of these are for us
at this point. Can anyone shed some light on this?
> >
> > * * *
> >
> > Ideally, our Nov 17. Announcement says that everything is available
right away.
> >
> > Is there anything else that you think we absolutely need to do to make
a “Developer Preview” of CouchDB 2.0? Do you see any significant issues
with the things I outlined above?
> >
> > Is there anything in particular that I should be mentioning in (one of)
my ApacheCon EU talk(s)?
> >
> > Thank you for your help!
> >
> > Jan
> > --
> > [1]: This doesn’t have to be an ASF, capital R, Release,d putting up
some tarballs somewhere would be fine.
> >
> >
> >
>
Re: The ApacheCon EU Announcements
Posted by Jan Lehnardt <ja...@apache.org>.
marketing@ only:
In addition to the technical discussion:
Cloudant have offered to issue a press release on our behalf about the Developer Preview.
Is this something we’d like to take them up on?
I’m +1. My only concern was to not use too many of these because we want to make a big splash with the 2.0 release, but they said it is fine because CouchDB hasn’t been the most vocal project in the past, so we can do this one.
Best
Jan
--
> On 10 Nov 2014, at 15:01 , Jan Lehnardt <ja...@apache.org> wrote:
>
> Hey all,
>
> it is ApacheCon EU next week and CouchDB has a full track of great talks, yay! \o/
>
> http://events.linuxfoundation.org/events/apachecon-europe/program/schedule (Click on “Couch” in the side bar, to see all our talks :)
>
> I’m scheduled to give a CouchDB State of the Union kind of talk on Monday morning and aside from where we are coming from and what the current state is, I’d like to make some forward looking statements about CouchDB 2.0.
>
> The big news, of course, is the BigCouch merge and what this means for CouchDB. There are also various bits and pieces about things that we have done outside of that (the rcouch merge), as well as a few things that are part of the merge, but are worth pointing out on its own (like better compaction e.g.).
>
> I’d like to do one better though. Bob Newson has already mentioned a few times that master is now in a state that it would be great if more people could get their hands on it.
>
> To that end, I’d like to propose that we announce and release[1] a “Developer Preview” version of master as it is now (+ the _cluster_setup) work.
>
> If you agree, what do you think we should be doing for this?
>
> My short list is:
> 1. finish /_cluster_setup & Fauxton UI, to make cluster setup easier
> 2. write “how to get started” guide
> 3. figure out what a “release” could look like.
>
>
> For 1. I can do on the backend, but I’d love some help from the Fauxton team for the UI bits.
>
> For 2. I can do too, but any volunteers are welcome :D
>
> For 3. I only have a vague understanding of how to do releases in a 2.0 world overall. Ideally, we have something that puts out nightly tarballs (easily done on our CI server). A next step and nice to have would be pre-built binaries for people to run, as well as a Docker/Vagrant/etc setup that people can use. But I don’t know how far away all of these are for us at this point. Can anyone shed some light on this?
>
> * * *
>
> Ideally, our Nov 17. Announcement says that everything is available right away.
>
> Is there anything else that you think we absolutely need to do to make a “Developer Preview” of CouchDB 2.0? Do you see any significant issues with the things I outlined above?
>
> Is there anything in particular that I should be mentioning in (one of) my ApacheCon EU talk(s)?
>
> Thank you for your help!
>
> Jan
> --
> [1]: This doesn’t have to be an ASF, capital R, Release,d putting up some tarballs somewhere would be fine.
>
>
>
Re: The ApacheCon EU Announcements
Posted by Jan Lehnardt <ja...@apache.org>.
re-CC dev@.
Thanks Robert!
> On 10 Nov 2014, at 19:37 , Robert Kowalski <ro...@kowalski.gd> wrote:
>
> Hi Jan,
>
> I have a list with potentially blocking issues that are affecting Fauxton:
>
> https://issues.apache.org/jira/issues/?jql=project+%3D+COUCHDB+AND+resolution+%3D+Unresolved+AND+priority+%3D+Blocker+ORDER+BY+key+DESC
>
> I can try to help you whenever I have time this week with
> /_setup_clusters, however my main project will be some internal stuff
> this week.
>
> Best,
> Robert
>
> On Mon, Nov 10, 2014 at 3:01 PM, Jan Lehnardt <ja...@apache.org> wrote:
>> Hey all,
>>
>> it is ApacheCon EU next week and CouchDB has a full track of great talks, yay! \o/
>>
>> http://events.linuxfoundation.org/events/apachecon-europe/program/schedule (Click on “Couch” in the side bar, to see all our talks :)
>>
>> I’m scheduled to give a CouchDB State of the Union kind of talk on Monday morning and aside from where we are coming from and what the current state is, I’d like to make some forward looking statements about CouchDB 2.0.
>>
>> The big news, of course, is the BigCouch merge and what this means for CouchDB. There are also various bits and pieces about things that we have done outside of that (the rcouch merge), as well as a few things that are part of the merge, but are worth pointing out on its own (like better compaction e.g.).
>>
>> I’d like to do one better though. Bob Newson has already mentioned a few times that master is now in a state that it would be great if more people could get their hands on it.
>>
>> To that end, I’d like to propose that we announce and release[1] a “Developer Preview” version of master as it is now (+ the _cluster_setup) work.
>>
>> If you agree, what do you think we should be doing for this?
>>
>> My short list is:
>> 1. finish /_cluster_setup & Fauxton UI, to make cluster setup easier
>> 2. write “how to get started” guide
>> 3. figure out what a “release” could look like.
>>
>>
>> For 1. I can do on the backend, but I’d love some help from the Fauxton team for the UI bits.
>>
>> For 2. I can do too, but any volunteers are welcome :D
>>
>> For 3. I only have a vague understanding of how to do releases in a 2.0 world overall. Ideally, we have something that puts out nightly tarballs (easily done on our CI server). A next step and nice to have would be pre-built binaries for people to run, as well as a Docker/Vagrant/etc setup that people can use. But I don’t know how far away all of these are for us at this point. Can anyone shed some light on this?
>>
>> * * *
>>
>> Ideally, our Nov 17. Announcement says that everything is available right away.
>>
>> Is there anything else that you think we absolutely need to do to make a “Developer Preview” of CouchDB 2.0? Do you see any significant issues with the things I outlined above?
>>
>> Is there anything in particular that I should be mentioning in (one of) my ApacheCon EU talk(s)?
>>
>> Thank you for your help!
>>
>> Jan
>> --
>> [1]: This doesn’t have to be an ASF, capital R, Release,d putting up some tarballs somewhere would be fine.
>>
>>
>>
Re: The ApacheCon EU Announcements
Posted by Jan Lehnardt <ja...@apache.org>.
re-CC dev@.
Thanks Robert!
> On 10 Nov 2014, at 19:37 , Robert Kowalski <ro...@kowalski.gd> wrote:
>
> Hi Jan,
>
> I have a list with potentially blocking issues that are affecting Fauxton:
>
> https://issues.apache.org/jira/issues/?jql=project+%3D+COUCHDB+AND+resolution+%3D+Unresolved+AND+priority+%3D+Blocker+ORDER+BY+key+DESC
>
> I can try to help you whenever I have time this week with
> /_setup_clusters, however my main project will be some internal stuff
> this week.
>
> Best,
> Robert
>
> On Mon, Nov 10, 2014 at 3:01 PM, Jan Lehnardt <ja...@apache.org> wrote:
>> Hey all,
>>
>> it is ApacheCon EU next week and CouchDB has a full track of great talks, yay! \o/
>>
>> http://events.linuxfoundation.org/events/apachecon-europe/program/schedule (Click on “Couch” in the side bar, to see all our talks :)
>>
>> I’m scheduled to give a CouchDB State of the Union kind of talk on Monday morning and aside from where we are coming from and what the current state is, I’d like to make some forward looking statements about CouchDB 2.0.
>>
>> The big news, of course, is the BigCouch merge and what this means for CouchDB. There are also various bits and pieces about things that we have done outside of that (the rcouch merge), as well as a few things that are part of the merge, but are worth pointing out on its own (like better compaction e.g.).
>>
>> I’d like to do one better though. Bob Newson has already mentioned a few times that master is now in a state that it would be great if more people could get their hands on it.
>>
>> To that end, I’d like to propose that we announce and release[1] a “Developer Preview” version of master as it is now (+ the _cluster_setup) work.
>>
>> If you agree, what do you think we should be doing for this?
>>
>> My short list is:
>> 1. finish /_cluster_setup & Fauxton UI, to make cluster setup easier
>> 2. write “how to get started” guide
>> 3. figure out what a “release” could look like.
>>
>>
>> For 1. I can do on the backend, but I’d love some help from the Fauxton team for the UI bits.
>>
>> For 2. I can do too, but any volunteers are welcome :D
>>
>> For 3. I only have a vague understanding of how to do releases in a 2.0 world overall. Ideally, we have something that puts out nightly tarballs (easily done on our CI server). A next step and nice to have would be pre-built binaries for people to run, as well as a Docker/Vagrant/etc setup that people can use. But I don’t know how far away all of these are for us at this point. Can anyone shed some light on this?
>>
>> * * *
>>
>> Ideally, our Nov 17. Announcement says that everything is available right away.
>>
>> Is there anything else that you think we absolutely need to do to make a “Developer Preview” of CouchDB 2.0? Do you see any significant issues with the things I outlined above?
>>
>> Is there anything in particular that I should be mentioning in (one of) my ApacheCon EU talk(s)?
>>
>> Thank you for your help!
>>
>> Jan
>> --
>> [1]: This doesn’t have to be an ASF, capital R, Release,d putting up some tarballs somewhere would be fine.
>>
>>
>>
Re: The ApacheCon EU Announcements
Posted by Robert Kowalski <ro...@kowalski.gd>.
Hi Jan,
I have a list with potentially blocking issues that are affecting Fauxton:
https://issues.apache.org/jira/issues/?jql=project+%3D+COUCHDB+AND+resolution+%3D+Unresolved+AND+priority+%3D+Blocker+ORDER+BY+key+DESC
I can try to help you whenever I have time this week with
/_setup_clusters, however my main project will be some internal stuff
this week.
Best,
Robert
On Mon, Nov 10, 2014 at 3:01 PM, Jan Lehnardt <ja...@apache.org> wrote:
> Hey all,
>
> it is ApacheCon EU next week and CouchDB has a full track of great talks, yay! \o/
>
> http://events.linuxfoundation.org/events/apachecon-europe/program/schedule (Click on “Couch” in the side bar, to see all our talks :)
>
> I’m scheduled to give a CouchDB State of the Union kind of talk on Monday morning and aside from where we are coming from and what the current state is, I’d like to make some forward looking statements about CouchDB 2.0.
>
> The big news, of course, is the BigCouch merge and what this means for CouchDB. There are also various bits and pieces about things that we have done outside of that (the rcouch merge), as well as a few things that are part of the merge, but are worth pointing out on its own (like better compaction e.g.).
>
> I’d like to do one better though. Bob Newson has already mentioned a few times that master is now in a state that it would be great if more people could get their hands on it.
>
> To that end, I’d like to propose that we announce and release[1] a “Developer Preview” version of master as it is now (+ the _cluster_setup) work.
>
> If you agree, what do you think we should be doing for this?
>
> My short list is:
> 1. finish /_cluster_setup & Fauxton UI, to make cluster setup easier
> 2. write “how to get started” guide
> 3. figure out what a “release” could look like.
>
>
> For 1. I can do on the backend, but I’d love some help from the Fauxton team for the UI bits.
>
> For 2. I can do too, but any volunteers are welcome :D
>
> For 3. I only have a vague understanding of how to do releases in a 2.0 world overall. Ideally, we have something that puts out nightly tarballs (easily done on our CI server). A next step and nice to have would be pre-built binaries for people to run, as well as a Docker/Vagrant/etc setup that people can use. But I don’t know how far away all of these are for us at this point. Can anyone shed some light on this?
>
> * * *
>
> Ideally, our Nov 17. Announcement says that everything is available right away.
>
> Is there anything else that you think we absolutely need to do to make a “Developer Preview” of CouchDB 2.0? Do you see any significant issues with the things I outlined above?
>
> Is there anything in particular that I should be mentioning in (one of) my ApacheCon EU talk(s)?
>
> Thank you for your help!
>
> Jan
> --
> [1]: This doesn’t have to be an ASF, capital R, Release,d putting up some tarballs somewhere would be fine.
>
>
>
Re: The ApacheCon EU Announcements
Posted by Andy Wenk <an...@apache.org>.
inline ...
On 14 November 2014 13:59, Jan Lehnardt <ja...@apache.org> wrote:
> Hey all,
>
> I’d like to announce a CouchDB 2.0 Developer Preview at ApacheCon EU on
> Monday.
>
YAYYYYY :)
>
> I think the best we can do is put up a tarball for people to build, with
> instructions.
>
> I’m happy to do the legwork of producing the tarball and writing up the
> install doc and put it all up on the site. E.g. at
> http://couchdb.apache.org/developer-preview/2.0/ or something.
>
> This would *not* constitute and official ASF Release, but a convenience
> code dump for people who want to try things out.
>
+1
>
> If nobody objects, we could also set up a preview@couchdb.apache.org
> email list, so people have a safe space to talk about this and are not
> overwhelmed by the rest of dev@, or we do this on users@, what do you
> think? :)
>
I am not sure if it is needed to add a new ML. We saw historically, that
new ML's are not that good accepted. And I don't really see problems with
asking the people to join user@. I think user@ is perfectly fine.
>
> Also, the tarball, are we good to simply zip up the couchdb.git repo after
> ./configure? Or would all users run ./configure? My main concern is of
> course the time to setup, as cloning all the small repos tends to take some
> time. If there are things that are tied to the Erlang version of the person
> who runs ./configure, then we can’t circumvent this (without some extra
> work). Can anyone shed some light on this? Thanks! :)
>
> Best
> Jan
> --
Cheers
Andy
>
>
> > On 10 Nov 2014, at 15:01 , Jan Lehnardt <ja...@apache.org> wrote:
> >
> > Hey all,
> >
> > it is ApacheCon EU next week and CouchDB has a full track of great
> talks, yay! \o/
> >
> >
> http://events.linuxfoundation.org/events/apachecon-europe/program/schedule
> (Click on “Couch” in the side bar, to see all our talks :)
> >
> > I’m scheduled to give a CouchDB State of the Union kind of talk on
> Monday morning and aside from where we are coming from and what the current
> state is, I’d like to make some forward looking statements about CouchDB
> 2.0.
> >
> > The big news, of course, is the BigCouch merge and what this means for
> CouchDB. There are also various bits and pieces about things that we have
> done outside of that (the rcouch merge), as well as a few things that are
> part of the merge, but are worth pointing out on its own (like better
> compaction e.g.).
> >
> > I’d like to do one better though. Bob Newson has already mentioned a few
> times that master is now in a state that it would be great if more people
> could get their hands on it.
> >
> > To that end, I’d like to propose that we announce and release[1] a
> “Developer Preview” version of master as it is now (+ the _cluster_setup)
> work.
> >
> > If you agree, what do you think we should be doing for this?
> >
> > My short list is:
> > 1. finish /_cluster_setup & Fauxton UI, to make cluster setup easier
> > 2. write “how to get started” guide
> > 3. figure out what a “release” could look like.
> >
> >
> > For 1. I can do on the backend, but I’d love some help from the Fauxton
> team for the UI bits.
> >
> > For 2. I can do too, but any volunteers are welcome :D
> >
> > For 3. I only have a vague understanding of how to do releases in a 2.0
> world overall. Ideally, we have something that puts out nightly tarballs
> (easily done on our CI server). A next step and nice to have would be
> pre-built binaries for people to run, as well as a Docker/Vagrant/etc setup
> that people can use. But I don’t know how far away all of these are for us
> at this point. Can anyone shed some light on this?
> >
> > * * *
> >
> > Ideally, our Nov 17. Announcement says that everything is available
> right away.
> >
> > Is there anything else that you think we absolutely need to do to make a
> “Developer Preview” of CouchDB 2.0? Do you see any significant issues with
> the things I outlined above?
> >
> > Is there anything in particular that I should be mentioning in (one of)
> my ApacheCon EU talk(s)?
> >
> > Thank you for your help!
> >
> > Jan
> > --
> > [1]: This doesn’t have to be an ASF, capital R, Release,d putting up
> some tarballs somewhere would be fine.
> >
> >
> >
>
>
--
Andy Wenk
Hamburg - Germany
RockIt!
GPG fingerprint: C044 8322 9E12 1483 4FEC 9452 B65D 6BE3 9ED3 9588
https://people.apache.org/keys/committer/andywenk.asc
Re: The ApacheCon EU Announcements
Posted by Jan Lehnardt <ja...@apache.org>.
> On 14 Nov 2014, at 17:38 , Jan Lehnardt <ja...@apache.org> wrote:
>
>>
>> On 14 Nov 2014, at 14:01 , Alexander Shorin <kx...@gmail.com> wrote:
>>
>> On Fri, Nov 14, 2014 at 3:59 PM, Jan Lehnardt <ja...@apache.org> wrote:
>>> Also, the tarball, are we good to simply zip up the couchdb.git repo after ./configure? Or would all users run ./configure? My main concern is of course the time to setup, as cloning all the small repos tends to take some time. If there are things that are tied to the Erlang version of the person who runs ./configure, then we can’t circumvent this (without some extra work). Can anyone shed some light on this? Thanks! :)
>>
>> In case of no ./configure usage e.g. all deps are included we couldn't
>> provide any fixes just in time.
>
> that’s a great point
>
>> But the preview will be in safe if we'll broke something (:
>
> and we’ll have more people tell us :D
>
> I’m just worried about this because asfgit was rather slow this past few weeks, to the point of erroring out during configure.
>
> Maybe we switch this to the GitHub mirror for the preview?
>
> Here’s a cursory comparison:
>
> ASF:
>
>
> GitHub:
>> time ./configure
> real 2m21.935s
> user 0m12.495s
> sys 0m10.272s
>
> git-wip-us:
>> time ./configure
> real 9m40.581s
> user 0m12.512s
> sys 0m10.535s
Current attempts at providing a tarball without all the .git directories shows that our current system isn’t really set up to be build from a source-only release:
make
ERROR: sh(git describe --always --tags)
failed with return code 128 and the following output:
fatal: Not a git repository (or any of the parent directories): .git
Any suggestions as what to do?
If there is no easy solution, the two options for Monday are:
1. a tarball with all .git directories included (~70MB .tgz or .tbz)
2. Or the installation instructions are pointing at our git repo.
Currently leaning towards #2, but would love to hear everyone’s input.
Best
Jan
--
Re: The ApacheCon EU Announcements
Posted by Jan Lehnardt <ja...@apache.org>.
> On 14 Nov 2014, at 14:01 , Alexander Shorin <kx...@gmail.com> wrote:
>
> On Fri, Nov 14, 2014 at 3:59 PM, Jan Lehnardt <ja...@apache.org> wrote:
>> Also, the tarball, are we good to simply zip up the couchdb.git repo after ./configure? Or would all users run ./configure? My main concern is of course the time to setup, as cloning all the small repos tends to take some time. If there are things that are tied to the Erlang version of the person who runs ./configure, then we can’t circumvent this (without some extra work). Can anyone shed some light on this? Thanks! :)
>
> In case of no ./configure usage e.g. all deps are included we couldn't
> provide any fixes just in time.
that’s a great point
> But the preview will be in safe if we'll broke something (:
and we’ll have more people tell us :D
I’m just worried about this because asfgit was rather slow this past few weeks, to the point of erroring out during configure.
Maybe we switch this to the GitHub mirror for the preview?
Here’s a cursory comparison:
ASF:
GitHub:
> time ./configure
real 2m21.935s
user 0m12.495s
sys 0m10.272s
git-wip-us:
> time ./configure
real 9m40.581s
user 0m12.512s
sys 0m10.535s
Best
Jan
--
Re: The ApacheCon EU Announcements
Posted by Alexander Shorin <kx...@gmail.com>.
On Fri, Nov 14, 2014 at 3:59 PM, Jan Lehnardt <ja...@apache.org> wrote:
> Also, the tarball, are we good to simply zip up the couchdb.git repo after ./configure? Or would all users run ./configure? My main concern is of course the time to setup, as cloning all the small repos tends to take some time. If there are things that are tied to the Erlang version of the person who runs ./configure, then we can’t circumvent this (without some extra work). Can anyone shed some light on this? Thanks! :)
In case of no ./configure usage e.g. all deps are included we couldn't
provide any fixes just in time. But the preview will be in safe if
we'll broke something (:
--
,,,^..^,,,
Re: The ApacheCon EU Announcements
Posted by Jan Lehnardt <ja...@apache.org>.
Hey all,
I’d like to announce a CouchDB 2.0 Developer Preview at ApacheCon EU on Monday.
I think the best we can do is put up a tarball for people to build, with instructions.
I’m happy to do the legwork of producing the tarball and writing up the install doc and put it all up on the site. E.g. at http://couchdb.apache.org/developer-preview/2.0/ or something.
This would *not* constitute and official ASF Release, but a convenience code dump for people who want to try things out.
If nobody objects, we could also set up a preview@couchdb.apache.org email list, so people have a safe space to talk about this and are not overwhelmed by the rest of dev@, or we do this on users@, what do you think? :)
Also, the tarball, are we good to simply zip up the couchdb.git repo after ./configure? Or would all users run ./configure? My main concern is of course the time to setup, as cloning all the small repos tends to take some time. If there are things that are tied to the Erlang version of the person who runs ./configure, then we can’t circumvent this (without some extra work). Can anyone shed some light on this? Thanks! :)
Best
Jan
--
> On 10 Nov 2014, at 15:01 , Jan Lehnardt <ja...@apache.org> wrote:
>
> Hey all,
>
> it is ApacheCon EU next week and CouchDB has a full track of great talks, yay! \o/
>
> http://events.linuxfoundation.org/events/apachecon-europe/program/schedule (Click on “Couch” in the side bar, to see all our talks :)
>
> I’m scheduled to give a CouchDB State of the Union kind of talk on Monday morning and aside from where we are coming from and what the current state is, I’d like to make some forward looking statements about CouchDB 2.0.
>
> The big news, of course, is the BigCouch merge and what this means for CouchDB. There are also various bits and pieces about things that we have done outside of that (the rcouch merge), as well as a few things that are part of the merge, but are worth pointing out on its own (like better compaction e.g.).
>
> I’d like to do one better though. Bob Newson has already mentioned a few times that master is now in a state that it would be great if more people could get their hands on it.
>
> To that end, I’d like to propose that we announce and release[1] a “Developer Preview” version of master as it is now (+ the _cluster_setup) work.
>
> If you agree, what do you think we should be doing for this?
>
> My short list is:
> 1. finish /_cluster_setup & Fauxton UI, to make cluster setup easier
> 2. write “how to get started” guide
> 3. figure out what a “release” could look like.
>
>
> For 1. I can do on the backend, but I’d love some help from the Fauxton team for the UI bits.
>
> For 2. I can do too, but any volunteers are welcome :D
>
> For 3. I only have a vague understanding of how to do releases in a 2.0 world overall. Ideally, we have something that puts out nightly tarballs (easily done on our CI server). A next step and nice to have would be pre-built binaries for people to run, as well as a Docker/Vagrant/etc setup that people can use. But I don’t know how far away all of these are for us at this point. Can anyone shed some light on this?
>
> * * *
>
> Ideally, our Nov 17. Announcement says that everything is available right away.
>
> Is there anything else that you think we absolutely need to do to make a “Developer Preview” of CouchDB 2.0? Do you see any significant issues with the things I outlined above?
>
> Is there anything in particular that I should be mentioning in (one of) my ApacheCon EU talk(s)?
>
> Thank you for your help!
>
> Jan
> --
> [1]: This doesn’t have to be an ASF, capital R, Release,d putting up some tarballs somewhere would be fine.
>
>
>