[00:00] which items are affected? [00:35] chfoo: list coming right up [00:41] chfoo: http://termbin.com/7yit -- 36 items [00:47] is there a good way to fix them up in bulk? [00:49] chfoo: if you make any change to the metadata, that should trigger a re-creation of the torrent. [00:50] You should be able to script that with the python library. I can try to come up with something you can cut-n-paste if you'd like. [00:50] oh ok. i was thinking of using the python library too [00:51] if only editing the metadata works then it should be quite simple. [00:54] cool, thanks for doing it [00:56] let me know when you have and I'll try re-downloading them [01:06] *** ix has joined #urlteam [01:07] hey need any help or something [01:10] *** Odd0002 has joined #urlteam [01:41] ix: sure, we can always use help analyzing more shorteners [01:42] The easiest way to help is to look over the unsorted list on the wiki page, and check whether each one is alive or not. [01:42] There are other ways to help, if that doesn't appeal. [03:25] *** Odd0002 has quit IRC (Remote host closed the connection) [04:14] *** Sk1d has quit IRC (Ping timeout: 250 seconds) [04:21] *** Sk1d has joined #urlteam [08:23] Some of Microsoft's URL shorteners: http://helps.ms/a0WK2h http://msft.social/1bwpHT [08:26] http://msft.it/6013BxcB9 [10:52] hook54321: Are you sure that helps.ms is Microsoft? http://helps.ms/ redirects to https://www.po.st/, which is owned by R1Demand according to the footer, which seems unrelated to Microsoft. [10:52] Same for msft.social, by the way. [10:57] On the other hand, almost all links seem to go to Microsoft websites. [10:57] I only found one that doesn't so far: msft.social/5YEy0w [10:58] helps.ms is registered by Microsoft though. [10:58] This is weird. [12:11] Somebody2: I'm thinking about splitting up the warrior project table on the wiki page. I think it would make more sense to have the general information about a shortener (number of URLs, incremental, what the short codes look like, example URL) in the relevant section below (Alive or Dead/Broken), and only keep the information about the warrior project in that table. It would also be nice to have individual r [12:11] ows for each scrape rather than lumping them together, and then replace the date columns with "start date" and "end date". [12:11] What do you think about that? [16:02] JAA: I like the idea of separating out each scrape. I'm less enthusiastic about duplicating entries between the warrior table and the Alive/Dead table. [16:04] In my conception, the Alive table is a step *before* the warrior table -- ideally, everything in it should eventually have a warrior job. [16:04] And the Dead table is a memorial -- the ones we didn't catch in time. [16:05] Splitting up the info seems like it would just make it harder to update. [16:06] Now, if you wanted to *combine* the three tables, with a column: Warrior/Alive/Dead -- I might support that. [16:17] i mean it makes some sense to track whether a site is alive or dead even when it does have a warrior job [16:17] but you do whatever you think makes most sense [16:17] don't let me stop you :) [16:17] bc you're already KILLING IT [16:17] fantastic job, urlteam. i'm always super impressed. [16:18] *** astrid sets mode: +oo JAA Somebody2 [16:18] *** astrid sets mode: +o hook54321 [16:30] we do track if a site with a warrior job is still generating new short codes (it gets a gray background if not) [16:30] (and a pink background if we are currently scraping it) [16:30] This is likely not the best way to track this stuff, though [16:33] ah kool [16:33] i mean however that takes the effort out [16:33] so we dont need to track its fate manually once it's warriored [16:41] oh, it's all still manual [16:42] If someone wanted to write code to automatically update the wiki page based on the tracker stats, that would be very welcome [18:28] Somebody2: I thought about throwing everything into one table as well. That would be quite neat, but the already somewhat crowded table would need several additional columns, so I'm not sure how well it'd work. [18:29] Yeah. Feel free to try it out and see if it works better. [18:29] I don't object to any of your suggestions enough to block them -- so if you will take the time to make the update, go for it! [18:31] Yeah, I think I'll take a subsample of the data and try out a few different things on a test page. I'll let you know how that goes. [18:31] Sounds good -- I'm glad to look it over if you'd like. [19:13] JAA: I saw it used on their support twitter page [19:27] *** mls has quit IRC (Ping timeout: 250 seconds) [20:18] https://nyti.ms/2vrH6Wl [21:39] bit.ly alias, added, thanks. [22:02] *** Odd0002 has joined #urlteam [23:12] *** Odd0002_ has joined #urlteam [23:13] *** Odd0002 has quit IRC (Quit: Leaving) [23:13] *** Odd0002_ is now known as Odd0002