[Baypiggies] November talk?

[jim stockford]

    alec's suggestion is for a 30 to 40 minute hand-waving
intro followed by expert commentary for an hour or two.
    seems great to me, can we arrange to get the time from
google wrt using their meeting room(s)? if not, is it
possible to split the talk into two, one for one month and
one for the subsequent month?


On Oct 27, 2007, at 6:38 PM, Shannon -jj Behrens wrote:

> On 10/27/07, Alec Flett <alecf at flett.org> wrote:
>> Here's an idea: a roundtable on concurrency. Rather than trying to do 
>> a
>> single, monolithic top-to-bottom of every approach, we cover the 
>> basics and
>> then delve into a more interactive session where people discuss 
>> specific
>> issues that they care about.
>>
>> I personally think we could have one person do a handwavy talk for 
>> 30-40
>> minutes about the 3-4 major Python concurrency approaches:
>> - forking - basic process behavior, IPC approaches
>> - basic threading - what's a thread, what's a lock
>> - async I/O - how it works, scheduling issues
>> - stackless - what is it, what's a channel?
>>
>> Leaving coroutines and erlang as being too advanced for the average 
>> person
>> interested in Python concurrency...
>
> Looks like we agree.
>
>> Then, you have some "experts" (maybe one per pattern) sit and have a
>> discussion in front of the rest of the group for an hour or two.
>>
>> I think I'd be willing to even be the dummy who handwaves about these 
>> 4
>> approaches if I could find 4 willing experts to correct my naive
>> simplifications in a constructive way, through the roundtable 
>> discussion. I
>> am by NO means an expert in any of these areas but I understand the 
>> concepts
>> well enough, and have dabbled just enough in each of them to get 
>> myself in
>> trouble. As a bonus, I think I have a guy who would be willing to 
>> claim some
>> expertise in Stackless.
>
> Libor and I are both experts in stackless.  In fact, egroups,
> IronPort, Slide, and the open source version of stackless all came
> from the same lineage.
>
>> Here's the catch: I just can't justify the time commitment (to myself 
>> or my
>> family) to truck down to Mountain View from Berkeley without a car, 
>> and
>> without living near BART! (I don't know how jj does it from his 
>> place, if he
>> does!)
>
> I volunteer a ride ;)  (925) 209-6439.
>
>> If there is no SF-based venue that people usually use, I can probably 
>> offer
>> up space at my company (Metaweb) at Montgomery & Howard in SF (near 
>> the
>> Montgomery BART and Transbay terminal) - This doesn't have to be one 
>> of the
>> monthly BayPiggies meetings, just a one-off SF-based meeting with a 
>> specific
>> focus.
>
> Since I can give you a ride, I think it's best to have it at google.
> Everyone knows where it is, and I think it's a pretty good venue.
>
>> Is this something people would be interested in?
>
> Yep.
>
> Drew, can you give a 10 minute overview of how Twisted uses callbacks
> and continuation passing style on top of select, etc.?
>
> Maybe it's just because I'm a Web guy, but I think it's helpful to
> give the talk from the point of view of someone writing a Web server.
> That's something that's well known and understood.  Hence, we
> shouldn't cover Twisted's Web server implementation, but rather how
> you would implement a Web server in Twisted vs. how you would
> implement it with threads.
>
> Best Regards,
> -jj
>
> -- 
> I, for one, welcome our new Facebook overlords!
> http://jjinux.blogspot.com/
> _______________________________________________
> Baypiggies mailing list
> Baypiggies at python.org
> To change your subscription options or unsubscribe:
> http://mail.python.org/mailman/listinfo/baypiggies
>