Re: for the "love of HURD"

[arnuld]

> On 10/11/06, Richard Braun <address@hidden> wrote:
>
> The problem is simple : we can't begin anything until we know what we
> want, until goals and ways to achieve them are designed and apply well.
> This is still not the case. It looks like we're not far from this point,
> but we haven't reached it yet.

i thought that HURD architects already know this. now you made it
clear that we do not even know about the 1st phase. from experience i
know nothing fruitful can happen untill we do the things you mentioned
here. tell me how can help in phase-1 of HURD development.

> Please understand that development isn't a matter of "I suggest we write
> our own microkernel" or "Coyotos is sexy, let's use it !". This
> mailing-list exists to allow people to talk about design decisions,
> implementation details, that sort of things. If you don't include
> technical details about writing a new microkernel and why it's a good
> idea, it is somehow useless... Of course many people already thought
> about writing a new microkernel... But just imagine one of the L4
> kernels gets better, suitable enough for the Hurd, and that hurd-l4
> is restarted, this would be far better than beginning a new kernel...

well, that will save years of our time & will be much better & a thing
like that never happend in 2 decades.

but situation is much, much different now , as we can see microkernels
are in better stage nof development now, so we can expect something
useful.


> If you don't know much about system development, I suggest you get better.

right. i guess it will take 1 year for me to understand "what exactly
a microkernel based OS is & how it operates"

> Study, work on existing projects,

recommend some.  untill now, nobody recommended me "GNU Mach/HURD".
most of the time i got suggestions for MINIX & L4.

what do you say?

> create your own if you want.

recepie for disaster (for a newbie)

> But please,
> don't ask others to do what you want if you don't know what you're talking
> about enough to make such suggestions.

i am *not* asking others to do what *I* want. HURD is not my baby, it
belongs to RMS. i am just a small man. to put it short, i was asking
to complete phase-1, that you mentioned in the beginning of email (
but i said that indirectly)

you are right i dont know much about HURD. i know of only one goal
"extensibily" with some ideas i understand from "Toward a new strategy
of OS design". that article hooked me to OS programming.

thanks for the reply :-)


-- arnuld
http://arnuld.blogspot.com/