[00:02] acf_: thanks [01:02] *** jlgaddis has joined #arpnetworks [01:02] *** jlgaddis has quit IRC (Changing host) [01:02] *** jlgaddis has joined #arpnetworks [01:04] *** dwarren has joined #arpnetworks [01:33] *** vissborg has quit IRC (Remote host closed the connection) [01:39] *** vissborg has joined #arpnetworks [03:41] *** meingtsla has quit IRC (Ping timeout: 258 seconds) [03:43] *** tellnes has quit IRC (Ping timeout: 258 seconds) [03:44] *** meingtsla has joined #arpnetworks [03:49] *** tellnes has joined #arpnetworks [04:25] *** medum has joined #arpnetworks [12:00] mnathani: If it's short term windows use for testing something, I normally just spin up an image from modern.ie [12:01] mnathani: And you really shouldn't be deploying anything XP based for real use any more :) [12:02] The client has an ancient dos based engineering software she needs to use [12:02] Which apparently works on xp, but nothing newer [13:00] mnathani: DosBox? Or FreeDOS in a VM? [13:17] dosemu? [13:17] is dosbox better than dosemu now [14:58] dosbox is all I hear about nowadays. Perhaps dosemu is dead? [14:58] it was pretty dead when i last looked [14:58] it was really slow in 64 bit mode withotu v86 support [14:58] I know that dosbox is used when Steam sells old DOS titles [14:58] i wonder if qemu/kvm is a better way to go [14:59] dosbox is pretty lean and fast [15:00] cpus have got faster. [15:00] i was playing with bbs door games, and just starting a dos program took 1+ seconds. [15:00] which seemed long to me considering the age of dos stuff [15:00] msdos was faster than freedos for startup time iirc [15:01] no-one uses bbs's anymore though :) [15:01] Frankly both msdos and freedos start too quick for me to measure. [15:01] Which is to say, as soon as it's loaded, it's running [15:02] this was core2duo [15:02] and starting dosemu with a program .. [15:02] I'm running DOS on Atoms primarily [15:02] And booting directly [15:02] os/2 was faster on my 486dlc which is like 386dx 40 mhz [15:02] (as far as msdos/freedos start times) [15:02] but it's always running etc. [15:02] yeah [15:03] it's a bit different when you're starting it to run a program [15:03] vs booting [15:03] i was going to try and do bbs stuff again in like 2000 [15:03] which i suppose was a bit late. [15:04] and i managed to get something going on os/2 that would execute programs called from a linux box :) [15:04] but yeah bbs's were too dead by then really. [15:04] i just had more money in 2000 than prior, and had renewed interest. [15:04] i used to run a bbs until like 98 [15:07] with custom bbs software on os/2. [17:05] woot, i got a ssh latency tester program working [17:06] you just run a program on the command line, and it gets the current cursor position ten times telling you how long it took each time [17:20] trying to think if i'm rsyncing one system over to a new system, which is currently vacant but for an encrypted LVM setup [17:20] i think i just need to rsync everything save for the fstab and grub cfg [18:56] anyone familiar with usenet administration? nntp / peering / providing binary groups with their ever increasing storage requirements [18:58] people still do nntp? [18:59] * brycec does, but not for "news" [19:10] how much storage do usenet servers use now? 40tb? [19:10] i assume the binarys are quite big :) [19:13] http://commons.wikimedia.org/wiki/File:Usenet-total-storage.jpg [19:13] hmm [19:13] i'm quite far off it seems [19:14] what is boneless? [19:33] I think it has to do with Warez / Apps / Movies etc [19:33] nothing legal [19:33] Presumably it's the name of a releasegroup [19:34] http://en.wikipedia.org/wiki/Alt.binaries.boneless [19:34] Alt.binaries.boneless :: alt.binaries.boneless is a Usenet discussion forum primarily used to transfer binary data content, rather than being used for textual communications. It has the unusual distinction of currently being one of the largest and most active binary newsgroups on all of Usenet, typically receiving tens of gigabytes of new data and over a million new postings each day, yet almost nothing is known of its origins or why it is so... [19:37] well it was mesured in pb's [19:37] most of the posts in there aren't labeled properly [19:37] so yeah i have no idea how to have a big server with enough retention to be good [19:37] so I think you need an alternate source to get an nzb file for something you want to download [19:37] i imagine it's mostly megaupload type custom solutions [19:38] its only a handful of providers that store the data [19:38] everyone else is just a reseller of the same service [19:38] ahh [19:38] i played with nntp 10 years or so ago [19:38] and it was full of spam and useless [19:39] I guesss you could look into a moderated group that keeps spam out [19:39] but not many folks use it to have meaningful discussions today [19:40] even in 98 it was pretrty bad [19:40] everythiing has shifted to mailing lists now [19:42] mnathani: want to try my console latency test? :) [19:43] sure [19:43] what do I need to setup / configure on my end? [19:43] nothing [19:43] you just run it in a shell [19:43] http://202.49.140.24:24/cl [19:43] it's a 64 bit linux binary [19:44] and http://202.49.140.24:24/cl.c is source [19:44] which needs improving :) [19:44] how can I compile it myself? [19:44] clang cl.c -o cl [19:44] or gcc [19:45] it's proof of concept so far ... [19:45] ok [19:45] but each of those msec is ping [19:46] should I run this from a machine in Toronto? [19:46] it needs to do longer testing, and not scroll down the page, and show minimum/average/max etc. [19:46] yeh [19:46] anywhere [19:46] i've been doing it in various locations [19:46] does it need root? [19:46] as long as you ssh somewhere it'll give good numbers [19:46] nope [19:47] it's just asking your terminal where the cursor is [19:47] and timing how long it takes to get an ESCAPE key back [19:47] ie hackish, it actually returns ESC[4;3R etc. [19:48] I ran it in a shell on an ubuntu system [19:48] and terminals could in theory give it back one character a time etc. [19:48] yip [19:48] how does it compare to your normal ping? [19:48] shell stopped responding [19:48] oh [19:48] it's probably not resetting the terminal back properly [19:49] uhh [19:49] I do use tmu [19:49] s/tmu/tmux [19:49] I do use tmux [19:49] i use zsh [19:49] oh i wonder if tmux will give fake nuimbers [19:49] hangon [19:49] yeah [19:49] tmux gives fake numbers [19:49] because it returns it locally [19:49] does the box I am testing on have to be on a different network? [19:49] or can it be on my LAN [19:49] yeah [19:50] to get good numbers. [19:50] wow tmux is slow for cursor locaiton [19:50] well i tested on lan too [19:50] for me it was about .15 msec on lan [19:50] and about .0012 msec on same computer [19:50] and 5 msec to closest internet host [19:50] its' more interesting for remote areas :) [19:51] i'm using zsh [19:51] maybe zsh is resetting the terminal back [19:52] it should be resetting it back fine now [19:52] i updated the binary and source [19:53] damnit [19:53] The program 'zsh' is currently not installed. You can install it by typing: [19:53] sudo apt-get install zsh [19:53] yeah it's working better now [19:53] just get the new .c [19:55] seems to work on openbsd too. [19:55] so like my ping to my openbsd vm is 133 msec, my ssh ping is 135 msec. [19:56] but then i ssh in again and it's 161 msec [19:57] there's some kind of load balancing, it's consistently differnet with different sessions [19:58] how do I get my ssh ping results? [19:58] ssh somewhere not running tmuxc [19:58] tmux [19:58] and run it [19:58] ok [19:58] there may be a way to get results inside tmux but tmux does cursor location itself [19:58] installing zsh on my ARP VPS [19:58] you don't need zsh [19:58] just get the new version [19:58] i fixed the bug [19:58] ok [19:59] zsh is nice anyway :) [19:59] try it sometime [19:59] outgoing port 24 is firewalled on my system :-) [20:00] oh [20:00] it should work on https too i think [20:00] https 443? [20:00] hmm [20:00] can't see it [20:02] yeah https isn't working it seems [20:02] umm [20:02] I scp the file [20:02] ahh ok :) [20:02] but cant compile it now [20:02] what's it doing? [20:02] /tmp/ccaZmHPG.o: In function `main': [20:02] cl.c:(.text+0x10b): undefined reference to `clock_gettime' [20:02] cl.c:(.text+0x185): undefined reference to `clock_gettime' [20:02] cl.c:(.text+0x224): undefined reference to `clock_gettime' [20:02] freebsd? [20:02] collect2: ld returned 1 exit status [20:02] centos [20:02] hmm [20:03] i wonder what library it wants [20:03] compiled fine on my ubuntu system [20:03] add -lrt [20:03] when compiling it [20:03] ok, that worked [20:03] time taken: 76.0990 msec. [20:04] your ping is about 76 msec? [20:04] also i have a feeling there's a 0 in front of those numbers :) [20:04] Minimum = 78ms, Maximum = 85ms, Average = 80ms [20:04] ahh [20:04] you have better tcp than icmp :) [20:06] is there a udp version of this test? [20:06] not yet :) [20:07] but this should test real world ping [20:07] like if you use dropbear bouncing [20:07] it can go up higher or come down lower [20:07] too bad you need a shell on the remote system to test [20:07] well yeh. [20:07] i thought it was a nice easy place to start though [20:07] and ssh has some level of overhead too etc. [20:08] at least it means if you have a laggy tcp connection you can get a number for it :) [20:08] and that %04d is meant to be %03d [20:08] that's why ther's a leading 0 [20:09] 76.990 << supposed to be like that? [20:09] also gcc doesn't like me declaring a varabile inside a fr loop [20:09] yeah [20:09] you can change the printf [20:10] i was thinking 1000ths of a second [20:10] and counted 4 for leading zero [20:10] but all the numbers between 0 and 999 is of course 3 places. [20:11] i was also thinking should show chunks of text of various sizes etc. [20:12] although terminal could also slow down ... xterm is about 10x as high ping on localhost when it has to scroll [20:12] localhost is insanely low anyway, but some terminals may get slow themselves if sending too much data [20:15] how does windows display on your 4k monitor? Have you tried it? [20:15] at 30 hertz [20:15] getting a new video card, it should come tomorrow i think [20:16] I would assume everything gets shrunk down [20:16] with such a high resolution [20:16] it was in the city at 9:15 am on saturday. [20:16] which is too late for courier :( [20:16] amd or nvidia? [20:17] r9 290 amd [20:18] windows doesn't scale very well on 4k i think [20:18] but chrome has problems with scrolling problems with radeonsi :( [20:19] firefox seems to work better for scrolling [20:19] evince works even better [20:25] time taken: 728.364 msec. [20:25] nz->uk->dallas->nz->dallas [20:26] it's pretty hard to use with this much latency [20:29] if you run it on a link with packet loss it should show spikes [20:30] time taken: 30.477 msec. [20:30] time taken: 261.944 msec. [20:30] time taken: 29.986 msec. [20:30] like that [20:41] http://pastebin.com/rb3ZZnim [20:41] that's the kind of thing i was hoping to pickup on [21:45] *** novae has quit IRC (Ping timeout: 265 seconds) [21:49] *** novae has joined #arpnetworks