[00:07] Not sure why I enjoy watching wget grab a site. [00:13] Watching the little bar go up? [00:20] it is the enjoyment of processing [00:23] every time that bar goes up, that little bit of data comes down, and you know you have saved it from dying an early death [00:24] You can be a rebel and use the alternate output graphs [00:35] I should totally pipe the stdout into a script that feeds https://github.com/holman/spark or something [00:39] There's probably some way to pipe stdout to a row of Christmas lights [00:39] And you can watch the string light up as wget runs [00:57] shaqfu: oh my god [00:57] :D [00:57] that's a great idea [00:57] I'm doing that for the next grab [00:57] shaqfu: That is brilliant [00:58] http://www.geholidaylighting.com/holiday-lighting/led/color-effects [00:58] Individually addressable/colorable lightbulbs [00:58] My friend has a set [00:58] there's an arduino code example thing floating around the net too [01:00] Sigh; I was going to apply to jobs, but now I'm going to look up piping stdout to Christmas lights [01:01] http://www.theregister.co.uk/2012/06/30/amazon_cloud_storm_outage/ [01:01] so one of amazon's clowns stumbled in the dark last night [01:02] Y'know [01:02] they apparently haven't heard of UPS and generator backup power [01:02] Evernote announced that it's trying for 100 years of data storage, even in case of company failure [01:02] As as soon as I get a little faith in cloud services, boom [01:02] Literal clouds take out the virtual ones [01:05] brunswick bowling makes a lighting product for glowbowl use, where each narrow capping has "addressable" RGB light modules. each capping string connects to an under-approach control/power box [01:06] but the modules have no address settings, and just have a five-pin connector on each end, going to the next or previous module (except for the last in the chain, which is left unconnected on one side) [01:08] I don't know what protocol they use, though. when I was working in bowling, I took home one of the dead modules, though. I just never got around to really REing it [01:10] isn't this the second northern virginia clown outage in as many weeks? [01:12] June 15, iirc, caused by power issues yet again [01:12] There was one "sorry, your passwords were in cleartext" [01:12] Ah, yeah, AWS outage [01:12] I meant AWS N. VA clown stumble [01:12] Spritzed too much seltzer into the servers [01:21] (moved from non-bs) [01:21] wait a minute [01:21] if (relatively) long wavelength EM passes bidirectionally, why can't you SEE through a stargate? light is just very high frequency EM [01:22] Coderjoe: You'd have to ask RDA [01:23] And blech; nothing like applying for a job that the last person quit due to sexual harassment issues [01:24] shaqfu: use that fu of yours and kick balls if it happens [01:24] BlueMax: Luckily, I'm a guy, so it won't be problematic [01:24] But still awkward to know it happened [01:24] I would likely get the "William Shatner SNL sketch" response [01:25] "Get a life"? [01:25] yes [01:27] Hm, what's the limit on distance of microwave? Does it fade out, or is the curvature of the earth the problem? [01:30] http://www.youtube.com/watch?v=FeS4tVFbNNk [01:48] shaqfu: or just enjoy it [01:48] <.< [01:48] >.> [01:49] :P [01:56] I wonder, does IA have any scans of player piano rolls? [02:02] I don't believe so [02:03] Do they scan well? [02:04] the reason I wonder is the uploader's description on http://www.youtube.com/watch?v=n73lXCzJUZ4 [02:04] and one of the comments suggest there are digital scanners for handling them (though usually for reproducing them) [02:06] Hm, that'd be neat. Scan sheets in, have software play it back [02:07] there are apparently different roll encodings, some of which include expression, some which don't [02:07] (Sounds like the software from Dirk Gently) [02:16] so Michigan just had a law pass which allowed fireworks like firecrackers, bottle rockets, and consumer-grade mortars. the law also prohibited local governments from banning the use of them on the day before, of, and after serveral holidays. [02:17] some local governments near me recently banned use except for on those days. [02:17] I'm rather disappointed in those locales [02:17] if they had waited 6 months to a year, the novelty would have worn off, and they wouldn't have them going off at all hours of the night anymore [02:18] (I'm also disappointed in the people that were lighting them off so late (sometimes past 3am) that caused these laws to be drafted in the first place, but that is another matter) [02:18] Or just make them legal year-round, so there's no novelty [02:19] they were legal year-round until the local govts clamped down [02:19] but there was novelty because they had been prohibited for a long time [02:19] Ah [02:19] (and people would go down to indiana or indian reservations to get them) [02:23] whoops, /nicked on the wrong network [02:24] good morning, data superheroes [02:24] Hooray, 8.5h later download done! [02:25] \o/ [02:27] Still irked at Montreal Mirror for pulling their webpage the day they announced they were shutting down, but glad they weren't clever enough to take down the archive of past issues. [02:27] Phew. I was wondering what happened to that [02:28] Yeah. [02:28] So, I got 1997 - 2010. 2011-2012 are MIA. [02:29] do indian reservations not have laws? [02:29] underscor: Different laws [02:30] Have you ever heard of Indian casinos? [02:31] the state of michigan apparenly does not actually have any legal jurisdiction on indian reservations within the state of michigan [02:31] Yep - they're semi-autonomous [02:32] You can go there and gamble [02:32] I think federal law applies, but not state [02:33] No killing people on reservations, but gamble and shoot off fireworks all you want [02:44] damn [02:44] that's p cool [02:44] shaqfu: As long as you're friends with the indians ;) [02:45] On a side note, if anyone actually does want something from IA on a HD and mailed to them for whatever reason, as long as you're willing to pay shipping to your country, I can work that out. [02:45] i kinda would like my own drive back at some point [02:46] (Looking at BlueMax specifically, but it applies to anyone within reason) [02:46] Coderjoe: ? [02:46] * BlueMax looks away shyly [02:46] lol [02:46] i mailed a usb 1tb drive to IA filled with friendster data, along with a return shipping label [02:47] well, s/mailed/shipped/ [02:48] but apparently, the XFS file system has been blocking attempts to copy the data off, or something [02:48] oh really [02:48] Coderjoe: What's your name, I'll go look around in Jason's pile [02:48] oh [02:48] feel free to pm it [02:50] i'm afraid UPS has forgotten about the tracking number for the shipping label I included [02:55] found it, I think [02:55] 2.5" external? [02:55] aw dang, totally uploaded Montreal Mirror without "archiveteam" in the item id [02:55] yes, with a label on it with my name and phone number [02:55] hm, don't see that... [02:55] oh, on the bottom! [02:55] yep [02:55] cool [02:56] a WD Elements SE [02:56] yep [02:56] should also have the cable with it [02:56] mhm [02:56] and a UPS label [02:59] lol [02:59] Coderjoe: Uh oh. Hopefully these were uploaded somewhere else/differently [02:59] -rw-r--r-- 1 alex alex 0 Aug 19 2011 friendster.000117000-000119999.tar.xz [02:59] -rw-r--r-- 1 alex alex 0 Aug 19 2011 friendster.000310001-000320000.tar.xz [02:59] -rw-r--r-- 1 alex alex 0 Aug 19 2011 friendster.003985001-004000000.tar.xz [02:59] (all other files are ok/intact) [02:59] no, I fucked those up and was chronicaling my shame [03:00] look in the csv/sfv/whatever file I think I included [03:00] there is no additional file [03:00] hmm [03:00] perhaps I only uploaded the file via rsync [03:00] ah [03:00] * Coderjoe pokes around [03:02] I think the drive with that data is offline, and I'm about 10 miles away from it at the moment [03:02] ah [03:02] np [03:03] man, it's nice to upload to fos at 14MBps [03:21] underscor: essentially, I was doing a multipart command and messed up, effectively overwriting the source files (by piping into them). >_< [03:22] oic :( [03:22] (all because I wanted to use PV to show the progress of the transfers) [04:50] I wonder if anyone in the position of building servers has eyed the backblaze pod [04:50] (for IA, that is) [05:03] we've talked about it [05:03] but they're not as easily hotswap [05:04] and also our datanodes have dual hex-core xeons with 192GB RAM [05:04] so the motherboards are too big [05:05] they definitely don't do easy hotswap. and I think they only support ATX borads, not EATX. :-\ [05:08] 01:04 <@underscor> and also our datanodes have dual hex-core xeons with 192GB RAM [05:08] Nice hardware porn [05:09] well, that's because they're not just datanodes anymore. they also run one or more VMs as well [05:10] Hm? I thought most of IA nodes were just wads of storage [05:10] not on the beefy supermicro-cased nodes [05:11] the "thumper" red box ones with via processors were just storage [05:11] Ah [05:11] with separate "power nodes" doing the heavy lifting [05:12] I was about to say - I thought it was divided between nodes doing the "thinking" and nodes that just held data [05:12] Glad to see I wasn't horribly far off [05:13] for instance, this one node I picked because it was in my browser history is hosting three vms [05:14] one of which is either the remnants of batcave or is fos [05:14] i think it might be the former [05:15] yeah, remnants of batcave [05:20] this other node is running one of the worker VMs [06:17] We have nodes that are both data and workers [06:17] and worker only nodes [06:17] known as "kvmhosts" and "petaboxes", inverse respectively [06:20] ia6002 - ia6008, ia6012 - ia6014 are currently datanodes; ia6000 ia6001 and ia6011 are kvmhosts and some other miscellany. oh, and ia7000 ia7001 are also kvmhosts [06:20] everything else is paired storage [06:34] however, the "worker" role is still performed in a VM [06:37] yeah, no derive work is run on bare metal any more [08:54] http://wegetsignal.org/tmp/lettertoeditor_20120628.png [09:16] Coderjoe: ... [09:34] yeah... for the fireworks part, there was a non-covert financial motive: sales tax income. [09:34] but I love the first paragraph [09:34] we should ban his car, then. [10:11] is it normal to feel like a jurk if your not uploading to archive.org? [10:12] i have a ton of 3d world shareware discs i have to upload [10:12] using ftp to know how long it will be [10:56] I feel a little bad that I do not upload but my connection is slow [10:57] I have been looking into if the archive can just fetch things directly and how that process works. [11:02] i have a slow connection too [11:02] its only around 250 kbytes [11:08] uploaded: http://archive.org/details/cdrom-3d-world-134 [11:10] * winr4r cheers [11:16] uploading 3d world 136 [11:16] \o/ [11:17] the cdrom [11:17] not the issue [11:36] found COTS Journal [11:36] newsnet is amazing [11:36] when you can't find something [12:20] uploaded: http://archive.org/details/cdrom-3d-world-136 [14:03] uploaded: http://archive.org/details/cdrom-3d-world-137 [14:05] you're busy! [14:07] uploading 3d world cd 112 [14:09] i have uploaded over 14gb of 3d world cds [14:09] :-D [14:21] :D [14:24] :) [14:49] uploaded: http://archive.org/details/cdrom-3d-world-112 [15:21] :))) [17:00] I was thinking of renting a remote box for doing uploads and transfers. Anyone have experience with this for uploading to ia? [17:00] Most of what I am backing up I can ftp from other sites. IA does not have FXP as far as I know so this would be a replacement for that [17:05] Well, I guess I know why both my Ubuntu boxes were completely wedged last night. Apparently Ubuntu's codebase goes something like if (seconds > 59) { kernel->shitYourself(); } [17:06] It's not Ubuntu's fault mate. [17:06] But yeah, that leap second sure did some fucked up shit to a lot of systems [17:06] Nice kernel lockup [17:07] Yeah, it seems to actually be kernel level, not OS level. [17:07] DoubleJ: the issue is apparently with ntpd [17:07] and sure, I guess one could argue that the kernel team should've noticed and fixed it prior to that, but yeah - AFAIK it might be ntpd that the fault really is at [17:07] damn joepie91, shooting in a quickie like that :) [17:08] I've seen it described as a livelock, which I had to google up to see what they were talking about. [17:08] So NTPD feeds the system something it doesn't expect and it chases its own tail trying to fix it. [17:08] Since NTPD doesn't have much choice in the matter -- leap seconds happen -- I'm going to blame the kernel :) [17:09] ersi: :D [17:09] let me pull up the thread that discussed the issue [17:09] one moment [17:10] joepie91: If it can describe the problem in something resembling plain English I'll be much obliged. [17:10] This is caused by a livelock when ntpd calls adjtimex(2) to tell the kernel to insert a leap second. See lkml posting http://lkml.indiana.edu/hypermail/linux/kernel/1203.1/04598.html [17:10] http://serverfault.com/questions/403732/anyone-else-experiencing-high-rates-of-linux-server-crashes-during-a-leap-second [17:10] it's a SA site, so should be a clear enough answer :) [17:10] er [17:10] SO* [17:10] .. [17:10] SE* [17:10] wow, I'm having some issues with abbreviations today it seems [17:11] yuppers [17:16] So it wasn't the leap second itself that made it lose its mind; it was the warning that one was coming up. [17:17] Now I just have to remember to upgrade (and thus get a non-mind-losing kernel) before we have another one. [17:17] Or not be archiving when it happens, since systems under load seem to have been the most vulnerable. [18:47] Well the next leap second is probably a year or more away [18:47] unless the fuckers at the ITU decide to do away with it [18:48] So you've some time to spare :) [19:21] It's a good fire drill for when the world blows up in 2032 [19:21] Er, 2038 [19:53] Oh, has that been rescheduled again? [22:48] I want to stream this but there's no derivatives :( http://archive.org/details/tmp_nc_617_netc [22:56] I just rana re-derive but I think there's a better version in there. [23:53] http://www.youtube.com/watch?v=qWjCkcAmzDc [23:54] I wonder if any other episodes ever made it out ... [23:54] chronomex: I think that was the only one broadcast; dunno how many others were shot [23:54] Only the pilot was ever screened, although eight episodes were planned and a number were recorded in which a story arc was about Adolf and Eva's attempt to kill the Goldensteins without the Goldensteins knowing it's Adolf and Eva. [23:55] Ah (I really need to stop Chrome from crashing) [23:55] At this point, they were probably shot over [23:55] that's ok, I'm happy to paste wikipedia into irc [23:55] that was 1990, not 1970...