[00:27:00] Hello, anybody here? [00:27:31] I need an urgent help [01:06:14] boom [01:08:20] drobbins, i create a new extension to SMW but i need to change the output page in MediaWiki .. [01:08:33] even better, the simple page in Media Wiki has a tab with READ right? [01:08:45] when you click there, the mediawiki create a template based on Category page [01:08:58] but i need to change this template [01:09:23] probably, there is a hook that help me .. [01:10:06] don't know. [01:11:00] oh .. ok, thanks! [01:52:42] is there a way to check which concepts have a certain page? [01:59:57] yaron: Can we check which concepts contain certain pages? [02:05:35] saptch: well, you can check it manually, by going to the concept page. [02:06:04] but there are thousands of concept pages [02:06:27] yaron: and I dont know which ones contain the page I am interested in [02:09:03] I have put {{#arraymap: [[Concept:some_concept]] |,|x| [[concept members::x]] }} [02:09:13] Oh, I get it now. [02:09:20] on all concept pages [02:09:42] then i query [[concept members::_pageIaminterestedin]] [02:10:00] But thats ugly, and it has problems, caching etc [02:10:12] That's a weird #arraymap call. [02:11:08] A better way to phrase the question would be, "what concepts does a certain page belong to?" [02:11:15] I just want to get it clear in my mind. [02:11:43] Yes [02:11:58] {{#arraymap: {{#ask: [[Concept:{{PAGENAME}}]] |link=none}}|,|x|{{#set: Concept members=x}} }} [02:12:07] it is like this on every Concept page [02:12:27] Oh - that looks better already. :) [02:13:01] But is there any better solution [02:13:10] Of course, you need to resave every concept, every time its set of pages changes, for this to work. [02:13:22] Yeah, exactly [02:13:36] And that is inefficient and ugly [02:13:51] Isn't there way to just mimic the query that that concept uses? [02:13:56] Also you actually have to PUT IT on every concept page :( [02:14:24] Nope, we dont know what query the concept is executing [02:14:31] Er, what? [02:14:41] what? [02:14:54] How do you not know? [02:15:08] Because we haven't reached the concept yet [02:15:14] and there are thousands of them [02:15:20] I don't understand. [02:15:34] (Also, I don't know why you would have thousands of concepts, but that's another story.) [02:16:12] You said to mimic the query that that concept uses [02:16:22] Right. [02:16:36] What do you mean by "we haven't reached the concept"? [02:16:38] But firstly how do we know which concept's query we wanna mimic? [02:17:47] Well, I guess I don't really understand your site's structure. [02:18:03] It's rare for wikis to use concepts at all, let alone thousands of them. [02:18:06] And I also dont know what "mimicing the query" would accomplish, [02:18:20] because the end goal is to Select the concept page [02:19:27] Well, let's say a concept A is defined by the query "[[B::C]]". [02:20:07] Then, you can call {{#show:PageX|?B}}, and if it's equal to "C", you know it belongs to concept A. [02:20:54] But the problems is how would you know the premises? [02:21:16] Well, that gets back to my question - how do you not know what each concept holds? [02:21:17] i.e. it should have B::C [02:21:37] I simply don't understand what your wiki is all about. [02:22:23] Well I would say my wiki is a bit complicated [02:22:49] That part I can believe. :) [02:22:59] But is there any solution [02:23:19] "what concepts does a certain page belong to?" as you said [02:23:25] I offered one solution... [02:23:45] that is not possible though [02:23:52] If you go to a concept page, and click "Edit", you can see the query it uses. [02:24:36] We dont even have to click edit... [02:25:08] That's probably true, yes. [02:25:17] But you are not understanding, it is an inline query which returns all the concepts a certain page belongs to [02:25:56] I have employed this: {{#arraymap: {{#ask: [[Concept:{{PAGENAME}}]] |link=none}}|,|x|{{#set: Concept members=x}} }} [02:26:06] on the concept page [02:26:13] Yes, but it doesn't have to be a single query. [02:26:33] and then [[Concept members::{{PAGENAME}} on the page [02:26:42] Sorry to harp on this, because it doesn't really get to your question, but - why do you have thousands of concepts? [02:26:42] but I want a better solution [02:27:11] There is some kind of inferencing [02:27:15] and AI stuff [02:27:52] Last time I went down my model, I couldn't find any other solution... [02:28:11] So it must have to be done like this [02:29:06] Yeah, I don't think there's a good solution - maybe the best solution would be to create a result format for it, in PHP - somewhat similar to the "incoming" format defined in SRF. [02:29:40] (Not that similar, but similar in concept, in that it queries the data structure, not actual data.) [02:30:18] Ok I'll look into it, though I am not experienced [02:30:42] If I could hire one of you, how much would it cost? [02:30:57] It depends on which one of us you hire. :) [02:31:21] I'll write you privately. [10:20:52] Hi everyone [10:27:40] hi! [10:28:39] Well it seems I am not the only one fiddeling with SMW at the moment [10:30:23] many ppl are fiddling with SMW [10:30:31] not so many of them are in this channel though [10:32:32] Or acutally awake as it seems ;) [10:33:03] most channels have more lurkers than chatters [10:33:57] How long have you been working with SMW? [10:34:33] some 3 years now [10:34:44] started early '09 [10:35:23] I heard about for the first time in mid 2010 [10:36:03] It's been a process learning to implement and use it, thats for sure [10:36:49] almost all people will agree on that [10:43:35] How often does it happen to you that you ask yourself: "Why doesn't it work as planned it?" [10:45:11] rarely [10:46:09] I've got a relatively simple closed-world information model, and I avoid overstretching the wiki beyond it's wiki character [10:50:00] Currently I have got several Queries with template formats and these templates have other queries with template format. at one point a template just does not appear anymore. Can't figure out why yet. It:s not this one: http://en.wikipedia.org/wiki/Help:Template#Expand_limits [10:53:32] It is mostly supposedly random stuff like this that keeps me occupied that just pops out of nowwhere. [10:55:34] not exactly random [10:55:42] seen http://semantic-mediawiki.org/wiki/Help:$smwgQMaxDepth ? [10:55:56] Ha! [10:56:21] also http://semantic-mediawiki.org/wiki/Help:$maxRecursionDepth [10:56:34] the latter has a default of 2 [10:57:17] Note that the value is not chosen just to annoy users [10:57:50] "The default of 2 is chosen since the total length of pages grows exponentially [10:57:50] in this number. If the wikitext is too large, PHP will run out of memory, the [10:57:50] page will appear blank and be hard to edit to correct the problem. This should [10:57:50] be avoided by default. If you know what you are doing, feel free to increase [10:57:50] the number to a higher value." [10:58:35] Sir, I am much oblieged! [10:58:43] you're welcome [10:59:50] I'd advise creating smarter templates, smarter queries, and perhaps amend your information model, before increasing the value of the recursion depth param [11:00:12] That is also why I said "supposedly" random. Sometimes it is hard to find the exact help page if you are not sure what you are actually looking for [11:00:41] and that's where the SMW community comes in handy [11:09:22] I think I will try default=3, PHP Memory shouldn't be a problem. But you are right, usually I do not need this many templates this overview page was requeste for an existing information model that would need a lot of tweeking, However thank you very much, I will keep these restrictions in mind in future. [11:10:18] k [11:18:06] so with properties I can add semantic information to a text [11:18:24] can I do the same with a subobject instead of a property? [11:18:42] Domerich: essentially yes [11:19:07] the difference is: a property can be a direct annotation in the text, a subobject is more of a programmatical declaration somewhere on the page [11:19:37] but if you use templates to annotate and show information, that difference isn't a hindrance [11:21:22] that sounds interesting but I cannot fully understand yet. What would be in the template et cetera? [11:22:12] Saruman you have helped indirectly this wiki: https://energypedia.info/index.php/Main_Page [11:23:17] Troika good to know [11:23:56] if you're satisfied with the quality of the (semantic) work, you could nominate the wiki for Wiki of the Month on smw.org [11:24:25] is there an example of how this could be accomplished? [11:25:20] Domerich: accomplish what? [11:27:34] okay an example follows [11:29:29] [[marked text:: This is a text passage I want to annotade|start line=2|end line=4|topic=topic1|topic=topic2]] [11:30:11] so afterwards it will show me article with text partials that contain topic1 for example [11:31:40] prolly need to create a template {{textmarking|field1= |field2= }} etc in which the fields are annotated [[MyProperty::{{{field1|}}}]] and shown [11:32:06] within the template you can also construct subobjects for more elaborate data constructs [11:32:21] what/how the template shows your information is up to you [11:33:01] main thing for the template is, that the semantic work is contained in the template, so that users aren't encumbered by it [11:33:31] added bonus: it's easy to create forms so that users can create/edit the information; the templates still handle all annotation and markup [11:33:54] that sounds really nice but I have no idea how to start :D [11:34:30] I know forms, templates et cetera but well now I have no know what to actually do here... [11:35:45] OK maybe start with this: http://workingwithmediawiki.com/ [11:36:04] lotsa MW and SMW information in there, written by SMW dev Yaron [11:37:20] ok thanks [11:57:59] what exactly would field1 and field2 be in your example? [11:59:36] ah okay that was the call for the template [12:01:13] so it would go like this {{Textmarking|actualtext|property1|property2... }} [12:02:31] and in the template I could create a subobject with that properties [12:05:01] possible, but if you leave out field names, you're going to have to work with unnamed parameters in the template [12:05:38] in your example, in the template {{{1}}} would be "actualtext" and {{{2}}} "property1" [12:05:42] very obtuse [12:06:11] rather have {{Textmarking|maintext=actualtext|field1=property1|field2=property2... }} [12:06:15] yeah Im going with named parameters [12:06:59] Yaron's book page 27 [12:08:06] I wanted to use named parameters all along I just said it wrong ;) [12:08:20] with acutaltext being my named variable ^^ [12:08:51] the book only comes free until page 19 ^^ [12:09:26] well you really need to invest in learning MW [12:10:01] if you can't invest the $20 for the ebook, you're going to need to set up a sandbox and create some very simple wiki's yourself, [12:10:15] and that takes a lot of time [12:11:13] People in this channel help with problems, but they generally don't feel inclined to providing a free online MW+SMW training [12:11:46] at least not that I know [12:15:43] that's only fair [12:46:19] Just ordered the paperback [12:49:56] Might in come handy to browse through some of the later chapters [12:50:53] well I'll tell you it's even handy in the first couple of chapters [12:51:04] little snippets here and there that were news to me regardless [12:53:24] Yeah, I think I will read it carefully there is a chance that I missed some basics and that I could be doing some things more efficiently [12:56:14] mhh so I built myself a sandbox [13:00:17] Saruman: is that anything like you meant it? http://domerich.funpic.de/wiki/index.php/Template:MarkText a sort of query http://domerich.funpic.de/wiki/index.php/Test with pages like http://domerich.funpic.de/wiki/index.php/Test1 [13:04:34] it seems to work somehow [13:07:48] I found a book by Denny Vrandečić, maybe this will help [13:10:19] Domerich: did a little improving on your template [13:12:19] thank you, let me check it out :) [13:20:49] that's awesome, also your code makes a lot more sense ^^ [13:21:29] heh, good [13:21:56] in the end, learning MW/SMW is often a case of learning lotsa little tricks [13:22:05] :D [13:22:10] also, go check out ohter semantic wikis, see how they did things! [13:22:52] I do in the help docs but usually they are too complex for me to understand [13:23:11] {{{startline|1}}} does that set a default value if nothing is entered? [13:23:20] it sets default value 1 [13:23:39] and {{{startline|}}} simply sets nothing [13:24:10] without the pipe, the variable "shows up" when left empty [13:24:12] okay but when some value is delivered, then that will be taken instead of 1 [13:24:20] exaclty [13:27:55] great. is it possbile [13:29:12] to also state the assoziated pagename to the query I made? that should make an actual link because these pages do exist [13:29:47] maybe that info has to be stored in the subocject first [13:30:08] ? [13:31:31] I have that query that returns all pages which have "objects" that contain things. now on my overview I can query those pages [13:32:39] oh now it works :D [13:33:11] so now I can navigate to the according article http://domerich.funpic.de/wiki/index.php/Test [13:35:20] so later if someone wants all lines that have the topic Kulturanthropologie they can have a list and also go to that article [13:39:15] Hi! [13:39:41] its semantic forms broken? [13:40:15] I just installed latest stable version of SMW and Maps [13:41:08] and when I try to active semantic forms extension, i get this: "Fatal error: Call to a member function setInputTypeHook() on a non-object in /var/www/p2pfoundation.net/html/extensions/SemanticMaps/includes/forminputs/SM_FormInputs.php on line 95" [13:41:33] This is my wiki installation: http://p2pfoundation.net/Special:Version [13:41:42] can you give me any help? [13:44:27] Yaco: is your SF enabled after the enableSemantics line? [13:46:01] and is SM enabled after SF? [13:57:35] Saruman, SM is enabled before SF [13:57:55] that may be the cause [13:58:28] so i should sort them like: [13:58:39] enablesemantics [13:58:41] "Some functionality of Semantic Maps is obtained by hooking into other extensions and is only available if these are installed as well. " [13:58:47] include Semanticforms [13:59:02] include semanticmediawiki.php [14:00:02] saruman SM = SemanticMaps? if so, i got SF after SM [14:00:14] SM = maps, yes [14:00:17] order should be [14:00:26] require_once("$IP/extensions/Validator/Validator.php" ); [14:00:45] require_once("$IP/extensions/SemanticMediaWiki/SemanticMediaWiki.php"); [14:00:45] enableSemantics('somestring'); [14:01:00] require_once("$IP/extensions/SemanticForms/SemanticForms.php"); [14:01:11] require_once("$IP/extensions/SemanticMaps/SemanticMaps.php"); [14:01:22] as far as I can tell, since I don't use SM [14:01:27] so, SF before SM [14:01:31] * Saruman never was much into SM [14:01:42] probably, yes [14:01:55] although each should function without the other as well [14:02:10] I am experienceing trouble with multible attributes here [14:02:30] if you enabled validator and SMW, you can first enable SF but not SM, and see if SF turns up in the Version page [14:02:48] {{MarkText|maintext=Das ist der 1. Testtext|startline=3|endline=5|topic=Kultur|Sprache}} it wont assign both Kultur and Sprache to the property::topic [14:03:28] ok Saruman [14:03:35] SF works if i disable SM [14:03:39] :Yaco in my experience disableing all extensions and then enabling one after another always helps [14:04:29] Saruman, Domerich: if i enable SM and SF together, it brokes [14:04:32] in any order [14:04:43] tryed SM after and before SF [14:05:19] Yaco: prolly one of them is too new [14:05:33] you need to check with JeroenDeDauw or yaron [14:05:35] Saruman, i am using latest stable version [14:07:23] Domerich: it can't, because you only have one field "topic"; you can't assign two values to one parameter [14:07:37] unless you use "topic="Kultur, Sprache" [14:07:49] but that means you can't search properly [14:07:59] gotta go back to the information model for your site [14:08:48] mh I have to think this through. in the docs it says {#subobject:mysubobject [14:08:48] |property1=value1|value2 [14:09:03] so I will have to modify my template maybe [14:09:07] as you said [14:10:12] maybe I should just assign topic1... topicn [14:10:28] but that's not dynamic and a bit of a pity ^^ [14:11:10] Domerich: #subobject does support multiple values, but templates don't [14:11:36] ah that I didnt know [14:11:50] in your call, Kultur winds up in parameter "topic" but Sprache in parameter {{{1}}} as the first unnamed param [14:13:19] ah because it thinks Sprache is the first unnamed parameter in my call [14:14:44] so if topic is the only variable parameter it could go like this in the template maybe [14:14:57] |topic={{{topic|no topic}}}|{{{1}}}|{{{n}}} [14:17:03] yes, but how are you going to create a form for that? [14:18:45] well I will only allow 4 topics or so [14:18:59] and that number could be left empty, right? [14:19:11] *the fields could be left empty [14:19:27] well now it sort of works :) [14:19:53] but I doubt AND , OR searches will work ^^ [14:20:50] They do, up to a point: http://semantic-mediawiki.org/wiki/Help:Selecting_pages [14:23:54] bookmarked. thats more than I need :) [14:24:17] Yaco: are you still having trouble with SF and SM? [14:24:40] hi yaron [14:24:42] yeap [14:24:49] yaron to the rescue! [14:24:55] :) [14:24:55] Well, we'll see. :) [14:24:59] btw yaron compliments on the book [14:25:11] ah, yeah [14:25:13] Oh, great, thank you! [14:25:30] i bought your book [14:25:34] :) [14:25:44] Thanks to you too! I hope you find it useful. [14:25:47] didnt get it yet [14:25:58] so [14:26:08] Well, I guess I'll have to answer your questions directly in the meantime. :) [14:26:16] So what versions of SF, SM and SMW are you using? [14:26:41] Semantic Forms (Version 2.5.1) [14:27:02] Semantic Maps (Version 2.1 alpha) - I just installed this latest repository version, not working either [14:27:09] Semantic MediaWiki (Version 1.8) [14:27:25] shall i install the stable version of SM? [14:27:36] Sure - stable is always better. [14:27:40] ok [14:27:42] just a sec [14:27:57] If you're still having problems after that, please describe the problems. [14:28:19] ok [14:29:49] yaron [14:29:56] i installed stable version of SM and enabled it [14:30:08] now i get this error, see here: http://p2pfoundation.net/Special:Version [14:30:38] That's weird! [14:31:35] Oh... did you put SM in LocalSettings.php after SF? [14:31:44] You should - I bet that's the issue. [14:32:06] SM is after SF [14:32:43] yaron, my last line in LocalSettings: http://pastebin.com/YyFmRgg2 [14:33:50] Hm, weird. I have no idea, then. [14:34:20] I would ask Jeroen, either by email or if he shows up here. [14:34:33] if i disable SM, it works [14:34:44] That's not surprising. [14:35:30] another issue yaron [14:35:33] i go to Special:CreateClass [14:35:37] and try to create a form [14:35:40] then i get his [14:35:53] Fatal error: Call to a member function getNamespace() on a non-object in /var/www/p2pfoundation.net/html/includes/job/JobQueue.php on line 353 [14:36:21] Also strange. [14:36:41] Can you disable SM, so I can see what's on your Special:Version page? [14:36:54] Oh, you did that already [14:36:55] . [14:37:02] yeap [14:38:32] Does that error happen every time you hit "Save" in Special:CreateClass? [14:39:34] testing again with other type of form [14:40:17] the error only appear if i enable the checkbos "This template can be included multiple times on the page." [14:41:39] Well, that's... weird. [14:47:56] i am totally lost [14:49:56] Saruman: Do you think something like this is possible with subobject? http://flockdraw.com/39wjyk [14:55:13] Yaco - JeroenDeDauw is here now; you should re-enable SM, then show him the error. [15:04:31] Hi Samuran: http://semantic-mediawiki.org/wiki/Help:$maxRecursionDepth and http://semantic-mediawiki.org/wiki/Help:$smwgQMaxDepth unfortunately didn't do the trick [15:05:47] yaron: aaaaah [15:06:24] I cranked them up pretty high but nothing changed, I am going to implement a differen layout for the template know that should work fine as a workaround... a.k.a. tactical withdrawal [15:08:07] JeroenDeDauw: sorry. :) [15:08:27] He has a weird problem, though - I've never seen it before. [15:10:14] yaron, JeroenDeDauw you can now see the error again [15:10:23] http://p2pfoundation.net/Special:Version [15:11:08] Oh great, that stuff broke again >_> [15:13:34] I assume the null object there is $sfgFormPrinter - but I don't know why it's not working. [15:14:46] yaron: I am fixing a huge mess already on the 3.0.x branch, only some of which was caused by me [15:15:00] Not very keen on fixing more breakage on 2.0.x that I did not cause [15:15:55] Alright. [15:16:00] Hopefully this is just an isolated incident (sorry, Yaco). [15:17:56] Could be inclusion order [15:19:53] Yeah, very likely a config or compat issue [15:20:01] Other people don't have this problem [15:20:56] JeroenDeDauw, this is the inclusion order in the LocalSettings: http://pastebin.com/YyFmRgg2 [15:25:22] Could it be that there's a conflict because there's a Maps and a SemanticMaps? [15:27:01] or that Maps needs a specific place between validator/smw/sf/sm? [15:27:19] Saruman, Maps is a dependency of SM [15:27:56] so could it be you need validator/smw/sf/maps/sm then? [15:29:24] lets try! [15:29:37] a desperate plan is better than no plan :-D [15:30:07] hey [15:30:15] seems that worked! [15:30:26] thanks Saruman ! [15:30:27] see [15:30:28] http://p2pfoundation.net/Special:Version [15:30:35] hah, I want my SemanticMaps certified logo now! [15:31:18] heheh [15:31:33] Special:CreateClass keeps giving me an error: Fatal error: Call to a member function getNamespace() on a non-object in /var/www/p2pfoundation.net/html/includes/job/JobQueue.php on line 353 [15:31:41] JeroenDeDauw: I've not seen instructions on the order of things on the SM page; a) is this order indeed as you expect, and b) shall I add the instructions then? [15:31:51] JeroenDeDauw, maybe you can help with that other issue? [15:32:37] Namespace error prolly unrelated [15:36:29] Saruman: Yaco: the order is specified in the INSTALL file [15:38:31] thanks JeroenDeDauw [15:38:37] didnt see that [15:39:07] JeroenDeDauw: I see it. I'll put it in the wiki [15:47:49] Saruman: awesome, thnx [15:48:16] A lot of the docs still need to be migrated from the old wiki [15:48:25] * JeroenDeDauw has other priorities though [15:50:43] I'll see if I can chip in on some semantic documentation this weekend [15:51:05] I saw some videos on #subobject that said things that are not in the docs [16:37:57] New review: Nikerabbit; "This might have caused https://bugzilla.wikimedia.org/42830" [mediawiki/extensions/SemanticMediaWiki] (1.9.x) - https://gerrit.wikimedia.org/r/33592 [16:56:48] :) [21:13:48] hi [21:18:20] hi [21:23:48] I have angered the Semantic Mediawiki gods and wish to make a sacrifice to appease them [21:24:14] I'm trying to execute an inline query to call up a calendar and it's erroring out on SMW::executeHierarchyQuery [21:24:33] With this error: 1044: Access denied for user [21:27:08] oeh, that's gonna take some sacrifice [21:27:20] maybe even perform some rituals [21:27:53] like editing the config or replace code or do database work [21:28:28] Harrumph. Well it's trying to access the slave database using a read only user [21:29:25] slave database? [21:29:58] I get enough traffic to merit a master/slave database configuration [21:30:30] well done! [21:30:49] but can MW be run from a read-only DB? [21:31:11] uhm, it's not [21:31:33] I've got it set up with a $wgDBservers array [21:31:33] I thought even read requests from a user result in writes, e.g. updating the has-been-accessed field [21:31:57] Naw, I've got squid caching set up [21:31:57] * Saruman looks up exciting stuff [21:33:02] advanced config [21:33:13] can't advise for this situation [21:33:20] Fooey. Thanks anyways [21:33:39] works fine on my monodatabase setups, but I *really* don't feel like dealing with a cascading replication failure on a friday [21:34:39] I always thought "it hasn't been Friday until I've seen a good cascading replication failure" but maybe I should get with the times :-) [21:35:28] pretty complex, low-tested config, I hope jeroendedauw_ or yaron see this and can chip in [21:35:48] but it's already 22:36 in Europe [21:36:04] Saruman: You'd think so, but restoring from backups is absolutely no fun. Everyone's like, where are all my edits?! [21:36:23] I have to work *so* hard to not go "You're lucky there's a website right now" [21:36:30] edits are highly overrated anyways! [21:36:53] sorry I can't be of real use here [21:37:31] meh. you helped me a *bunch* before [21:37:39] I think your'e still ahead in the karma department [21:38:09] Hi, guys. [21:38:22] Ulfr: that error only happens on that one page? [21:38:46] yaron: Only when I query for a calendar, my other stuff still works fine [21:38:48] * Ulfr knocks on wood [21:39:02] So any calendar query causes this problem? [21:39:40] yaron: correct. It also happened when I tried to run a join query [21:39:49] Ah, okay. [21:39:59] yaron: It's trying to hit my slave database with a read only login, suspect it's trying to write [21:40:01] My guess is that those queries are overloading the system. [21:40:09] I don't think so... [21:40:21] The "calendar" format doesn't do any writing, in any case. [21:40:47] Make sure you're using #calendarstartdate and #calendarenddate in your query - those improve performance quite a bit. [21:41:03] yaron: SMW::executeHierarchyQuery is the actual error it throws. Would not having any results throw an error like that? [21:41:12] Because I never asked my user if they actually had calendar stuff in place [21:41:16] That, I don't know. [21:41:50] yaron: This is my query: http://pastebin.com/yKXwFpwS [21:41:52] I don't think it throws an error in that case - at least, it didn't in previous versions of SMW. [21:42:16] I'm running 1.7.1 [21:42:21] Hm, okay. [21:42:38] Well, what happens if you change "format=calendar" to "format=table"? [21:43:05] error [21:47:11] Oh - very interesting. [21:48:09] Ulfr: I would try to keep minimizing the query, to isolate the exact thing that causes it to crash. [21:48:41] yaron: this is the query it's running [21:48:41] CREATE TEMPORARY TABLE `wd_t2`( id INT UNSIGNED KEY ) ENGINE=MEMORY [21:56:01] yaron: Sorry to poke again but my browser crashed due to crummy work AV. you say anything? [21:57:34] There can be only one! [21:57:41] Confusing! :) [21:57:47] Ulfr: no, I didn't say anything. [21:58:06] It's strange that a simple query would create a temporary table. [21:58:35] Nevertheless, maybe the issue is just that the DB user lacks the permission to create temporary tables. [21:59:09] yaron: It's using a read only account, but i've got it set in wgDBservers to use a different account for writes [21:59:49] It's trying to create a temporary table on a slave db meant for fast reads [22:00:15] Well, you can still allow it to create temporary DBs there, can't you? [22:00:25] I mean, temporary tables. [22:01:13] yaron: I suppose I could when it's not under load, but I've got dozens of queries daisy chained that run perfectly fine in other apps [22:02:12] The part that's bothering me is it has write access to the master [22:06:13] I don't understand - I thought the problem was that it was crashing. [22:06:40] yaron: Correct. It's banging it's head off the sneeze guard protecting the read only slave [22:06:52] when the beans the query needs are available for use [22:07:40] I believe that metaphor is lost on me. :) [22:07:59] Urn [22:08:15] It's trying to create a temporary table on a read only database, and it has full access to the master the slave is cloned from [22:09:05] Yeah, but it's on the read-only DB when it wants to create the temp table. [22:10:25] I mean, I don't know how temp tables are usually handled in read-only DBs. [22:10:51] yaron: Based on my limited experience, they aren't [22:11:01] Alright. [22:11:09] If the issue is a general master-slave DB issue, you might be better off asking on the #mediawiki channel. [22:11:44] yaron: Genuinely appreciate the noodling you gave my problem [22:11:51] ...though perhaps SMW (and Semantic Drilldown, for that matter) are the only MediaWiki-related code that creates temp tables. [22:11:53] Yeah, sure. [22:12:14] Who knows! Maybe ignoring the problem until it goes away will work out for once [22:13:11] * Ulfr headdesks so hard the whole bench shakes [22:13:21] the user's query had the freaking category spelled wrong. [22:13:21] toniher Heiya [22:14:28] Saruman: Hey! I thought your palantir had a PEBKAC detector. [22:14:37] It has [22:14:49] but it's friday, so it's off [22:14:58] Saruman: You really should come with some sort of warning label [22:14:58] or else it'll go crazy when I sit down at my PC [22:15:06] I do [22:16:14] Well gentlemen, thanks a lot for the assistance, I'm going to see how much whiskey it takes to forget that I just spent four hours monkeying with a typo [22:16:47] depends: whisky, or whiskey? [22:16:56] or the totally misspelled bourbon? [22:17:02] :-) [22:17:39] Ulfr: still, if a category name spelled wrong means heavy errors, there might be something to look into [22:19:03] Aha! Yes, Saruman is correct. [22:19:42] I guess misspelling a category (i.e., using a category name that doesn't extist) leads to the creation of a temp table. [22:20:04] That seems plausible, actually. [22:23:16] hey kghbln [22:23:25] toniher :) [22:23:35] I just had a look at tw.n [22:23:48] tw.n ? [22:24:09] There are still some messages for SMW in Catalan which are not translated (tw.n - translatewiki.net) [22:24:32] :O [22:24:48] 31 :( [22:24:52] thanks for commenting, kghbln [22:25:03] this night I'll take a look :/ [22:25:20] toniher. do not feel pressurised [22:25:48] no, no problem. Anyway it'll enter in next minor release, I understand. But I'll try to have it soon. [22:26:41] * toniher goes back to the dinner table :D [22:26:50] That's so it has not to be done immediately. However, it would be cool to have Catalan among the languages fully localised [22:27:07] enjoy ! [22:44:19] Hi! [22:44:31] I need some help with SemanticMaps [22:44:45] any user with experience around? [22:53:09] jeroendedauw_?