https://bugzilla.wikimedia.org/show_bug.cgi?id=13209
--- Comment #19 from Roan Kattouw <[email protected]> 2008-12-15 20:53:41 UTC --- (In reply to comment #18) > Another things I should clarify is that all the diffs that such requests might > generate would have been generated anyway, just possibly in separate requests. > And at least for en.wikipedia RC patrolling, if an edit looks even in the > least > suspicious then it is almost guaranteed that a diff is going to be generated > for it fairly quickly because there is always at least someone patrolling. > That applies to your tool, but the API isn't restricted to your tool. If it's *possible* to blow stuff up, your tool might not use that 'feature', but someone else might. (In reply to comment #17) > It seems that the only person who actually wants this is me, so I guess don't > bother too much about it if it's a lot of work. Things work OK how they are, > at > least nothing has fallen over yet. > It's not that much work, but it's more than just a few lines, which is why I wanna get Brion's approval before I implement this. > But in my case, the answers to these questions: > > * Ideally multiple diffs, but single ones would still mean simpler client-side > code and cleaner error handling. > Single diffs are definitely possible; I'll write a module that returns them later, and leave the possibility of multiple diffs open until Brion's cast his verdict. > I am assuming -- quite possibly incorrectly -- > that it would be easier on the servers if the client was asking for, say, 2 > diffs every 6 seconds rather than 1 diff every 3 seconds. If that assumption > is > complete bollocks then please disregard anything about multiple diffs. :) That makes sense. Retrieval of cached diffs is fast compared to request overhead, generating diffs isn't (AFAIK). But since most diffs are in cache your assumption is valid (although it doesn't matter much in a worst-case scenario, but then those scenarios are rare). > The > trade-off here is minimizing the delay between the user wanting the diff and > having it, while also not being excessive with web requests. I keep said delay > close to zero by pre-loading diffs a few in advance of the one they're > actually > looking at, hence it would be easy to fetch them two or three at a time. > That makes sense as well. -- Configure bugmail: https://bugzilla.wikimedia.org/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are the assignee for the bug. You are on the CC list for the bug. _______________________________________________ Wikibugs-l mailing list [email protected] https://lists.wikimedia.org/mailman/listinfo/wikibugs-l
