On Jo, 22 iul 10, 22:51:32, Stefano Zacchiroli wrote: > On Thu, Jul 22, 2010 at 03:10:16PM -0400, Joey Hess wrote: > > I've noticed that linking to this page seems to kill threads, which is > > not my intent, but: http://kitenet.net/~joey/code/debian/cut/ > > > > The idea still seems reasonable to me, and the work required, beyond > > what is already done, not too much. Shame that nobody has tried to do it > > yet. > > I've always found CUT to be an interesting (thought) experiment. > > A particular nice feature of how you present it, is that it's modular > enough to be tried with little impact on the rest of Debian; as you > observe: « most of which [needed ingredients] are already being done on > a best-effort basis ». So, in principle, nothing stops you or anyone > else to actually form a CUT team and experiment with the idea. That's > probably the only way forward from CUT-is-a-thread-killer to have data > that support the feasibility / user appreciation of the idea. For me it's quite obvious that CUT would happen if the relevant teams: (Security, Release, d-i, ...) would have enough manpower. One possible approach would be asking through d-i if the user wants "plain" testing or a CUT snapshot, and then set the apropiate pinning. Security updates would not be possible via unstable migrations (as it is for testing), but I already mentioned the security team needs more people for this. And there's no need to call CUT differently, just call it beta1, beta2 and rc1, rc2, ... as soon as testing is frozen. With a proper announcement/press release users will jump to using it ;) Regards, Andrei -- Offtopic discussions among Debian users and developers: http://lists.alioth.debian.org/mailman/listinfo/d-community-offtopic
Attachment:
signature.asc
Description: Digital signature