> I am frustrated to say that people repeatedly miss
> the point.  To remind - Those drivers were written
> long back ago (before OpenSolaris) by single person
> for his own cause and he was kind enough to make them
> available.

Incorrect. Those drivers (by Murayama-san) are still in development, and 
support most modern, easily accessible and mainstream hardware. Please consult 
"Free NIC drivers for Solaris" on Google and you will see what I mean.

> How long did it take before they will be
> in Solaris? Is that kind of speed (I wouldn't call it
> speed for god's sake) acceptable? Why are not 100
> people like that guy rolling up their sleeves and
> writing/reviewing/committing more of them? Because
> people do not feel like it would be for their cause
> to make an effort. Because people feel there is
> nothing for them to gain there. Because they know
> they will have to work in somebody else's accordance
> and for someone else's benefits. That was the point -
> it needs to be looked into as to how we can increase
> community participation.

I fail to see how integration in Solaris proper stopped any of us, including 
yourself, to download the drivers from his web page and use them. I even made 
my own build system around his drivers that automatically compiles them and 
packages them for me. Works like a charm, with or without being integrated into 
Solaris proper or OpenSolaris.

My point: so long as software for Solaris is available, it is mostly irrelevant 
whether it comes with Solaris stock or not. Would it help / make things easier? 
Well sure it would. Is it absolutely essential  or a show stopper for you to 
roll up your sleeves and repeat what Murayama-san did? No, it is not.

So here is what I suggest: write some drivers of your own and post them on the 
web somewhere. Then you can take your sweet time and work out the kinks with 
the people who know Solaris and have them audit your code. And once you've 
worked the kinks out, it will show up in Solaris. Meanwhile, users can use the 
stuff downloaded from your web page, should they choose to do so.

In the end, Solaris gets rock-solid, high-perf drivers and users get a relief 
measure while that is being worked out. Everybody wins.

> That is of NO USE to ME - I cannot run it on the
> hardware I choose. I cannot wait till the hardware in
> question becomes obsolete for your so called quality
> and process enabled drivers to show up.  I do not
> want to buy Sun hardware for my own justifiable
> reasons. You forget all of that and continue
> trumpeting quality and process. Quality and Proceess
> - for what? Sun's business yes. Can I use that
> quality and process now, today? No. Does it work for
> me today? No. So what was your point again?

I find that hard to believe. Sure I've had some very stubborn / uncooperative 
hardware, but in the end I forced Solaris to run on it, one way or another. 
Sometimes the solution was as simple as -B acpi-user-options=0x2, sometimes 
not. But Solaris always worked.

Did you detail your problem with this hardware anywhere? Can we help you TODAY?

> I am trying to highlight the need for accelerating
> changes and "getting there".  

And that's all fine, but if by acceleration you mean ad hoc implementation, 
then perhaps Solaris is not for you. Solaris and ad hoc don't go together in 
one sentence.

Neither should acceleration mean "let's change root's shell to /bin/bash 
because that's what I'm confortable with from Linux". If you chose to be part 
of the Solaris community, then choose to learn Solaris the way Solaris is.

And most importantly, seek to understand why things are done the way they are. 
There are actually reasons why things are done the way they are, and they don't 
have anything to do with Sun Microsystems per se.

> o Solaris is about slowness - making drivers when the
> hardware becomes obsolete? Letting years pass by
> without making things work?

I'm truly sorry that I don't quite understand what you're trying to communicate 
to me.
I'm confused by your statements, since I know that Solaris supports most 
mainstream hardware on the i86pc platform. I run Solaris on *all* my i86pc 
systems, from my laptop to my desktop PC bucket to my i86pc servers.

> In other words if a
> restaurant makes high quality food in 4 days of time,
> each day people will flock away to other restaurants
> to satisfy their hunger as long the other restaurant
> does give them something reasonable. That's what has
> happened with Linux - it is good enough and does what
> people want it to do and it is free. Why do I need to
> wait for years just to make it run on my hardware
> when Linux runs on it today and if it doesn't run the
> way I like it - I can just fix it up and propagate
> those changes for the world to consume - easy, and
> makes me happy at the end.

Linux might satisfy some very limited scenarios, such as a desktop PC operating 
system, or a deployment with a few servers. But don't risk upgrading your 
kernel or your drivers -- you risk losing a working system. Yes, there are ways 
to do this without too much pain, but we come back to expertise and proficiency.
Also, Linux fails miserably in large enterprise deployments, because the thing 
is simply not designed for server farms with thousands of systems on them.
That's why your local ATM, or your bank or even your insurance will never be 
powered by Linux and why they will always either be run on the z/OS mainframe 
or on Solaris.

I'm under the impression your gripe is with lack of support for your desktop 
hardware. Perhaps I'm wrong. Please do bear in mind that Solaris MUST, with no 
exceptions, provide exceptional levels of service on ALL levels, so someone's 
desktop, while being important indeed, must share priorities with other tasks, 
such as the enterprise, which at this times pays the bills and funds Solaris 
development.

This is not a rebellion project like Linux was/is. This is an engineering 
project.
 
 
This message posted from opensolaris.org
_______________________________________________
opensolaris-discuss mailing list
[email protected]

Reply via email to