Hi everybody (or is that nobody at this stage?), I'll probably come to the next Koeln meeting. Is the date of 7 November already agreed upon now? Will someone give a talk, or are we just going to chat and relax? Arthur
Arthur Lemmens <alemmens@xs4all.nl> writes:
Hi everybody (or is that nobody at this stage?),
Well I just got this message from the archives. Otherwise I wouldn't have seen it, as I was indeed not yet subscribed when you sent this.
I'll probably come to the next Koeln meeting. Is the date of 7 November already agreed upon now?
I assume I can come with you as usual? Many thanks. :)
Will someone give a talk, or are we just going to chat and relax?
It'll be an interesting day either way, I think. ;) Kind regards, Dirk Gerrits
Hi All! Look at what I found here: http://labs.google.com/papers/mapreduce.html regards, Mario.
Hi Mario, mario.mommer@gmx.net (Mario S. Mommer) writes:
Look at what I found here:
At first glance this looks somewhat similar to Connection Machine Lisp, but at a larger scale. Which reminds me, I still have some more references to put on my site... In any case, I hope I speak for everyone when I say: thanks for yet another one. :) Kind regards, Dirk Gerrits
Dirk Gerrits <dirk@dirkgerrits.com> writes:
mario.mommer@gmx.net (Mario S. Mommer) writes:
Look at what I found here:
At first glance this looks somewhat similar to Connection Machine Lisp, but at a larger scale.
Aha... *lisp, you mean? This one seems to me particularily interesting, among other things because it is actually in use. If I understand it correctly, these folks (at google) are using one absolutely monstruous cluster of comodity boxes with this. Also, AFAICT, this is very close to the example given in the NetCLOS paper (longer version).
Which reminds me, I still have some more references to put on my site... In any case, I hope I speak for everyone when I say: thanks for yet another one. :)
Oh, you are welcome. :-) This is all very interesting for me. Regards, Mario.
mario.mommer@gmx.net (Mario S. Mommer) writes:
Dirk Gerrits <dirk@dirkgerrits.com> writes:
mario.mommer@gmx.net (Mario S. Mommer) writes:
Look at what I found here:
At first glance this looks somewhat similar to Connection Machine Lisp, but at a larger scale.
Aha... *lisp, you mean?
I mean this paper: http://fresh.homeunix.net/~luke/misc/ConnectionMachineLisp.pdf I don't remember it mentioning the term *Lisp...
This one seems to me particularily interesting, among other things because it is actually in use. If I understand it correctly, these folks (at google) are using one absolutely monstruous cluster of comodity boxes with this.
That, they do. :)
Also, AFAICT, this is very close to the example given in the NetCLOS paper (longer version).
I haven't gotten around to reading that yet.
This is all very interesting for me.
Same here. I'll update my references page with "MapReduce: Simplified Data Processing on Large Clusters", "Connection Machine Lisp: Fine-Grained Parallel Symbolic Processing", "Making Asynchronous Parallelism Safe for the World", and "Naming and Synchronization in a Decentralized Computer System". Should be up this evening, but Google if you can't wait. ;) Kind regards, Dirk Gerrits
Dirk Gerrits <dirk@dirkgerrits.com> writes:
mario.mommer@gmx.net (Mario S. Mommer) writes:
Aha... *lisp, you mean?
I mean this paper: http://fresh.homeunix.net/~luke/misc/ConnectionMachineLisp.pdf I don't remember it mentioning the term *Lisp...
I'm told that *Lisp was a follow-on that was less ambitious but easier to implement on the Connection Machine. AFAIK *Lisp was used for real programming and CMLisp wasn't. There is a *Lisp simulator in Common Lisp in the CMU AI Repository, but I haven't looked at it closely. They also made a *C language. P.S., I registered this list on Gmane as gmane.lisp.eurolisp. Cheers, Luke (very excited to see Data Parallel style ride again. Maybe I'll get a chance to use it some day :-))
Luke Gorrie <luke@synap.se> writes:
Dirk Gerrits <dirk@dirkgerrits.com> writes:
I'm told that *Lisp was a follow-on that was less ambitious but easier to implement on the Connection Machine. AFAIK *Lisp was used for real programming and CMLisp wasn't.
I see.
There is a *Lisp simulator in Common Lisp in the CMU AI Repository, but I haven't looked at it closely.
Yeah me neither. But perhaps I should, if only to see if there's anything worth stea... borrowing for my own endeavors. ;)
They also made a *C language.
:( Kould we please abolish the See-family of languages and get bak to Kommon Lisp? ;)
P.S., I registered this list on Gmane as gmane.lisp.eurolisp.
Ah cool.
(very excited to see Data Parallel style ride again. Maybe I'll get a chance to use it some day :-))
I second that emotion. :) Although for me, "excited to see Data Parallel style ride again" really equates to being "excited that perhaps you won't have to be actively searching anymore in order to find out that there are other ways to do parallelism than with shared-memory and locks". Kind regards, Dirk Gerrits
mario.mommer@gmx.net (Mario S. Mommer) writes:
Dirk Gerrits <dirk@dirkgerrits.com> writes:
mario.mommer@gmx.net (Mario S. Mommer) writes:
Look at what I found here:
At first glance this looks somewhat similar to Connection Machine Lisp, but at a larger scale.
Aha... *lisp, you mean?
I mean this paper: http://fresh.homeunix.net/~luke/misc/ConnectionMachineLisp.pdf
These folks had a helluva time writing this paper, I reckon. Indeed, the map/reduce mechanism seems to be a simplified version of a subset of CMLisp. I'm also having a helluva time reading this paper. Among other things I am happy to know that there exists prior art. Regards, Mario.
Hi all, tomorrow, Monday 25. October 2004 at 19:00 hours, the second monthly Berlin LISP User groups' meeting takes place at c-base (http://c-base.org/). Please join in to talk about LISP and associated topics. Die c-base befindet sich im Erdgeschoß des zweiten Hinterhauses in der Rungestraße 20 in Berlin Mitte. Falls die Tür im Durchgang abgeschlossen ist, muß man hinten am Ufer durch die Glastür gehen. Nicht verzagen. Greetings, Hans
participants (6)
-
Arthur Lemmens -
Dirk Gerrits -
Hans Hübner -
Luke Gorrie -
Mario Mommer -
mario.mommer＠gmx.net