On May 10, 2013, at 4:59 PM, Paul B. Henson <[email protected]> wrote:
> (I crossposted this to both discuss@ and developer@, please reply only to
> discuss@; I know some of the developers don't watch discuss@, but was hoping
> they might want to contribute this time).
>
> Historically OpenIndiana has been the primary illumos development
> environment. There's some discussion over on the OpenIndiana dev list about
> the current and future status of the project. Some people say it's dead,
> other people want to revive it; without commenting on or starting a
> discussion about that, I think it's fair to say it's mostly stagnant right
> now.
>
> There was some talk once of starting an illumos "reference" distribution for
> development purposes, did anything ever come of that? If not, is there any
> interest in putting together a "supported" configuration for core illumos
> development under omnios? ("supported" mostly just meaning being able to
> discuss issues or problems without worrying about the development environment
> itself causing the issue or problem, used by at least some of the core
> developers would also be a plus :) ).
Nothing came of it. It got bikeshedded into oblivion because people wanted
"reference" to mean desktops, GUIs, and a bunch of other stuff that fully
departed from what I was trying to achieve.
I'd still like to have a base environment as an accepted reference for
developing the illumos core itself. For those purposes, I don't care about all
the other gadgets and desktop goodies, office suite, etc. It needs to be able
to run and validate the things we have in illumos-gate… plus the minimal amount
required to self host that bit.
I think OmniOS is almost there -- the "problem" is that OmniOS has made some
changes (IMO good ones) that illumos has not -- specifically nixing a bunch of
the broken or ancient printing infrastructure. With these changes, it is not
possible to compile stock illumos on OmniOS (or at least it wasn't the last
time I tried).
IMO, the best solution for us is to adopt the same change in illumos -- but
this gets into another rat hole discussion -- because some people apparently
*like* those crappy crufty printing tools. I'm strongly of the opinion that
we ought to just go ahead in spite of those complaints, as the source for the
SystemV printing stuff will always remain available in tree for any distro or
site that really wants it. I think its time for the needs of the many to
outweigh the needs of the few … and those who want to stick with the SYSV
printing stuff are definitely the few (albeit a vocal few).
(There may be *other* changes that limit OmniOS from building stock
illumos-gate -- I'm not at the moment immediately aware of any aside from the
printing problem.)
The other problem with using OmniOS for this is that OmniOS is totally
controlled by a single commercial entity, which is distasteful for some people.
As for myself, I'd be willing to overlook that and accept that if Omni ever
decided to go seriously astray we could always fork *their* work. I don't see
that being likely any time soon though.
>
> Personally I don't really need a desktop environment based on illumos,
> although I wish the best to those that do and wish them good luck. I need a
> stable server environment, which I have found in omnios (thanks omniti!), and
> I'd like a stable development environment, whether that be on top of omnios
> or a separate reference distribution. Ideally I'd prefer not to try and
> maintain an OI environment just for development. Is that more or less what
> everybody else is still doing?
I think so. I am in total agreement here, modulo the above issue centering
around printing.
- Garrett
-------------------------------------------
illumos-discuss
Archives: https://www.listbox.com/member/archive/182180/=now
RSS Feed: https://www.listbox.com/member/archive/rss/182180/21175430-2e6923be
Modify Your Subscription:
https://www.listbox.com/member/?member_id=21175430&id_secret=21175430-6a77cda4
Powered by Listbox: http://www.listbox.com
