Brian Gupta wrote:
Various drivers behind an effort to produce OSH.
1) There is a desire for a minimal/core OpenSolaris distro, that other
distro packagers can leverage to create their own distros. Building a
distro from this core *may*, in the future, allow other distros to
also be hosted at OpenSolaris.org. This OS core must, for the sake of
practicality, allow layers to be added on top to get to Nevada.
2) Currently their is no Open development platform. The only allowed
development platform is, a closed source, proprietary platform: Sun
Solaris (Express). Ideally guardianship for this development platform
should be moved into the community. (Yes this seems in opposition to
bullet #1, but keep reading)
#1 and #2 seem to make a lot of sense.
#2 seems mostly a "free as in speech" issue, which I don't tend to care
about, but I understand it is very important to some people.
We all know, "free as in beer" is most important! :-)
(Note the smiley - let's not start a discussion on this in general.)
3) Currently the barrier of entry to getting involved with OpenSolaris
is high, both from a conceptual point of view, as well as a work
required point of view. (I can't just download a ISO and install an up
to date dev platform)
What does this problem have to do with OSH? It seems like a separately
addressable problem. Also, it is a problem Sun Solaris (express) knows
about (as 50 e-mail in my inbox the last two days will attest to!).
If you think this is fun, try OpenJDK.
This, like "closed sources", tend to be a high start-up cost for software
moving to the "free" domain.
4) Currently the barrier to entry for immigrants is high. Driver
support, that shall we say "does not lead the field". Installation and
configuration can be confusing to those that have not studied Solaris.
I think these are generally acknowledged as goals in the community,
but this is where we have a distro that focuses on this issue
specifically.
Ditto #3 - a problem that doesn't really motivate OSH (as the disclaimer
states).
5) Although AOL has proved that mass mailings of optical media can be
made into a scalable distribution model, this only worked because AOL
self updated upon install. OpenSolaris needs a quicker and simpler
distribution method. (Which allows free mirroring of ISOs, binary and
source packages, without lengthy legal reviews. This ties into the
desire to have a completely unencumbered base distro set)
Ditto #3.
6) Sun (Ian/Indiana/Others) wants a second Solaris distro that will be
the Fedora to Sun's RHEL (Solaris). They wish to do this and still
leave a clear migration path from one distro to another. (Unlike
Fedora). ;)
OK, you have allies. 8^)
7) Currently there isn't a method for non-Sun distros to offer
optional non open source bits. We would want to establish a standard
and precedent for doing so, that all distros could leverage. (I'm
thinking mostly HW support issues).
Huh? What prevents a distro from doing this? Please elaborate.
8) Currently, by continuing to use closed source bits as the standard,
there is minimal incentive to close (open?) the loop.
Not sure what you are saying. If its that the "closed" bits shouldn't
exist,
please work on that problem. I think we would all like nothing more.
This is an important requirement to having a complete, truly open
distribution, but not a reason to have OSH.
9) Current HW requirements are a bit on the high end. I can't run this
on an old 386, or for that matter on an old P/PII I have lying around.
And the thought is that Sun wouldn't include such projects in it's Solaris
distribution? I'm not sure, you might be right. We've been moderately
aggressive about removing such support (more on the SPARC side
actually) motivated mostly by a desire to minimize the test matrix. The
secondary reason was code bloat.
A harder issue is "minimum storage". Just how much memory and how
big a disk does that old 386 have attached to it? :-)
I guess its a valid reason for OSH - so that the community can determine
the supported hardware list, but I'd (perhaps wrongly) tend to think that
the need is actually more in unsupported ISA/PCI cards than processors.
Seems like these could plug into any Solaris. (Seems to motivate the
blastwave-like soultions we've discussed elsewhere more than OSH, but
it does motivate OSH somewhat (particularly if that card is a NIC)).
Discussion have been had elsewhere about including unsupported (by
Sun) drivers on the distribution CD. Don't know where this has gone.
Solaris used to run on 486. I'm sure we can find those sources and donate
them to the community. SunOS 4.x used to run on 386 (A Sun product
called roadrunner). However, having run a leaner Solaris (2.3) on a 486,
I'm not sure this is really useful. I'm not that patient.
------------
Summary: This seems like a laundry list of ills with the current situation.
I believe only a few of these ills motivate an OSH, particularly if you
consider how long it will take a polished OSH to come into existance.
Maybe I've verged too far accross the "let's not debate the need" line, but
this entire posting seemed to assert the needs and most of them seemed
orthoginal to an OSH.
- jek3
_______________________________________________
opensolaris-discuss mailing list
opensolaris-discuss@opensolaris.org
