[Design] Chandler in space

[Ted Leung]

On Jun 15, 2006, at 6:50 PM, hank williams wrote:

> If I may ask, where is the greatest source of the performance issue?
> Is python compiled, or interpreted, or compiled to a VM? Are there
> algorithmic issues such as perhaps data retrieval in the repository?

Python is bytecode interpreted (but not JIT'ed) , but I don't know  
that all (or even a substantial portion) of the blame can be laid  
there.   If you talk to different people you will get different  
answers about the source(s) of our problems.

If you're interested in our performance efforts I'd suggest the  
following resources:

1. This message <http://lists.osafoundation.org/pipermail/chandler- 
dev/2006-June/006130.html> will get you access to the monitoring we  
are doing on performance, which will give you a view into our current  
status.   There are some very disturbing lines on some of the graphs  
-- places where Mac time go way up, while lines for the other  
platforms remain flat.

2. This page <http://wiki.osafoundation.org/bin/view/Projects/ 
PerformanceProject> shows you the areas we are currently focusing on,  
what the times are on the test platforms, and our performance target  
goals.   The page includes a description of the target hardware,  
which is significantly slower than any Intel Mac.

3.  The best venue for following performance work is the chandler-dev  
mailing list.

To reiterate Katie's point, we are very concerned about the  
performance of the system.   Those of us using the system on a daily  
basis are *painfully aware* of how far we have to go.

Ted