Re: [Monotone-devel] Performance Issues With Version 0.40

[Zack Weinberg]

On Thu, Sep 4, 2008 at 10:35 AM, Timothy Brownawell <address@hidden> wrote:
> On Thu, 2008-09-04 at 09:59 -0700, Zack Weinberg wrote:
>> We should probably just go ahead and do that, but I'd like to know
>> what things we still have to decode_hexenc - the 0.40 database changes
>> ought to have meant we need to do that way less often.
>
> roster_t::parse_from() and parse_marking(). Try "mtn get_roster" in any
> workspace and look at all the pretty hex strings :)

Hm, so maybe we ought to be looking into a binary on-disk
representation for rosters.  I'd thought that had already happened but
I guess I was wrong.  We'd still be stuck with the hex format for
netsync, but we know we need a new netsync anyway.

>> A cleaner approach would probably be to cache the sqlite handle in the
>> app_state.
>
> Or just keep a shared_ptr to the database_impl (or probably a map to
> allow lookup based on filename, wasn't part of the reason for splitting
> db out from app_state to allow for multiple dbs in the future?).

That's basically what I meant.  There's no deep reason for the
scoped_ptr, it's just that at the time there was no need for a
shared_ptr.

zw