johansen-osdev at sun.com wrote:
>>> It doesn't make sense to me that a project can say, "We want to be open,
>>> but we don't want our reviews public." This is saying you're open
>>> without actually being open. In fact, I don't believe that OpenSolaris
>>> should be endorsing projects that aren't open.
>>>
>>> Maybe I've misunderstood something fundamental? Why should some
>>> projects get to have it both ways?
>>>
>>>
>> Because, otherwise, if you don't provide a forum for them to get ARC
>> input early, then they won't get it. If they then try to open up later
>> and ARC says, "no, you have to do it this way", the likely course of
>> least resistance is to abandon the process and keep the project internal
>> ... and the source and the benefit will probably never see the light of day.
>>
>
> You're making an assumption here that I don't understand. I'm in favor
> of getting ARC review early. If a project is waiting for legal
> approval, why not draft your design documents then and go for an open
> review once legal has said okay? It's not clear to me how this is
> prohibitive.
>
Legal review can often take as long, or even longer, than the
implementation. I think design review earlier is better, otherwise it
creates situations where the time to delivery may extend so far as to
miss significant opportunities. Have you ever worked with a project
that required legal review?
>
>> ARC input into the process early is _good_. If you can't do that, even
>> for projects which participants are not ready to advertise to the world,
>> then you will lose a large portion of participants who otherwise would
>> be contributors to the community.
>>
>
> Again, I reiterate that I'm in favor of review. I'm a bit concerned
> about projects who have participants that aren't ready to advertise.
> Does the ARC community have the bandwidth and the interest to review
> various and sundry designs for projects that haven't committed to being
> open? Should we be encouraging projects to conduct their reviews in
> private?
>
*Encouraging*? No. But I don't think we should be prohibiting it either.
> Having a closed process seems like an invitation for abuse of the
> review mechanism. Everybody who has responded to me on this thread
> seems to agree that I'm incompetent. Would I be allowed to submit
> closed ARC cases for early review because I'm embarrassed that people
> might find out that I'm stupid? My proposals might end up being
> better-formed when I finally decide to announce that I'm working on a
> project.
>
Maybe. And ultimately, even _that_ might not be such a bad thing.
But I also agree that some bar needs to be set for having a case be
closed. If nothing else, I think it would be reasonable for the ARC to
require justification for why a case is not open, with the ability to
refuse to review such a case if appropriate justification is not provided.
>
>> Ultimately, if we can't make it work for Sun projects, then Sun will
>> eventually decide that the process costs too much or is unworkable, and
>> will abandon the project.
>>
>
> Is it not working for Sun projects? Where do you forsee the difficulty
> with the current process?
>
No, it isn't working for all current Sun projects. Many projects still
go to ARC closed, even though they could be open. And there are some
that still _have_ to be closed that go to ARC. Still more just avoid
the ARCs altogether, but that's a different problem.
-- Garrett
