On Fri, Mar 21, 2014 at 01:56:26PM -0400, Michael Terry wrote:
>
> For example, my feature depends on fixes in Ted's indicator silo and Boiko's
> telephony silo. So as delayed as they are, I'm even further down the list.
I've said this before, and I'll say it again. You can't do massively
integrated system development and testing while burying your head in
the sand and pretending the system isn't massively integrated.
If all these features landed in -proposed, they could be tested and
integrated together, a blocking bug could keep some bits from migrating
early, if that was warranted, but the key is that it would all be there
together, to see if it all integrates together.
This is how the entire distro works, except for these few projects that
are required to land via silos, and I think we've proven which method
moves more quickly and with less frustration.
That isn't to say there aren't occasionally good reasons to stage
something in a PPA. The massive X ABI transitions, the Qt5.2 armhf
transition, etc. But, much like topic branches in a VCS, if those live
for too long, then you're just doing extra work on re-integrating in
both directions, and you'd have been better off pushing to trunk.
... Adam
--
ubuntu-devel mailing list
ubuntu-devel@lists.ubuntu.com
Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel
