incubator-wave-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Pablo Ojanguren <>
Subject Re: Wave Android App Landing (GSoC)
Date Fri, 21 Aug 2015 06:53:45 GMT
Hi, good work

1) and 2) Ok

3) I would prefer to provide an .apk file somewhere.

4) A miminim hacking guide to explain the basics. The user guide just
explaining how to configure server connection. To consider configure the
app with a default Wave that could be used immediatly.



2015-08-21 1:29 GMT+02:00 Ali Lown <>:

> Hi all,
> As many of you know, Roshan has been working on developing an Android
> client for Wave, as part of GSoC '15. The code base can currently be
> found on his github account[0].
> I will let Roshan provide the full details of its functionality, but
> the core features are all there (signup, signin, search, new waves,
> wave rendering).
> Sadly, the end of GSoC is almost upon us and we should probably decide
> how best to fit this amongst the rest of our files, and what state
> Apache Wave should provide the client in.
> So, for handy reference when responding, here is my set of discussion
> points:
> 1) Where should we store the android client code?
>   a) In a new repository? incubator-wave-android?
>   b) In our main repository incubator-wave, under a new top-level
> folder android/ ? Should this link in with our main build system then?
> My thoughts tend towards option (a), as it has very little in common
> with the main codebase - beyond pulling a 'core' library in as a jar.
> 2) How to version the android client?
>   a) Should it follow the main client/server numbering (so we start it at
> 0.4)?
>   b) Should it have its own versioning?
> I would again tend to option (b) here, which works particularly nicely
> if it is in its own repository.
> 3) How should we distribute the client?
>   a) Should we provide a compiled apk for easy loading?
>   b) Should we publish on the Play Store?
>   c) Do we leave the client compilation up to Wave server operators,
> and simply provide the source code?
> I am leaning towards option (c) here, as I see the Apache Wave's code
> as mostly a 'reference implementation', but I know that others here
> might like a more turn-key solution. (Such as the binary releases
> alongside the normal source releases)
> 4) I am imagining two sets of documentation for the client, a "user
> guide", and a "hacking guide", kept with the rest of the new
> documentation in the incubator-wave-docs repository. Do we need the
> "user guide"?
> Please do respond with your thoughts on any/all of these matters,
> (In the event of no responses, I will simply execute my thoughts as
> set out above ;) )
> Ali
> [0]:

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