#qi-hardware IRC log for Tuesday, 2012-01-17

wolfspraulah, I had to rebuild the planet cache and now email subscribers get all posts again ;-)08:48
wolfsprauloh well08:48
whitequarkwolfspraul: what about http://irclog.whitequark.org/ ?09:31
whitequarkI think it has proven itself to be pretty stable09:31
whitequarklast outage was four weeks ago, and it was due to flaky IPv6 connectivity at my hoster09:31
whitequarkwell, it was "beta" but I thought it would work09:32
whitequarkit would not.09:32
whitequarkthat is, each week it would mysteriously get partly down for a period of time09:32
whitequarkand my bot would reconnect until it'd run out of nicks09:33
whitequarkand then it died09:33
wolfspraulah yes09:33
wolfspraulnext time I do server work I should just install the software, right?09:34
whitequarkyeah, that'll work09:34
whitequarkor you can just point a cname at irclog.wq.org09:34
wolfspraulI would probably want to integrate it with some css or so, but I need to do the design cleanup first anyway09:35
wolfspraulI think it will definitely get done, just not very fast right now09:35
whitequarkcan your work be integrated into upstream?09:36
whitequarkbecause I'm tempted to add support for changeable themes09:36
whitequarkthat would be quite a nice feature09:36
whitequark"dark" and "light" one, for example09:36
wolfspraulI'm not working on it actively right now, but of course we are both in this channel, things can go 'upstream' (to you :-))09:40
kyakwhitequark: is there a way to change a year fast?09:44
kyakside question: where are the logs for 14th and 21st of Dec? :)09:44
kyakyep. it's much better than today's logs (increased usability, fast search)09:48
kyakthe search output would probablt look better when grouped by date rather than each line prefixed with date09:49
kyaknow, i might be wrong, but i can get to a specific date much faster with the current interface09:50
kyakwrong because it is unlikely that someone wants to go to a specific date09:51
kyakhw would rather search09:51
kyakthen search is very important09:51
whitequarkkyak: you can change year in the address line. that's why I made sensible URLs09:52
whitequarkI may think of a better way, through09:52
whitequarkthe logs for 14th and 21st of Dec? simple. IPv6 outage. afther that I've switched to plain old IPv409:52
kyakaddress line, of course :) but it09:52
whitequarkI'll refill these gaps09:52
whitequarkthe search output, hm09:53
whitequarkthat's a good suggestion09:53
whitequarkcan you write an issue on github so I don't forget it?09:53
kyaki can do this09:53
whitequark(Search) yes, it is important09:53
whitequarkthat's why I am maintaining a rather large index and a huge mysql runtime09:54
kyakyep, the search (speed) is great09:54
whitequarkit also has modifiers like "and" and stuff09:54
whitequarkand it's better than "just Google" in this case09:55
kyakcan i apply time constraints to a search? i.e. search within specific time range09:56
whitequarkI think no09:56
whitequarkany ideas for an interface for this?09:57
whitequarkDB is indexed by both timestamp and text (obviously)09:57
kyakmaybe introduce some special word for that09:57
kyakfrom the other hand, this would increase comlexity for nothing..09:58
whitequarkthat'd interfere with the plain querying, won't it?09:58
whitequarkyes, and that too09:58
whitequark"and" and friends are quite obvious09:58
kyakdon't bother :)09:58
whitequarkyou already have that in google09:58
whitequarkI'm not afraid of complexity09:58
whitequarkthe logger is, I dunno, like 200 lines of code09:59
whitequarkand maybe the same for logging bot09:59
whitequarks,logger,log viewer,09:59
whitequarkruby is similar to perl in the amount of code, but quite dissimilar for its readability ;)09:59
kyakthe best thing is that the author (you) is interested in this project.. Unlike the author of current irc logs web interface09:59
kyaki just submitted the issue btw :)10:00
whitequark<trolling-mode>that's because the current logger is written in python</trolling-mode>10:00
whitequarkthat's why I won't be a migen contributor10:01
whitequarkuser, maybne10:01
whitequarkbut not a contributor10:01
kyaki see you didn't specify any well-known license for irclogger - did you do this for a reason?10:01
whitequarknope, just a usual attention deficit10:02
whitequarkit is MIT10:02
whitequarkI'll add the LICENSE when I'll get to that10:02
whitequarkmaybe a few days10:02
whitequarkas per the time range, I think I may add "extended search"10:04
whitequarkwith two fields10:04
kyakor maybe do it google way10:05
kyak"last 24 hours", "last month" etc :)10:05
whitequarkwolfspraul: any ideas on this? what's more convenient?10:06
wolfspraulI need to use it and compare for a while, don't know now10:07
whitequarkI'll do both variants10:08
kyakwhitequark: another thing we already had discussion about here is what to log. Many were concerned by joins/parts/quits being logged, as it reveals the host name10:09
kyaki think the decision was only to log public chat10:09
whitequarkkyak: it's public10:10
whitequarkanyone can join and see the hostnames10:11
whitequark /who, etc.10:11
whitequarkI've already had an argument with some partici^Wtrolls on #ruby-talk10:12
wolfspraullogging the IP addresses does not create any value10:12
wolfspraulwhich value does it create in your opinion?10:12
kyakfirst, it provides no information. second, it may be found ocasionally by googling10:13
whitequarkto be honest, I've been replicating irssi's format10:13
whitequarkit can be removed, of course10:13
wolfspraulat the least it's clutter10:13
whitequarkbut I think that logging joins/parts _do_ has its value10:13
wolfspraulin which way?10:13
whitequarkI have found that useful quite a few times10:13
kyakthird, i have a vhost here; so i keep my privacy even if someone joins. But sometimes i can join unidentified and without vhost and wil be cought by logs10:13
kyaki don't want that10:14
whitequarkkyak: as I've said, I have nothing against removing hostnames in joins/parts10:14
whitequarkno problem with that10:14
whitequarkand it's clutter, yes10:14
kyakremoving hostnames would be great10:15
whitequarkdo you have any experience with mysql?10:15
whitequarkI have something like 2 million of rows10:15
kyaknope :)10:15
kyakwhat is your problem, maybe i still could help?10:16
kyakor someone else here10:16
wolfspraulremoval of hostnames and ips would be welcomed by most I think, I fail to see the value10:16
whitequarkkyak: http://pastie.org/320042710:20
whitequarkit's my "storage format".10:21
whitequarkand it sucks (I know)10:21
kyakprobably you can have different tables for different channels10:22
kyakother than that, why it sucks? it has int id for what matters :)10:22
whitequarkI can remove the hostnames in the logger bot10:26
whitequarkbut I'd need to run some sort of regexp on 500k rows10:26
kyakupdate [table_name] set [field_name] = replace([field_name],'[string_to_find]','[string_to_replace]');10:28
kyakthis is what google says :)10:28
kyakprobably 'replace' supports regexp or there are other functions supporting regexp10:29
kyakyou can also filter on nick == NULL to find those joins/quits10:29
whitequarkthat's how I highlight them in viewer10:42
whitequarkwell, I'll look into it10:45
kyakwhitequark: do you think it is hard to make bot connect via SSL?10:50
kyakit occured to me that some of my channels are ssl-only10:50
whitequarkit's ruby10:52
whitequarkjust a matter of substituting one class for another10:52
kyakoh god, time to update my distro.. my RubyGems is tool old to install "bundle"10:54
whitequarkoh yes10:55
whitequarkthat's a _really_ old rubygems install10:55
kyak1.3.5 :)10:55
whitequarkokay... trying to talk with four or five people simultaneously isn't a very good idea10:56
whitequarkespecially if you consider the fact this channel is in English and it's not native for me10:57
whitequarkso, latter10:57
whitequarkI'll be here at evening (something like 22:00 at UTC+4)10:57
kyakbye, and thanks :)10:57
whitequarkyou're welcome10:57
cladamw(VGA DDC failed :: 6 / 90) wpwrak, I'm thinking this problem...even you mentioned that intermittent problem, possibly12:04
cladamw caused by software.12:04
cladamwwpwrak, but i'm not sure and don't know if existing another potential reason, why i'm thinking this? since i reviewed R2 known issues and compared to R3 known issues about L3 and L19 shorted to fix frozen problem,  would it be also likely to "ground problem" to cause noises?12:09
cladamwwpwrak, sorry that wrong channel again., ;-)12:10
wpwraklovely news: nokia sells out ... to sisvel. http://www.theregister.co.uk/2012/01/16/nokia_patents/17:45
wpwrakthe reg puts a friendly spin on it, but not everyone would agree with that ...17:47
wolfspra1lwe can see - sisvel is doing great :-)18:14
wolfspra1lhey that's our friend Bill Ray, I would totally disagree with him on that one18:18
wolfspra1l"big friendly troll offering FRAND licenses to anyone who wants one"18:18
wolfspra1lha ha18:18
wolfspra1lcan't be more backwards18:18
wolfspra1lwpwrak: you know my guess is that Nokia is testing sisvel, giving them some starting business.18:21
wolfspra1lfor sure there will be payback to Nokia depending on how much money sisvel can make, and so forth18:22
wolfspra1land then nokia can decide later whether to expand busines with them or not, but then they have reliable historical sisvel performance data at hand :-)18:22
viricwhat is sisvel?18:32
viricah I read.18:33
wolfspra1lwaste of time actually, the kind of thing that after you learned about, you wish you never had18:38
larscor the kind of thing after you learned about it, you wish you had known about it before18:42
wpwraksisvel is the unfriendly face of philips and now nokia18:43
wpwrakthe reg has an interesting comment: this could be a ploy to outsource litigation. that way, there may be less risk of retaliatory lawsuits18:45
larscwell, you could still sue nokia nayways18:47
wpwrakyeah, but it may a) be less clear whether they're really behind it, and b) it may also cause a different public perception18:49
wolfspra1llarsc: no, I'm sure. the first one :-)19:46
wolfspra1lthere was a nice NanoNote article today http://www.omgubuntu.co.uk/2012/01/the-ben-nanonote-the-worlds-smallest-linux-laptop/19:49
wpwrakhow's the tendency in ben sales actually ? stable ?20:01
wolfspra1lha, yes, you could call it like that20:01
wpwrakhopefully non-zero, too :)20:01
wolfspra1la few here and there, slowly moving20:01
wolfspra1lthe important part for me right now is to keep it in stock, and that's the case20:02
wolfspra1lpower forward the m1, then marry the two eventually :-)20:02
wpwrakheh :)20:02
urandom__the people should be educated that the nanonote lacks w-lan and slide keyboard cause of patent shit20:16
wpwraknanomist sounds like this: http://en.wikipedia.org/wiki/Prey_(novel)20:16
urandom__wpwrak: we have to many boooks like that, people are already overscared by new tech, we need them to believe in progress not in stagnation20:19
wpwrakoh, a little frisson now and then doesn't do much harm :)20:21
wolfspra1land for our English learners (including me), I looked this up - frisson: a moment of intense excitement :-)20:55
AylaI think that's a French word20:58
Aylabut it's not exactly the same meaning in French20:59
larschalf of the english language is french21:18
AylaI wouldn't say half21:19
wpwrakwordnet says "an almost pleasurable sensation of fright"21:19
Qwin1234Hey all22:48
Qwin1234I wonder if anyone can help me out here with an external drive problem22:48
Qwin1234after unplugging from tv the partition is unrecognizable ... 22:49
Qwin1234So i can format it but there is important data on it22:49
Qwin1234I tried getDataBack to repair the partition but it give me IO errors22:49
Qwin1234So I think it has to do with house of the harddrive but I am not sure. What do you guys think the problem is ? ( Also any recommended tools of fixing the problem )22:50
mthI think you have the wrong channel... this channel is about building open hardware (Ben NanoNote, Milky Mist)22:51
Qwin1234Hmmm true :/22:51
--- Wed Jan 18 201200:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!