Wikibooks:Reading room/Proposals/2013/January

Wikibooks-tan
Hello, Wikibooks! I'm autoconfirmed on WP. Wikibooks-tan will be our mascot in the future. My idea for her: for head and skin color;  for body, face and outfit. YellowPegasus (discuss • contribs) 21:22, 4 December 2012 (UTC)


 * The skin color looks odd to me. Don't like it, tbh.
 * What is there about this appearance that you feel expresses the unique character of Wikibooks amongst the sisters? (I'm not seeing it, off hand.)
 * Why that outfit? Hair color?  Hair style?
 * The name for a Wikibooks moe would be more likely "Books-tan", wouldn't it, parallel to Quote-tan for Wikiquote?
 * Over at Wikinews we have, btw, discussed what a News-tan would look like. Features of appearance are chosen for specific reasons; e.g., red hair is a connection to Commons-tan, as Wikinews and Commons have some things in common.  The design of these characters can really embody the spirit of a sister project, but one needs to put careful thought into the details.  --Pi zero (discuss • contribs) 07:51, 5 December 2012 (UTC)


 * This proposal has WP puzzle pieces on it (the old puzzle pieces, in fact). That's not very WB-ish in my opinion. Maybe a mascot with a book for a head will work better. Kayau 07:43, 4 January 2013 (UTC)
 * What on earth is a Wikibooks-tan or any other wiki-tan? Why do we need a mascot and why should we accept the mascot imposed on us by somebody who is new to the project?--ЗAНИA [[Image:Flag_of_Estonia.svg|15px]]talk 18:56, 17 January 2013 (UTC)
 * I think a mascot would be nice, just not this particualr design (and nothing anime-like because anime sucks). I know you're usually rather opposed to non-'constructive' things like barnstars, but I think a good mascot would be a great way to bring the community together, not to mention look nice on the main page. Kayau (talk · contribs) 05:30, 18 January 2013 (UTC)
 * There are three sister projects with these personifications now, and at Wikinews we did once have a discussion about how we would imagine a "News-tan" sister. There's a pretty good explanation at w:WP:Wikipe-tan.


 * A design for Books-tan &mdash;by Wikibookians, obviously&mdash; could be good (community spirit and whatnot), but we'd need to consider it carefully, to get it right. The personality of Wikibooks should come through clearly in all the details of the depiction.  (For example &mdash;I need to recover the details of the discussion while I still can&mdash; as I recall, our design for News-tan had red(?) hair, a camera, a press pass, and maybe a green eyeshade.)  --Pi zero (discuss • contribs) 13:59, 18 January 2013 (UTC)
 * While an anime girl-style mascot may work for some sisters (IMO it doesn't, and gives WP a rather negative image in fact), I think for WB, it's best to stick to stuff like a bookworm with glasses. It's about our image... Kayau (talk · contribs) 14:16, 18 January 2013 (UTC)


 * I've got pretty mixed feelings about them myself, and indeed about the concept of a mascot. Though it does seem a common-themed set of mascot-ish things for a bunch of different projects might, just perhaps, help to promote some sense of multi-project community spirit, which could be a good thing.  For me, that possibility sort-of nudges the idea over toward the "okay" column.  Even so, as I say, my feelings on it are mixed.  Without that cross-project theme, a mascot would have to clear a higher bar for me to support it.


 * A bookworm with glasses? How would you envision that, in more detail?  --Pi zero (discuss • contribs) 14:56, 18 January 2013 (UTC)
 * This kind of thing. :P Kayau (talk · contribs) 15:59, 18 January 2013 (UTC)


 * I read the info on tan on the page you mentioned. Seems somewhat interesting although not really very effective if I'd never seen any mention of it before.  I'm thinking that an anime-style looks best as it's more universal whereas something more traditional may make us seem backwards or American.  Think we should steer clear of stereotypical things like glasses and a bookworm and aim for something more modern like an e-reader!  Maybe starting with a Wikijunior-tan first is a good idea as kids would certainly notice such a mascot more than others?  --ЗAНИA [[Image:Flag_of_Estonia.svg|15px]]talk 16:01, 18 January 2013 (UTC)


 * Come to think, I've never really thought "mascot" about the wikimedia sisters; Wikipe-tan isn't officially a mascot of Wikipedia, after all. More like, "personification" for, as it were, social occasions where it's useful to have one.


 * Btw, it occurs to me Wikipe-tan was one of the earliest-added pictures in the coloring book &mdash; Wikijunior:Maze and Drawing Book/Wikipe-tan. --Pi zero (discuss • contribs) 16:55, 18 January 2013 (UTC)
 * I think as anime is mainly enjoyed by people born during the last two decades of the 20th century, it is frowned upon by most older people. Plus, they haven't exactly gained a reputation of being a healthy form of comics/cartoons. That is enough to put off some schoolteachers and librarians from using WB for students. Although I think an e-reader is a great idea, we may have to change it every few years or so - technological innovations occur ever so often and it could be that e-readers ten years from now don't even remotely look like e-readers of the present. Therefore, I think a mascot with traditional books will be better. Perhaps a book with a limbs and a face on the cover? Kayau (talk · contribs) 22:48, 18 January 2013 (UTC)
 * It's not actually anime, of course, but an interesting suggestion about conservative educators. Although... I also do wonder, sometimes, how old my fellow wikimedians imagine I am.  --Pi zero (discuss • contribs) 00:01, 19 January 2013 (UTC)

Proposal to implement metadata on Wikibooks
Hi all - I'm Max Klein. I would like to make it possible to use metadata - in the form of LRMI and Schema - on Wikibooks. Yaron Koren and I have developed a Mediawiki extension and a library of templates to use in conjunction with it that would allow this to happen. LRMI and Schema are ways to describe the information contained on a page in a way that is easily intelligible to machines. Schema was developed by Google, Bing, Yahoo, and Yandex, while LRMI was developed by a broad coalition of people and groups interested in open educational resources headed by Creative Commons and the Association of Educational Publishers. You can see significantly more detail about Schema and LRMI in the collapsed proposal below, or on their websites (linked above,) and the LRMI specification.

This is a simplified version of the proposal that I initially posted on this page. Hopefully, it will better convey why we think this idea is important. I've left my original post directly below this one in a collapsed box. It contains a lot more details than this does. I know that it's unusual to do this to a proposal that has already initially been posted, but I think it is for the best in this case. If anyone strongly feels that this is inappropriate, please let me know.

Why would we want to do this?
Machines can't always understand information that is contained in text, even if humans are able to. Metadata of the nature we are suggesting adding helps machines understand what a page is about. The biggest effect this would have would be increasing the accessibility of Wikibooks' content via search engines. Schema.org was formed collaboratively by Google, Yahoo!, Bing, and Yandex with the aim of making it easier for people to find high quality relevant results in their searches. Most major search engines use metadata in their search result placement, as well as use metadata to generate better previews of pages. Improving the accessibility of Wikibooks content via search engines would allow more people access to free knowledge. The addition of metadata to Wikibooks would also allow other people to build tools that better catalog Wikibooks content in ways not currently possible.

What would this involve for Wikibooks editors?
Editors could add the template to books that they happened to be editing already. The template would not be in any way required - it could gradually be added to existing Wikibooks’ content over time. Every page tagged would present an incremental benefit. The template could be added either by cut and pasting it, or by clicking the LRMI button that we have added to the editing window.

Here’s what the template looks like:

The tags that are currently in the template are taken from LRMI's specification. It would be easy to add other tags from schema.org's specification that would be useful to describe Wikibooks content as well. Not all pages would have to be tagged with this template, and on pages that do use this template, not all parameters would have to be filled out. Although many of the parameters in the template are self-explanatory, not all are - we would copy over descriptions of what each parameter is for from LRMI's specification, and include further details about how to fill them out. From a technical standpoint, doing all of this would involve installing the HTML Tags Mediawiki extension on Wikibooks, and porting over the templates we have developed.

I think that the use of metadata has significant potential to benefit Wikibooks without adding much extra workload to Wikibooks' contributors. People could tag books with metadata as they go through their normal editing process, and there would be no urgency in doing so. As an increasing number of pages were tagged, the benefits of the added metadata would grow, eventually being quite substantial. You can read further details in the collapsed original proposal below. Thanks, Maximilian.Klein.LRMI (discuss • contribs) 01:14, 18 January 2013 (UTC)

Discussion
I'm very much in favour of adding metadata to Wikimedia projects (I'm the founder of Wikipedia's [[Wikipedia:WikiProject Microformats|WikiProject Microformats; and have initiated the use of microformat on Wikimedia Commons, Wikispecies Wikisource, Wikivoyage and here on Wikibooks). I know schema.org to be a worthwhile and well-respected project. Likewise, I've worked on other projects with Max, and hold him in high esteem.

I do have one concern about the proposal made here; I'm not saying we shouldn't proceed, but we should be mindful of the issue I have in mind, and see what we can do to address it.

The examples given above involve what's called "hidden metadata". The values which are to be read by other systems are not shown on the page, for example:

"This OER is appropriate for"

renders as:

"This OER is appropriate for Grades 6-8"

in other words, the values "11-14" are not shown on the page, and if "21-34" had been entered in error (or by a vandal), there would be no way to tell, from reading the page.

A better solution would be to mark up "Grades 6-8" as a value, or have a template in which one set of values were entered, but which emitted metadata of "11-14" and a visible value of "Grades 6-8" (we do this with Template:Birth date on Wikipedia, for instance, where the input is, say,  but the output for machines is "1993-02-04" and for humans "February 4, 1993".

The same applies to the example on http://lrmi-demo.referata.com/wiki/Biology - for instance, the values "P30M", "0-12" and "non-interactive" are not visible on the page., This is analogous with PERSONDATA on Wikipedia; I have a draft essay on the issues this causes, at w:User:Pigsonthewing/Persondata. The solution may be to make the details in the template on that demo page visible, in a footer box.

Also, as a separate point, we need to consider how this will interact with the new Wikidata project.

Finally, this section is already long; should we move it to its own sub-page?

I look forward to working with Max and other Wikibooks editors, to resolve these matters and start emitting extremely useful metadata. We should enable the HTML tags extension in the meantime. Pigsonthewing (discuss • contribs) 15:18, 17 January 2013 (UTC)
 * As far as I know, there are no current plans to implement Wikidata on Wikibooks, so using this extension here shouldn't present any problems in terms of that. If Wikidata does eventually include Wikibooks, I don't think integration would present especial problems.  You may be right that it would be best to display the details of the metadata in a footer box - as long as no one objected to the added visual element, it would reduce the likelihood of meaningful vandalism occurring. Maximilian.Klein.LRMI (discuss • contribs) 01:48, 19 January 2013 (UTC)
 * Tried to understand what this is all about and read the coutless information mentioned but even after all that this goes way over my head. I really have no idea what is being proposed here.--ЗAНИA [[Image:Flag_of_Estonia.svg|15px]]talk 17:30, 17 January 2013 (UTC)
 * Likewise. As a proposal, this is a miserable failure, because it isn't accessible.  The first paragraph is indecipherable by assuming expertise, to the point where anyone who doesn't already know what the proposal is about isn't going to find out this way.  If we understood this, would we want it?  If so, then it needs to be presented to us with way more thought into making the presentation understandable.  --Pi zero (discuss • contribs) 20:49, 17 January 2013 (UTC)
 * The "why" part is buried pretty deep. It's in the FAQ section titled "Why should we use metadata/HTML Tags?"  That portion of the proposal should have been placed a lot earlier and more prominently (i.e., not in a hidden section). Nevermind.  That info does come in an earlier section.  Basically, it's to make WB content show up more prominently to search engines.  I think it's a worthy goal. --Jomegat (discuss • contribs) 21:07, 17 January 2013 (UTC)
 * Really? Yeah, the first sentence should have said that.  It does sound like a worthy goal; the next questions would be, what are the downsides of this approach to the goal, and will the approach promote the goal sufficiently to make up for them.  Hm.  --Pi zero (discuss • contribs) 21:55, 17 January 2013 (UTC)
 * Off hand, I'd hazard a guess at one downside: it's probably bloody complicated.  Anything whose proposal is such a sprawling mess (sorry to be blunt, but it is) would likely appeal to people who're kind of tone-deaf about simplicity.  --Pi zero (discuss • contribs) 21:59, 17 January 2013 (UTC)
 * You can see an example of the template used in the collapsed 'example template' box above. We can, if desired, simplify the template further - and in copying over the template, we'll include a description of what each line of the template is used for.  The whole process of adding metadata would pretty much be hitting a button on the editing toolbar to dump in the template, and then filling out whatever lines of the template you find apply to the book you are working on. It is about as complex as the citation templates used on ENWP. Maximilian.Klein.LRMI (discuss • contribs) 00:31, 18 January 2013 (UTC)
 * I saw the Why part but it didn't help. So this is a method of search engine optimization along the lines of what we did in the 1990s when we'd fill out webpages with white text keywords to fool the search engine crawlers?  I guess it seems like a good idea - are there any disadvantages?--ЗAНИA [[Image:Flag_of_Estonia.svg|15px]]talk 22:18, 17 January 2013 (UTC)
 * Disadvantage? Well, every page has to be edited to add the right tags. Given the size of the community it'll be 2100 before the whole project is updated... QU TalkQu 22:27, 17 January 2013 (UTC)
 * That'll never happen. No easier way around it?  I'll read all of the above blurb again when I get chance to get a better understanding.--ЗAНИA [[Image:Flag_of_Estonia.svg|15px]]talk 22:29, 17 January 2013 (UTC)
 * Pages can be updated incrementally, not all will need to be updated at once. No one would need to go through all existing books and tag them; people could just add tags to pages as they edit them, and could choose not to add metadata if they did not want to do so. People who were interested in doing so could add metadata templates to books they were already editing, taking perhaps a minute or two to do so.  Maximilian.Klein.LRMI (discuss • contribs) 00:31, 18 January 2013 (UTC)
 * It'll never happen. Less than 5% of the pages here are regularly edited. In the 2 years since FlaggedRevs was implemented there are still less than half the pages reviewed, and that really is just clicking of a button. Most of these reviews were done by a handful of people. I can't see more than 1% of the pages having metadata added in a year. As I noted to Darklama on IRC today, 95% of editors need help just getting their books into the right subjects and page structure and the most active books are student collaborations who struggle with simple editing, let alone this more complex area. QU TalkQu 10:55, 18 January 2013 (UTC)


 * Oppose: Yet another thing for people to learn, understand, use, and agree on, which makes content creation all the more difficult, increases the difficulty of jumping right in, increases the complexity of book maintenance, increases the skill set required, and will likely decrease the potential pool of people willing to even try to write a book, all for the sake of a theoretically higher search ranking. --dark lama  22:50, 17 January 2013 (UTC)
 * Since metadata wouldn't be required, it shouldn't increase the difficulty of jumping right in. For people uninterested in doing so, installing HTML Tags and adopting this set of templates would not change the Wikibooks editing experience at all.  For people who were interested in doing so, adding metadata to existing books would gradually improve the accessibility of information on Wikibooks. Maximilian.Klein.LRMI (discuss • contribs) 00:31, 18 January 2013 (UTC)
 * Stylesheets aren't required for a website either, but when you embed CSS through the use of the style attribute on every html tag, maintenance can be a nightmare. Websites are easier to maintain when style and structure are separated, and accessibility can also be improved. I think for meta data to be practical and keep content creation accessible there needs to be a clean and clear separation of meta data from content creation. --dark lama  18:45, 19 January 2013 (UTC)

I know it's unusual to do so, but would anyone object to me replacing my initial post with a greatly abbreviated and less technical version? (I would subpage the existing text so that it would still be viewable to those interested.) Maximilian.Klein.LRMI (discuss • contribs) 00:31, 18 January 2013 (UTC)
 * I've gone ahead and replaced the original proposal with a greatly abbreviated version that I think is a lot more clear. I've collapsed the full original proposal immediately under the new version.  I know this is an atypical approach, but think that in this situation it is for the best.  If anyone feels it is an inappropriate thing to do, please let me know - I think that having a more clear intro is important, and thus feel like it's a good idea. Maximilian.Klein.LRMI (discuss • contribs) 01:14, 18 January 2013 (UTC)


 * Oppose: I completely agree with darklama. Good content will rank high in searches because people link to it not because it is tagged in some clever way. (Also: isn't the hype about the semantic web already over?) --Martin Kraus (discuss • contribs) 21:15, 18 January 2013 (UTC)
 * Good content is certainly a first, crucial, requisite step - but it's not always sufficient. Taking an example I used earlier, take a look at this book.  I think it's a pretty good book.  It is not, however, one that is easily accessible via a lot of search terms, and only gets a couple hundred of hits a month.  There are a significant number of search terms that it doesn't currently rank on, that the addition of LRMI tags would likely rank it for.  Additionally, once a sufficient quantity of material is tagged across the internet, customized search tools can be developed (Google already allows custom searches for schema.org parameters) that would allow someone to look explicitly for, say, a book whose subject was the sun that was aimed at kids under twelve - thus turning up this book. Maximilian.Klein.LRMI (discuss • contribs) 01:48, 19 January 2013 (UTC)

I support implementing this suggestion as long as it isn't required or anything. I think it can be up to the individual book contributors to decide whether and how to use the metadata. There's nothing wrong with a new option that could make search results more accurate. Kayau (talk · contribs) 22:53, 18 January 2013 (UTC)


 * Oppose, for now. Even though something like this may be optional, when it does get used, it adds to the perception of complexity by those who don't use it, as well as to the difficulty of book maintenance.  If, sometime down the road, we come up with a plan to mitigate those problems, we can adopt the extension then.  --Pi zero (discuss • contribs) 00:09, 19 January 2013 (UTC)
 * Oppose I'm not convinced of the benefits of this and the additional work involved seems to outweigh the possible benefits.--ЗAНИA [[Image:Flag_of_Estonia.svg|15px]]talk 10:16, 19 January 2013 (UTC)
 * Support as I don't see any harm in having the extension, and it does have the potential to do a lot of good. Contributors are not required to add the metadata, and any work required to take advantage of it is elective - it is not forced upon anyone. --Jomegat (discuss • contribs) 00:57, 20 January 2013 (UTC)
 * I did address, in my opposition stance above, how such an "optional" complication does harm even to segments of the project community who don't use it. --Pi zero (discuss • contribs) 02:11, 20 January 2013 (UTC)
 * Yes, I read that. --Jomegat (discuss • contribs) 03:24, 20 January 2013 (UTC)

It's a well-intentioned proposal and Wikibooks certainly could benefit from greater exposure. But looking at the demo, the button on the editing toolbar just inserts the raw template code and does indeed add complexity to the editing process. A JavaScript form would be more beneficial, much like the toolbar buttons for adding citations on Wikipedia. Furthermore, the template's inputs can be reduced, with many of the parameters being identical for all books. Those can be built-in to the metatemplate being called by LRMI-object and advanced users could override if really necessary. Parameters like interactivityType and inLanguage for example.

Installation of the HTML tags extension could allow partial application of metadata through existing templates such as reading level. A realistic use case I can see is to change the name of the LRMI-object template, only make use of it on the root page of a book, and have the parameters that are unique to each book provide output such as that seen in recipe summary and other infoboxes. You could also combine functionality from collection to link to a PDF/print version. Combining more functionality, the name of the book could be entered and call alphabetical as well as specify the LRMI "name". Then this would provide a visible benefit for book authors/readers. – Adrignola discuss 21:43, 26 January 2013 (UTC)

Adding a 'share' feature on the side
Howdy! Just cruising through here, and noticed that there's no 'share' features on the pages! It would be awesome to have a quick link to Facebook, Twitter, Google Talk, or stuff like that! I'm pretty sure it's easy code to implement~ —Preceding unsigned comment added by 198.60.121.1 (discuss • contribs) 23:23, 23 January 2013


 * We do something very like this on en.wn, by automagically transcluding at the bottom of every published article a template for sharing.  --Pi zero (discuss • contribs) 02:56, 24 January 2013 (UTC)


 * I agree that this would be a very useful function but wouldn't we have to include such an option for lots of different social networks to avoid favouring one commercial website? I really wish Wiki sites would try their absolute best to avoid linking to any commercial sites.  Maybe a simple email this page to someone function would be simpler?--ЗAНИA [[Image:Flag_of_Estonia.svg|15px]]talk 12:23, 24 January 2013 (UTC)


 * You may note the large number of different services on the above-linked Wikinews template. The first of which is email.  We do have occasional maintenance tasks as some of the linked services have to be handled differently, or are discontinued, or as new services come into existence and should be added to the list.  --Pi zero (discuss • contribs) 12:43, 24 January 2013 (UTC)


 * I know the linkage to commercial sites is what has created opposition to this on other wikis. There are AddThis extensions available for the major browsers if one would like to have the same versatility for social media sharing (see the bottom of the page). – Adrignola discuss 21:56, 26 January 2013 (UTC)