[00:18] up_the_irons: sweet, time to splurge on some new electronics! =) [01:26] *** HighJinx has joined #arpnetworks [01:29] *** tabthorpe has quit IRC (*.net *.split) [01:40] *** tabthorpe has joined #arpnetworks [01:54] *** dzup has quit IRC (Ping timeout: 260 seconds) [01:59] *** tabthorpe has quit IRC (*.net *.split) [02:10] *** tabthorpe has joined #arpnetworks [02:18] *** tabthorpe has quit IRC (Read error: Connection reset by peer) [02:18] *** tabthorpe has joined #arpnetworks [02:19] jlgaddis: yeah srsly ;) [02:19] those HF radios are _not_ cheap [02:37] *** Ehtyar has joined #arpnetworks [03:08] good thing xmas is coming up [05:11] heh [05:31] *** heavysixer has joined #arpnetworks [05:31] *** ChanServ sets mode: +o heavysixer [07:01] *** heavysixer has quit IRC (Remote host closed the connection) [07:37] *** cmeiklejohn has joined #arpnetworks [07:43] *** heavysixer has joined #arpnetworks [07:43] *** ChanServ sets mode: +o heavysixer [07:45] heh. support.arpnetworks.com/discusssions/... claims a 10mb file limit yet I just (in theory) uploaded a 13.3mb iso image ;-) [08:51] *** cmeiklejohn has quit IRC (Quit: Computer has gone to sleep.) [08:57] *** dzup has joined #arpnetworks [09:01] *** heavysixer has quit IRC (Remote host closed the connection) [09:07] *** cmeiklejohn has joined #arpnetworks [10:01] *** dj_goku has joined #arpnetworks [10:01] *** dj_goku has quit IRC (Changing host) [10:01] *** dj_goku has joined #arpnetworks [10:04] *** HighJinx has quit IRC (Quit: Computer has gone to sleep.) [10:12] *** dj_goku_ has joined #arpnetworks [10:14] *** dj_goku has quit IRC (Ping timeout: 264 seconds) [10:31] *** HighJinx has joined #arpnetworks [10:37] *** sako has joined #arpnetworks [10:43] *** cmeiklejohn has quit IRC (Quit: Computer has gone to sleep.) [11:00] *** sako has quit IRC (Ping timeout: 244 seconds) [11:12] *** sako has joined #arpnetworks [11:15] *** dj_goku_ has quit IRC (Read error: Connection reset by peer) [12:20] *** mnathani has quit IRC (Ping timeout: 256 seconds) [12:25] *** sako has quit IRC (Ping timeout: 245 seconds) [12:38] *** cmeiklejohn has joined #arpnetworks [12:56] *** sako has joined #arpnetworks [13:14] *** mnathani has joined #arpnetworks [14:15] *** cmeiklejohn has quit IRC (Quit: Computer has gone to sleep.) [15:32] *** cmeiklejohn has joined #arpnetworks [15:37] *** cmeiklejohn has quit IRC (Client Quit) [16:14] *** niner has joined #arpnetworks [16:19] ever have one of those days where you wonder why the powers that be ever give you administrative rights on a network? That was my day. [16:48] *** niner has quit IRC (Quit: Leaving) [17:30] *** jloks has joined #arpnetworks [17:36] *** dzup has quit IRC (Read error: Connection reset by peer) [17:46] *** jloks has quit IRC (Ping timeout: 268 seconds) [17:46] *** dzup has joined #arpnetworks [18:08] *** sako has quit IRC (Ping timeout: 260 seconds) [18:09] *** HighJinx has quit IRC (Quit: Computer has gone to sleep.) [18:12] *** sako has joined #arpnetworks [18:22] *** sako has quit IRC (Ping timeout: 264 seconds) [18:22] *** sako has joined #arpnetworks [19:46] *** cmeiklejohn has joined #arpnetworks [19:54] well you appear to be onlne... [20:11] I seem to be too young to have experienced this myself, but perhaps someone in here can explain.. [20:11] How does USENET work? My Google foo is yet to lead me to some explanation [20:12] Also do arrangements need to be made similar to how the internet works ie: people pay for transit / premium usenet [20:17] That's not exactly on-topic.... But who cares :) [20:17] mnathani: In short, yes. [20:17] There are providers, who who massive server/storage farms, and they host USENET. They sync new articles among each-other, so they usually all have the same content [20:18] *who own [20:18] Some providers keep old content (called retention) longer than others. [20:18] And not all providers "carry" the same (much less, all of the) newgroups [20:19] brycec: err they don't usually all have the same content :) [20:19] they're meant to [20:19] otu of sync is common [20:19] err ar leats was [20:20] usenet is kind of dead now, i kind of wish it came back [20:20] Once upon a time, ISPs would often offer USENET/newsgroup access. This meant faster speeds for their users, and less b/w cost. (Think: caching proxy) However A) They rarely carried the binary newsgroups since those were enormous and usually filled with warez and other unsavoury material, and B) "nobody" uses USENET anymore. [20:21] Which leaves the big providers now (e.g. Astraweb/news() [20:21] it was really the 90s that usenet died [20:21] mercutio: I haven't had any out-of-sync issues lately... at least, nothing that wasn't sync w/in an hour or two. [20:21] oh you have nntp access now days? [20:21] last time i checked out usenet it was full of spam [20:22] i think the problem before was with the interconnections [20:22] mnathani: Nowadays, USENET continues to live on, but not for sharing news - those forums have generally been moved on to the Web. Instead, USENET is the #1 alternative to torrents. [20:22] like there could be gaps somehow [20:22] mercutio: ^ [20:22] brycec: yeah, well binary usenet is blah [20:22] heh mercutio remembers when USENET actually had more people than spam... You've gotta be OLD [20:22] i'd prefer messages being on usenet or such than heaps of mailing lsits [20:23] (or forums) [20:23] forusm are even worse [20:23] aye [20:23] i hate forums [20:23] aye aye [20:23] mailing lists at leat you can sign up to and stick procmail on [20:23] but it's still more farting around than usenet [20:23] acutally [20:23] my main gripe with mailing lists is i want them to send me a few years of old posts sometimes [20:23] mnathani: hopefully that answered your query. [20:23] but that shouldnt' be terribly complicated to do. [20:23] brycec Kind of [20:24] i suppose some mail servers wont' liek that though :( [20:24] I am still confused how the content syncs regardless of which server it was uploaded on [20:24] ha [20:24] mnathani: magic. [20:24] mnathani: there are nntp sync daemons... but unless you're running a news server, you don't have to care about it. [20:24] like do some providers just transfer data 24/7 multiple times to other providers [20:25] Pretty much. [20:25] there has to be some kind of map [20:25] i think usenet is like fidonet? [20:25] Though, it's usually just new posts that get pushed to their peering partners. [20:25] to show who is upstream [20:25] and who peers with whom etc [20:25] you have a whole lot of noeds and leaf nodes [20:25] mnathani: Yeah, the providers peer with other providers... It's a big mesh. [20:25] *** cmeiklejohn has quit IRC (Quit: ["Textual IRC Client: www.textualapp.com"]) [20:26] can you get free usenet peering? [20:26] lmao... no [20:26] well, yes, sorta? [20:26] but its not like bittorrent where a partial file is sent to multiple nodes at the same time and by the time 1 unit of that file is transferred, everyone has it? [20:26] like say i wanted to start alt.vps [20:27] or something [20:27] also [20:27] I couldn't say, never been in the business of admin'ing a news server [20:27] say i wanted to start alt.vps [20:27] i meant [20:27] how would that be done? [20:27] I wanted to start alt.vps [20:27] there I said it [20:27] lol [20:27] and how do you make sure other people have it [20:28] is the protocol similar to SMTP in any way? [20:28] NNTP I mean [20:28] mercutio: As I understand it, you either ask a provider to start it, or host it yourself. If a provider starts it, it will usually trickle out to the others. [20:28] mnathani: extremely similar in many ways. [20:28] question becomes, what if I want to become a provider and start it myselft? [20:28] are there discussions on usenet anymore? [20:29] mercutio: in the case of hosting it yourself, you then need followers who would request their provider carry your feed. [20:29] mercutio: no. :p [20:29] it's all binaries now :) [20:29] (not literally... but that's all it's good for imho) [20:29] mnathani: $packagemanager install innd [20:29] http://www.giganews.com/peering.html [20:29] hmm [20:30] so where do people go to have internet conversations now? [20:31] mailing lists, forums, twitter, facebook, the loo... [20:31] oh, ans iex [20:31] *irc [20:32] it's not the same [20:32] *shrug* [20:36] http://www.aioe.org/index.php?how-to-setup-a-feed-with-aioeorg [20:37] fwiw, there are a couple of free-access news servers I use to supplement my Astranews connections, newszilla6.xs4all.nl and weathergirl-ipv6.tele2.net (both ipv6 only as I recall) [20:37] heh, that's pretty much what I described, aint it?\ [20:38] (very nice walk-through though) [20:39] hmm [20:39] is slrn still the best reader to use? [20:40] tin was like the old one wasn't it [20:42] I've only used KNode recently, and that's because news:// was associated with it somehow (silly KDE). Before that, many years ago or so, I used Thunderbird. [20:42] oh? [20:43] i've never liked netscape [20:43] well it was the only newsreader I knew of at the time, and I was curious was nntp was all about [20:43] ahh yip [20:43] i wonder if slrn can use http proxy [20:51] ok got it [20:51] it loaded way faster than dialup used to load :) [20:51] wow there are binaries [20:52] you sound surprised [20:52] well it's free nntp [20:53] Dear user, This group is not locally stored, but now due to your [20:53] request it is being downloaded. Articles should be available in a [20:53] ok that's lame [20:53] it had fetched some porn binary group [20:53] i suppose it depends what the users had requested already [20:56] a cache to a bigger site would seem more sensible [20:56] than this fetch group as reading thing [21:17] the one use of usenet that was not mentioned above is that some mailing lists are proxied into the usenet arena and some people actually post to mailing lists from newsgroups and/or read mailing lists from the newsgroups [21:18] I should enhance my rss2email perl program to also have an alter ego rss2usenet .. [21:19] yeh [21:19] i think i found one like that [21:19] this non instant thing bugs me though [21:19] well what you need to understand [21:20] is that usnenet is the internet modern incarnation of uucp [21:20] uucp -> unix to unix copy [21:20] err i meanhaving to read the message that says that messages will be pulled [21:20] various unix nodes did not have direct connectivity nor did they have continual connectivity [21:20] e-mail works fine for mailing lists... [21:20] yeh i used to use uucp for all my internet email [21:20] so there were various dialup connections established so data could be copied w/out calling long distance [21:20] static: you don't go back intime nicely though [21:20] so things eventually trickled from point a to point b [21:20] yeh [21:21] like i checked a post from NZ to NZ on usenet [21:21] and it was like 6 points [21:21] usenet is continually connected server farms still doing the uucp style store and forward [21:21] (and went out of the country) [21:21] hmm [21:21] you can consider usenet propogation delay slightly slower than mailing list propogation delay [21:21] that's not the problem [21:21] the problem is that the stupid nntp server i'm using doesn't subscribe to things [21:21] consider if you're at the tail end of a long list of people on a mailing list .. you'll not get instant email of course [21:21] until people read messaeg in that gruop [21:22] and it's pretty slow at gettign new stuff [21:22] and it's messy/clumsy to get it to realise there are new messages on the client [21:22] toddf: even if not the tail end but a very large list like linux-kernel [21:22] sounds like its a caching server instead of a host of sorts with a crontab to fetch news from groups people are interested in [21:22] anyway [21:22] it's "dnews" apparently [21:22] one application you may not realize is that 'if you eventually get your content' then thats awesome for things like .. traveling to space and inhabiting other planets [21:23] heh [21:23] Dear user, This group is not locally stored, but now due to your [21:23] request it is being downloaded. Articles should be available in a [21:23] few minutes. To see a progress report you will have to re-read [21:23] hmm i wonder hwo to get slrn to reread [21:24] there is actually a sccs (space communications system) protocol you can use that is eerily like uucp with security and 4 dimentional coordinates for antenna pointing added .. (time + 3 dimentional location of satellites etc) [21:24] i heh tin is still availble [21:24] I used tin way back in the day [21:24] idk what you guys are talking about :( too young for this :p [21:26] staticsafe: well usenet kind of died when the internet went mainstream [21:27] that is true [21:27] like usenet was never really mainstraem [21:27] i've used usenet (only for binaries) [21:27] i used usenet via uucp for a bit [21:27] err [21:27] reai t [21:27] i don't think i actually have posted on usenet [21:28] i don't post that much in gneeral though [21:31] look for tfries@umr.edu .. I did a ton of stupid posts attempting to learn from the internet and even participated in some flame wars .. those were the days .. starting around 1994'ish [21:32] heh [21:33] you contributed to linux toddf ? [21:33] stupid posts >.> [21:33] for xt hard disk controllers :) [21:33] I modularized the xd driver, woot! [21:33] oh xt, sorry, bad memory [21:33] it's called xd [21:33] one of them ancient things I had back in the day. I may still have it, no clue if it works though heh [21:33] but the comment says for xt [21:34] oh boy, now we need to fix it, talk about years old bugs being fixed! [21:35] haha [21:35] oh you're an openbsd developer [21:36] yes, I got tired of trying to build linux back in 1998'ish timeframe and turns out 'cvs up && make build' is easy vs 'download individual packages and build your own build tree' .. I was hooked lined and sinkered from that point on. [21:37] heh [21:37] i start using openbsd in like 2001 [21:37] well i played with it a little in 2000 [21:38] which was like 2.9 or something? [21:38] hmm maybe 2.7 :) [21:39] the web will have better info than my brain at this hour of night at the end of a long week ... ;-) [21:39] heh [21:39] do you use openbsd as desktop toddf? [21:39] I'm just hoping up_the_irons manages to make an appearance and during said appearance manages to flip an iso image for my new vps .. [21:40] indeed I do [21:40] what window manager do you use? [21:40] wow...my first UNIX-like OS usage was Ubuntu 8.04 [21:40] Wow, can anyone say Round Robin DNS? http://dns.winvive.com/propagate.php?Domain=news-out.nntp.giganews.com&QueryType=A [21:40] cwm ftw [21:40] >.> [21:40] <.< [21:40] staticsafe: how old are you? [21:40] $ xlsclients | grep xterm | wc -l [21:40] 172 [21:40] mercutio: 19 [21:40] haven't tried cwm [21:40] staticsafe: oh [21:41] i was 19 in 2001 [21:41] it started out life as a 9wm clone [21:41] i think [21:41] yeh [21:41] i used to hop between twm and enlightenment [21:41] and now all the code is re-written .. you can do anything with the keyboard w/out a mouse and its quite minimalistic, I even have 1 pixel boarders on all my windows [21:41] until i found ion [21:41] is ion that tiling window manager? [21:41] yeh [21:42] if you can stand / and or work with tiles all day long you have to checkout spectrwm [21:42] I can't stand them thus I don't use it but those who do tend to like spectrwm ;-) [21:43] *** dzup has quit IRC (Read error: Connection reset by peer) [21:43] hmm every time i investigate other ones they're not as good as ion [21:43] like they'll try to automanage your windows [21:43] or other annoying things [21:44] I like KDE ;) [21:44] XFCE too [21:44] DE wise [21:44] i only have 1920x1200 resolution [21:44] i find that it's much easier when dealing with limited resolution to have tiling window manager [21:45] with tabbing [21:46] it's weird, back when ion came out nothing really compared [21:46] like twm was kind of similar speed [21:46] but had worse ui [21:46] and enlightenment was kind of better default ui but a bit slow [21:47] but even now when computers have got faster, alot of desktops can still feel sluggish [21:47] partially from finding where windows "hid" [21:47] at least with twm you could iconify [21:49] I'll try to notice when you report back how spectrwm works out for you. I've been asked to try it out, and promptly got tired of not having 80x25 terms unless I manually resize and decided tiling wm's were not for me [21:49] do you need 80x25? [21:50] i find 100 columns nicee [21:50] someday if I ever get the time I want to do a truly 3d window manager. not one that decorates windows in fancy 3d shadows and overlays and transparency. I mean all apps get displayed in a 3d space managed by the window manager. someday. [21:50] pretty much 100x30 is like basic working size for me [21:50] after that it can be nice to have extra [21:50] or things on the side [21:50] *** dzup has joined #arpnetworks [21:50] hmm [21:50] for irssi with 200 windows yeah I make a wide term. but for the rest, 80x25 seems usable, I maximize terms I code in usually, but then I shrink them as I move on to other distractions [21:50] i want a non window manager [21:50] err [21:50] i want somethin gthat filters/aligns information for me [21:50] from many sources [21:51] font of 5x7 is also my preference everywhere [21:51] where like i can put filters on stuff in real time etc [21:51] like i don't like having to type zcat blah.gz | grep ... [21:51] | less [21:51] blah blah [21:51] and then resize the window cos the input is too wide [21:52] i think the whole way at looking at things could be improved [21:52] rather than just have a new window manager [21:54] my brother pointed out to me the other day that openbsd's man page rendering is 80 coluns regardless of the size of the term. he's used to linux auto generating things based on the size of the term. probably has a point. [21:54] if you want your window to resize based on your input I bet thats doable. [21:55] if you want to write a 'smart less' that reformats and wraps output to match terminal widths thats quite another. [21:55] well i don't always [21:55] it's like when tcpdumping [21:55] sometimes i don't care about having as much width for all the timestamps and shit [21:56] urgle, you know you have too many packages installed when... [21:56] Filesystem Size Used Avail Capacity Mounted on [21:56] /dev/sd1e 9.8G 9.4G -95.7M 101% /usr/local [21:56] but don't want it to be too wdie [21:56] but lik [21:56] dynamic filters [21:56] where you can specify to munge input appropriately [21:56] and jump back to it would be col [21:56] todd: wow [21:56] what do you have in there? :) [21:56] what don't I have installed... [21:57] I've trialed a lot of stuff and I use openoffice and latex and java and firefox and test thunderbird and evolution .. name something big I probably have it installed [21:57] now I'm wondering if someone dropped a core or something because indeed that seems excessive to have filled that up even for me [21:57] thats after I just removed jdk-1.6.x since I have 1.7.x installed [21:58] time to cull the /var/db/pkg/.lib* subdirs [21:59] s/subdirs/packages/ [22:00] https://bbs.archlinux.org/viewtopic.php?pid=1160979 [22:00] that says awesome is the most popular [22:00] for tiling window maangers [22:00] way more so than ion [22:00] but curiously openbox has more keen users than all of them [22:01] i assume openbox is a fork of fluxbox? [22:01] depends on if you are license sensitive. awesome = GPLv2+, spectrwm = BSD [22:01] i'm not that sensitive for a wm [22:02] but i didn't think much of awesome [22:02] i'm more concerned about license for base system stuff [22:03] irssi is gpl toddf :) [22:04] find me something that is bsd licensed that works even remotely similar [22:04] gcc is gpl also [22:04] yeah [22:04] gcc i think is bigger issue myself [22:04] especially seeing it's gnu [22:04] gnu code is scary [22:06] hmm ircii is bsd [22:06] biggist dir in /usr/local is share/ and biggest dirs in there are untouchable ;-( [22:06] 184300 gcompris [22:06] 211162 gtk-doc [22:06] 411636 doc [22:06] 495968 locale [22:06] 655720 icons [22:06] 999934 texmf-dist [22:06] wtf [22:07] mv texmf-dist to remove? [22:07] remote [22:07] gtk-doc?! [22:07] actually just purge gtk-doc? :) [22:07] do you do gtk development? [22:07] it is populated by any gtk package that includes documentation [22:08] /var/db/pkg/atk-2.6.0/+CONTENTS:share/gtk-doc/html/atk/up.png [22:08] someone should teach them to share 'up.png' icons though [22:09] hmm [22:09] I have and expect to coninute to contribute to xombrero, originally a gtk2 now a gtk3 + webkit based browser [22:09] having 'devhelp' with fully populated docs really helps [22:10] # du -s /usr/local/share/* | sort -nr | head [22:10] 39512 /usr/local/share/vim [22:11] vim is my biggest [22:11] 21500 /usr/local/share/doc [22:11] 10956 /usr/local/share/locale [22:11] 9700 /usr/local/share/gettext [22:11] gettext?! [22:11] gah [22:11] i hate gettext [22:11] there's c code in there [22:12] wc -l vasnprintf.c [22:12] 5568 vasnprintf.c [22:12] epic c code :/ [22:12] without tabs [22:12] # endif [22:12] # endif [22:12] #endif [22:12] and wtf is that [22:13] some people think C preprocessing is a programming language and it deserves its own style of indentation, of course! [22:13] it's gnu [22:13] it's hideous [22:13] # if NEED_PRINTF_LONG_DOUBLE [22:13] they have stuff like that [22:13] then the whole code block isn't indented [22:14] and the gaps on it [22:14] make it like the code isn't seperated out properly [22:14] like single blank line before the #if and after that #if [22:14] if the #if was condensed into the code block itd' be more obvious that code block was only used if that #if was being used [22:15] then go down a page and it's got the very same if need blah [22:15] which straight away suggests it needs modularisation [22:15] and maybe moer than one file [22:15] so that it's not 5000+ lines long [22:17] * mercutio has finished ranting [22:18] toddf: oh don't know if you noticed before, but beta-vm openbsd works with smp [22:18] beta-vm ? [22:18] err kvr276 [22:18] -6 [22:19] so kvm isn't as screwed with openbsd now [22:19] I had a beta vps until I roasted it and now with my home inet being unstable I've decided to not rely on home inet for serving various web pages (was backing afs cache from home to a http server at arpnetworks) .. so I converted my beta vps to a paid vps .. [22:19] I think what you mean is [22:19] mpbios is not screwing with uniprocessor kernels [22:20] no [22:20] mpbios is working [22:20] mpbios0 at bios0: Intel MP Specification 1.4 [22:20] cpu0 at mainbus0: apid 0 (boot processor) [22:20] cpu1 at mainbus0: apid 1 (application processor) [22:20] etc [22:20] cpu0: QEMU Virtual CPU version 0.12.3, 2600.87 MHz [22:20] cpu1: QEMU Virtual CPU version 0.12.3, 2600.55 MHz [22:20] or are you saying I should have asked to test smp on the beta vps and failed to do so? [22:20] heh [22:20] i dunno maybe you can still get extra cpu? [22:20] or did the beta vps'en have multi cpus and I just didn't notice? [22:21] they don't by default [22:21] i dunno what policy is about smp in general [22:21] I do [22:21] $$ for additional cpus [22:21] just like $$ for more mem/disk/net [22:21] is it much/ [22:21] I do not recall ever seeing the pricing [22:21] I just recall seeing the discussions on here randomly [22:21] ahh [22:22] i have benchmarks [22:22] not sure why its not officially advertised in the 'powerups' section but it does make reference to supporting multiple cpus in a couple of places [22:23] and i reckon that 2 cpu cores is prob enoguh [22:24] to be helpful without overkill when compilin gor such [22:24] cos you still have disk waits etc [22:24] compiling is not going to benifit a lot unless you use '-j2' [22:24] i did [22:24] i did straight -j2 and -j5 [22:24] haven't tried with non smp kernel though [22:24] but with modern openbsd and rthreads you get multi cpu benifits from multithreaded apps (mysql) not just multiprocess apps (postgresql) [22:24] http://support.arpnetworks.com/kb/vps/why-does-my-vps-show-only-1-cpu [22:24] i mean i tried on non smp on my other vm [22:25] and my other vm was faster with straight make [22:25] additional CPU cores for $2.00 USD / core / month. A total of 8 cores is supported. [22:25] We're sorry, but something went wrong [22:25] We've been notified about this issue and we'll take a look at it shortly. [22:25] it's still doing that [22:25] mnathani: hmm that's not too bad [22:25] mnathani: wow, nice. someday that should be listed in the powerups ;-) [22:25] although it's normal to be included in bigger plans [22:26] with other places [22:26] i wonder if diff people have diff priority [22:26] toddf: is there aio? [22:26] i hardly run any threaded apps [22:26] I asked earlier and the answer was no, 1 core no priority even with the bigger plans [22:27] i hardly do anything cpu itnesneive evne [22:27] well, if you look at the powerups page and such I actually traded memory for more disk space on my new vps (equivalent cost etc) so I'm sure the majority of people hosting at arpnetworks use 1 cpu or else it'd be different default ;-) [22:27] i've always figured that multiple cores will hbelp if you ever compress files ora ynthing [22:27] to not slow down other things running [22:27] is basically my way of looking at things [22:27] heh [22:28] aio is some lib wrapper with threads right? not sure that exists just yet, but with rthreads you get the separation of a process (can have blocking fd's in various threads) w/the behavior of a thread [22:29] i think mysql is the only threaded app i'm running [22:29] and it's not doing much anyway [22:30] *** sako has quit IRC (Ping timeout: 276 seconds) [22:31] hmm, something tells me I should go to bed and wait on my iso flipping for morning. ;-( [22:33] lots of times the benefits of threading can be achevied in other ways anyway [22:33] the only pause i hit regularly is if i do bgp feed [22:33] indeed [22:33] and type bgpctl show summary [22:33] it'll pause for ages [22:33] when loading new bgp routes [22:34] but if it just interjected every now and then [22:34] it'd allow that to work more promptly [22:34] learning lots of bgp routes tends to use a lot of cpu [22:34] for some reaosn [22:34] arp only has 100k routes though [22:34] until the new routers come in [22:34] and i think there are 3 people doing bgp :) [22:34] i notice it more on work machines [22:35] but it's when you lose a bgp peer that you want the summary :) [22:40] can't say that I've really played with bgp. suppored a couple routers for a short while but someone else set them up and I really didn't want to touch it for fear of breaking them. ;-( [22:40] heh [22:41] overall i find openbgpd pretty good [22:41] i just wish there was a linux version :) [22:41] vendor lockin! [22:42] heh [22:42] openbsd doesn't run under xen :( [22:42] well not well [22:43] it only takes someone who cares to make it domU capable at least [22:48] well I guess up_the_irons is not going to appear tonight. no worries, will need to wait until monday to have time to deal with the new vps properly anyway. cio. [22:50] later [23:37] *** HighJinx has joined #arpnetworks