User:CBM/NFCC discussion 2007-9-17
< User:CBM
* --- Log opened Mon Sep 17 18:58:09 2007
* 18:58 carl-cbm (n=cbm@wikipedia/cbm) has joined #BetacommandBot
* 18:58 «Users» | 3 nicks (2 ops, 0 halfops, 0 voices, 1 normal
* 18:58 «Irssi» | Join to #BetacommandBot was synced in 1 seconds
* 18:58 carl-cbm . seems like I'm early to the party
* 19:03 @ Betacommand . yeah
* 19:03 @ Betacommand . Ive been here for over 12 hours
* 19:04 @ Betacommand . With the logs I will edit them for security reasons
* 19:04 @ Betacommand . IE remove hosts
* 19:05 carl-cbm . That seems reasonable, just the nicks are enough. It might be worthwhile to ask people to identify their enwiki name if they want to
* 20:01 wikidemo () has joined #BetacommandBot
* 20:03 wikidemo . Hello. Are people here?
* 20:03 @ Betacommand . hat
* 20:03 carl-cbm . Yes, hello
* 20:03 @ Betacommand . hay
* 20:04 @ Betacommand . wikidemo: lets get your pov on NFCC
* 20:04 wikidemo . Yay. Greetings....Oh, I see the participant list now in the sidebar. I brought a nice Merlot but I don't think I can upload liquids.
* 20:04 @ Betacommand . you could DCC it...
* 20:04 wikidemo . POV, meaning where I'm coming from?
* 20:04 @ Betacommand . yeah
* 20:05 @ Betacommand . your current thoughts
* 20:05 wikidemo . My main interest is to make sure we have an orderly transition and don't end up with a lot of holes in the article-space where images used to be. I usually work on articles about companies, food products, etc.
* 20:06 wikidemo . I often run into articles with missing images and it's a chore to find, re-upload, etc. I know those images were fine, if only the use rationale, sourcing, etc., were correct. It seems there's got to be a way to automate more smoothly.
* 20:06 @ Betacommand . I have several ideas, that need to be heavly restricted
* 20:06 @ Betacommand . I agree
* 20:06 @ Betacommand . My thoughts are this,
* 20:07 wikidemo . Go ahead.... (you might have noticed I floated a very preliminary proposal earlier)
* 20:07 @ Betacommand . we create 3 specific templates
* 20:07 carl-cbm . I am looking at WP:LIP right now
* 20:07 carl-cbm . in front of me
* 20:07 @ Betacommand . and they are the only three that will be valid
* 20:08 @ Betacommand . One rationale for album covers on said album article
* 20:08 wikidemo . (listening to both of you....aside to Carl: sorry it's so sketchy, more like the outlines of a proposal than an actual proposal)
* 20:08 @ Betacommand . that rationale is VERY specfic
* 20:08 @ Betacommand . you cant use it for anything else
* 20:09 @ Betacommand . any other use of the image would need a specific hand written reason
* 20:09 @ Betacommand . same thing for logos and the main company page only
* 20:09 @ Betacommand . and for book covers and book pages
* 20:10 @ Betacommand . do we have agreement there?
* 20:10 carl-cbm . Most people on WT:NONFREE agree to that
* 20:10 wikidemo . On those specific uses, yes. A little tweaking perhaps, but I like the idea of having specific allowed rationales for specific cases.
* 20:11 @ Betacommand . and only those uses
* 20:11 carl-cbm . Betacommand: have you looked at WP:LIP ?
* 20:11 @ Betacommand . carl-cbm: glanced
* 20:11 carl-cbm . The issue is how to schedule getting older images into compliance
* 20:11 wikidemo . well, as an example of a tweak a logo about a company could apply to the company. A logo about a product could apply to the product article. A logo that is a highway sign could apply to an article about that highway, etc.
* 20:12 wikidemo . But I think that's too fine-level implementation to need talking about right now (the "tweaks")
* 20:12 carl-cbm . it should be possible to write a description that captures those uses
* 20:13 @ Betacommand . wikidemo: we would need to HEAVILY control those as we dont want the use of template rationales to become trash
* 20:13 @ Betacommand . http://tools.wikimedia.de/~betacommand/old_images.txt
* 20:14 wikidemo . And as per WP:LIP I'm going to suggest we leave the door open and have a process for uses other than those three. Even those three will have to go through an approval process as to the exact template, wording, rules. I don't want to cloud the discussion now by considering more than those three but I can see some others people could propose in the future. Some might be more controversial,...
* 20:14 wikidemo . ...some might not fly, but others could work out. EX: a software title cover works much the same as an album cover, and that would be a separate one.
* 20:14 @ Betacommand . is a list of images that I have ID'ed as having problems
* 20:14 carl-cbm . that's not too many
* 20:14 carl-cbm . wikidemo: one issue with the recommendations of WP:LIP is that it requires a lot of editing of compliant articles
* 20:14 wikidemo . well, they would be templates. they should be protected as high-risk templates to avoid vandalism, well-meaning but bad changes, etc.
* 20:14 carl-cbm . as in, every image description page has to be formatted
* 20:15 @ Betacommand . and EVERY template would need an article field
* 20:15 @ Betacommand . so that we know what page the rationale is for
* 20:15 wikidemo . Carl, that issue just came up today when someone tried to edit the master fair use rationale template. They added an "article=" parameter and give a red warning message when people forgot to mention the article, a good idea, but it made all the old transclusions obsolete.
* 20:15 carl-cbm . that would go a long way towards machine readability. But the source template recommended by WP:LIP doesn't need an article field, just the rationale ones
* 20:16 carl-cbm . wikidemo: and some people put multiple articles in there
* 20:16 wikidemo . It's an open issue whether to allow multiple articles under a single rationale. I see both sides but I think the answer is no because that's hard to manage.
* 20:17 @ Betacommand . wikidemo: exactly
* 20:17 carl-cbm . I had a good example on the wiki of an article where the rationale was completely generic
* 20:17 @ Betacommand . the template rationales are one per page
* 20:17 wikidemo . A template would have to have parameters like article1=xxx, article2=xxxx, article3=xxxxx, and that's stuffing a one-to-many relationship into what should be a relational style database
* 20:17 wikidemo . plus the problems people mention that editors get sloppy when they are allowed to do that
* 20:18 carl-cbm . one possibility would be to move the generic parts - small size, mostly, into another "checklist" template
* 20:18 @ Betacommand . wikidemo: we should allow only one artilce per template
* 20:18 carl-cbm . so that the rationale templates can actually talk about the rationale in the sense of the foundation
* 20:18 wikidemo . okay, i'm agreed. that might need to be sold to get buy-in by people who want to do it the easy way. The three initial templates we mentioned aren't susceptible to being used on more than 1 article though. They're about using the image on the main article.
* 20:19 carl-cbm . well, this is hardly the easy way. it means lots of work to standardize image description pages
* 20:19 carl-cbm . Betacommand: can you estimate how many uses might fall into the big 3 standard rationales?
* 20:20 @ Betacommand . carl-cbm: http://tools.wikimedia.de/~betacommand/NFCC_templates.txt
* 20:20 @ Betacommand . that is about 3 days old
* 20:21 carl-cbm . it looks like a large number are "standard" but a lot aren't. the "promotional" and the "non-free fair use in" category will take a lot of work
* 20:21 wikidemo . so to summarize the info from there, 73,742 non-free album covers, 16,941 book covers, and 70,366 non-free logos. That looks to be about 40% of all non-free images. That might be overlooking some alternate copyright tags used for these three main cases.
* 20:21 @ Betacommand . dont forget that if we move to standard templates for the standard uses I can write a bot for that
* 20:22 carl-cbm . wikidemo: I was including game covers and dvd covers in my mental count
* 20:22 wikidemo | Television screenshots (29,000) and film screenshots (13,679) are going to be tough. I'll reserve judgment on whether we can help much with those or not if someone proposes a good template. Certainly not a one-size-fits-all template.
* 20:22 @ Betacommand . we have ~325680 non-free images
* 20:23 carl-cbm . that seems like a lot of images to convert to a new metadata system
* 20:23 @ Betacommand . wikidemo: templates for those are a bad idea
* 20:23 wikidemo . We'll probably never get any headway with promotional images (16,563)....we won't get them all but we can make the problem smaller.
* 20:23 @ Betacommand . we already have massive abuse of TV images
* 20:24 carl-cbm . let's leave the question of abuse to another day. the question here is how to tag them correctly
* 20:24 carl-cbm . if they do belong
* 20:24 @ Betacommand . TV images should have a good ratiional or be deleted
* 20:24 wikidemo . Sure....I dont want to actually propose anything beyond the three main uses right now. But hypothetically, someone could come up with some subset within a category that can't be completely solved.
* 20:24 carl-cbm . well, that's true of all nonfree images
* 20:25 @ Betacommand . as for NFCC 10c that cannot be removed. We agree on that?
* 20:25 carl-cbm . but... 300k images and 6 months is 50k images per month, about 16000 a day
* 20:25 wikidemo . Just to throw out an example, "non-free poster" is very generic and we don't want to allow posters of models, cars, etc., in a single template. But we *may* decide at some point that posters of movies used to illustrate the article about the movie is an acceptable template.
* 20:26 @ Betacommand . we also need to create a very strict method of approving such templates
* 20:26 wikidemo . I agree that we should keep 10(c), and even if we go to a simpler or automated rationale system we need something recorded for every single use. Not everyone agrees, and not everyone knows the issue. I would guess there's a consensus-like majority of concerned editors who agree but some people who don't.
* 20:27 @ Betacommand . if we remove 10c we might as well give up on the NFC control
* 20:27 wikidemo . I agree that there should be a strict list of approved templates, and if it isn't approved it doesn't fly. We definitely need a process but I'm not sure what the process is. I guess some kind of proposal system, discussion, and acceptance. But the last thing we need is an edit war betwen peopel who say a template has consensus and others who say no.
* 20:27 carl-cbm . wikidemo: agreed
* 20:28 carl-cbm . Betacommand: well, we had a free-for-all for a long time
* 20:28 @ Betacommand . wikidemo: agreeed
* 20:28 @ Betacommand . carl-cbm: that is the fuck mess that Im attempting to clean up
* 20:28 wikidemo . Also, templates should be approved or rejected based on how well they fit 10(c), the foundation resolution, other policies, our image process, etc. The template approval process shouldn't be a place to rehash debates over what kind of images are okay and which ones are not.
* 20:28 carl-cbm . is that avoidable?
* 20:29 @ Betacommand . yes
* 20:29 carl-cbm . I mean, to approve a generic rationale is the same as accepting it as valid
* 20:29 @ Betacommand . as long as there is 100% need for said images for said use
* 20:29 carl-cbm . well, yes, we can make these the opposite of CSD - a use so valid it is unlikely to be disputed
* 20:30 carl-cbm . that does reduce the fuss at CSD
* 20:30 @ Betacommand . the templates will need to be bulletproof uses
* 20:30 carl-cbm . sure
* 20:30 carl-cbm . wikidemo, I am still concerned about implementation time for WP:LIP
* 20:31 wikidemo . I'm not sure if it's avoidable. There have to be some ground rules with templates, and this whole process. If we start with the premise that 10(c) stands, that some templates are okay, etc., and that any argument for or against a template based on arguments outside of that is a non-starter and can be discounted, maybe we can keep it in line. Also, the rule that if it's not approved the...
* 20:31 wikidemo . ...template can't be used will render many of these off-topic arguments moot. The only issue would be obstructionists who simply don't like templates so argue that they're all non-consensus.
* 20:31 carl-cbm . No, I think there has been a lot of support for them on WT:NONFREE
* 20:32 wikidemo . WP:LIP is a proposal in several parts, probably some better ideas than others.
* 20:32 carl-cbm . well, I agree with large parts of it
* 20:32 wikidemo . For instance, the idea about putting "source" in a template once per image so that bots/machines can grab it, that's just a pet issue of mine. But maybe a useful issue too.
* 20:33 @ Betacommand . wikidemo: I agree
* 20:33 carl-cbm . I think it's a good idea, but I don't think we can do it by March
* 20:33 carl-cbm . not 16,000 images a day
* 20:33 @ Betacommand . or maybe have a source template
* 20:33 wikidemo . The idea of creating a wrap-around template that I called a "use statement" (as opposed to use rationale) with basically two fields - article=xxx and rationale=yyy, is just a kludge to address all the existing rationales.
* 20:34 carl-cbm . Ah, I see. That would be very easy to set up. And really, copying the same rationale three times isn't hard since we're using computers to edit with
* 20:34 wikidemo . 16,000 issues per day? Isn't that 1,600? There's no rush at all with respect to legacy images that are compliant and just need reformatting. We can move those to a new format for consistency or just leave them alone.
* 20:34 carl-cbm . let me check my math...
* 20:35 @ Betacommand . at last count we had less than 170k images with problems
* 20:35 carl-cbm . Betacommand: but we're talking about templates that have to be added to all images
* 20:35 carl-cbm . problematic or not
* 20:35 carl-cbm . wikidemo: yes, 1600 images a day
* 20:35 @ Betacommand . carl-cbm: I can bot the album covers
* 20:36 carl-cbm . Betacommand: not the source template, I bet
* 20:36 @ Betacommand . correct
* 20:36 wikidemo . In a perfect world we would have {{article=xxx } rationale=yyy}}. xxx="unknown" and yyy="not added" would stick out like a sore thumb and those could just about be deleted on sight.
* 20:36 carl-cbm . wikidemo: so you are suggesting that imags that don't meet #10c have to be formatted to the new format right away, and older ones on a delayed schedule?
* 20:37 @ Betacommand . what we could do is create a master license /source template
* 20:37 wikidemo . Many of the sources are simply not there. Nobody bothered to add them. But these are "self-sourcing" in the sense that for all or nearly all, the source is the record company.
* 20:37 carl-cbm . (we'll ignore the question of whether that meets the NFCC requirement)
* 20:37 @ Betacommand . IE we could have {{NFC License}}
* 20:38 @ Betacommand . that has a field for the class
* 20:38 wikidemo . From a copyright sense, a photo, scan, copy, reformat, etc., of an album cover is a "slavish reproduction" that introduces no new copyright. So for copyright purposes it is sufficient to say the source is the copyright holder. For good housekeeping purposes we still want people to say where they got it and how. That gives us a double check. But for legacy images, I don't think that info...
* 20:38 wikidemo . ...is there.
* 20:38 @ Betacommand . book covers
* 20:38 @ Betacommand . albumcovers
* 20:38 @ Betacommand . logos
* 20:38 @ Betacommand . and then a source class
* 20:38 carl-cbm . wikidemo: yes, I'm sure it often is not there, because there was little double-checking
* 20:39 wikidemo . Okay, I think I'm interrupting Betacommand. (listening)
* 20:39 @ Betacommand . that would merge all the license templates into one
* 20:39 carl-cbm . hmm.. it's easier for people if they have different names
* 20:39 @ Betacommand . (the main would call the sub templates)
* 20:39 carl-cbm . I think parameters are less user friendly
* 20:39 @ Betacommand . Class=bookcover
* 20:40 @ Betacommand . Class=logo
* 20:40 carl-cbm . Betacommand: that also is a maintenance headache, since it will be a protected template
* 20:40 wikidemo . Perhaps using separate templates versus mashing choice of template into a parameter field is just an implementation issue.
* 20:40 @ Betacommand . carl-cbm: we can do that
* 20:40 carl-cbm . Betacommand: I cringe at the infobox classess because they are such a pain to maintain
* 20:41 @ Betacommand | that would help with future images
* 20:41 carl-cbm . OK, so to move forward with a proposal onwiki, what do we need?
* 20:41 @ Betacommand . we set that up and then slowly move over to that
* 20:41 wikidemo . I don't understand the mention of infoboxes. In pratice, for those three uses (logos, album covers, book covers), the vast majority are already in infoboxes.
* 20:42 carl-cbm . wikidemo: I just mean that because there is so much functionality in a single template, the code is a pain to maintain
* 20:42 wikidemo | Oh. The infobox is probably a hard palce to keep the use rationale information. But a simple thing like keeping the use rationale template fields consistent with the accepted infobox would be a help (e.g. publisher= | title=)
* 20:43 carl-cbm . wikidemo: I shouldn't have mentioned infoboxes, just ignore me
* 20:43 @ Betacommand . wikidemo: use rationales are seperate
* 20:43 wikidemo . okay....
* 20:43 @ Betacommand . we were thinking about merging license templates
* 20:44 carl-cbm . I'd like to get some ideas for a proposal to make on the wiki about this.
* 20:44 carl-cbm . a minimal start would be to tag just the big three with a templatized rationale
* 20:44 carl-cbm . and let it develop from there
* 20:44 @ Betacommand . the issue about the big three is over use
* 20:44 wikidemo . I'm not sure that the rationale should be part of the copyright tag.
* 20:45 carl-cbm . wikidemo: no, it shouldn't
* 20:45 @ Betacommand . 80% of album covers are probably used in 2+ articles
* 20:45 carl-cbm . Betacommand: I checked books today, only 1450 are used in multiple articles
* 20:45 wikidemo . They were used in 2+ articlues until we got rid of galleries and discographies....
* 20:45 @ Betacommand . wikidemo: correct it would be seperate
* 20:45 @ Betacommand . wikidemo: they are still used in 2+
* 20:46 wikidemo . but a template's approval would be limited to specific copyright tags. you can't use an album cover rationale on a promotional image, for ex.
* 20:46 carl-cbm . wikidemo: that's why it must have an article= parameter
* 20:46 @ Betacommand . tag and article class
* 20:46 carl-cbm . but a bot can check the license tag matches the rationale that way
* 20:46 @ Betacommand . IE abum articles
* 20:47 @ Betacommand . and I can cross check article classes with use classes
* 20:47 wikidemo . Do you mean the bot would go to the article and make sure the article has a category indicating it's a music album?
* 20:47 carl-cbm . wikidemo: no, I meant that it would prevent people from claiming the rationale when they shouldn't
* 20:47 carl-cbm . but yes, it would be possible to do what you just said
* 20:48 wikidemo . okay....so the primary thing is that the bot makes sure if you use an album rationale it's for an image with an album copyright tag?
* 20:48 carl-cbm . yes
* 20:49 @ Betacommand . wikidemo: correct
* 20:49 wikidemo . That may expose an error we haven't talked about much, images with more than one copyright tag or that are mistagged.
* 20:49 carl-cbm . Betacommand: do you already look for those?
* 20:49 wikidemo . Many people think that if you take a picture, scan, etc., of an album or book cover, then donate it to commons, you have a free image.
* 20:49 carl-cbm . well, if it's on commons, that is another issue altogether
* 20:50 @ Betacommand . carl-cbm: no but it wouldnt be that hard
* 20:50 @ Betacommand . I can create a master list of album covers and cross check that with image rationale listings
* 20:50 wikidemo . I guess that won't be an issue with the 3 uses (albums, logos, books) but it is with, say, pictures of sculptures.
* 20:51 carl-cbm . well there are always mistagged things, but I think we should focus on correctly tagged things - how should they be tagged - not on incorrectly tagged ones
* 20:51 wikidemo . Okay, agreed that this is a side issue for this discussion.
* 20:52 carl-cbm . invalid license tags have a CSD criteria and can always go to IFD for discussion
* 20:52 @ Betacommand . I also have [[User:Betacomand/FUIR]] (http://en.wikipedia.org/wiki/User:Betacomand/FUIR) that can used
* 20:53 @ Betacommand . or I can write another simple pywiki tool for us to use
* 20:53 carl-cbm . So I think the issue here is to get the uses that are valid correctly tagged. And among book covers and logos, most uses are valid
* 20:54 wikidemo . What about process? should we move on to that?
* 20:54 @ Betacommand . carl-cbm: unless they are on several articles
* 20:55 @ Betacommand . wikidemo: explain
* 20:55 carl-cbm . Betacommand: most book images are on 1 article only
* 20:55 carl-cbm . wikidemo: I think so
* 20:55 @ Betacommand . carl-cbm: not album covers
* 20:55 carl-cbm . I am willing to believe that
* 20:55 carl-cbm . but I said books and logos for a reason
* 20:56 wikidemo . Oh, I'll save that comment. Once we're finished talking about what an image page should look like and how the bots can work, we need to think about how to implement it, encourage people to hold off on bot-tagging legacy images, etc.
* 20:56 @ Betacommand . logos are abused too
* 20:56 carl-cbm . wikidemo: right.
* 20:56 @ Betacommand . there are often logos with 20 uses
* 20:56 carl-cbm . So I think an image page should have something about the source/copyright holder and then a list of article/rationale pairs
* 20:56 wikidemo . logos are occasionally used in a gallery-like fashion.
* 20:57 carl-cbm . does that description sound reasonable?
* 20:57 @ Betacommand . carl-cbm: yeah
* 20:57 wikidemo . In the corporate/business articles I usually see logos are rarely used more than once or abused. I see a lot of logos used multiple times in articles about games, fictional worlds, etc. The "warrior's guild tartan" (to make up an example) will appear in ten different game articles.
* 20:58 wikidemo . but yes, agreeing.....a source/copyright holder statement, once per image.
* 20:58 carl-cbm . so the question is how to implement that
* 20:59 carl-cbm . in a way that the community can agree to
* 20:59 @ Betacommand . and rationale/article pairs too
* 20:59 wikidemo . Article / rationale pairs is good. Could be a bear to implement where people have not been clear about which rationale applies to which use.
* 20:59 carl-cbm . well, at first you just copy the whole rationale once per article. later you can change it if it is challenged
* 21:00 wikidemo . For newly uploaded images, is it reasonable to expect people to put ALL rationales inside a template like I proposed, if only a simple pairing of article=xxx and rationale=yyyyy?
* 21:00 carl-cbm . probably, so long as it is easy to use
* 21:00 carl-cbm . we could even add it to the upload wizard given time
* 21:02 carl-cbm . we could even make "newly uploaded" mean Jan 1, 2008
* 21:02 wikidemo . For legacy images, if we see 1 rationale but no article information, is it okay to pull that form the "file links" info and simply add it?
* 21:02 carl-cbm . we can add a category asking for verification, as well
* 21:02 «Quit» | Betacommand (i=Betacomm@unaffiliated/betacommand) has signed off (Nick collision from services.)
* 21:03 carl-cbm . Don't worry, he'll be back soon
* 21:03 carl-cbm . So what's the best way to move forward onwiki?
* 21:03 wikidemo . Will he see any comments he missed?
* 21:04 wikidemo . "onwiki" means implementation? Sorry, just don't know that lingo...
* 21:04 Betacommand (i=Betacomm@unaffiliated/betacommand) has joined #BetacommandBot
* 21:04 carl-cbm . wikidemo: I mean on the wiki rather than on irc. what matters is consensus there; otherwise we can devise all sorts of grand schemes
* 21:04 Betacommand . sorry I got disconnected
* 21:05 carl-cbm . you didn't miss anything
* 21:05 wikidemo . Oh....I do wish there were more participants so it's not just a gang of three.
* 21:05 carl-cbm . well, we need to write up a proposal and try to be persuasive
* 21:05 Betacommand . wikidemo: I agree
* 21:05 Betacommand . and a ideal image discription page
* 21:05 carl-cbm . really, I have brought up templated rationales several times and I think it has some traction
* 21:06 anetode () has joined #BetacommandBot
* 21:06 carl-cbm . wikidemo: ask and receive
* 21:06 anetode . ah, finally
* 21:06 carl-cbm . anetode: one hour late
* 21:06 anetode . yeah
* 21:06 anetode . i suck
* 21:07 anetode . didn't check wt/nfc until fifteen minutes ago
* 21:07 anetode . was there any progress made?
* 21:07 wikidemo | How many other bot operators are out there? A few days ago I noticed one who tagged a couple thousand logos, from 0000 through ARxxxxxx alphabetically, for having no source. If we could get everyone to agree to concentrate on the new (1/1/07 and beyond) images for a few weekswhile we seek consensus that would be a big help. Also get people to buy in that all the bots / tagging / deletion...
* 21:07 wikidemo . ...efforts should be based on a communal decision, not just everyone deciding for themselves what to do.
* 21:07 wikidemo . anetode, we have decided that Wikipedia is no longer a free content project, and we will be accepting advertising :)
* 21:07 carl-cbm . wikidemo: there are lots of bots, but I don't think many do image tagging
* 21:08 Betacommand . wikidemo: that bot should never have been apprved
* 21:08 anetode . wikidemo: now that's something i can get behind, do we get dividends?
* 21:08 Betacommand . carl-cbm: Im the only bot op that does it with half a brain
* 21:09 carl-cbm . anetode: we talked about the end goal for image description pages
* 21:09 Betacommand . the others are clueless about NFCC
* 21:09 carl-cbm . anetode: we think they should have a source/copyright, a license tag of some sort, and then a list
* 21:09 anetode . carl-cbm: good, is there a consensus towards convergence?
* 21:09 carl-cbm . the list has rationales and article names
* 21:09 carl-cbm . one rationale per article name
* 21:10 carl-cbm . convergence of?
* 21:10 Betacommand . anetode: convergence?
* 21:10 wikidemo . Okay, I think that answers the question about other bots. As long as they're all incidental, we'll just announce this is the process under consideration. If anyone strikes out on their own, we'll just catch them and ask them to join, right>
* 21:10 anetode . the rationales
* 21:10 anetode . toward one shared format
* 21:10 carl-cbm . we were talking about making templates for the big three
* 21:10 Betacommand . anetode: not quite yet
* 21:10 carl-cbm . we were talking about templated rationales for the big three
* 21:11 anetode . coverart and logos?
* 21:11 carl-cbm . with a future possibility for more tempalted rationales
* 21:11 carl-cbm . cover art for books and albums, on articles about the subject
* 21:11 anetode . basic, safety zone fair use stuff, i like that approach
* 21:12 Betacommand . anetode: and they are not per image they are per use
* 21:12 wikidemo . those three are about 40% of the fair use images, not sure what % of all the noncompliant ones.
* 21:12 carl-cbm . anetode: yes, safety zone like "book cover on the article about the book"
* 21:12 anetode . beta: ok, it doesn't make sense for multiple instances of usage anyway
* 21:12 wikidemo . and in practice, those "big 3" are only used in the safe way once per image.....
* 21:13 carl-cbm . that's one that that makes it safe to use generic rationales
* 21:13 Betacommand . anetode: what do you mean?
* 21:13 wikidemo . nobody uses the same book cover image in an article about more than one book (unless I play devil's advocate....Time-Life book series?)
* 21:13 Betacommand . doesnt makes sense?
* 21:13 anetode . beta: as in, i agree with the "use only at the home article" type of rationale
* 21:14 anetode . carl-cbm: one concern i had, that you briefly touched on at wt/nfc, is the difference between basic rationales that just say "lo-rez, no free alternative" and more substantial ones that identify the necessity for use
* 21:15 carl-cbm . anetode: I run into those all the time, but for these three it should be safe enough
* 21:15 anetode . is there any chance to split these up? have all the basic details templated and the real necessity for usage explicitely spelled out
* 21:15 anetode . outside of the big three, that is
* 21:15 carl-cbm . anetode: I suggested that earlier in this conversation
* 21:15 anetode . cool :)
* 21:16 carl-cbm . make a "checklist" for the basic things like low-res
* 21:16 Betacommand . I think the substantial rationale is needed
* 21:16 carl-cbm . well, it's the goal at least
* 21:16 wikidemo . If we were ever to approve a template for a use where that's not obvious, we could always require a parameter for that, e.g. "significance=". But we can leave that to the future, for if and when people propose a new template.
* 21:16 Betacommand . anetode: we could work with non-free media rationale
* 21:16 anetode . beta: that's a decent, if simplistic, foundation
* 21:17 carl-cbm . what is that?
* 21:17 anetode . at least as far as fair use laws go, the two main concerns are impact on market value and critical commentary
* 21:17 carl-cbm . oh, that one. no, please, not that one
* 21:17 carl-cbm . for example, the source only needs to be listed once per article
* 21:18 anetode . and those two concerns are not often addressed in rationales
* 21:18 carl-cbm . and so does "low res"
* 21:18 carl-cbm . and "portion used"
* 21:18 carl-cbm . anetode: the issue is that "market value" is very generic
* 21:19 wikidemo . Well, for the "big three" the argument is the exact same in each case. That's why they're the easiest cases. Portion used is always the whole thing. Resolution is standard or should be. Impact on market value and critical commentary are the same, etc.
* 21:19 anetode . carl-cbm: true. but there are clear counterindications. for example, corbis and "rights-managed" images
* 21:20 anetode . also press agency photographs, etc.
* 21:20 carl-cbm . so here's my suggestion. remove the fair-use-rationale-like parts from "non-free book cover" and make "non-free book cover rationale" that includes them, and requires an article name parameter
* 21:21 anetode . as in, {{non-free book cover used in|article}}?
* 21:21 carl-cbm . exactly
* 21:21 carl-cbm . but that title for the template isn't good, since it looks like any article name fits
* 21:21 wikidemo . anetode, I don't think we'll run into Corbis on books, album covers, and logos. Press photos and stock photos are probably never going to be templated for the reasons you mention....we can find as smooth as possible to tag and delete them but we'll probably never find anything better to do for a use rationale than just having people write it out.
* 21:22 Betacommand . {{non-free book article|NAme}}
* 21:22 carl-cbm . that's more clear
* 21:22 anetode . wikidemo: you're right, sorry, i didn't mean that in the context of the big three. just jumping off to a pet peeve
* 21:22 wikidemo . Well, we can at least force people to say what article they're talking about, and to use one rationale per article, that's already an improvement.
* 21:23 Betacommand . brb
* 21:23 carl-cbm . well, we can force it only with agreement on the wiki, but yes
* 21:23 carl-cbm . so what concretely are we proposing? 1) the format of image description pages
* 21:23 anetode . the other category to consider, outside of covers and generally promotional material, is images of deceased people
* 21:24 carl-cbm . that's a hard issue
* 21:24 wikidemo . I think naming the templates is a detail we can figure out later, being precise but avoiding wikispeak.
* 21:24 carl-cbm . true, but if we make a new template we should pick a descriptive name to start with
* 21:25 carl-cbm . 1) image pages and 2) three standard rationale templates
* 21:25 carl-cbm . don't those templates already exist?
* 21:25 anetode . these are to be seperate from the copyright tags, correct?
* 21:25 carl-cbm . yes
* 21:25 carl-cbm . they have to be, because they name an article, but copyright is an absoluteproperty of the image
* 21:26 anetode . but the rationale also requires the source and copyright status of the image, does it not?
* 21:26 wikidemo . Were we going to include the {{source=}} tag or field somewhere in the proposal?
* 21:26 carl-cbm . anetode: yes, but that isn't in any template right now
* 21:26 wikidemo . Source and copyright status are once per image logically, not once per use.
* 21:27 carl-cbm . wikidemo: let's talk about that the {{source=}} tag
* 21:27 carl-cbm . one easy thing to do is add it to the copyright tag
* 21:27 anetode . yeah
* 21:27 carl-cbm . and then slowly fill in the data over time for legacy images
* 21:27 wikidemo . They've been stuffed into the master use rationale template for some reason. They should be standing alone or in the copyright tags. Too much effort to add a field to copyright tags.
* 21:27 carl-cbm . it's easy to edit the templates
* 21:27 carl-cbm . to add another field
* 21:27 wikidemo . Well, we could go and add a field to every existing copyright tag and require it of all new copyright tags.
* 21:27 carl-cbm . that's what I meant
* 21:28 anetode . usually this info is stuffed into the "summary" heading, the source details could be merged
* 21:28 wikidemo . And when we can, see if we can integrate it into the upload wizards.
* 21:28 carl-cbm . once it's in place in the template, they should be able to make the upload wizard handle it
* 21:29 wikidemo . For images uploaded to date could be in the summary, or in a random stray comment on the page, or in a use rationale.
* 21:29 carl-cbm . yeah, it has to be manually dealt with for existing images. But those can be grandfathered for a period of time
* 21:29 wikidemo . It would be nice if we could organize it, but can that be automated? If we can't automate it, it's not much use for purposes of tagging and deleting old images. Is anyone seriously planning to delete the legacy images for lack of source identifications, or is it one errant bot?
* 21:30 carl-cbm . don't know. I would think it's a very hard thing for a bot to do
* 21:30 anetode . some copyright tags and rationales are self sourcing, cover art is usually copyrighted by the distributor or artist and logos are owned by their companies
* 21:30 carl-cbm . because I couldn't do it
* 21:31 carl-cbm . anetode: the policy as it stands isn't that liberal
* 21:32 anetode . carl-cbm: there's a difference between liberal and practical
* 21:32 Betacommand . carl-cbm: I cant do a source bot
* 21:32 carl-cbm . I agree; just pointing out that it requres a change in policy
* 21:32 wikidemo . Yes, for book covers, album covers, and logos, we can add a source automatically if we can't find one. That does open up two concerns. One is, does it need verification by hand on an image per image basis before we trust it? The other is what if we add a source by bot, even though there is an existing one we failed to notice because it was stated in a way that was not machine readable? ...
* 21:32 wikidemo . ...Then we have duplicate info.
* 21:33 anetode . it requires a clarification in policy
* 21:33 anetode . otherwise it will create a lot of useless work
* 21:33 carl-cbm . I am all for avoiding useless work
* 21:33 carl-cbm . or making bots do it
* 21:33 Betacommand . carl-cbm: now play nice and lets not abuse BCBot
* 21:33 carl-cbm . I think we have some concrete ideas here that we can move forward on
* 21:34 carl-cbm . I think we should get a proposal on the wiki, so that other people can comment on it
* 21:34 wikidemo . Instead of going through the motions of adding a source, we could add a policy / guideline statement that in cases of logos, book covers, and album covers (and others, as we come up with them) the source will be inferred from the copyright owner....and the copyright owner will be inferred too.
* 21:34 anetode . providing that there's still a field which could be corrected if the basic source info is insufficient
* 21:34 wikidemo . (or we could incorporate that assumption into the master rationale for each class of use)
* 21:35 wikidemo . Yes, source is inferred if not explicitly stated. Not a perfect idea going forward, but perhaps a practical solution for legacy images. Going forward it coudl be a checkbox, or a default to be overridden.
* 21:36 wikidemo . Okay...concrete ideas, proposal on wiki, comments....
* 21:36 carl-cbm . I think that it's worth getting comments, since there's only four here
* 21:37 Betacommand . so lets say thursday same time same place?
* 21:37 anetode . sure
* 21:37 Betacommand . I'll post this and lets see the results
* 21:37 wikidemo . Well, if betacommand, carl, and I jointly propose something and say we're actually in agreement, I think everyone else at the NFCC talk page is going to faint, and we'll get consensus by silence.
* 21:38 wikidemo . (for people reading this, that was humor to suggest that we have been at opposite ends of a few debates, so perhaps our agreement here is indicative of a broad span)
* 21:38 carl-cbm . well, this is the easiest part of the problem
* 21:38 anetode . and a huge chunk of it, i'm glad you've mulled over the big three
* 21:39 carl-cbm . so what needs to be done? should we write the rationale templates so that people can look at them? Or just write up a detailed proposal on WT:NONFREE?
* 21:39 Betacommand . carl-cbm: make it a subpage
* 21:40 wikidemo . On the question of what's yet to be done / talk about / implement, what about the notion that when we do come up with a plan, we sort and then tag all the images at once then set a schedule for processing. If there are 70,000= album covers and perhaps 30,000+ are noncompliant, we do have to have either a bot or process for adding the template rationale to all of them, then decide whether...
* 21:40 wikidemo . ...each needs to be checked by hand before declaring them safe. Should we leave that for the next discussion?
* 21:41 carl-cbm . I think so. no need for an implementation plan before the general strategy is sound
* 21:41 wikidemo . Personally, I think we may want to wait on proposing specific templates until people have agreed to the idea of templates. Then it becomes an implementation detail, not a second front to oppose the proposal.
* 21:41 Betacommand . wikidemo: [[User:Betacommand/FUIR]] (http://en.wikipedia.org/wiki/User:Betacommand/FUIR)
* 21:41 carl-cbm . I agree it's an implementation detail, but not everyone might agree
* 21:42 wikidemo . But I could be wrong. (Yes, I'll take a look there). For some people a real template is less scary than the idea of one.
* 21:42 Betacommand . wikidemo: its a program for working with images
* 21:43 wikidemo . Is this a real language? It looks like my old professors' pseudo-code....
* 21:43 Betacommand . wikidemo: [[m:Pywikipedia]] (http://en.wikipedia.org/wiki/m:Pywikipedia)
* 21:43 Betacommand . its python
* 21:43 carl-cbm . wikidemo: no, it isn't a real language, it's python
* 21:43 wikidemo . okay, thanks.
* 21:44 Betacommand . but written for wikipedia
* 21:44 carl-cbm . real languages have syntax like $nsNumber{${$namespacestemp}{$ns}{'content'}}
* 21:44 Betacommand . carl-cbm: PHP?
* 21:44 carl-cbm . perl
* 21:45 Betacommand . pywiki is better
* 21:45 carl-cbm . so... someone needs to write the proposal
* 21:45 Betacommand . Im not that go at it
* 21:45 carl-cbm . I would like to volunteer wikidemo, if you're willing
* 21:45 anetode | i can at least help with the actual writing, since i don't know shit about code
* 21:46 wikidemo . I was afraid of that. Perhaps I'm too new and controversial for anyone to listen to me.
* 21:46 Betacommand . anetode: I can do the coding
* 21:46 carl-cbm . code can always be dealt with... the real issue is to get consensus to do something, then the code is easy
* 21:46 Betacommand . wikidemo: you can get support from us
* 21:46 anetode . wikiedmo: wikipedia's a meritocracy, if you make sense, people will listen
* 21:46 carl-cbm . wikidemo: or not old and controversial enough
* 21:47 wikidemo . Okay. As I tell everyone, I always make sense. Just that not everyone understands me.
* 21:47 Betacommand . or hated like I am
* 21:47 Betacommand thinks that I am the most hated person on wikipedia
* 21:47 wikidemo . I'll have to review this to figure out exactly what we're proposing. How about the templates? Are they part of the proposal?
* 21:48 carl-cbm . I think they are. the image page format and the templates
* 21:48 Betacommand . wikidemo: the big three only
* 21:48 anetode . beta: controversial does not necessarily equal hated
* 21:48 carl-cbm . I can have a log ready in very short order, so you can review it
* 21:48 wikidemo . I mean, will we be developing specific templates to propose at the same time?
* 21:49 Betacommand . anetode: have you seen my archives?
* 21:49 carl-cbm . I thought your idea to put it off was OK. We can write them if there is demand for it, writing a template doesn't take much time. Debating the wording does, though
* 21:49 Betacommand . carl-cbm: remember to remove the hostmasks
* 21:49 carl-cbm . Betacommand: OK
* 21:50 anetode . beta: guess not. from what i've seen, you have enough people supporting you to keep going
* 21:50 wikidemo . Okay. I'll model it after a subset of my LIP proposal. The "rationale" and "background" might be important for gaining consensus but I don't want to write the Magna Carta just by way of introduction.
* 21:51 carl-cbm . I think it is a subset of LIP, just a few points of it
* 21:51 Betacommand . anetode: I have support but also a lot of F*&k you's
* 21:51 wikidemo . What's the timing? Maybe I should put it on my user page first so the four of us can make sure it's what we really mean to propose, then when we agree, announce it to....NFCC, village pump, etc?
* 21:52 anetode . beta: that's people venting, and it sucks. conjure up a couple of ultra-polite canned replies, and you'll see many of them take their words back
* 21:52 anetode . wikidemo: could i please get a link to this LIP proposal?
* 21:52 carl-cbm . yes, I think that's reasonable. Just let me know and I'll comment on it
* 21:52 wikidemo . It's at [[WP:LIP]] (http://en.wikipedia.org/wiki/WP:LIP)
* 21:52 anetode . *smacks head*
* 21:53 wikidemo . And to repeat an earlier caution, it's more of a discussion docuemntn than a proposal. Nobody is exactly proposing we adopt it in anything like its present form.
* 21:53 carl-cbm . true. it's a "proposal" in the sense that we're not actually deciding anything here, only deciding on what to propose
* 21:54 carl-cbm . but once it's on wiki I hope it gets good discussion there
* 21:54 wikidemo . What about timing? Thursday seems a little fast unless I get paid extra for priority service.
* 21:54 carl-cbm . wikidemo: If you are occupied, I can write up a proposal tomorrow
* 21:54 Betacommand . wikidemo: that was just a date tossed out
* 21:55 wikidemo . Either way. Realistically, I haven't been too great at meeting deadlines even in my paid professional life. I would hate to let anyone down.
* 21:55 anetode . then do away with the deadline. start working now, and when you're finished, put it forth
* 21:55 Betacommand . what about 16:00 Tuesday?
* 21:56 carl-cbm . to have a draft?
* 21:56 carl-cbm . that's 14 hours
* 21:56 Betacommand . carl-cbm: no for the next meeting
* 21:56 Betacommand . carl-cbm: a week from now
* 21:56 carl-cbm . ah. sure
* 21:57 wikidemo . OKay, that I can do for sure.
* 21:57 carl-cbm . I'll work on the proposal tomorrow, and let you all know where it is, and we may be able to make it live on Wednesday
* 21:57 wikidemo | Carl, why don't you take a first pass and I'll help expand, comment, etc, as necessary?
* 21:57 carl-cbm . yes, that would be very helpful
* 21:58 carl-cbm . I'll let you know when I put it on the wiki
* 21:58 carl-cbm . Are we done for the evening then?
* 21:59 Betacommand . carl-cbm: I think we have gotten as far as we can without getting more input from the wiki
* 22:00 anetode . g'night
* 22:00 «Quit» | anetode (n=n@unaffiliated/anetode) has signed off ()
* 22:03 carl-cbm . OK, I will post a log through now.
* 22:03 carl-cbm (n=cbm@wikipedia/cbm) has left #BetacommandBot ()
* --- Log closed Mon Sep 17 22:03:27 2007