apache spark, anyone?

[gil]

hi

Il 27/07/2015 21:26, Haïkel ha scritto:
> This is a good case to show that we need layered products on top of Fedora.
> If we want to keep Fedora relevant in the big data field (and many
> others!), we have
> to accept "second-class" repositories with less restrictive guidelines (hence
> quality).
i disagree, if it's at the expense of quality
> I'm in favor that we allow approved SIGs to ship in separate
> repositories packages
> that does not strictly follow some guidelines. Granted that fesco/FPC allow them
> to lift *identified* guidelines for valid reasons.
but this repository are already present? i mean "rpmfusion" ...
> ---
>
> It's not realistic to think that we could ship big data stuff without allowing
> bundled libraries. It's not how the java/hadoop ecosystems work at all,
> and it's unlikely to change.
i disagree, bundle libraries are really a crap ... sorry. this to me 
means that the
project or the developers who use them are low-level, and the software that
uses it loses value and my interest
> Requesting bundle exceptions from FPC for every big data package would
> jam the process. I would suggest that the big data SIG coordinate to maintain
> a copr repository until we find a satisfactory solution.
>
> H.
this is the rule, god bless FPC :)
regards
gil
p.s. some build dep for Sparl 1.4.x are already available here:
https://gil.fedorapeople.org/cassandra/
and
https://gil.fedorapeople.org/lz4-java-1.3.0-1.fc22.src.rpm
this last bundle part C code of the xxHash project 
https://github.com/Cyan4973/xxHash
> _______________________________________________
> bigdata mailing list
> bigdata at lists.fedoraproject.org
> https://admin.fedoraproject.org/mailman/listinfo/bigdata

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.fedoraproject.org/pipermail/bigdata/attachments/20150727/f3b1442b/attachment.html>