opinions about lisp on 64 bit 8 cpu machine
gavino
Are there alternatives to get a program to use 8 cores?
Any one care to share opinons about which strategies are working for
them?
jos...@lisp.de
Gavino, today no questions about SAP and Lisp, Oracle and Lisp, the
best Lisp for cloud computing, Lisp for space/time travel?
Why not finish one of the chapters of a Lisp book first?
Maybe you could also troll in your garden?
Have a nice day.
gavino
wrote:
I am considering digging up the gentle intro by touretsky, or perhaps
successful lisp by lamkins..
I am excited about teepeedee2. fast web!
PS I don't like non free software.
for database I am more fascinated by www.prevayler.org or happstack
do you write lisp for a living?
awesome!
what interesting lisp programs have you written?
I saw someone once made a dns server in common lisp.
Then the pages disappeared from google.
weird eh
guy said he made it in very few lines of code.
Pillsy
[...]
> I am considering digging up the gentle intro by touretsky, or perhaps
> successful lisp by lamkins..
Considering?
Look, Touretzky's book is quite good, but it's also exactly what it
says on the tin: "gentle". It's short (maybe 250 pages), clearly
written, has copious examples and exercises, and is geared towards
people without much background in programming. Download and install a
Lisp implementation, and start reading it and doing the exercises.
If you have trouble with the exercises, or don't understand what you
read in the text, that's the time to come back and start asking
questions. You've been posting these questions for years now; if you
spent all that time and effort learning the language, you'd be able to
answer them for yourself.
Cheers,
Pillsy
fortunatus
Why not make an implementation that dispatches the bindings in LET to
individual threads? That's why there's LET and LET*. It's the same
as explicit threads library but simply taking advatage of syntax
already available.
Otherwise no reason not to scan code to find independent expressions
and dispatch.
Pillsy
> On Dec 16, 1:58 am, gavino <gavcom...@gmail.com> wrote:
> > Is is all about bordeaux?
> > Are there alternatives to get a program to use 8 cores?
> > Any one care to share opinons about which strategies are working for
> > them?
> Why not make an implementation that dispatches the bindings in LET to
> individual threads?
[...]
Because the initialization forms in a LET are specified to occur in
the order they appear.
http://www.lispworks.com/documentation/HyperSpec/Body/s_let_l.htm
I find this restriction on implementors to be sort of strange, but
there you go.
Cheers,
Pillsy
Tamas K Papp
> On Dec 16, 3:21 am, gavino <gavcom...@gmail.com> wrote: [...]
>> I am considering digging up the gentle intro by touretsky, or perhaps
>> successful lisp by lamkins..
>
> Considering?
>
> Look, Touretzky's book [...] is geared towards people without
> much background in programming.
Then gavino certainly won't be overqualified.
> Download and install a Lisp
> implementation, and start reading it and doing the exercises.
Hey, hold it right there! What do you think you are doing, asking
gavino to actually PROGRAM, let alone read a book or do exercises? If
he does any of that, one of these days someone might actually ask him
to THINK. This is one of them slippery slope situations. Careful,
gavino, this is a trap.
> If you have trouble with the exercises, or don't understand what you
> read in the text, that's the time to come back and start asking
> questions. You've been posting these questions for years now; if you
> spent all that time and effort learning the language, you'd be able to
> answer them for yourself.
What makes you think that he wants to learn anything?
Tamas
Rahul Jain
> Maybe you could also troll in your garden?
Maybe his garden only has gnomes in it, and it would upset the sense of
community if a troll suddenly started setting up shop there.
--
Rahul Jain
rj...@nyct.net
Professional Software Developer, Amateur Quantum Mechanicist
Rahul Jain
> On Dec 16, 1:58 am, gavino <gavcom...@gmail.com> wrote:
>> Is is all about bordeaux?
>> Are there alternatives to get a program to use 8 cores?
>> Any one care to share opinons about which strategies are working for
>> them?
>
> Why not make an implementation that dispatches the bindings in LET to
> individual threads? That's why there's LET and LET*. It's the same
> as explicit threads library but simply taking advatage of syntax
> already available.
No, LET cannot work like that. That is why QLET was defined. (and
QLAMBDA of course).
Vinay
I remember reading an article in which Dan Weinreb, David Moon and some
one from Azul were discussing something along these lines ...
I don't have the link handy, but I'm sure Google should get it for you ...
Captain Obvious
g> Is is all about bordeaux?
g> Are there alternatives to get a program to use 8 cores?
Sure. You can run 8 processes.
This discussion does not really matter until you have
a concrete task.
Then there are different approaches. You can
do amazing stuff with functions and macros,
but that onlyt makes sense
if you need it.
g> Any one care to share opinons about which strategies are working for
g> them?
I can say that for web serving thread-spawning or multile processes
work well.
For some data processing I've wrote few functions/macros
which automatically spawn worker threads and schdule work
on them.
Captain Obvious
to
individual threads?
Because dispatching to threads have siginifcant overhead. In most cases
you will spend more time dispatching than running code.
Parallel version of mapcar actually makes sense, but it should
be separate from ordinary mapcar bcause it changes semantics.
f> That's why there's LET and LET*. It's the same as explicit
threads
library but simply taking advatage of syntax already available.
Parallel LET will have somewhat different semantics, LET* probably is not
doable
at all.
--- Fidolook 2006 Xmas Edition (HL) 6.0.2800.95 - 24/12/2006 14:46:33
* Origin:
Jaseem. V V
fortunatus
Thanks for straightening me out on that - I had mistakenly thought
that since the bindings in LET are "parallel", the evaluations could
be as well - but I see the clear language to the contrary. Clearly
this is to resolve ordering of side effects in the expressions. I see
that LET and LET* are only provided to allow referencing either the
outer bindings or the inner bindings.
(SETQ X 7)
(LET ((Y (SETQ X 5))
(Z (+ X 2)))
(COND
((= Z 7) 'INDEED)
((= Z 9) 'I-GUESS-NOT)))