| *** ambv has quit IRC | 00:19 | |
| *** hajs has joined #europython | 03:11 | |
| *** zeth has quit IRC | 03:22 | |
| *** zeth has joined #europython | 03:23 | |
| *** hajs has quit IRC | 03:43 | |
| *** ambv has joined #europython | 10:02 | |
| ambv | http://www.jihadwatch.org/2010/04/may-20-everybody-draw-muhammad-day.html | 10:58 |
|---|---|---|
| *** Aiste has joined #europython | 11:14 | |
| *** MrTopf has joined #europython | 13:00 | |
| *** aa_ has quit IRC | 15:11 | |
| *** ronny_ has joined #europython | 15:11 | |
| *** ronny has quit IRC | 15:13 | |
| *** aa_ has joined #europython | 15:14 | |
| *** Aiste has quit IRC | 15:45 | |
| zeth | Hi | 17:18 |
| _dboddie | Hi | 17:18 |
| zeth | How are you? | 17:18 |
| *** povbot has joined #europython | 17:30 | |
| zeth | what should it be? | 17:31 |
| zeth | since I don't have a polish keyboard, can you paste here what it should be? | 17:31 |
| ambv | zeth: sure thing, here goes: | 17:32 |
| zeth | ? | 17:33 |
| zeth | I got nothing | 17:33 |
| ambv | (typing) | 17:33 |
| zeth | did you send the name | 17:33 |
| zeth | ah sorry | 17:34 |
| ambv | Łukasz Langa is a rookie father and husband based in Poznań, Poland. In his free time he's working as a contractor for companies using Python including STX Next, a company developing solutions for the banking sector in Poland. Łukasz is a perfectionist, big fan of readable and testable code, building stuff is his thing since the first box of Legos. | 17:35 |
| ambv | this is it, in UTF-8 | 17:35 |
| ambv | slightly redacted so my other employers won't complain (I have a couple of those) | 17:36 |
| zeth | Weird | 17:38 |
| zeth | the L turns into a question mark | 17:38 |
| ambv | how about ń in Poznań | 17:39 |
| zeth | Incorrect string value: '\xC5\x81ukas...' for column 'bio' at row 1 | 17:39 |
| zeth | This may be some mistake how we set up django | 17:40 |
| ambv | what DB are you using? | 17:40 |
| zeth | mysql | 17:41 |
| ambv | :( | 17:41 |
| zeth | I will check the encoding | 17:41 |
| ambv | yup, it's probably the default | 17:41 |
| ambv | which is ISO-8859-1 | 17:41 |
| ambv | Polish special characters aren't in it, they're in ISO-8859-2. but you're better off with UTF-8 anyways. | 17:42 |
| ambv | but beware not to screw all other special characters on the whole website. | 17:42 |
| ambv | all those ISO-8859-1 ones are currently okay. | 17:42 |
| ambv | maybe the safest thing to do will be to use some HTML entities for the unicode stuff | 17:43 |
| ambv | and use UTF-8 next year | 17:43 |
| zeth | DEFAULT CHARSET=latin1 | 17:43 |
| ambv | yup, latin1 is ISO-8859-1 | 17:43 |
| ambv | so, leave it | 17:43 |
| ambv | try this: (html entity encoding in progress, please wait...) | 17:44 |
| ambv | zeth: in the mean time, you can remove my evil twin brother from the listing ;-) | 17:45 |
| zeth | I tryuy | 17:45 |
| zeth | alter table submission_speaker modify bio longtext character set utf8; | 17:45 |
| zeth | That seemed to work | 17:45 |
| ambv | no difference on the website, caching? | 17:46 |
| zeth | alter table submission_speaker modify name varchar(32) character set utf8; | 17:46 |
| zeth | website is generated | 17:47 |
| ambv | ah OK | 17:47 |
| zeth | hold on a sec | 17:47 |
| ambv | sure thing, I'm waiting :) | 17:47 |
| zeth | It has updated in the django | 17:49 |
| zeth | for reasons I don't particularly agree with | 17:50 |
| zeth | the stuff goes into django | 17:50 |
| zeth | and then goes into pickles | 17:50 |
| zeth | then goes into a completely different system | 17:50 |
| zeth | which generates the website | 17:50 |
| zeth | because john hates django basically | 17:50 |
| zeth | so now it is correct in the Django part (my bit), I might leave it until someone next regenerates the website | 17:51 |
| ambv | in the django the lion sleeps tonight | 17:51 |
| ambv | zeth: BTW, I'm covered in terms of the photo (I have uploaded a square one) but what about most of the other speakers who have their faces warped? | 17:52 |
| zeth | last year we had someone tidy them up | 17:53 |
| ambv | and Grzechu Jakacki has a huge one | 17:53 |
| ambv | ;-) | 17:53 |
| zeth | I don't think the person who tidied the photos up is involved anymore | 17:53 |
| *** paul_boddie has joined #europython | 17:58 | |
| paul_boddie | zeth: Did you see my mail to europython-improve about the talk submissions creating a docutils error? | 18:08 |
| zeth | I have it here | 18:09 |
| zeth | what is the nature of the medical emergency? | 18:09 |
| zeth | Which talk is the problem? | 18:09 |
| paul_boddie | Emergency? It's just a docutils error! http://www.europython.eu/talks/talk_abstracts/#talk124 | 18:10 |
| paul_boddie | The PyPy 1.2 talk. | 18:10 |
| zeth | "please state the nature of the medical emergency?" - is star trek joke | 18:10 |
| zeth | whenever they turn on the doctor (who is computer generated hologram, he says that) | 18:11 |
| zeth | I see the problem | 18:11 |
| aa_ | zeth: that guy was my idol when I was at medical school | 18:12 |
| aa_ | I even had a star trek badge for my white coat | 18:12 |
| aa_ | ah, umour is so lost on ill people | 18:12 |
| aa_ | humour | 18:12 |
| paul_boddie | aa_: Are you another person who did medical school but didn't go on to be a medical person? | 18:13 |
| aa_ | paul_boddie: I did practise for 6 years after medical school | 18:13 |
| aa_ | paul_boddie: but the same outcome in the end :D | 18:13 |
| paul_boddie | aa_: I guess the answer is "no", then. :-) | 18:13 |
| aa_ | I needed to save money to pay off the loans and go back to uni to study computer science | 18:14 |
| zeth | I thought about becoming a barrister, then I found out the average salary of barristers is less than the average programmer | 18:14 |
| zeth | [PyPy is updated in the datasbase, so that system error should be gone next time they generate it, I will look at the other rst errors] | 18:15 |
| paul_boddie | I was worried that the pickles were all you had. | 18:16 |
| paul_boddie | I also found that the blog software wanted to double-escape everything. Apostrophe became &39; and reST links became encoded twice. Nasty stuff! | 18:17 |
| zeth | Only other error I could find on the talk page was on "Speed thrills" | 18:17 |
| paul_boddie | I only saw one error, myself. | 18:18 |
| zeth | The data goes into the django system | 18:18 |
| zeth | but it is not published from django | 18:18 |
| zeth | that would be too simple | 18:19 |
| zeth | It goes Django > Pickle > Something else > generated HTML | 18:19 |
| paul_boddie | Yes, it makes for a bit of a mess. | 18:19 |
| zeth | plus the wiki and blog and so on | 18:22 |
| paul_boddie | Well, I don't see the point of this Pyramid stuff, and I've said so a number of times now. It isn't easier than having a Wiki, and even the stated benefits of writing HTML by hand don't really seem to come through when you look at the kind of HTML on the site. | 18:24 |
| zeth | Well I don't think it matters too much now | 18:24 |
| zeth | as the whole thing will be binned in two months | 18:24 |
| paul_boddie | Nope. I guess the Italians will keep doing their own thing for next year. | 18:25 |
| zeth | and the world will run on whatever PyConIT uses now | 18:25 |
| zeth | but for PyConUK 2011, we should rationalise the system perhaps | 18:25 |
| paul_boddie | And in a couple of years, people running a conference (maybe even EP) will say, "This existing stuff is just too much. Let's write something simpler!" | 18:25 |
| zeth | Well I think with this system, the talks and speakers need to be shown in Django | 18:26 |
| zeth | or the submission stuff needs to be done in the other system | 18:26 |
| paul_boddie | I have to agree. There's no point integrating stuff for the sake of integration itself. | 18:27 |
| zeth | Yeah, we did try to run with the idea that each hosting country would be required to sort out the venue etc, but the EuroPython society would keep control of the website and so on | 18:28 |
| zeth | but that seemed to have run out of steam | 18:28 |
| zeth | The benefit of europython soc running the site would be that you would get some CRM benefits | 18:29 |
| zeth | as people could keep a constant login from year to year | 18:29 |
| zeth | and people can get rebooked in semi-automatically | 18:29 |
| zeth | but because PyConIT is running it, it maybe makes more sense for them just to do their stuff | 18:30 |
| zeth | and I will just push my pram around florence eating ice cream | 18:30 |
| paul_boddie | I advocated this a long time ago, but people aren't motivated to make this happen. Another thing is that there's always this constant urge to redevelop everything, so that people who were involved (in addition to being worn down by the work they've already done) feel that their experience isn't being taken seriously. | 18:31 |
| zeth | a lot of the problem is that the people who run the conference (like us) get demob happy at the end of the conference | 18:32 |
| zeth | we ditch everything and catch up on our normal lives | 18:32 |
| zeth | I think an approach might be just to develop software as generic conference hosting content system | 18:33 |
| zeth | and if EP in any given year wants to adopt it, then fine | 18:33 |
| zeth | if other conferences want it then fine | 18:33 |
| paul_boddie | Well, for local organisers, it's a lot of work. The problem is that there are other people who could be taking a lot of the load, but they don't have much influence. I understand this: if you're trying to just get things done, it's easier to grab a bunch of local people and put them in a room until you've implemented something, but then there's no remote involvement and the load has to be borne by the local people. | 18:34 |
| zeth | Yeah indeed. The idea of making the EuroPython Soc keep constant development of the software, website, data etc is still a good idea | 18:35 |
| zeth | but for that I think we would need to reboot the EuroPython soc | 18:35 |
| zeth | At the moment the EuroPython Soc just means Laura | 18:35 |
| zeth | in practical terms | 18:35 |
| zeth | We would need to create a committee etc | 18:36 |
| zeth | that would be semi-detached from the local activity each year | 18:36 |
| zeth | I think already some people have been talking to the PSF about making a European-PSF | 18:37 |
| zeth | so one idea would be to fold the EuroPython Society into that | 18:37 |
| paul_boddie | Despite the complaints, Indico permitted people to get on with the local stuff, and there are other systems out there. I guess the PyCon Tech stuff could be used similarly. I still think that these solutions focus on the wrong stuff - things like ultra-pedantic talk review - instead of the important stuff like registration, useful information. | 18:37 |
| zeth | true | 18:38 |
| zeth | I agree | 18:38 |
| zeth | the really important thing is delegates | 18:38 |
| zeth | not speaker | 18:38 |
| zeth | s | 18:38 |
| zeth | the CRM type stuff should be with delegates | 18:38 |
| *** MrTopf has quit IRC | 18:38 | |
| zeth | delegates should be able to login and rebook for next year etc | 18:39 |
| zeth | get info | 18:39 |
| paul_boddie | If the Italians and others don't get bored soon, there may be no major need for EuroPython any more. Even the incubating powers of EuroPython seem quite limited, really. That's arguably what people should be focusing on. | 18:39 |
| zeth | ? | 18:40 |
| zeth | Explain more? | 18:40 |
| paul_boddie | Helping people start their own conferences. | 18:40 |
| zeth | yep | 18:41 |
| zeth | and perhaps more importantly, their own local user groups | 18:41 |
| zeth | BTW, have you seen this? http://wiki.python.org/psf/MembersAgenda | 18:41 |
| zeth | under "Discuss setting up a European PSF branch" | 18:41 |
| paul_boddie | Yes. Has it been updated recently? | 18:41 |
| zeth | no | 18:41 |
| zeth | so I think EuroPython type stuff could be folded into that | 18:42 |
| zeth | because the US Pycon is organised by PyConUS | 18:42 |
| zeth | Akin to what you said above. In a way EuroPython doesn't really exist. | 18:42 |
| zeth | It is whatever conference that would have happened anyway, just moved to July and labelled EuroPython | 18:43 |
| zeth | PyConUK and PyConIT happen anyway | 18:43 |
| zeth | this year the first is EuroPython | 18:43 |
| paul_boddie | I guess having a local Python-wide target for donations and sponsors could be useful. Currently, it's all about event-specific sponsorship which doesn't seem to get onto anyone's budgets in any permanent way. | 18:43 |
| zeth | next year PyConIT is EuroPython | 18:43 |
| zeth | One alternative to EuroPython, if there was this PSFEurope getting more regular sponsorship, would be for | 18:44 |
| zeth | the PSFEurope to sponsor the best few speakers of X conference to speak also at Y conference | 18:45 |
| zeth | So someone who gave a great talk at PyConIT then goes and gives it in the UK or wheverever | 18:46 |
| zeth | Depends on the volcanos | 18:46 |
| zeth | Volcanos permitting, people like to go to new places | 18:46 |
| zeth | people like to go hang out in florence | 18:46 |
| zeth | maybe even they like to hang out in birmingham | 18:46 |
| paul_boddie | Probably the biggest benefit is just locality: European sponsors would find it more justifiable to sponsor a European organisation, and it'd be easier from a practical perspective, too. | 18:46 |
| paul_boddie | I wouldn't mind going to Iceland again. | 18:47 |
| zeth | so for people like us, who enjoy conferences, we don't mind going over to florence | 18:48 |
| zeth | but maybe most people are not like that | 18:48 |
| zeth | and maybe in the longer run, developing more local stuff is better use of time and money | 18:48 |
| zeth | I hope one day they sort out roaming charges in the EU | 18:54 |
| zeth | if there was ubiquitous and cheap 3G | 18:55 |
| zeth | then people who want internet bring 3g with them | 18:55 |
| zeth | If in say 5 years time, laptop battery life is far longer | 18:56 |
| zeth | then that would make organising a geek conference much easier | 18:56 |
| zeth | no need for power or wifi install | 18:56 |
| zeth | so choice of buildings would be much wider | 18:57 |
| zeth | just have any old shed | 18:57 |
| zeth | and setup your own local conference | 18:57 |
| ambv | zeth: not with 3G you won't | 19:02 |
| ambv | there are a couple of tens of available data channels for each 3G BTS | 19:03 |
| ambv | don't remember the exact number | 19:03 |
| ambv | but even a hundred is far too small for a conference with a couple of hundreds of attendees | 19:03 |
| ambv | plus, data rates with that load are dramatically low | 19:04 |
| ambv | so wi-fi is still the way to go | 19:04 |
| ambv | BTW, I also have to go now. | 19:04 |
| * paul_boddie still doesn't understand why people go to conferences only to spend all their time surfing the Internet. | 19:04 | |
| ambv | I hope someone will unpickle those pickles | 19:05 |
| ambv | and the amount of question marks will drop in my bio | 19:05 |
| ambv | bye now guys! | 19:05 |
| paul_boddie | ambv is The Riddler! | 19:05 |
| paul_boddie | See you! | 19:05 |
| zeth | bye ambv | 19:05 |
| zeth | paul_boddie: I agree | 19:06 |
| zeth | spending all your time on internet is a bit much | 19:06 |
| zeth | however, people do need a little bit | 19:06 |
| zeth | to sort themselves out | 19:06 |
| zeth | handle crises at their workplaces | 19:06 |
| zeth | and so on | 19:06 |
| zeth | and hotel internet is already nonsense | 19:06 |
| paul_boddie | Would like to see that bandwidth quota idea implemented. | 19:06 |
| zeth | What was that idea? | 19:07 |
| paul_boddie | Quentin and someone else were talking about it. Limit people to 3G speeds in order to stop people thinking that the fat pipe is for their movie downloading pleasure. | 19:07 |
| *** ambv has quit IRC | 19:09 | |
| zeth | I didn't see it as much of a problem last year | 19:11 |
| zeth | people socialised with me at least | 19:11 |
| * _dboddie kicks stupid Wiki "parsers" that can't handle things like trailing whitespace. | 19:15 | |
| zeth | oh right, what wiki "parser" in particular? | 19:17 |
| aa_ | even worse are language parsers that can't handle traiing whitespace /me hides | 19:17 |
| _dboddie | zeth: Textile | 19:18 |
| zeth | oh right | 19:18 |
| _dboddie | Still, it's not as bad as using Markdown as a Wiki syntax... | 19:18 |
| zeth | I never really used it | 19:19 |
| zeth | I know that the Django markup module supports textile, markup and restructuredtext | 19:19 |
| zeth | but I don't like any of them really | 19:19 |
| zeth | they are all over the top | 19:19 |
| zeth | They start with a great idea | 19:19 |
| zeth | lets have a markup that looks nice when read in text form | 19:20 |
| zeth | but then they make hundreds of tags that look ugly and lose the simplicity | 19:20 |
| _dboddie | Maybe the problem is that I'm usually converting some other format to a Wiki format. | 19:21 |
| zeth | yeah | 19:21 |
| zeth | and that is the other thing I hate | 19:21 |
| zeth | they often have really poor one way parsers | 19:21 |
| zeth | so you can go from restructuredtext to HTML ot PDF or whatever | 19:22 |
| zeth | but you cannot go from HTML to restructuredtext | 19:22 |
| zeth | very easily | 19:22 |
| *** Wu has joined #europython | 19:22 | |
| zeth | Maybe one just has to do things in HTML | 19:22 |
| _dboddie | My mistake: it's OK with trailing whitespace - it's just that it doesn't like links to be enclosed with markup for italic text. | 19:23 |
| zeth | oh thats weird | 19:23 |
| _dboddie | Normal Wiki markup uses [ and ] for this, and there's a good reason for that, which people tend to forget. | 19:24 |
| paul_boddie | Although Moin's table syntax could be better, you don't want to stare at the reST table markup for very long. I found that things like URLs just can't go in the "simple" tables in various places at all. | 19:46 |
| zeth | the reSt url syntax is too hard | 19:47 |
| zeth | Link_ then | 19:47 |
| zeth | .. _Link: http:www... | 19:47 |
| zeth | no one would guess that | 19:47 |
| _dboddie | It's a nice idea, but it doesn't scale very well. | 19:48 |
| paul_boddie | Yes, but if you look at the EP front page, there are relative links in the not-particularly-nice-looking table. Try making some begin with "http:" and docutils gets really upset. | 19:48 |
| paul_boddie | Because you've used a colon. | 19:48 |
| zeth | yeah | 20:18 |
| aa_ | I know this is a really silly question, and sorry, but are _dboddie and paul_boddie related? :) | 20:19 |
| zeth | yeah | 20:19 |
| zeth | they are both from the superclass Boddie | 20:19 |
| aa_ | hah | 20:19 |
| zeth | or both instances of the class Boddiw | 20:19 |
| zeth | Boddie | 20:19 |
| zeth | not sure which | 20:20 |
| zeth | paul = Boddie() | 20:20 |
| zeth | david = Boddie() | 20:20 |
| zeth | big hierarchy or derived classes and superclasses is a bit Java-ish | 20:21 |
| zeth | so we keep it flat | 20:21 |
| aa_ | zeth: are you speaking this year? | 20:21 |
| zeth | ? | 20:24 |
| zeth | what year? | 20:24 |
| aa_ | 2010? | 20:24 |
| zeth | are you talking about new style classes? | 20:24 |
| aa_ | no I am talking about spekaing at europy | 20:24 |
| zeth | they are there but people don't use them in the same extreme way in java | 20:24 |
| zeth | ah okay | 20:24 |
| zeth | sorry | 20:24 |
| zeth | completely got the sense wrong | 20:25 |
| aa_ | it's ok, I can see my lack of punctuation was confusing | 20:25 |
| zeth | I submitted a talk today | 20:25 |
| aa_ | cool | 20:25 |
| zeth | might fit somewhere | 20:25 |
| aa_ | I'll be there to heckle you | 20:25 |
| aa_ | zeth: and have we got pyconuk again next year? | 20:25 |
| zeth | I couldn't decided whether to keep on with the same as before or write a new talk from scratch | 20:25 |
| zeth | went for a new talk | 20:25 |
| aa_ | cool, about what? | 20:25 |
| zeth | Using OWL in Python | 20:26 |
| aa_ | OWL as in RDF schema stuff? | 20:26 |
| zeth | I originally called the talk "Who the Fuck are you" but I changed it to "What does it all mean" | 20:26 |
| zeth | yeah | 20:26 |
| aa_ | oh cool, I'll definitely be there | 20:26 |
| aa_ | have had some interesting wanderings through all that semantic bollocks | 20:26 |
| zeth | I will probably rewrite what I have done so far | 20:27 |
| aa_ | by bollocks I do of course mean "stuff" | 20:27 |
| zeth | since someone else has already submitted a semantic web talk | 20:27 |
| aa_ | that's a django talk | 20:27 |
| zeth | but is doing RDF with practical working stuff | 20:27 |
| zeth | I am doing more of a "lets go on a journry to see if this all nonsense or can we make it do something useful out of it" | 20:28 |
| aa_ | call it "RDF is the 'java' of data storage" | 20:28 |
| paul_boddie | It's all nonsense! ;-) | 20:28 |
| paul_boddie | Spoiler alert! :-) | 20:28 |
| zeth | well yeah | 20:29 |
| zeth | My default position is that the semantic web etc is all nonsense | 20:29 |
| zeth | but I have a go to use OWL | 20:29 |
| zeth | aa_: tell me about your interesting wanderings | 20:33 |
| aa_ | zeth: medical data storage | 20:34 |
| aa_ | massive distributed data sets | 20:34 |
| aa_ | where literally old granny publishes her blood glucose, and then gp queries it | 20:35 |
| aa_ | or some drug company publish their research and clinicians use it to make decisions | 20:35 |
| aa_ | then the data is indexed and searched like a search engine | 20:35 |
| aa_ | it's actually not too bad an idea | 20:36 |
| paul_boddie | What technologies, aa_? | 20:36 |
| aa_ | paul_boddie: just plain rdf | 20:36 |
| aa_ | with that querying thing, rasql? | 20:36 |
| zeth | I am trying to use it with manuscripts, my work is paying me to find out about it | 20:36 |
| paul_boddie | But the indexing and stuff? | 20:36 |
| aa_ | paul_boddie: ah, no idea on that | 20:37 |
| paul_boddie | It's a magic box?! | 20:37 |
| aa_ | paul_boddie: I was thinking something custom | 20:37 |
| aa_ | paul_boddie: it never got to the build it stage | 20:37 |
| zeth | so the talk will hopefully appear as part of the process | 20:37 |
| paul_boddie | I heard that the Allegro Lisp people do RDF stuff these days. They do like their niches. ;-) | 20:37 |
| zeth | so if ambv comes back we can tell him that his name is fixed: http://www.europython.eu/talks/speakers/index.html#langa_lukasz | 20:38 |
| zeth | he has a little line through the L rather than the question mark | 20:38 |
| aa_ | paul_boddie: for a start any of the indexing engines should be fine I think | 20:38 |
| zeth | the site has been updated | 20:38 |
| paul_boddie | I wrote an RDBMS-backed triple store once. Wouldn't recommend it, really. | 20:39 |
| aa_ | paul_boddie: triplestores are popping up over and over | 20:39 |
| aa_ | paul_boddie: did your one make release? | 20:40 |
| paul_boddie | I'm skeptical about a lot of this stuff. | 20:40 |
| paul_boddie | http://www.boddie.org.uk/python/sqltriples.html | 20:40 |
| aa_ | paul_boddie: why skeptical? | 20:41 |
| paul_boddie | I think it's tempting to use RDF because "we might not know what we're going to store", but there are some awkward consequences of dumping stuff into a triple store, rather than specifying a proper schema and doing it all relationally. | 20:42 |
| paul_boddie | This thing used RDF underneath: http://www.boddie.org.uk/python/WebOrganiser.html | 20:42 |
| zeth | good point | 20:42 |
| zeth | I'll take that for my talk | 20:42 |
| zeth | For me, RDF and OWL etc | 20:43 |
| paul_boddie | I've already taken some stuff discussed today for my talk. ;-) | 20:43 |
| zeth | are more for exchanging data | 20:43 |
| zeth | rather than storing it permentantly | 20:43 |
| aa_ | paul_boddie: the attraction for me was having an inherently scalable situation, where everyone publishes as much data as they need | 20:43 |
| aa_ | "it's scalable like the internet" is how I tried to explain it to the money people | 20:44 |
| paul_boddie | I think the query optimisation is pretty awkward. I've learned a lot about that since making that solution, and I'd probably want to think a lot harder about it if I had to do that stuff again. | 20:44 |
| aa_ | paul_boddie: also, if you have a real graph of data, sometimes squeezing it into rdbms is a pain | 20:46 |
| paul_boddie | Not that you couldn't make it work, but I think people don't really appreciate that one major strength of RDBMS solutions is the work they do to find a hopefully optimal query strategy. | 20:46 |
| aa_ | many man hours gone into that though in the history of the universe :P | 20:46 |
| paul_boddie | aa_: Agreed. You can have a basic triples table, which is what I did, but it doesn't lend itself to good performance. | 20:46 |
| paul_boddie | The main reason for writing that stuff was because rdflib had a dependency on the awful bsddb and its continual readiness to lock itself into oblivion. | 20:48 |
| aa_ | paul_boddie: I found rdflib really bad | 21:02 |
| aa_ | well, way better than I could make, but still :) | 21:03 |
| paul_boddie | It may all be better now, but bsddb in particular was annoying. | 21:04 |
| * paul_boddie has to go. | 21:04 | |
| paul_boddie | See you all! | 21:04 |
| *** paul_boddie has left #europython | 21:04 | |
| aa_ | pygtkhelpers | 22:49 |
| aa_ | oops wrong window | 22:49 |
| aa_ | wrogn screen and wrong keyboard too, if anyone was wondering :D | 22:49 |
Generated by irclog2html.py 4.0.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!