573603 - Speed up Page Info > Media

Status: RESOLVED FIXED

(Firefox :: Page Info Window, defect)

Description

[Tanner M. Young [:tmyoung]]

Attachment: Patch v. 1.0

From the discussion in Bug 515549, let's do two--possibly three--things to speed up Page Info's media.

1. Lower the timeouts to the minimum of 10ms
2. Increase the number of elements processed at once
3. (Optional) Replace timeouts with postMessage as described here: http://dbaron.org/log/20100309-faster-timeouts

This patch does the first two, and it can do the third if needed.

http://commons.wikimedia.org/wiki/Category:Crater_Lake is a decent example of speed testing if you run "View Image Info" on the last image.

Comment 1

[Daniel Brooks [:db48x]]

We need longer testcases. Page Info (trunk) has finished with every page I've tried so far before I could even click on the Media tab.

Comment 2

[Tanner M. Young [:tmyoung]]

Attachment: Patch v. 1.1

Setting to 500 elements per loop.

Comment 3

[Daniel Brooks [:db48x]]

Comment on attachment 452905 [details] [diff] [review]
Patch v. 1.1

r=db48x

Comment 4

[Daniel Brooks [:db48x]]

http://hg.mozilla.org/mozilla-central/pushloghtml?changeset=a32ea115521e

Comment 5

[Tanner M. Young [:tmyoung]]

Comment on attachment 452905 [details] [diff] [review]
Patch v. 1.1

From bug 579717 this appears to fix hangs and performance problems on Mac OSX--more than it does on Windows.  I propose that we port the change to 1.9.2.

I will quickly make sure that the patch doesn't conflict with anything, but it shouldn't touch code that's 2.0-branch specific.

Comment 7

[Mike Beltzner [:beltzner, not reading bugmail]]

Comment on attachment 452905 [details] [diff] [review]
Patch v. 1.1

Please renominate for branch landing once you've confirmed that it compiles on that branch, preferably with tryserver builds and test runs! :)

Comment 8

[Tanner M. Young [:tmyoung]]

Comment on attachment 452905 [details] [diff] [review]
Patch v. 1.1

This works on 1.9.2 like it does on Trunk.  The patch doesn't need any branch-specific changes.

Comment 9

[Mike Beltzner [:beltzner, not reading bugmail]]

Comment on attachment 452905 [details] [diff] [review]
Patch v. 1.1

a=beltzner for mozilla-1.9.2, simple safe perf improvement for non-critical code path

Comment 10

[Tanner M. Young [:tmyoung]]

This needs check-in on the 1.9.2 branch.

Comment 11

[u88484]

Can the same type of fix be applied for bug 513092?

Comment 12

[Tanner M. Young [:tmyoung]]

(In reply to comment #11)
> Can the same type of fix be applied for bug 513092?

It's all a question of the code that populates that list.  Either the code is using timeouts or it's waiting on something else.  If it is the former, then yes, this same philosophy can be applied, but if it is the latter, I'm not sure.

Since I don't know that code or what would be causing it, I couldn't say, but someone like Dao or Gavin would have better answers.

Comment 13

[Dão Gottwald [:dao]]

http://hg.mozilla.org/releases/mozilla-1.9.2/rev/3b265279e11b