[Stackless] Stackless debugging

Andrew Francis andrewfr_ice at yahoo.com
Fri Oct 26 17:49:50 CEST 2007


Hello Richard and Colleagues:

> I think what would be killer is a debugger with the
> ability to analyse and manipulate the scheduler.

>I understand.  Unfortunately no-one has had the time
>and interest to step forward and implement one.  But
>that is of course the way stuff works when people do
>it as a hobby.

I feel the "I implement feature X because I need it"
approach will not grow the Stackless community as
quickly as a more coordinated effort. I feel Stackless
deserves to grow. 

In the meanwhile, I can focus on my own efforts :-(

Very soon, I will start running tests to see how well
my prototypes perform and better understand
Twisted/Stackless issues. 

I am very interested in examining the Stackless
scheduler. About two months ago, I started to look at
Stackless code. I am trying to decide whether the
addition of queues will help performance. Again, I
need to run tests. I need to be able to properly
profile Stackless programmes. 

If I do this in November, I can perhaps get the help
of one of my operations research professors to analyze
the algorithms. I have to keep in mind that the
professor is very busy.

What would help (speed things up) is:

1) Can I have a pseudo-code description of the
scheduling algorithm.

2) What pieces of code should I be focusing on? 

3) How do I expose new functions to a debugger?

Cheers,
Andrew





 

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 




More information about the Stackless mailing list