|
From: | Graydon Hoare |
Subject: | [Monotone-devel] Re: Improving the performance of annotate |
Date: | Wed, 19 Jul 2006 11:34:16 -0700 |
User-agent: | Thunderbird 1.5.0.4 (Windows/20060516) |
Eric Anderson wrote:
Any guess as to how difficult that would be? Given your description, I don't think I would be qualified to make the change. If it would take a while to make this change, is it worth transitioning through a relatively ugly hack to make it tolerably fast while work on the correct fix progresses?
Difficulty? Not very high. I could probably sketch out the basic parts in code in a couple days of hacking. Assuming I can digest njs' email describing problems related to the storage of markings.
Finding a couple days to do it in might be a challenge. I'll try; that's the best I can do.
If you don't have write access to our venge.net server yet, please ask njs to grant it; you should at very least land your work on a "eric-super-performance-hacks" branch. It's very helpful to get concrete numbers and confirmation-of-hunches.
As far as committing it to the mainline, well.. that's pretty subjective. I'm sort of passive these days so I'd just as soon wait for something pretty. But if you're in a hurry, well, I see in the patch two categories of change: the hacks to parsing, and the global flags that disable verifications. Both are reasonably easy to "merge now, revert later"; but I'm hesitant to accept disabling verification even if it's temporary. The parsing hacks seem harmless enough (especially since only annotate uses them).
-graydon
[Prev in Thread] | Current Thread | [Next in Thread] |