Xiao-Feng Li wrote: > On 6/25/07, Mikhail Loenko <[EMAIL PROTECTED]> wrote: >> 2007/6/25, Xiao-Feng Li <[EMAIL PROTECTED]>: >> > On 6/25/07, Mikhail Loenko <[EMAIL PROTECTED]> wrote: >> > > Let's create snapshots and then decide (based on testing results?) >> > > what to release >> > >> > Why not to have them all? I don't understand. Sorry if I missed >> >> It depends on testing results. For example, it does not make much sense >> to release IPF build as a stable build (IIRC it currently runs apps >> like "hello world" >> only) >> >> If we have excellent testing results on the first platform and pretty >> modest ones on >> the second one (or no testing for the second platform) I'd prefer to >> release only the first one. > > I agree with the criteria, well I think we are talking about > Linux-X86/Win-X86/Win-X86-64 per Stepan's question, and these three > configurations are rather stable (?) in my understanding. I wish I am > right since that's what the feature/code freezes were/are intended > for. Otherwise, we might want to reexamine the M-series release > process.
Yes, the Milestones are our public statement about the quality of those builds. We should only publish a Milestone for platforms that are behaving well. Regards, Tim
