Account Options

  1. Sign in
The old Google Groups will be going away soon, but your browser is incompatible with the new version.
Google Groups Home
« Groups Home
Message from discussion Places and Firefox 2
The group you are posting to is a Usenet group. Messages posted to this group will make your email address visible to anyone on the Internet.
Your reply message has not been sent.
Your post was successful
 
From:
To:
Cc:
Followup To:
Add Cc | Add Followup-to | Edit Subject
Subject:
Validation:
For verification purposes please type the characters you see in the picture below or the numbers you hear by clicking the accessibility icon. Listen and type the numbers you hear
 
Brett Wilson  
View profile  
 More options Apr 24 2006, 4:51 pm
Newsgroups: mozilla.dev.planning
From: Brett Wilson <bre...@gmail.com>
Date: Mon, 24 Apr 2006 13:51:36 -0700
Local: Mon, Apr 24 2006 4:51 pm
Subject: Re: Places and Firefox 2
We need to decide soon about sqlite being in or out. We need to change
some of the APIs around and make some things safer. If it's shipping in
2.0, I need to make this a priority. Otherwise, it will stay on the back
burner.

> I am wary of shipping complex platform-level features that do not get
> widespread "natural" testing, and have not received (IMO) sufficient
> targeted testing in the field. The SQLite performance issues that were
> turned up by the places work amplified my initial worry several times
> over.

I would come to exactly the opposite conclusion you have. Places
performs fine now, in some ways better than Mork. Any system requires
tuning to perform well. The previous history system no doubt had a lot
of performance work done on it to make it fast, and sqlite is no
different. I don't see this as a warning sign at all. (Perhaps some of
this feeling is a result of the early impression that was floating
around that sqlite would solve all our problems, which is clearly not
true.) We have produced a system that has been tested it to perform well
under constant use while you browse.

> There are outstanding unresolved issues about how we plan to version the
> SQLite data structures over time within profiles (most of which are
> admittedly app-level issues, not platform-level issues),

I would also like to see this discussion take place. At least we need a
plan, even if we don't actually do anything.

 > as well as

> questions about whether and to what extent we should be supporting
> multi-process access to the data.

The decision is that we don't support this. Supporting it would kill a
lot of the performance. Maybe this decision wasn't made explicitly, but
it will take a lot of work that hasn't been planned and would give a
measurable performance hit no matter what, so I think the decision has
been made implicitly.

Brett


 
You must Sign in before you can post messages.
To post a message you must first join this group.
Please update your nickname on the subscription settings page before posting.
You do not have the permission required to post.