[00:08] *** zhongfu has quit IRC (Remote host closed the connection) [00:10] *** zhongfu has joined #internetarchive [00:17] *** zhongfu has quit IRC (Remote host closed the connection) [00:41] *** zhongfu has joined #internetarchive [02:19] *** Stilett0- is now known as Stiletto [03:44] *** qw3rty113 has joined #internetarchive [03:50] *** qw3rty112 has quit IRC (Ping timeout: 600 seconds) [03:54] *** odemg has quit IRC (Read error: Operation timed out) [04:05] *** odemg has joined #internetarchive [16:58] does anyone know the exact set of characters you can't use in an archive.org item? (if you want a working directory listing) [16:59] so far I have figured out that "+" is incorrectly converted to " " probably by using the wrong php quoting function, and that you can't start or end filenames with whitespace because of how browsers interpret hrefs [17:14] ah # and % are also bad because they aren't encoded [17:33] also ? and \ [17:35] I accidentally discovered that you can add a /? to the end of a /download/ URL to get redirected from the broken PHP directory lister to a working nginx directory lister [18:26] *** Lord_Nigh has quit IRC (Ping timeout: 252 seconds) [18:33] *** Lord_Nigh has joined #internetarchive [19:05] "you can't start or end filenames with whitespace because of how browsers interpret hrefs" Really? Can't you just encode it as %20? [19:06] ivan: "An identifier is composed of any unique combination of alphanumeric characters, underscore (_) and dash (-)." [19:06] JAA: well, of course you can, but IA doesn't apply any sane escaping [19:06] https://internetarchive.readthedocs.io/en/latest/metadata.html#archive-org-identifiers [19:06] So basically avoid anything else, I guess. [19:06] it's got just enough escaping to stop XSS but not to do the right thing [19:06] Ah, right. [19:07] would help if archive.org put their PHP spaghetti on github and took PRs :-) [19:08] Oh yes, I'd love to see their software open-sourced. [19:08] would you? you haven't looked inside yet [19:14] Oh, I'm absolutely sure it's horrible. [19:15] I'm a strong supporter of open-source software, so I consider horrible open-source code better than any closed-source code. [19:15] And once it's open-source, others can contribute and make it better. [21:50] *** Coderjo has quit IRC (Read error: Operation timed out) [21:58] *** Coderjo has joined #internetarchive [22:18] *** Lord_Nigh has quit IRC (Read error: Operation timed out) [22:20] *** Lord_Nigh has joined #internetarchive