>>>>The problem that I see (And I am in
>>>>no way asserting that I'm smart enough
>>>>to be right.) is that Solaris only works
>>>>well on a tiny, tiny, tiny, tiny portion
>>>>of the hardware that's out there.
[...]
>>>I have been running Solaris on x86 for about as
>>>long as it existed and it wasn't always easy to work
>>>with, but you really have a large large large
>>>portion of the hardware that's out there to pick
>>>from.
[...]
> > I wonder if both the OP and you aren't spinning a
> bit.  x86 does indeed
> > run on
> > quite a few platforms including older ones.  But
> there are doubtless also
> > plenty
> > it doesn't run on
> 
> spinning ?
> 
> come on over to my place and I'll show you Solaris
> 2.5.1 x86 still running
> on a Pentium P90 as well as OpenSolaris snv_98
> running on a mother board
> that is nailed to the wall here. As well as my HP
> Kayak from 2000 that has
> been running with four SCSI controllers since 2002,
> and yes, one very
> small old IBM Thinkpad .. that too.
> 
> there is no spin here ... why would there be?
> 
> quite frankly .. get the CDROM and boot it. Most
> likely it will work well
> enough.

Given the use of the appropriate version for the hardware
(i.e. that support for _really_ old hardware that was once present
may have been removed, and some support for relatively new hardware
has only become available quite recently, in part perhaps due to licensing
or other such issues), you are perhaps closer to the truth than the OP.
But if "tiny, tiny, tiny" == 5% or less, and "large large large" == 95% or 
more, then I wonder
if both of you aren't a bit off base.

Back around build 50ish or earlier, I'd run it on a Ferrari 4000, and not  been 
very
thrilled.  It worked, but at the time, the available graphics support fell far 
short
of taking advantage of resolution, let alone acceleration;  the advanced 
features
of the touch pad didn't work, forget about the BlueTooth mouse (which still 
wouldn't
work AFAIK although that is on the way), the SD card reader didn't work.  More 
of
that (the graphics, and one or both of the SD card reader and the advanced touch
pad support) might work now, although perhaps with some scrounging rather than
_all_ out of the box.  I don't know - the Ferrari died, and I haven't found a 
cheap local
option for getting it fixed yet.

I do have a Mac Mini now, and if its internal disk were a little larger, I 
might try x86
on there for kicks.  I wonder if x86 could boot off of a USB drive (AFAIK MacOS 
can,
but without some unsupported hacky stuff, XP can't; don't know about any 
others).
I don't really want to run MacOS off the USB drive for performance reasons, and 
I doubt
I'd be running x86 on it all that often; if I want to run Solaris, I've got an 
IPC, a SPARC 10,
a Voyager, a Sun Blade 100, and a Sun Blade 2000 (the latter two running 24/7) 
around
the house.  (as you might gather, I've used Solaris since 2.0, but 99.9% on 
SPARC; before
that, SunOS 4.x on SPARC and mc68k, and even some SunOS 3.x on mc68k I 
think...not
to mention Domain/OS on Apollos, SVR2 and 3 on 3B15's, version 7 on PDP-11's, 
...)

I may play with running x86 under VirtualBox on the Mac; I expect that should 
work
ok, subject to the limitations of  VirtualBox as to USB support and such (and 
some rumors
that Leopard's USB support is a little flaky - BitPim doesn't much like it for 
instance).
Although I don't usually use Linux, I had tried Ubuntu Studio under VirtualBox 
on the Mac
so I could run Rosegarden, but the performance made it less that useful for a 
more or
less real-time app like that; I ended up paying for Harmony Assistant running 
native on
Mac OS.  (I'd love to see something like Rosegarden running on Solaris; alas, 
it's tied to
jack+ALSA, and while OSS4 may solve a lot of audio problems, it doesn't 
necessarily
help getting apps that use something else ported.  OTOH, one of the Rosegarden 
developers
reportedly wants to work on a Mac OS port, having found the whole 
audio-on-Linux situation
to be just too darn frustrating.  Since OSS4 _does_ run on a lot of OSs 
_other_than_ just Linux,
I _hope_ it's not too late for it to catch on enough to get a decent level of 
app support.)

It frustrates me when folks like the OP, from one bad experience, imply that 
Solaris
doesn't do a lot of things that it does.  OTOH, when other folks generalize 
their range
of happier experiences in a way that suggests it _already_ works as many places 
as
(for example) Linux does, I'm sorry, but that sort of implied (or readily 
infer-able,
if you didn't intend to imply it) exaggeration doesn't seem to me to be helpful 
either.

The situation has gotten much better over time (just the one example of the 
Ferrari that
I mentioned shows that).  And there's every reason to believe it will continue 
to improve.
Will Solaris on x86 ever support as wide a variety of hardware as Linux does?  
I don't know,
although unless individuals doing unpaid work help with writing and porting 
drivers, I
tend to doubt it, since in  all probability businesses will invest mainly in 
work that results
in hardware or support contract sales, and sometimes further for the sake of 
mindshare
that might get future sales.  All of which will tend to favor newer hardware, 
some perhaps
modestly priced, but not the sort of freebie hand-me-downs that broke students 
and such
might want to try out.

Then too, some of the difference may be a matter of semantics or expectations.  
Does it
constitute "working" if a system boots and is more or less usable, or do you 
expect most or
all functionality of all the peripherals to be supported?  Look at all the 
complaints about
RAID cards, various Ethernet and WiFi NICs, etc.   Some of these just take some 
tweaking,
but even that most people probably expect to just work without that.  Others, 
sorry, no
driver.  (never mind that there were some really crap RAID cards sold, not to 
mention
flaky BIOS support, and such.  No, I don't work with peecees much, but 
sometimes I sit
near some folks that do, and the horror stories I've heard, even with Windows 
or Linux
theoretically having driver support...)

So what I'd welcome is people saying "it's a lot better than it's been, and 
continues to
improve; here's some of the recent improvements, and here's some that have been
mentioned as forthcoming".   That, along with a pointer to the HCL, seems to me 
to
be a balanced response, sufficiently free of blatant advocacy that if it  turns 
out that
whatever the complainer wanted to do still doesn't work, at least they might 
realize
they could try again later, rather than feeling mislead and going away for good.
--
This message posted from opensolaris.org
_______________________________________________
opensolaris-discuss mailing list
[email protected]

Reply via email to