From mboxrd@z Thu Jan 1 00:00:00 1970 From: lsworkemail112 at gmail.com (Luke SanAntonio) Date: Sat, 29 Sep 2012 15:14:46 -0400 Subject: [PATCH] Added code to output the age as seconds instead of "0 min." In-Reply-To: References: <1348878616-10065-1-git-send-email-lsworkemail112@gmail.com> Message-ID: Okay, I understand now. I guess my take on the matter is a user who selects caching already has accepted that the data generated might not correctly reflect the actual data up-to-date... these users know it going in, so I don't think it's a huge concern... Sorry about my previous misunderstanding... - Luke On Sat, Sep 29, 2012 at 2:51 PM, Jason A. Donenfeld wrote: > On Sat, Sep 29, 2012 at 8:43 PM, Luke SanAntonio > wrote: >> Hi Jason, >> >> First of all, good call with the cache, it hadn't even crossed my mind... >> Second I think the cache isn't something we need to worry about... >> mostly because >> cgit is very lightweight, and I think for now, we don't have to be too >> worried about >> performance, correct me if I'm wrong though... Also it seems to me >> that with or without the cache, >> every cgit page I've ever come across seems to load in much less time >> than a second... > > > Hey, sorry, just to be clear -- my concern isn't about performance, > but about this: > > If you set the cgit cache to 1 minute, and the granularity of > timestamps is only 1 minute, then for the most part, the pages, though > cached, will see up to date. > > However if you set the cgit cache to 1 minute, and the granularity of > the timestamps is 1 second, then for the most part, the pages will > seem out of date. > > But this is just how caching is, so I'm not sure it's such a big > concern. Just throwing it out there, in case anyone had some elegant > solution or attitude about it.