[Stackless] Re: [Python-Dev] Stackless Design Q.
Samuel Smith
smithsm at samuelsmith.org
Tue Feb 26 01:15:14 CET 2002
>
>There should be a non blocking way for tasklets to communicate.
>That is one writer but many readers
>With potentially thousands of tasklets to communicate with each
>other a publish subscribe paradigm is more appropriate than a
>message passing paradigm.
>Blocking should be the exception rather than the rule.
>
>Component based architectures for multiple autonomous agents are based on a
>framework where data is shared transparently and one agent (tasklet) doesn't
>need to know about any other agent(tasklet) only about the structure
>of the data
>shared and what portions of the data share a given agent can write too.
For an example of what I mean see the enclosed article
--
**********************************************
Samuel M. Smith Ph.D.
360 W. 920 N.
Orem, Utah 84057
801-226-7607 x112 (voice)
801-226-7608 (fax)
http://www.samuelsmith.org (web)
*********************************************
-------------- next part --------------
A non-text attachment was scrubbed...
Name: %dstewart.tse97.pdf
Type: application/applefile
Size: 128 bytes
Desc: not available
URL: <http://www.stackless.com/pipermail/stackless/attachments/20020225/4ba80e1a/attachment.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: dstewart.tse97.pdf
Type: application/pdf
Size: 883357 bytes
Desc: not available
URL: <http://www.stackless.com/pipermail/stackless/attachments/20020225/4ba80e1a/attachment.pdf>
More information about the Stackless
mailing list