[01:06] damn that's awesome [01:35] did he give a list of what cds he got? or is he just gonna surprise upload them? [02:29] Surprise! [02:36] well ain't that gonna be fun [02:46] cmake is so maddening [02:46] I'm trying to figure out how to shell out as part of a custom target [02:47] add_custom_target and execute_process are both cmake commands, but composing them together is full of SURPRISE [02:53] oh, appearently the answer is to not use execute_process, because that's meant to be used to execute a program at configure time [03:30] Anyone know a decent place to get old Windows versions? Looking for Windows 2.0ish [04:00] Trying to collect as I can [05:05] used to know a good place for old windows versions but now I'd be damned if I could remember [05:43] http://wdl2.winworldpc.com/Abandonware%20Operating%20Systems/PC/Microsoft%20Windows/ [09:53] stupid robots.txt [11:08] on the subject of old software versions, this site is pretty comprehensive: http://vetusware.com/ [11:08] there's a group trying to re-create Opera v12 here: http://otter-browser.org [12:41] dashcloud: yeah, there's no easy way to archive that site though [12:41] even with the highest membership you only can do 10 downloads per day [14:27] there was a mirror at http://files.wehack.net/Vetusware/ that seemed pretty comprehensive [14:27] it seems down now but archivebot grabbed it [14:27] https://archive.org/download/archiveteam_archivebot_go_024/files.wehack.net-inf-20140128-223321-bfv4w.warc.gz [16:40] how was archiveteam getting 4chan? [16:42] I don't think we are specifically, jason has some older archives that I think he was given [16:43] I used to archive sma with https://code.google.com/p/libchan/ and it no longer works, looking for a new automated method [17:00] Ouch, no development since 2012 [17:04] there was this story recently https://library.stanford.edu/blogs/digital-library-blog/2014/01/sdr-deposit-week-4chan-forum-archives [17:05] maybe these yotsuba society folks have a tool [17:21] DFJustin: i think it's more a combination of their crawler as well as sites such as 4chan archive [17:21] thought* [17:50] This is a very Cool Thing [17:50] http://pywb.herokuapp.com/ [17:50] Check out the twitter scrolling that works! [17:50] Even facebook too [17:50] https://groups.google.com/forum/#!msg/openwayback-dev/MAFY4Q0Jo8Y/nsHKReRwfyAJ [17:58] wow [17:58] ay idea how they are doing that? [17:58] any* [17:59] hmm [17:59] http://pywb.herokuapp.com/pywb/*/https://pbs.twimg.com/profile_images/2179744751/IIPC_Twitter_ProfilePic_bigger.gif [17:59] same as wayback [18:00] not he interface, but the way of the urls [18:00] with the * [18:01] https://github.com/ikreymer/pywb [18:01] It's a (nearly) feature-complete rewrite of the wayback machine in python by the engineer who wrote a lot of the old wayback components [18:02] including the new "save-page-now" feature and the new api that was released on 10/24/13 [18:02] wow [18:02] (pywb is fully unaffiliated with IA, though) [18:02] gosh, IA could look at the code and implement facebook and twitter crolling in heritrix... [18:03] but it creates warc.gz files [18:03] so it should work with wayback [18:09] the big issue is that java wayback doesn't support domain specific rules [18:10] which you really need to playback weird ajaxy content [18:10] certain headers, certain string manipulations, etc [18:10] isn't it doing something custom for youtube [18:11] kinda [18:11] but it's very hacky [18:11] and not scaleable at all [18:12] also youtube doesn't require nearly as many "special" things, since it just replaces the yt player with a custom jw one [18:12] jwplayer* [18:13] it also seems to redirect you if there is a grab of the same video with different url parameters [18:24] is there an example in the wayback machine with a youtube page that actually plays a video? [18:24] I actually have never found such a video [18:29] hmm the one I saved before doesn't work anymore, it seems kind of inconsistent [18:35] yes [18:36] but man the wayback machine is showing pages better and better [18:36] http://web.archive.org/web/20140312182903/https://www.uber.com/ [18:36] looks great [18:37] DFJustin: they are crawling youtube videos: https://archive.org/details/youtubecrawl [18:37] but just blocking the [18:37] m [18:38] I don't think they're blocking them so much as the wayback hack stuff is flaky [18:45] you can see how it works with the archive-it ones https://wayback.archive-it.org/4399/20140301235240/http://www.youtube.com/watch?v=n1Q1p7Oc_5g [18:46] Ah I see, thank you [18:47] but why don't they just add that to the IA wayback machine too then? [18:51] the code is in place it's just not working, probably because they have no manpower to fix everything all the time [18:52] ah, well, it's good to see the information is saved, even though it's not playable [19:02] another really annoying problem with infinite scrollers is that some sites (i.e. patch.com) shove a timestamp into the URL [19:03] so recording the request/response is insufficient [19:03] this is the result of voodoo web programming by people who think that cache-busting needs to be done with fucking the query string [19:39] yeah seen a bunch of ?nocache=137891278941 stuff in archivebot [19:43] http://www.reddit.com/r/IAmA/comments/2091d4/i_am_tim_bernerslee_i_invented_the_www_25_years/ [19:45] well, URL-based cache busting is sometimes needed to get around stupid caching proxies that don't respect the cache control headers [19:52] oh myspace brought old photos back [20:02] DFJustin: did they bring journal content back too? [20:03] the email only mentions photos [20:08] Coderjoe: I maintain that the blame is still on shitty web programmers who expect that they will always be able to expect a request from a browser [20:09] and designing their applications on faulty assumptions [20:18] I hate sites where all I get is a blank page until I start blessing javascript [20:59] I hate a lot of things [20:59] I practice hate-driven development [21:00] i hear that's good for your health [23:43] http://redmine.replicant.us/projects/replicant/wiki/SamsungGalaxyBackdoor [23:43] dashcloud: :)