You (Giovanni) wrote:
> I really do hope it it like you say.
>
> Nevertheless, I have t say it is very difficult to work this way.
> I am talking OpenSolaris (but many things apply to Solaris as well).
>
> 1) No roadmap (do you have seen one recently)
Has been mentioned many times here: Oracle has a policy of NEVER publishing
roadmaps. So, that fact, that you did not SEE any roadmap does not mean, that
there is no such roadmap... Anyone of the engineers or people inside Oracle
talking to stuff on those roadmaps are under the threat of being thrown out
immediately. So, sorry, get used to that fact...
> 2) Support for "new" hardware is still in development branches (new SAS 2
> controllers from LSI in svn_134)
Which "new" hardware? Oracle's? Your's? Anybody's? You can contribute as well,
and add the driver you'd like to see yourself...
> 3)svn_134 has still lots of bugs/issues (at least in the GUI): NIC
> configuration is a pain; better go via command line
That's one of the reasons, why OSOL 2010.X isn't out yet... ;-)
> 4) some key components (i.e. ramdisk implementation) have severe performance
> issues (a ram disk running at 500MB/s on DDR3 1333 is slower than working on
> a striped physical disk set; same ramdisk on Linux on same hw runs at several
> GB/s)
>
> We all know no perfect software is there (the perfect one is the one that
> never comes out....), but short releases cycles (with roadmap) allow
> community to test and contribute and to make things more stable and better
> performing.
You can iterate that topic ad infinitum, you won't change Oracle's way of
doing business w.r.t. roadmaps... ;-) And, as others mentioned, RTSL, there
are infos up to build 142...
> Instead no svn releases after 134 are out there and no idea if/when they will
> come and what they will contain...
Can we get a new chorus line, please? And, again, check the source, there are
infos up to 142 already...
> Ok, I know one can go, grab source, compile and test: this makes test base
> much narrower and more error prone...
So, you're placing yourself into the position of beggar with no intent to
help... ;-)
> If you develop applications relying on certain OS features (such as ZFS,
> RAMdisk, COMSTAR, ...) you expect to test those features and give feedback on
> bugs/improvements/perf issues, not to start developing/debugging those
> features (you can, but again, how may will have experience to do this ?
> Developing a kernel driver is a complete different story from developing an
> application layer...different experience, competencies, dev languages, ...)
In former times, these things were called Alpha- and Beta-tests. And were VERY
limited. Now we do have OpenSolaris, and that's open to everyone, what's your
complaint here? Even with precise Alpha- and Beta-tests there have NEVER EVER
before been 2-weekly updates... So, what we now have, even with the slight
delay, is way better than what was available years ago... Or as Monty Python
put it:
Always look on the bright side of life!
> I really do hope the story with Solaris/Opensolaris goes on, but I must admit
> (and I think all the posts in the newsgroups confirm this) that working this
> way is very difficult for us and -as far as i can see - for many others.
No-one said otherwise! Yes, it might be difficult, but, changes require BOTH
PARTIES to change. So, try to get accustomed to the change, and look at the
bright side!
> I am not talking about having the latest fancy GUI with 3D effects (even if
> maybe this is important for the ones who adopted Osol as a serious desktop
> replacement) but also having new fundamental server hardware support.
Why should a PAID ORACLE employee develop drivers for hardware that will never
show up in an Oracle product? They do it on their spare time, just like you
and me!
> If everything will be tied up to Sun hardware....we are back to a proprietory
> solution for custom hardware, the Apple way I think most of folks here around
> do not appreciate very much....
No, there's a fundamental difference! You can add, you#re allowed to, and you
have the COMPLETE source!
> Anyway, we all have to sit down and wait for feedback to our knock knock at
> Oracle door: we do not know if anybody will open and what the'll tell
You're not knowcking Oracle's door, you're beating at a door to the
COMMUNITY. Wrong door!
> But-sadly- time is running out, Oracle .....
Why? There were fantastic FY numbers from Oracle last week...
Matthias
--
Matthias Pfützner | mailto:pfu...@germany | "It's all very cool, but
@work: +49 6103 752-394 | @home: +49 6151 75717 | I wonder, what it really
SunCS, Ampèrestraße 6 | Lichtenbergstraße 73 | means." Anon. MIT Student
63225 Langen, FRG | 64289 Darmstadt, FRG | in Byte's 4/96 Editorial
_______________________________________________
opensolaris-discuss mailing list
[email protected]
