commons-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Luc Maisonobe <>
Subject Re: [math] TLP
Date Thu, 14 Jan 2016 10:58:47 GMT
Hi Phil,

Le 14/01/2016 01:50, Phil Steitz a écrit :
> I would like to propose that we split [math] out into a top level
> project at the ASF.  This has been proposed before, and I have
> always come down on the side of staying in Commons, but I am now
> convinced that it is a good step for us to take for the following
> reasons:
> 0) We have several committers who are really only interested in
> [math], so being on the Commons PMC does not really make sense for them
> 1) The code base has swollen in size to well beyond the "small sharp
> tools" that make up the bulk of Commons
> 2) We are probably at the point where we should consider splitting
> [math] itself into separately released subcomponents (could be done
> in Commons, but starts smelling a little Jakarta-ish when Commons
> has components with subcomponents).
> The downsides are
> a) [newPMC] loses Commons eyeballs / contributors who would not find
> us otherwise
> b) Migration / repackaging pain
> c) Overhead of starting and managing a PMC
> d) Other Commons components lose some eyeballs
> Personally, I think the benefits outweigh the downsides at this
> point.  New better tools and ASF processes have made b) and c) a
> little less onerous.  I don't think d) is really a big problem for
> Commons, as those of us who work on other stuff here could continue
> to do so.  It is possible that a) actually works in the reverse
> direction - i.e., we are easier to find as a TLP.
> What do others think about this? 

I also think it is now time for us to grow up and leave parents home.
[math] has become big, really big by now. It looks more like a
standalone autonomous project than a shared component. Since a few
years, it started to becomes a singular component, not really
similar to the others. We almost monopolize the bandwidth on the
mailing list, which can be painful for non-math developers.

I think going TLP could also allow us to do somes things differently,
perhaps experimenting on less stringent constraints about releases,
mainly related to stuff that is not stabilized. We could also accept
some ideas that were rejected up to now as not fitting in commons
scope (higher level stuff like the expression parser that was submitted
twice by different people if I remember well).

So +1 for going TLP.

best regards,

> Phil
> ---------------------------------------------------------------------
> To unsubscribe, e-mail:
> For additional commands, e-mail:

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message