Brian Gupta wrote: > Joseph, I mixed together a bunch of your emails, and answer the points > below. > >> >> Actually, I wonder if this is Sun's to decide. I *think* its something >> in the CAB approved development ruleset. I suspect the CAB can >> revise it. (The whole idea was for Sun to put a stake in the ground, >> not that the stake couldn't be moved.) > > Agree that this is a separate issue. > > Could someone who is knows someone at SuSE, RedHat or Ubuntu please > reach out and see how they handle it. You are right it is a Sun issue, > but that doesn't mean OpenSolaris can't set it's own standard. > Finally, I think that the contributor agreement, should have a section > regarding wether or not, and how, one wants to be acknowledged. (Maybe > someone want sto be anonymous.) I think you misunderstood me. I was saying it was once a Sun decision (and we said no names), but it seems (at least to me) that when we handed the source over to OpenSolaris, we handed this decision with it, along with an initial "stake in the sand". I'm not sure of the OpenSolaris process to move that stake, but I'm sure it can be done.
But, bottom line is that this is a completely separate topic. >>> - When appropriate blastwave maintainers will also maintain SFW >>> - Blastwave will continue to maintain unstable S11 packages >> Nit: last two bullets could be less ambiguous. > > Ok. How about?: > > - Leverage the experience of the Blastwave package maintains. They are > already familiar with the packages, package authors, and stability. > They would be natural candidates for maintaining stable OpenSolaris > packages. First line seems sufficient. > - The S11 branch of blastwave will follow the OpenSolaris pathing and > packaging standards such that the full blastwave tree could be > considered the unstable repository/branch of the OpenSolaris > distribution. I'm lost on this one. There are no Nevada (I'm not supposed to put "S" and "11" next to eachother) patches. There won't be any until Nevada become a named GA release. I'm rather lost on this whole point. >> > 9) Woo upstream developers and maintainers, to join the cause. This of >> > course would be limited to smaller projects. (Once we have a coherent >> > procedure and policy in place.) (Brian) >> Why "of course limited to smaller projects"? I simply don't understand. > > rpm is supported on *ALL* Linux distros. Therefore, RPMs are the natural > choice for distribution. I say smaller packages, because they are > generally > maintained by one person. They are also more prone to see more widespread > adoption of their packages. Have you tried to install a rpm archive on Ubuntu or Debian? RPM is not universal. rpm is provided on Debian and Ubuntu, but you can't install rpm archives with it. I believe its provide so you can inspect (and with some additional effort) build rpm archives on their systems. And, of course, there is alien... Oh, I didn't translate "smaller projects" to "smaller packages". I agree. > It really depends on what you mean by support. If by support you mean > apply patches and what not. That would be the package maintainers. > > If instead by support you mean, who deals with end user "problems". A > mailing list should be setup. (Or forum.) If it is determined that it > is a usage issue, not an OpenSolaris specific issue, the user will be > forwarded to the appropriate user community discusion. (e.g - > vim at vim.org for vim) Short answer: the community members (an a tiime/interest available basis). That's probably fine. > If on the other hand you are referring to how Sun should handle > support for Solaris, it will be a multitiered approach. For packages > that are deemed to be stable/supported, Sun will support it. (This may > have the support tech reaching out to the open solaris lists.) In all > other cases sun support will forward the user to the new OpenSolaris > discussion forum. > > Of course this would open the door for some enterpriseing individual, > group or company to offer commercial support for the unstable > universe. Also fine, as long as you understand which packages this applies to is not an OpenSolaris decision. It a decision for each distro (where Sun Solaris is one distro) and is likely to be different for the distros. This is good stuff. It puts some substance behind "support" and firms up some rather blurry lines. > > P.S. - Please let me know if we still have some issues to address. I'm less and less comfortable with your bullet #5. As per my previous response, I think these categories need to be very firmly established before we (as a group) agree on anything. I'll note there seems to be a lot of discomfort around that bullet - I just got wordy. We also need to make sure the playing field is level for all potential distros. I think the proposals aren't that far from it, but some items are stated in terms of "Sun" which perhaps shouldn't be. - jek3
