import cost in bzr

[Robert Collins]

On Thu, 2006-05-18 at 18:35 +1000, Martin Pool wrote:
> On 16 May 2006, Alexander Belchenko <bialix at ukr.net> wrote:
> > I do some experiments with bzrlib in interpreter session and eventually
> > discover that execution of simple statement like:
> > 
> > from bzrlib.workingtree import WorkingTree
> > 
> > take a lot of time.
> 
> Yes, and thanks for bringing it up.  
> 
> Robert mentioned a tool used by optparse which preprocesses during
> installation several modules into one large module.  Loading one large
> module from bytecode is allegedly much faster than loading it piece by
> piece.
> 
> However, that may be true only if we actually end up using most of the
> code; for many invocations we possible will not.  Also preprocessing it
> in this way is just a bit kludgy.

this is true. However, I think that having to optimise ones source for a
slow compiler is equally kludgy. We should try to really clearly define
what it is we are trying to achieve : how often is 'bzr' invoked from
scratch for instance?

Rob

-- 
GPG key available at: <http://www.robertcollins.net/keys.txt>.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 191 bytes
Desc: This is a digitally signed message part
Url : https://lists.ubuntu.com/archives/bazaar/attachments/20060518/36134721/attachment.pgp