RFC: startup time - again

[Robert Collins]

On Wed, 2008-09-10 at 09:35 +0200, Matthieu Moy wrote:
> Alexander Belchenko <bialix at ukr.net> writes:

> >
> > I disagree: human reaction is about 200-300 ms. So 500 ms is too much.
> > Especially comparing to hg speed.
> 
> +1.
> 
> Startup time >= 2s would be a _problem_, but changing a startup time
> from 0.5 to 0.1s is a matter of comfort.

Not at all. Low startup time gives a buffer for performance where actual
work is needed.

E.g. to have bzr st be < 1 second on a mozilla tree, 0.3 second startup
time gives us 0.7 seconds for actual stat code; 0.1 second startup gives
us 0.9 seconds for actual stat code.

Right now, 'bzr st' of a bzr tree: 350 ms., 'bzr st' of a empty tree,
253ms - so 100ms to 'st' the bzr codebase, and 253ms to startup, load
all the needed code, and process a root-only tree.

When optimising one normally looks at the most expensive part of the
process - which startup is, for modest code bases (like bzr).

-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: 189 bytes
Desc: This is a digitally signed message part
Url : https://lists.ubuntu.com/archives/bazaar/attachments/20080910/305fb7f3/attachment.pgp