[00:00:12] um [00:00:32] Krenair, yes: http://wiki.novawebdev.webfactional.com/index.php [00:00:40] you should get something like this: https://aa.wikipedia.org/wiki/Special:Version [00:01:02] that looks pretty broken.. [00:01:19] very broken [00:01:34] i had that magicword problem before, but I found a patch that fixed it for me [00:02:15] you can get the version from javascript though [00:02:18] wgVersion [00:02:18] "1.20.2" [00:02:48] nicky22, I saw something about that, but it said that version 1.19 made the patch unnecessary. Since I'm at 1.20.2 I didn't want to go there without checking. [00:03:21] you realise that 1.20 was EOL before the end of 2013 right? [00:03:32] Krenair, Yeah. I already mentioned that in the beginning of my question. And PHP is at 5.2.17. [00:03:47] Hmm.. I had this on 1.21.3 and the patch fixed it for me [00:03:48] oh [00:03:51] right, yeah [00:03:53] that won't work [00:03:57] Krenair, I'm not the guy who set it up, nor the regular maintainer. [00:04:11] minimum PHP version was 5.3.2 [00:04:20] https://gerrit.wikimedia.org/r/#/c/122450/1/includes/MagicWord.php this is what i used [00:05:29] nicky22, Methinks I'm going to try that. And, Krenair, though I'm not sure WebFaction will respond, I'll try lobbying for a newer PHP so that I can upgrade. [00:05:59] yeah, that patch would've been released for 1.19 and 1.21, not 1.20. You can probably manually apply it to 1.20 but it won't be supported [00:06:01] you might aswell try, you can always revert if that doesn't work for you. [00:19:12] Krenair, nicky22, That was sufficient to fix it... for now. [00:19:21] Krenair, nicky22, thanks. [00:19:27] Did that resolve your login issue? [00:20:37] Krenair, the login issue was completely separate: I was attempting to register with www.mediawiki.org. It's a question for mediawiki's admins. [00:20:46] Oh, I see. [00:20:53] Yes, mediawiki.org is hosted by Wikimedia [00:21:00] And so has integrated login with the rest of Wikimedia's wikis [00:21:44] e.g. https://www.mediawiki.org/wiki/Special:CentralAuth/Krenair [00:21:58] you'll find mediawiki.org listed, along with several hundred other sites [00:22:40] if I change my email, password, etc. on any one of those, they're all changed [00:22:45] Krenair, yeah. I'm just wondering where else I've registered (or who else has the same name). Will the URL you gave work for me to find my doppleganger? [00:22:49] my global rights apply on all of those [00:22:58] Yep [00:23:14] If you put the username into the form at the top and submit, it will show you where the username is registered [00:24:14] And yes. I see it's me. Now to see if I have the password squirreled away. (Almost assuredly yes.) [00:24:48] if not you can try resetting the password [00:24:54] assuming you set an email address [02:18:31] Is the REST API publically accessible? [02:19:10] Yes. [02:19:15] the api should work unless the administrator turned it off [02:19:39] https://rest.wikimedia.org/en.wikipedia.org/v1/?doc [02:19:45] I'm looking at the doc and getting confused about how I actually send a query. [02:19:49] via wgEnableAPI = false [02:19:59] Krenair: REST probably means RESTBase? [02:20:07] Oh, maybe not. [02:20:14] Katie, I don't think so, most wikis won't have that [02:20:17] ResMar: Do you mean api.php? [02:20:22] https://rest.wikimedia.org/en.wikipedia.org/v1/?doc#!/Transforms/transform_html_to_wikitext__title___revision__post [02:20:33] What are you trying to do? [02:20:44] You probably don't need rest.wikimedia.org. [02:20:56] But both api.php and rest.wikimedia.org are public, as far as I know. [02:21:02] It provides capacities that the api.php lacks. [02:21:09] In particular I'm looking at HTML-to-wikicode conversion. [02:21:27] How do I work that into a script? I don't see any obvious instructions on where to send queries. [02:21:57] Like, give me an example of a working URL request. [02:22:04] You realise that this is #mediawiki, right? [02:22:29] I could ask this question on any of the tech channels and someone will eventually answer it. [02:22:34] This channel is fine. [02:22:37] RESTBase is an external, entirely optional and not at all bundled with MediaWiki service [02:22:39] You loiter on all of them anyway. [02:22:41] Wikipedia runs MediaWiki. :-) [02:22:55] Doesn't matter which pie I put my finger in. [02:22:58] I wouldn't call it too external. [02:23:11] I would call it very external. [02:23:19] How? [02:23:24] Okay, but the people working on it are Wikimedia Foundation employees. [02:23:34] It's explicitly developed for VisualEditor and Parsiod. [02:23:36] Parsoid* [02:23:38] And that. [02:23:41] Anyway, this doesn't matter. [02:23:43] What are you trying to do? [02:23:47] It's not even a MediaWiki extension [02:24:02] ResMar: Why are you converting HTML to wikitext? [02:24:02] Oh boo hoo hoo. [02:24:20] I do the opposite in a few scripts. [02:24:31] Katie: I am interested in exploring the capacities of the API, though I have a few concrete usecase examples. [02:24:31] Because HTML is significantly saner to work with than wikitext. [02:24:48] Disagree. [02:24:48] Well, there's Special:ApiSandbox to explore the capacities. [02:24:50] !apisandbox [02:24:55] Parsing links is much, much easier in wikicode. [02:25:07] Useful bot we have. [02:25:07] No it isn't. [02:25:10] wat [02:25:24] Wikitext links can contain all kinds of craziness. [02:25:33] is vastly simpler. [02:25:36] Hmm. [02:25:47] Ok I see your point. [02:25:47] Anyway, if you talk about your actual use-cases, someone may be able to help. [02:25:53] m: vs. :m: vs. meta [02:26:06] But I'd like to ask a simpler question first. [02:26:09] Which is how do I access this API? [02:26:22] Or [[{{template that includes "m:"}}page]]. [02:26:29] Via a data request, not a sandbox. [02:26:36] Or [[{{#if:m}}]] [02:26:44] Edge cases [02:26:55] You make an HTTP POST request to the API. [02:27:04] It lives at https://en.wikipedia.org/w/api.php [02:27:11] You send an HTTP POST request there and you'll get a response. [02:27:25] It's the response body content you care about. [02:27:40] But you generally don't need to go that low level. There are libraries and frameworks that have functions that simplify this. [02:27:53] https://en.wikipedia.org/wiki/Special:ApiSandbox [02:27:54] So that instead of sending raw HTTP requests, you send like "getWikiText(page)" or whatever. [02:28:09] But if you want to send raw requests, that's also fine. [02:28:14] Yeah well...to be honest I find those frameworks too enveloped to delve into. Besides, it's fun to work out the system yourself. [02:28:16] You can use curl. [02:29:03] So are you saying that this is more complicated than going to a URL? [02:29:11] You can go to a URL. [02:29:21] The API also accepts GET requests. [02:29:22] Again...how. [02:29:26] Parts of it, anyway. [02:29:29] That's the MediaWiki API. [02:30:04] Sure. [02:30:29] There's no dedicated sytem for RESTBase access? [02:30:40] So for instance HTML coversion is only available via: https://en.wikipedia.org/w/api.php?action=help&modules=flow-parsoid-utils [02:30:43] ? [02:31:18] If so I don't understand why there's an entire seperate sandbox and documentation page that's been set up [02:31:19] I imagine you'd bypass api.php in this case. [02:31:31] If you want to interact with RESTBase. [02:31:35] You'd just interact with it directly for now. [02:32:12] O...k...uh... [02:33:18] So to do API queries, or at least all of the ones that I have done so far, I have constructed strings and then sent them against `api.php?` using the Python requests library. [02:33:24] And then worked with my back-put. [02:33:35] Right. [02:33:42] So instead of using api.php as your base, you'll use... [02:33:47] https://rest.wikimedia.org/en.wikipedia.org/v1/ [02:33:55] Example: https://rest.wikimedia.org/en.wikipedia.org/v1/page/title/Barack%20Obama/ [02:34:19] Is it impossible for them to have said this somewhere in their documentation? [02:34:36] It's documented here, kind of: https://rest.wikimedia.org/en.wikipedia.org/v1/?doc [02:34:41] There are GET sections. [02:34:44] With some "Try it out!" buttons. [02:34:49] And then there are POST sections. [02:34:58] Which also have "Try it out!" buttons. [02:35:33] This for instance fails: https://rest.wikimedia.org:443/en.wikipedia.org/v1/transform/html/to/wikitext [02:36:06] Fails how? [02:36:17] File not found [02:36:19] what are you posting to it, and what does it return in response? [02:36:25] You have to send a POST request. [02:36:30] With the HTML you want to turn into wikitext. [02:36:32] Did you do that? [02:38:13] POST methods only huh. So I assume the browser doesn't do it. [02:38:30] But then I don't understand what the GET output for your second example request is either. [02:38:43] I see items and then a lot of numbers. [02:39:15] "Lists revisions" [02:39:19] Revision IDs then? [02:39:51] But only ones stored in RESTBase...whatever that means. [02:39:56] is there a php code that can do a ctrl f5 ? [02:40:07] because my counter is only refreshing when i press ctrl f5 on chrome , it works fine on IE [02:40:19] ctrl-f5 = refresh with cache? [02:40:28] i guess so [02:40:31] ResMar: which functionality or content are you looking for? [02:40:33] *without [02:40:40] since it only updates my counter when i press ctrl f5 [02:40:46] when i press f5 it stucks at 0 [02:40:51] on chrome and ff [02:41:07] There is probably a way to force your php not to use cache, rather than trying to do a refresh without cache [02:41:26] gwicke: Fine, is there a LIGHTWEIGHT library available for Python for handling RESTBase API queries? [02:41:31] ae. not named pywikibot [02:41:42] ResMar: any HTTP client will do [02:41:56] requests? [02:42:03] Because I'll be honest, it was easier for me to rewrite my own API request methods than it was to figure out what the heck is going on within pywikibot. [02:42:15] http://docs.python-requests.org/en/latest/ [02:42:31] do u know any? [02:42:35] because i already have this line [02:42:44] $parser->disableCache(); [02:42:51] Take a look at https://phabricator.wikimedia.org/P850 ResMar [02:43:24] particularly lines 84-89 and 113 [02:43:24] *whistles* [02:43:52] grr_: honestly, I know very little php [02:44:08] is ok. thanks though =) [02:44:10] but some people online say you can set some kind of header [02:44:14] i did [02:44:19] damn [02:44:19] apparently not working [02:44:32] If all requests are that complicated between a lack of understand of what's going on (sorry, not a CS major) and the complexity involved and the (typical) lack of a good high-level walkthrough, I doubt it'll be worth the effort for me. [02:45:15] Now, I've not done POST requests before, but I can't imagine they're much worse than GET, and I hope that Requests handles them. [02:45:25] I don't have a CS degree either [02:45:52] most of that paste is the ssl cli client being quite loud [02:46:20] > requests.post(url, data=None, **kwargs) [02:46:37] > Sends a POST request. Returns Response object. [02:46:40] Sounds like business. [02:46:41] ResMar: Do you understand the difference between HTTP GET and HTTP POST requests? [02:46:56] Not at all, I'm sure I could read up about it on...Wikipedia... [02:47:04] Yes. [02:47:08] Beyond that one takes data and the other posts it. [02:47:16] a post request sends data with the url, no? [02:47:22] GET requests are the equivalent of hitting enter and visiting a URL in your browser. [02:47:23] And that there's a third request method of some sort. [02:47:29] A POST request is like a form submission. [02:47:37] So your browser does POSTs all the time. [02:47:40] Every time you make an edit. [02:47:49] Or change your user preferences. Or search for something. [02:47:58] And you make a lot of GETs by browsing around. [02:48:41] Well sure I know that much [02:48:42] GET requests are supposed to be idempotent and not change state on the server. [02:48:47] But that's a high level description. [02:49:00] POST requests generally do change state on the server. [02:49:01] For me it's more a question of "How much worse is it going to be to make a POST request, than it is to make a GET one?" [02:49:15] They're both simple using the requests library. [02:49:23] Well there you go [02:49:49] You should try to be a bit less agitated and whiny. [02:49:59] It'll make people want to help you more. [02:50:05] ez. i've never used the requests library before and i've figured it out just by looking at the docs. [02:50:15] If you don't like the current documentation, write better docs. :-) [02:50:16] They have good docs [02:50:22] Unlike say...pywikibot... [02:50:28] If you don't like the current documentation, write better docs. :-) [02:50:38] Click edit. It's a wiki world. [02:50:42] See [02:50:56] But that assumes that I understand the tech at a masterfully low level. [02:51:12] A guy falls into a hole, see. [02:51:28] Which is a different thing from arriving there looking for a way to get something done, and being frustrated by a document written for people at the same level as the writer. [02:51:29] any idea how to get the author of the created page/article in mediawiki ?? [02:51:35] Or documentation that's just missing in places. [02:51:46] four: ummm [02:51:46] four: You can query the first revision of the page. [02:51:52] Using api.php. [02:51:54] ^ [02:52:11] ResMar: Patches welcome. [02:52:21] Patches? [02:52:38] Software patch. It's a thing. [02:52:45] I can barely use the software, how could I patch it? [02:52:50] ResMar: As they say on "House of Cards," you're entitled to nothing. :-) [02:53:04] You seem to have some grand sense of entitlement and expectation about how things should be. [02:53:13] Which is fine, I guess, but you'll then have to help make it that way. [02:53:26] Wow am I getting into an argument on the Internet? [02:53:36] hmm ok .. [02:53:45] four: https://en.wikipedia.org/wiki/Special:ApiSandbox#action=query&prop=revisions&format=json&rvprop=user&titles=User%3ASn1per [02:53:50] you can probably fiddle with the date parameters [02:54:05] You can change the sort order. [02:54:13] To sort by oldest revision timestamp. [02:54:20] And then limit to 1 entry. [02:54:54] Oh and use formatversion=2 [02:55:04] so i have to use sql ? [02:55:10] No. [02:55:13] You can use api.php. [02:55:19] four: Did you click GEOFBOT's link? [02:55:26] yup i did [02:55:30] looking @ it now [02:55:34] Okay. [02:56:45] four: using Katie's suggestion gets the first revision of my user page: https://en.wikipedia.org/wiki/Special:ApiSandbox#action=query&prop=revisions&format=json&rvprop=user%7Ccontent&rvlimit=1&rvdir=newer&titles=User%3ASn1per [02:56:55] and you can look at who made the revision and you get the creator [02:56:58] four: Just a small note, when you actually build your request I recco you include a `formatversion2` parameter, because otherwise you're going to have to go through a dictionary layer with an effectively random key---the pageid---to get to the data of the request. [02:57:12] Sorry, `&formatversion=2` [02:57:27] wow looks complicated [02:57:36] i didnt know i haveto fiddle with the url i hate this kind of stuff [02:57:51] actually what i want to do is , since i already retrieve the page title out [02:58:00] beside it , i would like to display (Created by XYZ) [02:58:06] still have to do the above u all mentioned? :) [02:58:32] you don't have to fiddle with the URL [02:58:50] APIsandbox gives you a nice interface to play with the parameters, then you get a finished URL [02:59:00] or you can use the parameter values to build a request [02:59:24] Here's a complete example: https://en.wikipedia.org/w/api.php?action=query&prop=revisions&format=json&continue=&formatversion=2&rvprop=user|content&rvlimit=1&rvdir=newer&titles=User%3ASn1per [03:00:24] alright, will look at it . thanks for the help guys [03:02:43] See, I feel like you can't get that easily from the documentation. [03:03:00] I didn't know about formatversion=2 until I was 500 lines into the code I was writing. [03:03:17] Why does it matter? [03:03:56] Without it there's an additional key in the dict, the pageid, which is essentially a random number. [03:04:36] can't you just grab the first item in a dictionary? or am i an idiot [03:04:50] Well I don't know about other languages [03:04:57] But in Python the dict is randomly sorted [03:05:07] And json.loads() arbitrarily resorts it AFAIK [03:05:24] It's not randomly sorted, it's unsorted. [03:05:27] But yeah. [03:06:06] I got past that key by listing the keys with .keys() and then picking it out of the lineup. [03:06:42] Sure. [03:06:42] It was poor work done all the way back at the beginning of the API's history that's loitered around since then for compatibility. [03:07:19] You can think Anomie for fixing that. :-) [03:07:25] Err, s/think/thank/ [03:07:38] Now that graphs are enables [03:07:40] enabled* [03:08:03] I am looking forward to every wiki writing its own local module encapsulations. [03:08:18] The de.wp Module:Bar chart! The en.wp Module:Bar chart! Oy... [03:08:38] Someone should write a bot to deal with it until interwiki transclusion finally shows up. [03:10:00] Sometimes I think about all the work that went into Template:Convert that was obliterated the moment Lua was enabled, and shudder. [03:11:36] Sure, but we're still much better off now with Lua. [03:12:11] hello [03:12:28] hi [03:12:46] Yeah I know [03:12:47] I mean [03:12:57] It's a no brainer; why did it take so long to get Lua installed? [03:13:13] Things are easier to see in hindsight [03:13:30] I'm suspicious of it. [03:13:38] It was a known problem the moment parsers showed up. [03:14:05] Well, maybe not the *moment*, but very quickly... [03:15:48] Same issue with SUL [03:19:25] Wikimedia has become...a large corporation. [03:20:49] Interestingly, when one googles "Wikimedia," an unsuccessful logo entry in the Wikimedia Commons logo competition is displayed [03:21:07] I tend to refer to it as the Wikimedia movement [03:21:19] Distinguished from the WMF and from Wikipedia itself. [03:23:38] ResMar: a large.. corporation? [03:24:08] doesn't feel like a corporation to me. i submit a patch, people tell to fix it or get rid of it, I do what they say. [03:25:52] Internal politics are consistently the least read part of the Signpost, but the one I find most interesting in writing my bit. [03:26:07] i love reading about internal politics [03:26:19] hi unicodesnowman, haven't seen you in a while [03:26:29] unicodesnowman: I still love your nickname. [03:26:40] hi GEOFBOT and ori :) ty [03:26:50] I know more about Erick Moller leaving than I could possibly let on in a neutral venue, and there is a lot more that I don't know, but can guess at. [03:27:01] *Erik [03:27:37] Mo^(**)ller [03:28:02] unicodesnowman: did you acquire your nick before or after VE used those snowmen? [03:28:17] visualeditor uses snowmen? [03:28:23] I thought it used pawns? [03:28:31] it started with snowmen [03:28:42] these days it's an entire zoo afaik [03:28:44] slippery slope! [03:28:45] gwicke, way before! [03:28:52] lol [03:29:15] unicodesnowman: ah, cool ;) [03:33:14] is it just me, or does the register feel like a tech tabloid? [03:33:25] it's not just you [03:34:35] I read Ars Technica personall.y [03:34:39] personally*. [03:34:45] "wealthy administrators at the Foundation devise schemes to spend the cash" [03:34:55] But even that I find gets on my nerves sometimes. [03:35:14] How many times can you reiterate how much you hate Comcast/Verizon/etc. etc. etc. [03:35:32] I get it, telecoms are E-E-E-E-E-V-I-L. Some real news, please... [03:37:28] Is slashdot still any good? Thought I might try it at some point. [03:37:48] I find it difficult to navigate, though. [03:38:27] reddit and hacker news have take over some of that audience [03:38:30] news.ycombinator.com and reddit.com/r/programming, highscalability.com, are decent [03:38:35] :) [03:39:27] Would a melting radioactive core would defy gravity and go straight through the Earth? [03:40:52] if you are looking for light entertainment, http://www.theonion.com/ can occasionally be fun [03:41:06] because the register apparently thinks fukushima is going to destroy the falklands [03:41:15] The Onion is great. [03:41:23] and the brits have a "secret evac cover-up" [03:41:28] not always equally, but often. [06:21:42] can anyone help me spot the error ? i am trying to do a query on database [06:21:51] i tried this (https://dpaste.de/RLXS) on phpmyadmin and it gets the result i want [06:22:08] then i tried to implement (https://dpaste.de/Xo3x) this to my php file, but it is giving me a blank page [06:22:15] anyonw knows where went wrong? [11:08:52] hey hey hey [11:09:46] is there anyone who could suggest a good theme for uh.. "static webpages" with mw? [11:10:12] we used to use GuMaxDD but I recently upgraded from 1.15 to latest, and it won't activate.. [11:13:49] mjau^, https://www.mediawiki.org/wiki/Category:All_skins [11:14:49] andre__: thanks mate, I've looked at some of those but.. I dunno [11:16:00] trouble finding something I think will work [12:34:24] i am trying to implement single sign on to my mediawiki and has made use of the extension remote_auth user [12:34:46] i am using apache (fastcgi/cgi) and the module (mod_sspi) for the single sign on [12:34:58] is the mod_sspi compatible with cgi fast cgi ?? [12:35:12] because it is working on my apache2handler but not working on cgi/fastcgi [12:35:16] am i missing out anything [12:38:05] is this a right place to ask? :) [12:38:13] urms....... [12:38:26] since related to mediawiki ? :p i hope someone knows though :/ [12:40:56] you have posted to serverfault, right? [12:41:09] just did but dun think anyone will reply though :/ [12:41:22] good question, it is not certain it works [12:41:36] oh [12:41:38] meaning ? [12:41:38] as far as I remember SSPI auth requires some additional request turnaround, correct? [12:41:51] i guess so [12:42:11] are there any other methods to configure sso for my mediawiki using windows credentials ? [12:42:12] :/ [12:42:15] (you can fix the formatting on your post - the config directives are glued) [12:42:36] yes, I think one of the ldap/ad extensions [12:42:45] will do fixed, gimmi a min [12:43:14] just fixed [12:43:18] for the ldap/ad extension [12:43:22] i did configure [12:43:28] i am able to login with the windows credentials [12:43:33] but it is not auto-logging me in [12:43:47] so i guess it is not working the way i want ? :/ [12:45:25] I don't know, I am not an expert on NTLM [12:46:17] ahhhhhhhhhhhhhhhh [12:46:23] been stucks for days , tired of t [12:46:27] thanks though :) [12:48:14] by the way [12:48:17] is there any way [12:48:29] to notify the admin that an article has been created [12:48:36] based on namespace [12:48:52] for example, Welcome: namespace? [12:50:12] !extension:email_notification [12:51:29] u mean this ? [12:51:30] https://www.mediawiki.org/wiki/Extension:Email_notification [12:51:57] should do the job [12:52:58] hmm [12:53:07] thanks saper :) [12:53:10] i hope it does.. [12:53:29] you can also use the RSS feed https://pl.wikipedia.org/w/index.php?title=Specjalna:Nowe_strony&feed=atom&hideredirs=1&limit=50&offset=&namespace=0&username=&tagfilter= and external database script http://cybergav.in/2012/03/31/new-page-notification-in-mediawiki/ [12:54:02] rss feeds?!? [12:54:07] how to achieve using rss feeds?!?1 [12:58:54] just use RSS reader to get notified [12:59:14] and there are ready tools to convert RSS to email, like feedburner or something [12:59:39] will look into it.. [12:59:42] thanks [13:49:52] everytime I upgrade, JS breaks for a few hours. Any ideas whay would cause this? [14:21:42] Is there a magic word that corresponds to the name of the current user viewing the page? [14:22:53] Not by default. It'd be cache-busting :) [14:22:57] Might be an extension [14:23:10] Fair enough. Thanks! [14:32:20] Nemo_bis: Why isn't there a way to disable the help links without using CSS? :P [14:38:42] hey guys, currently on my main page , i am displaying the recent articles (1. title1, 2.title2, 3.title3) [14:39:05] is it possible to remove (for example 1.title1) if it has been updated? [14:39:18] i am using the newestpage extension to display the recent articles? [14:39:29] or is there any other better methods? [14:54:36]