On Tue, Dec 21, 2010 at 11:45 AM, steve donovan
<[email protected]> wrote:
> On Tue, Dec 21, 2010 at 3:14 PM, Hisham <[email protected]> wrote:
>> I agree with that, but so far I haven't had a convenient way of
>> verifying this. One possibility is to require that people submitting
>> rockspecs should also submit the generated rock_manifest file.
>
> With a little automated help, this could become a straightforward part
> of the process. Another way is to informally distribute initial
> acceptance; e.g. I could try rocks out on Windows, someone on Mac OS,
> etc and whoever gets it integrated can submit the rock_manifest.
>
> Then we can do something interesting with the extra information, for
> example an extended search command that can look for module names, not
> just packages.

Sure, that would be great. However, I can't commit myself to actually
_run_ the rockspecs we receive. I often do so when the rockspec has no
external dependencies (and sometimes even when it does) to make sure
it builds well when something catches my eye. Rockspecs already take
too long sometimes to be uploaded (and I thank Fabio for helping me
out on this whenever I'm MIA) and adding an acceptance step when
someone else is required to actually run them could slow things down
even more.

-- Hisham

_______________________________________________
Luarocks-developers mailing list
[email protected]
http://lists.luaforge.net/cgi-bin/mailman/listinfo/luarocks-developers

Reply via email to