| 00:09.29 | *** join/#infobot Ismail_ (n=Ismail@41.251.75.48) |
| 01:27.31 | *** join/#infobot ravenbird (i=ravenbir@unaffiliated/simonrvn/bot/ravenbird) |
| 03:33.25 | *** join/#infobot TimRiker (n=timr@rikers.org) |
| 03:33.25 | *** mode/#infobot [+o TimRiker] by ChanServ |
| 03:34.30 | TimRiker | infobot: status |
| 03:34.31 | infobot | Since Sun Mar 16 03:34:46 2008, there have been 22 modifications, 80 questions, 0 dunnos, 0 morons and 48 commands. I have been awake for 23h 59m 44s this session, and currently reference 114515 factoids. I'm using about 19664 kB of memory. With 0 active forks. Process time user/system 1384.4/63.99 child 0/0 |
| 06:57.40 | simonrvn | holy smokes 100K+ factoids |
| 14:46.51 | *** join/#infobot TimRiker (n=timr@70-7-201-54.area7.spcsdns.net) |
| 14:46.52 | *** mode/#infobot [+o TimRiker] by ChanServ |
| 14:51.10 | troubled | TimRiker: hey |
| 14:51.59 | troubled | I ran those tests on the db normalised for users. it seems that your db just doesnt have enough duplicated infor to justify it |
| 14:52.52 | TimRiker | interesting. :) |
| 14:53.03 | troubled | while the normalized table with the *_by (created, modified etc) fields converted to an unisgned int using the rowid of a "users" table as the fkey, was ~10% faster than the current |
| 14:53.19 | TimRiker | faster for what? |
| 14:53.23 | TimRiker | adding new entries? |
| 14:53.37 | troubled | ...the left join on the table casused a ~30% drop in speed over the current table |
| 14:54.00 | troubled | so yes, normalised is faster by 10%, but the view that would be used in the bot to actually display is a 30%. |
| 14:54.27 | troubled | 30% slower i mean |
| 14:55.30 | troubled | you just dont have enough duplicated info to justify a full nick!ident@host table. although I didnt try seperating the nick/ident/host (ill call a NUH from now on k?) though |
| 14:56.30 | troubled | I think it would be better to seperate it into 3 fields than just 1 big field, but I think you still just dont have enough duplication still and would still be slower. of course, thats just a guess until I actually try it |
| 14:57.46 | troubled | that and the speed of sqlite is so blindingly fast, its not worth it until we get into the million+ row range |
| 14:58.52 | troubled | but only if a majority of the users are repeats. eg: if we had 1 million nicks make 1 million diff factoids, normalizing would be slower. if we had 1 user create 1 million factoids, it would be stupid not to :) |
| 15:00.38 | troubled | i suspect that the big problem is that when its all NUH as one big field, the ip changes of the NUH's cause too many distinct fields. And the fact you only have 20k suggests that you wouldnt gain much still. I can shoot you a copy of the db I tested with if you like though for your own testing |
| 16:15.00 | TimRiker | nah. I removed the NOT NULL on requested_by (iirc) and cleaned up my db a bit. it loaded fine then, and I'm running svn now. |
| 16:15.31 | TimRiker | I did NOT change the other sql inits, but I think we should. |
| 16:15.50 | troubled | pokes his head back in |
| 16:16.13 | troubled | ya, taking out the not nulls probably help a bit. one less constraint check |
| 16:18.29 | troubled | do you think we should remove the tables we arent using yet? Sure we may plan to use them, but the problem is that people would have them created despite being empty and unused. then if we ever do use them, there could be extra work for people if the schema differs |
| 16:21.31 | TimRiker | probably |
| 16:24.56 | troubled | TimRiker: from what I can tell, the following arent needed: connections, freshmeat, news and uptime. Agreed? |
| 16:27.35 | TimRiker | at present, yes. |
| 16:28.05 | TimRiker | users and chans should might want to be get converted to tables |
| 16:28.14 | TimRiker | s/should// |
| 16:45.40 | CIA-26 | infobot: 03djmcgrath * r1641 10/trunk/setup/ (16 files in 4 dirs): * Removed unused tables until needed |
| 16:46.05 | troubled | go figure, passwords at sourceforge actually expire :) |
| 16:49.34 | troubled | TimRiker: did you ever notice that the infobot project still appears to be waiting for an approval email? or am I just misreading the X trashcon icon purpose? |
| 16:51.42 | TimRiker | troubled: huh? where? |
| 16:52.45 | troubled | when you first login to sf.net in your sf.net/my/ page, it shows you the projects you belong to. blootbot has a trash can, but infobot has a trash can with an X on it. I cant remember where but I seem to recall it meaning it didnt have approval yet or something |
| 16:53.05 | troubled | (btw, i set the old infobot bugs to deleted in sf.net, cleanup) |
| 16:55.55 | TimRiker | troubled: that means you are an admin on that project. you can't leave a project that you admin. need to de-admin first. |
| 16:56.07 | troubled | ah |
| 16:57.21 | TimRiker | need to put up a new web site, as we still don't have access to infobot.org |
| 16:57.42 | troubled | sf hosted one? |
| 16:57.43 | TimRiker | blootbot site should forward over to the new one. |
| 16:57.55 | TimRiker | might as well be. should not need anything fancy. |
| 16:58.45 | TimRiker | http://infobot.sourceforge.net/ is a mirror of what is at infobot.org now. so we could just update what's there. |
| 16:58.49 | troubled | start with a mirror of the current infobot.org perhaps? |
| 16:58.56 | troubled | ah, nm then |
| 16:59.33 | troubled | quick decision. version number. |
| 16:59.55 | TimRiker | you'll see that the current one says infobot.org is "official". that's changed now. |
| 17:00.51 | troubled | btw, do we have access to anything fancy on source forget for www stuff like perl or frameworks? or they just html only? |
| 17:01.31 | TimRiker | well, the old infobot was 0.x.x and blootbot was 1.3.3 or something. should probably go 2.0.0 at some point. |
| 17:01.34 | troubled | i got into the Catalyst framework a while ago, wouldnt mind continued use of it if they support it. just for site consistency more than anything |
| 17:02.10 | troubled | ya, i would like 2.0 or something just so I can put up a release. but i would hate to release a 2.0 of trunk atm with that disconnect bug |
| 17:02.46 | TimRiker | troubled: php, python works, mod perl too iirc. but you can't connect to any outside machines. |
| 17:03.19 | troubled | wouldnt need outside machines, just internal is fine. I just want a template system if anthing. like TT |
| 17:03.27 | TimRiker | I don't see it with infobot. It's a ping/pong issue, yes? |
| 17:03.58 | TimRiker | TT? |
| 17:05.51 | troubled | ya, it seems to be ping/pong since the server reports a timeout as well |
| 17:06.09 | troubled | TT is a templating system for perl sites. goes well with catalyst framework |
| 17:07.06 | TimRiker | the less someone needs to learn to update the web site the better. |
| 17:07.13 | TimRiker | it's not like it's a huge site. |
| 17:07.27 | troubled | true, thats why i would settle for just TT :) |
| 17:07.46 | TimRiker | which then means that SomeoneElse would need to learn TT. |
| 17:07.47 | troubled | its pretty straight forward, but html is fine too i guess. just curious thats all |
| 17:08.09 | troubled | ya, but SomeoneElse would have to learn HTML to begin with too heh |
| 17:09.39 | troubled | should we keep imMute on the access list and keep his branch? Just seems like he commited a few rev's, got on don's bad side and never came back |
| 17:10.03 | troubled | he stills hangs in a channel with me though, so I could invite him back into the channel if he wants to stay |
| 17:25.58 | TimRiker | sure. I tend to grant access rather liberally. never know who's going to have time to help out. |
| 17:26.09 | troubled | Sorry! The wiki is experiencing some technical difficulties, and cannot contact the database server. |
| 17:26.10 | TimRiker | do something Bad, and then you loose access. |
| 17:26.12 | troubled | Access denied for user 'infobot'@'10.5.1.114' (using password: YES) |
| 17:26.31 | TimRiker | yeah looking at that now. it's been dead for a long long time. |
| 17:26.33 | troubled | hmm, wiki mysql account fubar'd at some point. i setup wiki while ago and it worked fine |
| 17:26.38 | troubled | ah k |
| 17:27.28 | troubled | i wonder if it was related to my password expiring since I setup the initial wiki |
| 17:28.00 | TimRiker | the mysql changed quite some time ago. looks like we lost the old data then. |
| 17:31.58 | troubled | its ok, wasnt anything in there cept some test data I think |
| 17:32.28 | troubled | i think was just copy/pasted description from main site or something if there was anything |
| 17:35.21 | TimRiker | http://web.archive.org/web/20050118135816/infobot.sourceforge.net/wiki/Current_events <- the only cached page I can find is empty. :/ |
| 17:36.47 | troubled | thats not the wiki i had setup then |
| 17:37.10 | troubled | thats a wikimedia site. the one I used with a sf.net built in one |
| 17:52.21 | TimRiker | ah. sf one might still work, don't know where that lives. |
| 17:53.26 | TimRiker | http://infobot.wiki.sourceforge.net/ seems live. |
| 17:56.09 | troubled | gnome is utterly terrible with regard to copy/paste.... |
| 17:57.05 | troubled | ya, thats the one I started |
| 18:06.10 | TimRiker | k. completely nuked mediawiki as the data was gone anyway. |
| 18:06.18 | TimRiker | changed the link to the SF one. |
| 18:09.31 | troubled | k |
| 19:07.19 | troubled | TimRiker: http://picpaste.com/infobot_logo.png seem ok? |
| 19:23.59 | troubled | take a look at how it looks on the wiki |
| 20:22.57 | *** join/#infobot Dotted (i=Dotted@bureaucrat.wowwiki.dk) |
| 20:30.18 | TimRiker | do we need yet-another-logo ? |
| 20:30.34 | TimRiker | http://infobot.sourceforge.net/infobot.jpg would seem a better choice. |
| 20:53.56 | troubled | thats not a logo, thats just text with an effect applied to it though |
| 20:54.34 | troubled | I like the circle deisgn myself. makes the favicon.ico easier too :) |
| 20:55.30 | troubled | take firefox for example. you dont click an icon thats a word, you click a picture. Anyways, just putting something out there to see what you guys think |
| 21:33.12 | simonrvn | troubled: that's cool |
| 23:45.44 | *** join/#infobot Dj_SiMo7 (n=Dj_SiMo7@41.251.83.78) |
