Whatcha think about using this brand new tool to get page access counts: [1]? (We should probably contact Henrik before writing a web scraper that might overload his website.)
Mindspillage showed it to me, I think it's effing awesome -- I don't trust article importance ranking done by individuals, now that we can assess page popularity maybe it's a more useful metric? At least we could add the information?
I was thinking of something like that, but I'm not sure how to go about it without being a parasite of [2]. I definitely want to include a function for article information like that at some point, I'm just not sure what, how, or when. – ClockworkSoul22:04, 12 March 2008 (UTC)[reply]
Quite the resource there! Igor is a standalone application, so we would have to have a server somewhere that can provide that much information in a computer-readable format. It's very nifty though: I'm definitely going to have to speak to Henrik! – ClockworkSoul22:25, 12 March 2008 (UTC)[reply]
About the file sizes ... cutting out the low count numbers makes for a much smaller file: from the only 10+ counts is 1/10th the size (198MB unzipped), 100+ is 67MB, and 1000+ is 11MB. Don't know if this helps... another problem is that if a page has moved you have to search under the old name to find the right pageviews. Madeleine✉✍20:31, 18 March 2008 (UTC)[reply]
I'm certain that I'll be able to do something with this data, but right now I'm not sure what. If I could find somebody with a server willing to provide me a little space and an open port I could whip up a little service that downloads this data automatically, puts it into a local database, and spits it out on request. It'll be some time before I can even really think about that, but it's definitely on the to-do list. – ClockworkSoul21:19, 18 March 2008 (UTC)[reply]
If I may propose a feature that allows resources (images, databases) and article sources sharing between project team members.--mrg3105 (comms) ♠♥♦♣ 00:54, 13 March 2008 (UTC)[reply]
An interesting idea. How do you envision this working, as the user is concerned? What do you see it providing that the Wiki itself cannot? – ClockworkSoul04:04, 13 March 2008 (UTC)[reply]
Often a single source will cover multiple articles. Physical sources are often not available to all the project editors. With Igor a source sharing is possible, however, one editor working on the source may not be aware of the needs of another editor. By having a project resource and source repository, the resources and sources can be readily shared. It would also provide greater consistency in source use, and assessment. Use of the sources can be scheduled for different editors depending on article priority, with sources assigned to article sections allowing section editors to work independently. I'm not sure how Igor will be implemented, but as a centralised point of project management, it seems it can do with the sort of tools real-world projects use.--mrg3105 (comms) ♠♥♦♣ 04:25, 13 March 2008 (UTC)[reply]
Interesting... If I'm understanding you correctly, then what you're proposing in no small task. At the moment, though, Igor is a standalone application that is downloaded and run on a client machine. It downloads all of its information from the Wikipedia servers on demand, more or less at the time of execution. However, I have a long term (i.e., post version 1.0) plan that may includes the use of a dedicated server that can provide a central location for certain non-Wiki resources that are not otherwise available. – ClockworkSoul04:52, 13 March 2008 (UTC)[reply]
Yes, I appreciate its not immediately deliverable, just a thought.
Version 0.1 should be ready in the next week or so, hopefully, but that's not carved in stone. I'm having some trouble getting page edits to work, but that's the only genuine technical issue right now. I've been getting a bit of help in that department from Tim Starling. Everything else is just typical grind until its done. – ClockworkSoul05:55, 13 March 2008 (UTC)[reply]
Just a quick update on the current state of Igor: I have all the technical issues related to uploads finally licked, and the article tagging feature is working perfectly. Note that in order to define the project tag templates (and other project-specific preferences later on), the project MUST add a small bit of XML to its project page that function as preferences and processing directives. Igor knows where each of the project pages live, and before re-tagging pages will grab the project preferences from there.
All I have left to do for the 0.1 release is finish up a few GUI behaviors related to the update process itself, and when those are done, probably within a couple of days, you'll have an shiny, new (and very much still a work in progress application to test)! There are a couple things I still need, though:
A splash-screen graphic: not vital, but it would be nice. If I can't get anything fancy in the next few days, which I'm pretty sure will be the case, I'll whip up something ugly as a temporary measure. Anybody know a graphic article artist who's part of Wikipedia?
A server to host the app. I'll find something soon, I hope. I'm a poor grad student, though, and I would hate to have to pay for something out of my pocket.
Graphic article, eh? Although I don't know her I've noticed LadyofHats makes really beautiful svg diagrams (look at her commons page), a number of which have been featured pictures. If she's not interested and you can't find someone better I could probably make something for you. Madeleine✉✍19:05, 15 March 2008 (UTC)[reply]
Doh. Silly tyops. I've seen a great deal of her stuff while working on some of the MCB articles. She does some truly excellent stuff! I'll send her a message and see if she might be interested. Thanks! – ClockworkSoul21:52, 15 March 2008 (UTC)[reply]
Version 0.1
Okay, version 0.1 is just about ready to go! In anticipation of the release of version 0.1 for testing and comment, I've created a SourceForgeaccount and put in a request for a new project. Once its up you'll be able to download an executable JAR, and for those of you who know how to read really messy and partially documented (at best) Java code, you'll be able to download my source as well. :) Igor is still in a very, very early phase, so it still has some odd behaviors and quirks here and there, but they'll all be smoothed out fairly soon. Stay posted here for more updates, coming soon! – ClockworkSoul02:15, 18 March 2008 (UTC)[reply]
First release available for download
Okay, folks, the first release of Igor (v0.1) is available for download! Please note that it is still in development, and lacks a number of features that will be included in upcoming releases leading up to the 1.0 release. The included documentation includes a short list of known issues, and there will certainly be at least a few more to come. Grab the ZIP file from here. Enjoy!
First of all, I think it's a really good idea, I'm sure it will encourage the rapid growth and improvement of existing projects, and spur the creation of new ones.
Is it Windows-only, or multi-platform? On a somewhat related note, would it make sense to incorporate it into any existing tools for use with WP?
Finally, a recommendation, in case you hadn't already thought of it: implementing the ability to browse existing WikiProjects, both for its own sake and in order to help someone classify a new project that one might be interested in creating.
Again, looks really good so far, and I'm in agreement re your proposed assessments.--Daniel11 (talk) 07:28, 18 March 2008 (UTC)
Igor is a multi-platform application. It's written in Java and will be distributed as an executable JAR file, so (in theory) it should run on just about anything with a new enough JVM. I'm not sure of any tools that we would want to incorporate into it, at least not offhand. Did you have anything in mind? As for browsing WikiProjects, I'm not sure what you mean, exactly, but at the moment you can easily scan and see (and modify) the articles tagged for any of about 1200 WikiProjects and workgroups. Take a look over at the Igor page. Let me know what you think. If you have any other suggestions, you can post them right on the talk page. :) – ClockworkSoul 02:15, 19 March 2008 (UTC)
That's good to hear that it's a Java program and will thus run on multiple platforms. This also negates some of my other questions. For instance, what I originally meant about incorporating it into other tools was to suggest that it might be easier to develop Igor as an addition to an already-existing tool (e.g. something like Wikipedia:AutoWikiBrowser or something from Wikipedia:Tools), but it might make a lot more sense to develop it as a standalone application, as it is now.
Regarding the WikiProject browser, I see that you currently have a drop-down menu to select which project is active. I can't see from any of the screenshots though how the projects are arranged in the drop-down menu: are they sorted hierarchically based on their appearance somewhere like Wikipedia:WikiProject Council/Directory? There's also a Category:WikiProjects, although that would probably be less useful for sorting the projects. What I had in mind was some easy way to navigate around the projects, and with easily accessible methods to jump directly from that navigation to creating a new project or managing an existing project. Not sure if that's any clearer, and at any rate it might be a bit excessive or otherwise unnecessary -- it's just a thought. I'd imagine it would be perfectly workable to use the existing drop-down menu to achieve the same functionality and simplicity.
One other thing that I thought of when I saw you'd added some processing directives to project pages... would it be possible, easier and/or better to grab that sort of information directly from some existing structured part of most wikiprojects, or does it make more sense to add those tags?
On the whole it looks like a great idea and well-executed. I'll have to try it out eventually and then I'll be able to give some better feedback. It's a great idea though just for the ability to do those massive assessments, let alone everything else.
Right now, the source I use for building the dropdown that is used to select the active project is constructed from the data built by WP 1.0 bot, specifically this list. This application really couldn't be built without that bot because it provides a nice standard list that is easily parsed. Most other aspects of wikiprojects aren't nearly so standard, though. The directives on the project page are a little awkward, but the best solution to this problem of the lack of standards for things like the names of project banners. This solution also has the added bonus of providing a very flexible vehicle for allowing each project to tailor the behavior of Igor with regards to itself.
I like your idea for navigating projects, though. With 1200+ projects, I've been finding the dropdown to be a little awkward to use, and if I can find a nice way to implement your idea, it would make Igor a heckuvalot easier to use! – ClockworkSoul03:54, 19 March 2008 (UTC)[reply]
I've just downloaded Igor and it's working great. Some immediate feedback:
Does it actually require any 1.6+ features? The class file versions prevent a 1.5 interpreter from even trying.
Not sure if this is just an artifact in my setup, but there's an issue with the list redrawing, so that if I scroll through any articles, every third one gets covered up by another entry and is illegible. Some refreshing does seem to occur, and I can cause it to redraw by clicking on an entry. Need any more details?
Actually, this isn't consistent: sometimes it's every third line, sometimes it's a whole lot more than that, and sometimes it's not at all.
When you call the browser, you seem to call specifically "firefox.exe" -- this doesn't work on non-Windows platforms where the file is located elsewhere, and I think some people still don't have Firefox installed. :) Can it call the OS's generic "open browser" instead?
Group-select works, but I don't see a way to modify a group selection all at once.
How about: a toolbar (including a menu with actions to apply to the selection, like Set priority->High); some buttons (including various actions to apply).
It seems I can't resize the columns.
Actually, it turns out I can resize only the Article column, and only in the direction of the other columns (thus expanding Article and contracting the other columns). What I'd wanted to do was to expand the other columns.
The automatically updating search is great!
This is somewhat trivial, but maybe the Refresh and Update buttons should be in different places.
Also small, but there could be a "clear" button next to the search field.
I actually don't find the dropdown too unwieldy, now that I've used it, but any improvements to project selection would still be terrific.
Egad, 666 articles for project Beer -- quick, somebody add or delete something!
The mad scientist image is great, and the application as a whole is very nice, it looks good and is quick, effective, etc.
You can't distribute GPL software without making the source available! ;)
<outdent>Thanks for the feedback, Daniel! The source is available on request, until the SourceForge account is up. I'm trying to take over an abandoned project name, so it'll take a couple weeks. I'll the rest of these point by point...
Does it actually require any 1.6+ features?
The 1.6 interpreter is required, I'm afraid: Igor takes advantage of some of the recent changes in the Swing implementation. The splash screen is one of them, and ther are a few other minor ones sprinkled here and there.
'Not sure if this is just an artifact in my setup, but there's an issue with the list redrawing
Redrawing is taken care of entirely by Swing, but I'll see if I can do anything about it.
When you call the browser, you seem to call specifically "firefox.exe"
Java 1.6 actually provided a means of calling the system browser. I meant to implement that, but hadn't gotten around to it... this'll be among some bug fixes I release in a between-release update.
Group-select works, but I don't see a way to modify a group selection all at once.
Yes, group selection is enabled, but I wanted to push out a working version before group-update. :) 0.2 will let you right click on a selected block to update the entire block all at once.
How about: a toolbar?
0.2 will include a toolbar, absolutely. :) I'm still putting together a layout for it, so if you have any ideas, I would love to hear them!
It seems I can't resize the columns.
No, I locked it because the table would resize the columns uneconomically when the window itself is resized. I can easily undo that, though.
The automatically updating search is great!
Yeah, that's one of my favorite bits. :) Glad you like it! 0.2 will actually see some enhancements to that, too.
This is somewhat trivial, but maybe the Refresh and Update buttons should be in different places. Also small, but there could be a "clear" button next to the search field.
Oh, definitely. They were put there as a temporary measure because I'm still designing some bits of the interface. They'll be better in 0.2.
I actually don't find the dropdown too unwieldy, now that I've used it, but any improvements to project selection would still be terrific.
I'm going to add your project browser idea for 0.2 (maybe 0.3). The new setup will include a tabbed pane with the browser on the first pane. The drop-down (or something like it) won't go away, though, because lots of projects aren't in the directory, so they won't be on the tree. I'm going to include a dynamic project search also, a lot like the one in the article manager.
Egad, 666 articles for project Beer -- quick, somebody add or delete something!
Mmm... beer.
The mad scientist image is great, and the application as a whole is very nice, it looks good and is quick, effective, etc.
Can you tell I wrote it on my free time in less than two weeks? A whole week of that was getting the darn thing to post updates to Wikipedia!
No need to worry about the source until the SF name clears, I was just teasing. ;)
I'll try to condense my responses here, so accept my apologies if I leave anything out:
No worries about requiring a 1.6 Java environment -- I'm sure the vast majority of people probably have one anyway. The default Java environment that shipped with most Linux distributions until recently was 1.5, because it was the free GCJ which only went up to 1.5, but Sun recently opened their JDK so I think they should be moving to 1.6.
It sounds like there'll be a huge number of improvements by v. 0.2, so I look forward to trying that one out (and maybe looking through the source once the SF site is up). "Yes, group selection is enabled, but I wanted to push out a working version before group-update." Ah, this is probably a wise approach! :)
My ideas so far for a toolbar are pretty much the standard fare that one would find in a toolbar, so I guess I'll hold off on providing any ideas there until I come up with something worth mentioning -- I'm sure you don't need any help in coming up with the idea of adding a "Help|About" menu item!
I don't know that the columns really have any need to be resizable, it's just that the current default ones aren't particularly suitable for me anyway. The end columns are all too narrow in my setup, limiting the visibility of the items in question (e.g. the article quality levels).
The tabbed design sounds great too, can't wait to see that in action.
While it might still be in early development (and done in spare time etc.), I like how it's clean and intuitive -- it's actually quite impressive.
Hey, this is a nice idea for an app. I'm interested if you plan to make it work for other wikipedias. How does the app loads the list of Wikiprojects? GerakiTL09:13, 23 March 2008 (UTC)[reply]
Unfortunately no, I don't. Part of the problem is that this application relies in large part on the output generated by WP 1.0 bot, which (as far as I know) doesn't exist on other wikis. There's simply no standard implementation of wikiprojects aside from the efforts made by the Wikiproject Council and WP 1.0. – ClockworkSoul14:37, 23 March 2008 (UTC)[reply]
Just a heads up: I just posted a new zip file with version 0.1.1. This version doesn't add any new features, but it does fix a few bugs. Enjoy! – ClockworkSoul02:05, 24 March 2008 (UTC)[reply]
I tried Igor. It works great. I had some login issues, but was able to login after a few tries. Does Igor support taskforces? For example, WP:CHENNAI is a taskforce of WP:INDIA. Is there a special processing directive for taskforces? Chennai project tagging works by adding "chennai=yes" parameter to the India template and "chennai-importance" for taskforce specific importance value. Please see Talk:Chennai. Thanks, Ganeshk(talk)02:52, 24 March 2008 (UTC)[reply]
Hi, Ganeshk. Thanks for trying Igor. Igor doesn't have support for taskforces yet, primarily because there's no standard for taskforce implementation, so each project implements them a little bit differently. I'm working on some ideas, though, and I hope to have such support very soon. – ClockworkSoul04:46, 24 March 2008 (UTC)[reply]
Igor uses the Java Wiki Bot Framework to handle logging in (specifically, via the MediaWikiBot.login method). Looking at the JWBF's source code, it seems that the framework itself performs a login as if it was a web browser by sending an HTTP POST to index.php and reads the cookies in the response. The problem with that, however, is that after one failed login attempt, Wikipedia tries to make its user read one of those fuzzytext images to ensure that the failure wasn't the result of some bot trying to hack the account, and keeps asking for about five minutes from the last failed request. Good for security, bad for us, because at the moment our users are simply locked out for five minutes after every login attempt. Short term, I'll put in a message saying that login is temporarily disabled after a failed login attempt. Less short term, I'll see if I can hack the JWFB to scrape the image from the login page so that I can present it to the user after a failed login. My advice for now is to try and log in correctly the first time until I can get a fix in there. – ClockworkSoul00:47, 25 March 2008 (UTC)[reply]
Actually if Connect is clicked, it logs you in with no name. However I will try to log in again correctly. I may have got the password wrong because my Wiki passward is automated so I hardly ever need to retype it.--mrg3105 (comms) ♠♥♦♣ 10:52, 25 March 2008 (UTC)[reply]
Okay, I managed to fix it. The application will now scrape the fuzzytext image from the wikimedia response and show it in the "password failed, try again" dialog. It's actually kind of neat. When I got it to work, I giggled a little. Also, empty logins fail now correctly now. :) – ClockworkSoul04:47, 31 March 2008 (UTC)[reply]
First of all, great idea for a tool; this has a lot of potential.
You really should re-think the splash screen. Splash screens are a pointless relic from the '90s that have no benefit today other than to be obnoxious. Do a Google search for "get rid of splash screen", "splash screens suck", "fuck splash screens", etc. Really, what is the benefit other than to get in the user's face? They know they opened the program, they don't need a giant graphic to inform them. —Werson (talk) 01:06, 31 March 2008 (UTC)[reply]
Hi, Werson. Thanks for the advice. I can't agree more that the current splash image is kind of overkill. The giant graphic that's in there now is very, very temporary... it's already been replaced with something much smaller. I can't agree that users always know that they've started an application. Without going into unflattering details, I'll just say that it's supported by my experience working with end users in my old development gig. Also, later versions of this app will have a slightly longer startup as it downloads various bits, on top of the initial firing up of the JVM that accompanies Java applications. – ClockworkSoul05:00, 31 March 2008 (UTC)[reply]
Yes, I've seen this before. The way Igor gets the class and importances of articles is by scraping the category pages, which it's been forced to do temporarily since a recent MediaWiki change made grabbing categories via the JWBF non-functional. It seems that on rare occasions articles will appear absent from categories they should be in or in categories that they should not. This seems to be a glitch in how WM places articles into categories in general, and it usually disappears pretty quickly. – ClockworkSoul00:52, 10 April 2008 (UTC)[reply]
Igor's SourceForge page is finally up, with the most recent build. There's not much to see from the user perspective, but there are a few new quality of life features, including the ability update multiple articles and to recover after a failed login attempt. If you wish, you can also finally grab the source there as well. As usual, let me know if you have or notice any problems, or even if you have suggestions. – ClockworkSoul00:41, 10 April 2008 (UTC)[reply]
Just discovered this and it is a great little tool. However, I've got a query, sorry to be a pain, but what does the article info option do? It doesn't seem to do anything my end. HidingT11:39, 22 April 2008 (UTC)[reply]
Ah, another issue, it won;t let me update, it says it can't find the project preferences or an <igor> start tag. What am I messing up? HidingT11:47, 22 April 2008 (UTC)[reply]
No pain at all. The article info in the current version in non-functional (I'm actually working on that right now). I intended to comment it out before I uploaded the current version, but I missed it. For the preferences, take a look at the README document in the zip file. It'll tell you how to define your project preferences in an <igor> tag nested in the project page. Which project are you using this for? – ClockworkSoul13:26, 22 April 2008 (UTC)[reply]
Please, please let me know when this feature is finished. I know of several projects I'm intending to set it up for, but am not at all good at setting up Igor myself. John Carter (talk) 15:39, 22 April 2008 (UTC)[reply]
I'll post all of my changes here, and if you like you can also sign up for the Igor mailing list here. The full-featured project page creation wizard may be a while, but I was thinking of making a mini-version that would create the <igor> tags and add them to the project page. In the meanwhile, if you point me to the projects, I'll be glad to help in setting up your preferences. – ClockworkSoul16:09, 22 April 2008 (UTC)[reply]
Ah, excellent. Igor doesn't yet know how to handle task forces, but I'm going to be adding such support in this upcoming release. That turned out to be quite a bit simpler than I thought it would be, but it'll require a bit more effort in term of setting up your <igor> tags. – ClockworkSoul16:25, 22 April 2008 (UTC)[reply]
I've been reading about this tool to see how it could help us at WP:TWP. One potential problem that I can see is that WP:WikiProject Rapid transit, which is a subproject of TWP, uses ((WPRT)) as it's project banner, which itself transcludes (rather than redirects) ((TrainsWikiProject)) with a couple of parameters preset. WPRT also changes the effect of the importance parameter to relate to rapid transit's importance and adds rail-importance to set the general TWP importance value. I've been indifferent on changing WPRT calls to TrainsWikiProject calls, but if we use this tool in its current release, it appears that we may have to.
On the other hand, would it be possible to add to the <project-template> directive to add something like importance="FOO" to specify the correct value for the given articles? Doing so, the rapid transit project could continue using the same banner scheme (and presumably also use Igor themselves) and we would have something like the following on the TWP project page:
An even more advanced change would be for Igor to recognize when WPRT is called with nonrail=yes (which produces a completely different banner appearance) and not include pages with this usage in the TWP list. The WPRT template is designed this way because the majority of pages within the subproject's scope are also within TWP's scope, but TWP does not include bus systems while WPRT does.
Hi, Slambo. The issue that you're pointing out is actually very common, especially among taskforces. For the issue of templates that use specific fixed parameters to identify task forces using the syntax "nonrail=yes", I've come up with what I think is a very straightforward and flexible solution that would allow a project to define an arbitrary number of these fixed parameters for any given tag. The format isn't 100% final yet, but this how it will likely be implemented, using nonrail=yes as an example:
This would produce a template like this: ((WPRT|nonrail=yes|foo)). This feature is among the additions included for task force support that'll be in the next release coming in the next couple of weeks. The first issue, however (importance="rail-importance") requires a restructuring of the importance field, which at the moment is hard-coded to use only the most common WP1.0 values ("top", "high", etc.). I've been flirting with a way to make these values customizable, but that feature won't be available for the upcoming release, as my current priorities are the WikiProject Browser (a huge task) and support for taskforces/subprojects. I also hope to squeeze the category browser in as well, as that mostly uses existing code. I hope that helps! – ClockworkSoul13:59, 23 April 2008 (UTC)[reply]
Is it possible for Igor to create some sort of charts (like this crappy pie chart) to provide an overview of a WikiProject's activity at a glance? It is really difficult to provide a graphical representation of a WikiProject's progress at the task force level from general WikiProject data. Igor seems like it will have category intersection built-in, so it would be a great tool for these statistical analyses. Titoxd(?!? - cool stuff)10:54, 29 April 2008 (UTC)[reply]
I tried running the tool on my MacBook (Running Java SE 6 version 1.6.0_05 on OS X 1.5.2), and I get the following error:
5/2/08 2:44:19 AM [0x0-0x271271].com.apple.JarLauncher[11279] Exception in thread "main" java.lang.UnsupportedClassVersionError: Bad version number in .class file
Odd. That's the error that's generated when the version used to compile the class files is newer than the version used to run it. I compiled with 1.6.0_05, so I'm guessing that the Mac implementation still doesn't fully support the newer bytecode. I'll be posting the newest version in the next few days. I'll try building it with an earlier JRE, and hopefully that'll do the trick. – ClockworkSoul02:03, 3 May 2008 (UTC)[reply]
The version check is built directly into the compiled class files, but it will compile just fine under an earlier version. If you do that, however, some of the GUI features won't be 100% functional and you won't get the splash screen, but it'll probably be perfectly usable otherwise. I haven't bothered setting up the CVS yet, but you can download a zip file of the source files by going to the download page and clicking "Browse all files for this project". Scroll down a bit to "file releases" and download "Igor-v0.2.0-src.zip". The easiest way to compile and execute the files would be to download and use a good IDE (Eclipse IDE for Java Developers would be an excellent choice). Let me know if you have any trouble with it, and I'll help you get it set up. – ClockworkSoul11:10, 21 May 2008 (UTC)[reply]
Actually, I figured out what the problem was, and it is on Apple's end. It's just that Java 6 is not included by default on Leopard, and to enable it, one has to download an additional JVM, and then modify the order in which JVMs are executed. The latter task was not documented anywhere... Titoxd(?!? - cool stuff)21:18, 21 May 2008 (UTC)[reply]
I posted instructions as to how to get around that in the bug report, but in essence, Mac users need to change their settings on /Applications/Utilities/Java/Java Preferences. This could probably be added to an INSTALL file somewhere, perhaps. Titoxd(?!? - cool stuff)03:26, 22 May 2008 (UTC)[reply]
I think it has the potential to be an awesome tool. I have one suggestion/request: could it create lists of wikiproject articles that are tagged with various maintenance tags? Like no references, expand, POV, OR, various disputes, clean up, etc etc? So for example, it would have: article abcd - E-class - Mid-importance - tagged as needed inline references since May 2007. I think that would be insanely helpful to focus the efforts because just the class does not alway tell the full story. Is it doable? Renata (talk) 22:39, 27 May 2008 (UTC)[reply]
Thanks! That would likely require all of the articles in a project to be individually downloaded and parsed, but I'll see if I can find a graceful way to do it. I'm sure I can work something out... – ClockworkSoul01:12, 28 May 2008 (UTC)[reply]
I tried to use Igor for a tag and assess drive. But Igor lacked the feature to pull links from a specific page. It was trying to pull the 30,000 articles that were part of the India project. I would like it to tag unassessed articles with the banner template. Can this kind of feature be added? Regards, Ganeshk(talk)14:55, 7 June 2008 (UTC)[reply]
Right now, Igor will near seize up if you try to edit the tags on article titles containing colons. Only a re-sort using the columns will allow you to continue and when you have a whole bunch of articles with colons within the titles it can cause issues! Something to support for the future, keep up the good work CS. treelotalk02:40, 8 June 2008 (UTC)[reply]
Ah, yes. I just checked your history and guessed that, too :) I see what the problem is: the stupid default sorter it's using right now is buggy for tables the change in size like Igor's does. I'm not sure why it affects your project worse then others, but I'll bump it's priority up. – ClockworkSoul03:38, 8 June 2008 (UTC)[reply]
Is there anything to help with the removal and/or addition of articles and other types of pages from/to projects? Noticed that "Drop from project" currently is inactive, hopefully to be implemented soon but addition would be very helpful. In connection could be using external lists like plain text files or aggregation from categories, watchlists or what links here much like AWB does when creating worklists. Just suggestions to pass your way. treelotalk03:48, 8 June 2008 (UTC)[reply]
Definitely. The development for the next release is focusing on these features, including tagging articles from categories, pages, and your watchlist. If you have any other ideas, feel free to let me know! – ClockworkSoul04:00, 8 June 2008 (UTC)[reply]
Hi, it looks like C-Class will be going ahead - can you add the relevant code to Igor at the appropriate time? See the conclusion of the poll. Also, I hope to promote the use of Igor in the rewrite of the scheme. Thanks, Walkerma (talk) 05:23, 19 June 2008 (UTC)[reply]
I just came accross a problem when the text of a template appeared twice on the same talkpage. Take a look at this diff. Would it be possible to get the program to ignore templates that are within nowiki tags? Jim MillerSee me | Touch me21:45, 3 August 2008 (UTC)[reply]
Hi, hope that progress with Igor hasn't fallen by the wayside recently. Anyway, currently I'm having issues with retrieving the article table for the WP I work on. All goes well at the article manager, loads them up according the progress bar but hands at 99% with the output telling me it got 0 articles. Tried the usual things such as clearing the cache and even starting from scratch with a fresh download of the .jars but no dice. Anything you reckon that could be done? treeloradda21:36, 5 September 2008 (UTC)[reply]
Hi, Trello. Igor has been on the back burner while I deal with an especially intense project here in what I derisively call the "real world". Nothing like 9am to 1am work days to squash everything else in your life! Which project are you trying to load articles for? – ClockworkSoul16:08, 10 September 2008 (UTC)[reply]
Careful with that real world, you could have an eye out with it. I'm trying to load articles for the Cartoon Network project but it's not specific to that project, I tried various different projects, even WP:MCB but get a nice response of "Got 0 articles for project X" no matter which I try to load. treeloradda16:58, 10 September 2008 (UTC)[reply]
Still no good but I can give you further things with which to diagnose the issue with. It's an odd problem as it does grab the list but fails to parse it and shrugs its shoulders at the issue. Clearly it's local only to seeing as nobody else has had issues with it but it's still odd for it to cease working for no reason. treeloradda20:49, 30 September 2008 (UTC)[reply]
Same problem here, for WikiProject Ohio and Ohio State Highways. Also, WikiProject Ohio Townships is null now, so it probably shouldn't be listed. §hep • ¡Talk to me!22:09, 9 November 2008 (UTC)[reply]
I opened Igor on Windows Vista and Firefox internet browser (not sure if they are the problems) and got this after being shown the Igor mad scientist logo thing:
Failed to read text configuration files:
Reason: XML resource directory does not exist: resources/XML/text
When I looked at this I was intrigued, although for some reason the project I'm interested in most (WikiProject_Java) is not listed in Igor. So, same question as all the other guys up there: what happens, you're still editing like crazy in Wikipedia, so YOU're not dead, but is Igor dead?...
-- A l a i n R 3 4 5 Techno-Wiki-Geek03:15, 24 October 2009 (UTC)[reply]