Re: what's next for the linux kernel?

Re: what's next for the linux kernel?

[Luke Kenneth Casson Leighton]

On Tue, Oct 04, 2005 at 05:37:59PM -0400, Chris wrote:

> You write:
> |  personally i find that i like a bit of a run-up and/or advance notice
> |  of major paradigm shifts.
> 
>  I maintain that one of the problems is that we simply do not know how
> to efficiently (in all senses of the word) program multi-processing
> systems and applications. Various people have various theories about how
> it should be done, but they are not well-proven.
> 
>  We do know that the current approach is difficult and error-prone, but
> terribly attractive because of an at least superficial simplicity. So
> far none of the alternatives have been convincing enough to supplant it.
> 
>  Efficient changing of paradigms requires a target paradigm. In this
> environment, a good target paradigm does not appear to exist yet and
> may not be clear for years. Starting engineering work seems a little
> bit premature in that situation.
 
 chris,

 your words paraphrase nicely the issues faced - for software engineers.

 however, if you were to speak to a hardware engineer - an embedded
 systems designer of ASICs - they would have a completely different
 take on it, because they are dealing with parallelism all the time,
 _and_ they even have some tools to help them do that.

 what those ASIC designers lack is an off-the-shelf affordable
 multiprocessor chip that makes their parallel algorithms run
 fast enough: they _have_ to go for custom ASIC.

 which is where the pricing of 90nm - $250k mask charges making up most
 of the $2m NREs - is so key.  even .13micron you're looking at $200k
 NREs [non-recoverable expenditure].
 
 and if you were to do a chip at .13micron you would be
 looking at 120 watts for something running at 800mhz with
 only 1 million gates.
 
 a pentium 3 800mhz in other words, and that won't exactly get you very
 far.


 what the software engineers you refer to above lack is the toolchain to
 assist them in developing for anything other than uniprocessor targets.

 such tools and techniques are being researched and developed (i sent
 references to a coupl and also a google search criteria in another
 earlier LKML email): however, yet again, it's chicken-and-egg.

 until the chips start going multiprocessor, the tools are going to
 remain in research labs.  until the tools come out of research labs,
 the chips are going to remain useless.

 where it all goes a bit pearshaped with that chicken-and-egg vicious
 cycle is if the bottom drops out of 65nm and 45nm processes, such that
 _even_ the top uniprocessor mass-market chip manufacturers are forced
 down a parallel processing line.

 my point is: we're starting to see evidence of that happening
 (small-scale, 2-cores, 2-hyperthreads, talk of 4-cores, etc.
  even the X-Box 360 PPC 3x2)

 _therefore_, i invite people who do linux kernel development
 to think ahead - to take a _lead_ for once instead of waiting
 for hardware to drop into their laps, at which point it is once
 again too late, the hardware design decisions will have
 already been made by someone else, and you will be treated
 like second class citizens.  again.

 l.

Re: what's next for the linux kernel?

[Jens Axboe]

On Wed, Oct 05 2005, Luke Kenneth Casson Leighton wrote:
>  _therefore_, i invite people who do linux kernel development
>  to think ahead - to take a _lead_ for once instead of waiting
>  for hardware to drop into their laps, at which point it is once
>  again too late, the hardware design decisions will have
>  already been made by someone else, and you will be treated
>  like second class citizens.  again.

You seem to be suffering the same fate as other 'pocket visionaries'
with grand visions on this mailing list. The combination of 'listen to
this awesome idea' (*) and someone who hasn't contributed anything
interesting is not very good. Add a pinch (or more) of 'I know better'
and you've got a great cocktail for a lengthy and tiresome thread, but
not for anything productive.

Why is that so hard to understand? Succesful contributions start at the
technical level, always have.

(*) Such an idea is often described in great text detail, but is often
greatly lacking in technical detail.

-- 
Jens Axboe

Re: what's next for the linux kernel?

[Luke Kenneth Casson Leighton]

On Wed, Oct 05, 2005 at 02:31:14PM +0200, Jens Axboe wrote:

> [i know better criticism and accusations deleted and not commented on]

> Why is that so hard to understand? Succesful contributions start at the
> technical level, always have.
 
 then we will have to agree to disagree, because i believe that
 successful contributions start with "what creative thing shall
 we do now / what problem shall we tackle today in a creative
 way?" and work their way down to the technical level, which,
 as you rightly point out, requires successful _technical_
 contributions.

 l.

Re: what's next for the linux kernel?

[Jens Axboe]

On Wed, Oct 05 2005, Luke Kenneth Casson Leighton wrote:
> On Wed, Oct 05, 2005 at 02:31:14PM +0200, Jens Axboe wrote:
> 
> > [i know better criticism and accusations deleted and not commented on]
> 
> > Why is that so hard to understand? Succesful contributions start at the
> > technical level, always have.
>  
>  then we will have to agree to disagree, because i believe that
>  successful contributions start with "what creative thing shall
>  we do now / what problem shall we tackle today in a creative
>  way?" and work their way down to the technical level, which,
>  as you rightly point out, requires successful _technical_
>  contributions.

It's not my opinion, I'm stating historical fact. I honestly can't
remember when a succesfull contribution was made based on a long
non-technical thread. I can, however, remember oodles of cases where the
reverse is quite true.

-- 
Jens Axboe

Re: what's next for the linux kernel?

[Luke Kenneth Casson Leighton]

On Wed, Oct 05, 2005 at 03:40:43PM +0200, Jens Axboe wrote:
> On Wed, Oct 05 2005, Luke Kenneth Casson Leighton wrote:
> > On Wed, Oct 05, 2005 at 02:31:14PM +0200, Jens Axboe wrote:
> > 
> > > [i know better criticism and accusations deleted and not commented on]
> > 
> > > Why is that so hard to understand? Succesful contributions start at the
> > > technical level, always have.
> >  
> >  then we will have to agree to disagree, because i believe that
> >  successful contributions start with "what creative thing shall
> >  we do now / what problem shall we tackle today in a creative
> >  way?" and work their way down to the technical level, which,
> >  as you rightly point out, requires successful _technical_
> >  contributions.
> 
> It's not my opinion, I'm stating historical fact. I honestly can't
> remember when a succesfull contribution was made based on a long
> non-technical thread. I can, however, remember oodles of cases where the
> reverse is quite true.
 
 ah.

 who's hosting the linux-visionaries mailing list, then?

Re: what's next for the linux kernel?

[Jens Axboe]

On Wed, Oct 05 2005, Luke Kenneth Casson Leighton wrote:
> On Wed, Oct 05, 2005 at 03:40:43PM +0200, Jens Axboe wrote:
> > On Wed, Oct 05 2005, Luke Kenneth Casson Leighton wrote:
> > > On Wed, Oct 05, 2005 at 02:31:14PM +0200, Jens Axboe wrote:
> > > 
> > > > [i know better criticism and accusations deleted and not commented on]
> > > 
> > > > Why is that so hard to understand? Succesful contributions start at the
> > > > technical level, always have.
> > >  
> > >  then we will have to agree to disagree, because i believe that
> > >  successful contributions start with "what creative thing shall
> > >  we do now / what problem shall we tackle today in a creative
> > >  way?" and work their way down to the technical level, which,
> > >  as you rightly point out, requires successful _technical_
> > >  contributions.
> > 
> > It's not my opinion, I'm stating historical fact. I honestly can't
> > remember when a succesfull contribution was made based on a long
> > non-technical thread. I can, however, remember oodles of cases where the
> > reverse is quite true.
>  
>  ah.
> 
>  who's hosting the linux-visionaries mailing list, then?

:-)

Don't think there is one, but if it will take the traffic of this list,
I'm sure davem can be quickly convinced to add one!

-- 
Jens Axboe

Re: what's next for the linux kernel?

[Steven Rostedt]

On Wed, 5 Oct 2005, Jens Axboe wrote:

> >
> >  who's hosting the linux-visionaries mailing list, then?
>
> :-)
>
> Don't think there is one, but if it will take the traffic of this list,
> I'm sure davem can be quickly convinced to add one!
>

Amen!

(noise to content is very high lately)

And the sad part is that I keep reading this crap.  It's like being sick
at home one day and getting stuck on a bad soap opera.  You know the stuff
you're watching is garbage, but you just can't get away and stop watching
it.

At least if it moved to another list I'd be strong enough not to
subscribe. :-)

-- Steve

Re: what's next for the linux kernel?

[Luke Kenneth Casson Leighton]

On Wed, Oct 05, 2005 at 12:38:35PM -0400, Steven Rostedt wrote:
> At least if it moved to another list I'd be strong enough not to
> subscribe. :-)
 
 and i'd feel less guilty about replying oh crap.

Request for starting a LVML (was: Re: what's next for the linux kernel?)

[Luke Kenneth Casson Leighton]

On Wed, Oct 05, 2005 at 05:51:39PM +0200, Jens Axboe wrote:
> On Wed, Oct 05 2005, Luke Kenneth Casson Leighton wrote:
> > On Wed, Oct 05, 2005 at 03:40:43PM +0200, Jens Axboe wrote:
> > > On Wed, Oct 05 2005, Luke Kenneth Casson Leighton wrote:
> > > > On Wed, Oct 05, 2005 at 02:31:14PM +0200, Jens Axboe wrote:
> > > > 
> > > > > [i know better criticism and accusations deleted and not commented on]
> > > > 
> > > > > Why is that so hard to understand? Succesful contributions start at the
> > > > > technical level, always have.
> > > >  
> > > >  then we will have to agree to disagree, because i believe that
> > > >  successful contributions start with "what creative thing shall
> > > >  we do now / what problem shall we tackle today in a creative
> > > >  way?" and work their way down to the technical level, which,
> > > >  as you rightly point out, requires successful _technical_
> > > >  contributions.
> > > 
> > > It's not my opinion, I'm stating historical fact. I honestly can't
> > > remember when a succesfull contribution was made based on a long
> > > non-technical thread. I can, however, remember oodles of cases where the
> > > reverse is quite true.
> >  
> >  ah.
> > 
> >  who's hosting the linux-visionaries mailing list, then?
> 
> :-)
> 
> Don't think there is one, but if it will take the traffic of this list,
> I'm sure davem can be quickly convinced to add one!
 
 davem, can we have a LVML, pleeeease?

 introductory text - capitalised of course:

 "the LVML - linux visionaries mailing list - is a place where linux
 kernel developers can gently but firmly push those people with
 'A Vision for Linux' so that they can get on with ...errmm...
 developing the linux kernel.  it's the place where 'i would love it if'
 can be frequently heard and revered, without fear of retribution
 or requests for an accompanying kernel patch."

 guidelines for posting on LVML:

 the list should be full of discussions beginning:
 
	 "wouldn't it be great if...", "has anyone considered ...",
	 "i have an idea".  "i am thinking of / considering ..."
 
 with intermediate stages comprising:

	 "that's a good idea because X but the problems are Y" or
	 "that is unrealistic because A, but a better or existing way
	  to do it is B"
 
 concluding in:
 
	 "okay, any takers for actually implementing this", "i've
	 implemented it", "i've started implementing it and really
	 now need to contact the LKML for a technical brain-picking
	 session: wish me luck".

 the list should _not_ be full of "that's a rubbish idea; don't even
 waste our time with it; prove it by implementing it and we won't speak
 or listen to you until you _have_ implemented it".
 

 you should really have a reasonable degree of expertise in at least one
 of the following areas (please advise LVML if you believe there should
 be any additions to this list.  the criteria for such items on this
 list is there to encourage people with a breadth or depth in
 specialist areas of expertise to get invovled.  for example, being a
 babbling idiot requires a _significant_ amount of talent):

	 * understanding of the history of kernel development:
	   unix, NT, netware, beos, mach, l4, gnu/hurd, you name it,
	   you've dissected it, met the developers, _are_ the developer,
	   or broken it so badly that people ran away screaming.

	 * high degree of exposure to a _significant_ number of
	   different programming languages.

	 * knowledge, expertise or exposure to a significant number
	   of futuristic, experimental, modern, ancient, obscure or
	   archaic hardware platforms, of any form: valves,
	   relays, transistors, gates, optical, magnetic -
	   you name it, you should have messed with it and
	   preferably received an overdose of some kind from it.

	 * the ability to absorb the contents of several high traffic
	   free software mailing lists: typically over 1,000 messages
	   per day.  people with adverse side-effects on brain function
	   of such a high degree of exposure to free software mailing
	   lists will be welcomed with open arms.

	 * the ability to converse happily with virtually any free
	   software developer, with or without beer, at conferences
	   and user groups, and to be able to make them think - even
	   if it's to think that you're a babbling idiot.

	 * talking complete and utter and very convincing nonsense.

	 * the ability to keep coming up with new ideas despite
	   clear and unequivocal contradictory statements from
	   virtually everyone around you that your ideas suck
	   goats.

 you should also have a high degree of tolerance for incoherent babbling
 idiots who at least show some some sparks of intelligence underneath.

 a full understanding of the linux kernel tree and its history
 is optional, except if you are a babbling idiot, under which
 circumstances you are expected to have a photographic recall
 of the entire linux source tree all the way back to its
 derivation.

 a liking for or tolerance of reactos, the gnu/hurd or any
 other experimental free software operating system is preferred.


 how about it?
 
-- 
--
<a href="http://lkcl.net">http://lkcl.net</a>
--

Re: what's next for the linux kernel?

[Dave Neuer]

On 10/5/05, Luke Kenneth Casson Leighton <lkcl@lkcl.net> wrote:
>
>  where it all goes a bit pearshaped with that chicken-and-egg vicious
>  cycle is if the bottom drops out of 65nm and 45nm processes, such that
>  _even_ the top uniprocessor mass-market chip manufacturers are forced
>  down a parallel processing line.
>
>  my point is: we're starting to see evidence of that happening
>  (small-scale, 2-cores, 2-hyperthreads, talk of 4-cores, etc.
>  even the X-Box 360 PPC 3x2)
>
>  _therefore_, i invite people who do linux kernel development
>  to think ahead - to take a _lead_ for once instead of waiting
>  for hardware to drop into their laps, at which point it is once
>  again too late, the hardware design decisions will have
>  already been made by someone else, and you will be treated
>  like second class citizens.  again.

With all due respect (and I do believe some is due); your comment
above makes no sense. Operating system designers design software to
operate on existing systems -- often doing as much as possible to
ensure that the design supports lots of different systems. However, _a
neccessary prerequisite_ for that activity is "hardware dropping into
their laps." You warn that "the hardware design decisions will have
already been made by someone else;" but that is the order in which it
_neccessarily_ works! What hardware designer out there spends his day
looking around for as-yet-unused (and hence untested) operating
systems that run on some hypothetical yet-to-be-designed hardware and
say to themselves, "ah, that looks nice -- I think I can design some
hardware that will run that software _really well_?"

Your argument defies logic, and it's that fact -- not your intentions
-- that makes this thread so tiresome.

Dave

Re: what's next for the linux kernel?

[Luke Kenneth Casson Leighton]

>  With all due respect (and I do believe some is due); your comment above
> makes no sense. 

 replying privately. l.