 Is this been live streaming this whole time? Who was that? Is this already live streamed? Oh! I can see it. I can see your shadow. Well, this goes on. Yeah, it's on the screen. Is it already live streaming? Yeah. But I can all cut off this part of it. We can start. This is the talk about bringing Enterprise Media Wiki to the Wikimedia Foundation. About me, I used to live in New York. Now I live in New Jersey. Recently. I do what I call Enterprise Media Wiki. I do what I call Enterprise Media Wiki. I do what I call Enterprise Media Wiki. I do what I call Enterprise Media Wiki. I do what I call Enterprise Media Wiki. Developing, consulting, hosting, and author. I run a consulting company called WikiWorks. And my book is Working With Media Wiki. And here's the second edition that's going to come out. Hopefully very soon. This is just a proof. But hopefully in the next few weeks. So watch for that. WorkingWithMediaWiki.com. I forgot to put the URL. I'm using Media Wiki Book right now. So, I hope you like it. What is Enterprise Media Wiki? It's a term I'm trying to popularize. Other people use it too. It's not, to me, it's not about corporate usage per se enterprise. The term enterprise gets a bad brat for various reasons. It's really just about the software. There are extensions that are used a lot within corporations, organizations, and various other wikis outside of wiki media that I think are very useful. The basic list, there's a lot of extensions that fit into this category and there's a lot more I could have added. These are some of the ones I plan to talk about. I guess I won't talk about data transfer, but especially the top three. There's one called Semantic Media Wiki. There's one called Cargo. There's one called Page Forms. There's one is various others that I'll get to in that list. I think it would benefit wiki media to make use of these extensions on what I would call its internal sites even though they're all public. The ones that are not massive and not wikipedia and the rest. That includes mediawiki.org, the meta wiki, various wiki media sites and so on. There's a few others. These are already somewhat in use. The two of these extensions, Semantic Media Wiki and Page Forms are in use somewhat on wiki tech, on translatewiki.net which is not technically wiki media but it's heavily affiliated or used by media wiki. There's a site called wiki apiary. I put it in italics because it is an independent site but my understanding is that it's going to be adopted by wiki media sometime in the future. In all these cases, potentially they could be making more use of the software. Just a brief history of Semantic Media Wiki especially because there's a lot of misconceptions or misunderstandings about it. This is an extension that was developed in 2005 mostly by Marcus Koch and Denny Vlennichich to German guys. The original vision was a way to tag data within wiki media. You can export it to the Semantic Web which was a big growing hot topic at the time. It still is but it doesn't have quite the buzz around it and also for internal query which is not to say that the Semantic Web is not in use. It is and we will get to that. Just for some interest, these are the two original logos that the first one is the second was in use for a while. Just to show that wikipedia was the focus of this software. It's basically the wikipedia logo but designed to look more structured and data like. The basic idea is you have a page called Germany and instead of just saying capital of Germany is Berlin, you put this tag around Berlin so that you semantically annotate the data and then what you're doing in Semantic Web terms is defining a triple of just a fact that Germany has capital of Berlin and then you can query that data internally using a parser function called ask and we'll get to some visualizations later. 2017 years later Semantic forms was released. I was the original author and that was the first Semantic media wiki spin off extension. Basically it lets you define forms to edit template calls. It I mean it caught on and I guess more importantly the vision the philosophy of it caught on which is instead of storing data with these this free form tagging you should put all the data within templates. So instead of a tag like that you would have an infobox template called country and you would put this tag within that and then you just go about using your infoboxes normally and you get for free this additional tagging and then of course you can query the data and so on. So in over the next few years it started getting usage and a whole bunch of other additional quote unquote Semantic extensions were created and some are still maintained quite a lot especially in those early years. Wikimedia developers remained dubious about Semantic media wiki for I'm not sure if this is still the case but for a very long time if you typed HSS that alias on the media wiki IRC channel you got this you pointed to this image I'm still not sure exactly what the thinking behind this was I guess there was some view that this was a kind of cult that you have to be careful not to get sucked into. We had the first conference in 2010 we had some events earlier but this was the first official Semantic media wiki conference which became a bi-annual thing around 50 60 people per event on average the project got a new logo in 2010 which I'm just putting this here to reflect the changing understanding that this is not intended for Wikipedia this is its own software for use within the enterprise because that's where all the real action was made on Semantic media wiki then in 2012 wiki data launched again it was founded by essentially by those same two people Danny and Marcus this was basically the second attempt at a Semantic Wikipedia and they did it what I would call the right way this time I mean there are few things that it has over Semantic media wiki as far as usage on Wikipedia probably the single biggest thing is that there's a single point of data entry so that you can query the data and display the data within all languages translation is built in whereas Semantic media wiki was just you know just naive tagging within a specific language skip right fast forward various other things happen during this time of course but fast forward to 2015 we released an extension called cargo which is meant to be a lightweight alternative to Semantic media wiki it actually it actually replicates not just Semantic media wiki but about five other extensions around Semantic media wiki so it's really a sort of all in one thing based on you know all the collective wisdom that had come up around Semantic media wiki and you know it's general usage then last year now this 2017 we had the first ever EMW con the North American SMW con events were renamed to the enterprise media wiki conference by the way a little plug the next event will be pretty soon in 2010 in Washington DC if anyone's interested in that I really mention it because I think it underscores the fact that these extensions what I guess you could call template data extensions extensions based around the idea of storing structured data within info box templates are at the heart of enterprise media wiki that's really the engine I guess that drives this alternate set of software and that's where the real action is around media wiki within the enterprise I mean you know just the fact that at this EMW con we had a lot of interest in these specific extensions and discussion of this specific kind of usage and then later that year very recently semantic forms, I renamed semantic forms to page forms to reflect the fact that it no longer requires semantic media wiki can now work with cargo or by itself so you can now have a small less semantic data focused wiki that doesn't have any semantic you know big S semantic extensions installed so that's the end of the history just a few more things about semantic media wiki versus cargo you know it's 10 years older 2005 versus 2015 it's better known right now and it has a lot more usage just over 10 years of cumulative usage and this just a small listing of I guess these are the big names around SMW but it's in use in a lot of organizations and companies NASA, NATO, World Bank Boeing, Phillips, GE and the wiki media foundation already so yeah I just want to make it clear I don't think there's anything wrong with semantic media wiki and I think if your name's not wikipedia or what I have here I think you as a wiki would benefit from having it installed but I do think cargo is the superior extension not to belabor the point because I guess some of you haven't heard of either extension but cargo is easier to install because it's one extension I said before that it replaces about five extensions but actually there's a bunch more library extensions that semantic media wiki requires which this one doesn't so it's just a lot easier to install and maintain the software for it's easier to set up data structures because everything is just defined within one template and we'll get to that later but it's more complicated for semantic media wiki it's easier for programmers on the developer side to maintain because it's code based it's smaller than semantic media wikis and it's more powerful I won't really get into the details of that but you can find documentation on that if you're interested on the cargo homepage basically what cargo does is instead of using those semantic web style triples it creates a standard database table where each call to a template defines a single row of that table so a field of a template is a column in that table and a single value within that template called will be a single cell within that table and hopefully this will make more sense to show some examples you have some parser functions here there's one to declare table which is these are all these all have SQL equivalents this is basically a wrapper around SQL so you have one that's basically like create table the store one inserts one row into the table and then you can query in a few different ways so not to get into the details here but basically you put everything in one template and you say I want to declare this table it might be too hard to see anyway but here's the name of the table the fields and their types and then and then you have the second one saying every time this template is called add one row to this table okay so and then finally you can query and again the details aren't that important but basically if you look at this it's a lot like an SQL select statement you specify which tables you want and which fields you want the conditions here's a more complicated one just showing that you can join on more than one table in this case you want to get all the cities in Europe so you want to find all the cities in a country where the country in turn is tagged or defined as being in Europe um and then there are various this is almost exactly like semantic media wakey but there are various ways to display or visualize the data it's just simple things like lists I'll get you examples of all these later but later we can do date based visualizations like calendars and timelines and maps and so forth other features it has you can do drill down browsing which I'll also show you but what's neat about that is that it's all generated automatically from the data so you don't have to define anything in order to get that interface it comes for free as soon as you create your data structure and populate the wakey and you can export the data via these standard data formats and various other things that I won't get to there's a few other little features you may be wondering what about security because you're essentially calling an sql call when you do that query thing could it be used to view data users are not supposed to or even do something more malicious like delete tables or drop star or whatever it is cargo has a lot of code to prevent sql injection but more importantly you can set it up to access a different database all together I mean by default it uses the main media wakey database so they have the amount of damage someone could do is minimal they could if this were poorly designed they could delete the data tables but then it would just be a matter of recreating those tables because the original wiki text where the data actually is stored it's not going to be affected again the other extension I want to talk about is page forms or talk about in detail it was known for most of its life as semantic forms I'm the main author although it's really gotten a ton of contributions from a lot of other people it's a framework for defining forms to edit template calls within pages also sections although that's kind of a minor detail but you can use it to pass by text areas for sections of a page pages that have a form associated with them get an edit with form tab right near the edit tab and those forms are defined via custom syntax in its own namespace which is form and there's just a lot of functionality there especially as far as the input types go all sorts of auto completion and standard stuff like text and drop downs and then you can have a date picker and an actual math based input to let you enter coordinates I don't know if you can see this at all but basically this is just a little example of what the form syntax can look like it's all just wiki text like text like and then yeah this just shows a bunch of a few field tags there's no additional parameters they're set but you can't set a lot of parameters for each field it tries to be smart so based on the cargo or some midi wiki data or whatever it is oftentimes it can figure out what the right input should be whether that's a drop down or a map input or anything else okay it's time for some demos how are we doing on time okay cool let's see so these are mostly going to be cargo examples in a few cases I'm using some midi wiki but it's the same thing there's just a lot more to choose from as far as some midi wiki right now here's just an example of a bulleted list here's an example of a table this is all generated from cargo data actually let me show the okay so here's a table and then here's a bar chart it's pretty cool and then if I go to just to show that this is legitimate if I go to the edit page here I can't find it in the interface you can see these are just calls to cargo query so all the data is coming from the back end let's see what else do we have here this is an outline this one's kind of not used that frequently but it's a way of showing an outline of data where the different levels have different relationships have a set of hierarchical relationships this is what's known as a dynamic table it's like a table but you can do more things you can sort you can show different amounts of values per page you can go from one page to the next and you can do a search within this is all done via a javascript library I didn't create any of this but here you can do a dynamic search it's a biomass I just show the ones that fit some dynamically entered value in here and all these it's really easy to set up all of these these are all just different formats that you specify within the cargo query call there's just an example of a gallery this again is just a query a gallery I think user images or something in this case this is a calendar this happens to be showing opinion items from newspapers and magazines this is a wiki iran but each of these links back to a page this is a timeline of events what else do we have here there were no really there were no good map examples for cargo there's a ton for some intermediate wiki this just happens to be a little demo I put up on my own wiki to show it can use google maps or open layers if you've heard of open layers that's like the open source slash open data of google maps and these are cities I guess that match different criteria and each one has has coordinates associated with it this is again from my wiki this is an example of the template format so lets you create a custom result format for each item so in this case because each one has a quote it does some nice formatting to show the quote and then the other information in a custom way and that's all done directly from within the wiki no coding is required just a bunch of wiki text yeah okay what do I want to show oh yeah okay well yeah as far as drill down here is an example this is again from my wiki I wish I could increase the font size here so these are all the opinion items there's 3,000 of them you can do a full text search and you can filter on various values so for instance I want to see all the ones from 2016 once I click on that okay yeah so now there are few enough values here that I can see them all individually and once I click on a year for instance it automatically sets the time period to be a month yeah it just tries to be smart about it basically I mean these are the kind of things that you would need regardless of the type of data so here let me click on Washington Post now and now it will be down to 2 2 items those are both about Puerto Rico so let me just randomly do a search a text search on the word Rico and it should show up well yeah okay so if you do a text search then it also shows what's known as the search context around the text within the page yeah so this can be this is drill down slash search which can be useful for just about any data set what else to do here's one that I didn't create same idea these values and does the right thing that's on yeah okay and then as far as page forms let me click on one of these pages I have here the whole all the data and then if I click on edit with form okay I'm not logged in but I can still edit I think there's a lot of there's a lot of built in functionality this thing lets you add additional authors because an opinion piece might have more than one author so if I start typing here I can specify well alright I'll keep it that big I can specify additional authors easily this is this is like a combo box drop down so I can choose other things there's a date input all this other stuff for more complex for more complex data you can actually have a table of data within the page so here each the set of opinions is really a table because each of these opinions has fields associated with it so I can add additional rows to this table and again all the all the all the functionality will be here this is pretty cool actually once you specify one value here it narrows down the options for I don't know if you remember this this was like 2006 a 2006 scandal it narrows down the set of options so there's like conditional a lot of completion you can do that kind of stuff yeah that's it and then you have your free text box that you can use for anything that doesn't fit within that structure data and this text area could be a lot bigger if there were more need for free text and then you just you have all the same inputs the standard inputs here like saving page and previewing and all that and let me preview it this might not work okay it worked yeah this is what the page would look like if I hit save yeah so that's how page forms work and let me just briefly show what the form definition looks like and that'll be the the end of the demo does anyone have any questions about this anything I've shown I search box and I search trigger on all the apps that I got for results you got where there are more results in one or the other or more in the other right box that may be because they're not all in that category when you do a search within the drill down interface it only shows the pages within that category slash cargo table so I don't know if you were doing a search within items or anything else yeah that may be it I mean it uses the same search is that true well it uses the really basic MySQL search but I think this this wiki uses the same thing because it doesn't have any cool search functionality like elastic search so that is probably the reason but I don't know for sure any other questions here is the here's what the form definition looks like you can ignore all this stuff basically it sort of looks like wiki text but with this other additional tagging and then the page forms extension takes care of parsing this on the fly and displaying a form and also parsing the page itself oh I didn't show you oh no I didn't the form can be used both for creating new pages and editing existing ones okay so in the context of the wiki media foundation what could these extensions be used for um one big thing is storing info box data on media wiki.org obviously for extensions potentially for skins also and who knows what else I think that would greatly improve both ease of data entry and finability yeah I mean media wiki.org has a real problem I think I mean you know simple question is what are the media spam extensions that work with media wiki 1.26 I have my own wiki it's on 1.26 and I want to I want to you know block spam what can I do that's sort of a rhetorical question because I mean there is an answer but it's a pain to find you basically have to go through each page you first have to find the category called spam extensions or whatever it's called if you even know to look for such a category and then you have to go through each one and see what it's maintained at all and what it's what's the word compatibility is having cargo installed and with that drill down interface would make that a lot easier in my opinion and then in turn the fact that you have that greater findability would encourage people to better maintain the info box data and potentially even expand it to add more fields to make it easier to use to increase the amount of data available and then of course if people did want to maintain info boxes better the page forms extension would make that easier to do oh the one thing I wanted to say about this I forgot to note it but there's all those translate tags within at least some info box calls within mediawiki.org the simple the basic answer is those would just show up in form fields so it wouldn't make the UI any worse the longer answers potentially could be replaced in some cases if you have a set amount of options for a certain field like a drop down or something then potentially you wouldn't need to do manual translation you could just have people choose the value within the English language one and then the template would automatically translate that into whatever language it was with a hard coded set of translations yeah displaying events in calendars there was a big discussion about this maybe six months ago or something the problem is there's no or maybe a year ago there's no there's no single calendar display of all wiki media events or media wiki related events it's just you know it's located in a bunch of different places mostly wikis with some fabricator and so forth this would be fairly trivial to accomplish with these extensions because you just create the form have a template for events and then you know as people add in the data then you can have one or more calendar pages to display all that information potentially you could even have well you could definitely have calendars matching different criteria like only development related events or wikipedia related events and so forth potentially you could even have personalized calendars all sorts of possibilities once there's tagging of that data tied in with this sort of is event and conference management I don't know if any of you have tried submitting an event for wiki mania browsing through submitted events but it's just copy and pasting it's really quite primitive and same for viewing and voting and all of that well forget I said voting that's a whole separate but potentially you could use form for voting too but that's really a whole separate discussion it could potentially also be used for events like this one the fabricator interface is somewhat clunky but this summit has sort of a task management feel to it project management so fabricator might still make more sense in any case but that is another option and then finally replacing custom development I really I don't want to single out this one thing but it's what happened to be familiar with this is a project called the wikipedia library card platform the goal is what stated as well the goal still is to create a site where users can request temporary access to online information sources I don't know Lexis nexus I don't know various ones where there's a limited amount of viewership allows and then you can accept or reject it's fairly basic stuff just in terms of the high level goals the project was supposed to be finished in about a year ago after five months right now it's up I don't know the URL still has tested it so I'm guessing it's not fully done so it's about a year behind schedule and I really I've hesitated to include this because I really don't want to to I'm not trying to imply anything about the people involved I know personal experience that creating usable software from scratch is very hard three minutes 30 minutes you know I've been working on page forms extension for over 10 years now so yeah creating usable software is hard and maintaining that software is hard as well it's my opinion that if they had gone with some combination of these extensions there would have been some custom development required probably but I think it would have been completed in a few months as opposed to you know in a year and a half and and you know I don't mean to harp on this one but I do hope that people will consider using something like this the next time there the next time it's the next time a custom solution is being considered okay so other potentially useful extensions there's one called widgets which lets you define what they what are called widgets little bits of HTML and JavaScript it's a lot like Lua modules if you know those but they're only editable by admins because JavaScript is inherently unsafe but it works great I mean there's a whole bunch of predefined widgets already and potentially within wiki media sites that can be used for displaying YouTube videos or other kinds of videos or like buttons that sort of thing there's another one called approved revs it's a deliberate lightweight alternative to the I guess better known extension called flag revs which is also the known as pending changes on wikipedia it just has a much simpler interface I mean flag revs is pretty heavy duty no matter how you look at it I think there's a whole assessment interface of quality and content and so forth here you just set one revision of the page as the approved one and then that's what people see by default and it's used a lot on smaller wikis and it may be useful for media wiki.org or meta or any place where flag revs used or flag revs was considered and rejected because it was too complicated there's another one called external data it can get data from various sources like APIs on the web or databases or local files you have to configure those it won't just go into your local directory and get anything and then you can display that data on the screen so potentially you could get data from fabricator or from garret or even from other wikis reducing you know reducing data redundancy or just making things easier to view at a glance I don't know what fabricators API is like but potentially that's something that could be done this is something that's come up why not use wiki data wiki data is so great let's just use that instead of bothering with all this other software it is great the idea is all your extension information you just put it all on wiki data and then and then you know display it on media wik.org or wherever else that's not what wiki data is intended for it's supposed to be a repository for general knowledge not for just dump your data on here so the basic problem is there's no guarantee that wiki data will accept all of this data or any of the data the other big problem is that it's just not nearly as nice a solution editing data but it's doable it's just a lot harder than simple form for the page to query the data you need to know sparkle and understand all this other stuff I'm not even sure if you can actually display the query data there's probably a way to do it or there probably will be a way to do it in the future there's definitely no way to browse or drill down on the data in the manner of the interfaces I showed before I'm not sure if anyone had this idea but potentially someone might say well okay we can't use wiki data but we can set up our own site that's like wiki data using the same software which is called wiki base but again you'll have the the data acceptance problem is solved but all the rest of the problems are still there as far as the difficult to use interface so yeah I think that's it that's my that's my spela yeah I don't know if anyone has any thoughts or comments I can also show more demos if anyone's not convinced yet I don't know I mean would yeah sorry go ahead I was involved with it was for managing ISO 9000 there were two big problems those were the people wanting security they were going to protect individual pages and I think that's a big thing in the enterprise world people sort of open to internet being editable by anyone but that's always a big issue and having to have separate wiki can be a bit of a headache yeah just to repeat it I don't know if I'm any more audible because I don't have a microphone the question was I guess about pushback from managers as far as security of just putting data on wiki in the first place I guess and then especially being able to put it on reviewing before publication right okay yeah yeah yeah yeah reviewing even edits that aren't malicious might still you know people still want to maybe give the okay on it before it goes live and then the other the other issue was multiple wikis um yeah well I guess that's just more about you know if you go to policy you can spread it across multiple wikis right yeah okay um yeah so for the first thing for security slash review I think the approved review is a good solution for that let me show an example here this is again on my wiki that I just happened to know that this works if I go here to the main page it says at the top here this is the approved revision of this page as well as being the most recent um but actually let me go to let me go to a page that where the approved revision is not the most recent um here's the the interface here okay all pages pages whose approved revision is not their latest this is just due to random testing and so forth okay fine so we have here a page about some opinion item uh it says this is the approved revision of this pages it's not the most recent view the most recent revision if I click on that I can see the live thing oh yeah okay I did this as a test at some point um yeah so you know somebody put in this malicious comment this is the worst case scenario um and uh but then because it had an approved revision no one will see it until it's approved and you can you can change the setting so that people don't see this little text message here then if I if I click on view history uh I guess because I'm not logged in as an administrator I can't uh I can't set the uh approval you can see the star here next to the approved revision if I were logged in I'd have administrator right so I could say I want to approve this or conversely I would undo this uh this uh edit um yeah so it's a it's a nice way to um to uh allow editing and uh you know not keep it closed off but but it would still with that with some confidence that uh get something you don't want on the page um and then as far as um as far as multiple wikis uh there there are ways to uh display data to query data from across multiple wikis uh if it's you know semantically annotated via semantic media wiki or cargo using the export functionality uh yeah and then you can use the external data extension to retrieve that potentially you can also you can then re-disk re-store that data within the central wiki or something it can get a little complicated but it's doable and I guess depends on the situation um yeah but that's that's that's that's uh potentially a big use of external data is to have like uh central hubs showing information from across a lot of different wikis and other sources yeah question Do you have any examples of some process which takes five steps? um um yeah that's interesting that the workflow is it's probably the thing that that could use the most uh attention that could use the most additional development on there's pretty basic workflow right now there's an extension that will let you get notified actually there's a few of them that will let you get notified you know you can specify if this field changes or if this field changes to this specific value then email the following people uh but there's nothing uh there's nothing there's no generic solution right now that says uh if if this data changes then do this other thing um it has to be a sort of a question done I mean that's a big thing on on wikipedia as well I mean anyone who's tried to nominate a page for deletion you know is that it's just uh uh it's it's a pain um yeah it would be great to have some kind of more generic solution for that kind of thing um um it's hard to I think it to some extent it's hard to have a solution to that that's generic um that's not really complicated to uh well first to develop and second to configure plus there hasn't really um the uh the requests for that kind of thing have been rather specific I think um I'm not sure that there's a uh a good easy to use generic solution for that but I could be wrong I may require having like a custom scripting language or something that's like it has ips and so forth but I don't I got Chris um so in past lives I've used many of the extensions you mentioned here extra wikis and the idea of bringing them to onto wikis that are in the directly to the wiki movement is an intriguing proposal um I think you have security concerns performance concerns what areas might benefit from using it is the next logical step then to take this to community on those wikis and say hey here's another proposal and say I guess that's something that a question that I would have for the audience is I really I don't know what the right steps are I don't know who really are the decision makers for any of these wikis um I don't know what the request process is so I guess I'll throw it out there but by the way I didn't actually mention performance you're too kind in thinking that I did but performance is fine actually cargo's performance is superior to some I didn't mention that but it's about for standard queries it's the testing I did show that it was about it took about 2 thirds the time it was about 33% faster yeah performance is fine and if you use if you use an alternate database then it won't affect the running of the main wiki even if the queries are slow yeah how many people in this room are interested in seeing the semantic media wiki and on media wiki.org that's a great question how many people would be interested in seeing some combination of these extensions on media wiki.org how many people in this room are interested in having the extension semantic media wiki installed for media wiki.org semantic media wiki or cargo or cargo and and perhaps paint forms yes so I would suggest that the next step in is to start a discussion on media wiki.org on the current issues flow then and then simultaneously perhaps open a task on fabric so I would just by not by saying what you do is because today is the coolest that you can install it as part of a specific thing so I want to improve how extensions are searched now that's a concrete task and we can identify that made from the cargo can improve that in certain ways that's my better protection than just technology so adding on to that there are all these things that developers could develop custom things for media wiki to a custom extension that would help organize all these extension information but here is something that is more flexible that could be done and that could be done now but yeah I want to know how its improved the way it was designed for the same bright period and I want to know how it simplewise or improves what we currently have as a useable sanity check for sure if this is actually a useful thing so can it solve some of the problems of our current solve if it can't do that then it will be hard to solve but it really can't do that maybe a better example I told people to keep that. The then is on nature might be a better example. The then's on man. Because the then's on nature is just random. There is an event standpoint to this. So it's a sort of... We're going to look at the calendar. ICS please. You know, but... There's something quite unusual. There's usually a throwaway way more less than the can for our average. Yeah, there was a big can. So the barrier entry for installing something there is really low. If you show it, it's actually a bit... So how do you just get... What do you do? Well, talking about this could be used to run the... The wishes quite easily. Instead of the collection of the box. In my experience, that's how things get done and how things get adopted is you start out the load back to where you're aging and see what happens. It grows with people who have good experiences and then they find some other tasks they like to use it for and it gets installed because of that task. Just... I think when you mention the wish list and community wish list, that's a perfect thing to figure out how I'm tying into that. And I'd say, you know, I think there's a lot of interest in getting more demo rules, right? So, to the extent that you can show that you can replace something else that you can replace, that you can install a hardware again and get rid of some of that. And it's more... Everyone is worried that we're going to have... If you can show that this replaces 15 inches and you get all re-installed and gets better features for X, Y, and Q states So, the next question and who's going to start the discussion and who's going to open an applicator against? Because we can all talk about this. Well, okay. Two things. I don't want to put anyone on the spot, but I'm actually quite curious about this. For people who didn't raise their hands and I really don't want to make anyone uncomfortable. But I'm curious what are the objections? Because I know there are objections, but I rarely hear them. So, I'd like to know what... Or for people who just aren't sure, what are the potential hesitations? Yeah? Well, from your own talk, I think there is a strong storing data inline. Data has data someplace else. Wicked data is the example, but even within the code race is like using JSON on JSS. So, I think... I think there is some general reluctance if you want to continue mixing data. Well, I think... I think it depends on what the intent of a particular piece is and not all of these are the same, obviously. So, the use case of media that we're talking specifically about the extensions in the same space. The extensions pages, by their very nature, are crying out for data. You know, and it doesn't... I think it doesn't ever be used in lots of places, like when you get the answer. Yes, it belongs there in that wiki as opposed to in some other source that can be cleared rather than this. The difference, and you know, we've developed a lot of wikis that users meant to be with your product. And the use cases for our wikis for our customers are generally they're building up the purpose of information about a product where you want to have the data embedded on the page. So, the examples of graphing in the previous talk, you made a road where the data was involved with comments or something that you'd carry from their display within the wiki. That's a very different use case from where your data becomes essentially a huge amount of no-single data by itself. If you use data, especially if you're using page forms that will allow you to edit and have that data tag done automatically as a relatively... I mean, use it to fill some form and it saves the page. So, each page you can save it's more data. Just as a devil's eye, why haven't our products actually protected them? I think what you're talking about is a feature that doesn't exist yet associated JSON with the wiki page instead of the infobox data. That's what you're talking about. I'm not a strong reader. I'm just trying to observe. This is my understanding. One of the differences is the opinion about where the data goes. And then how the data should be stored should be human readable first and I can always edit the wiki text and potentially just put it on my table. But, you know, there's a thing that gives you feedback or do I want to enforce a strict separation data is data and content is content and there's no way to edit the page to improve the data because... And we typically, for our non-SysOps there's a way in page forms to have it so that you only get the only edit tab is the tab that allows you to edit by form. And so you don't have people going in to edit the page. I'll just present the other example. I'm still really interested in whether or not this can be generalized for our work. That's another example where the foundation started this project to do with our information and take this process which is entirely content-based and replace it with this super, entirely data-based with no manual. And that didn't work. I don't think that question is settled. I think there is the option for saying some of these things or maybe this is a kind of growing, as I get more data and initially my data is stored in one of the text instead of talking, then I add some kind of extension like characters that can be made in structure and then maybe in the future when it gets reused in more places. I think there's a room for a spectrum but I want to see how they interoperate better. So, Anne, to your work-filled question we have two movies that we used that were created in part as Nefra Kira thought through and in part because we really needed something relatively low to manage what we were doing. So one is our planning with me and we set up the namespace of texts and each has a status that we use to sort the data and it's queryable and we can look at tasks with the date range and what are the statuses. So we use a wiki. Every page in that wiki in the task namespace is a task in that status and it states that people associate with that and we can slice and dice the areas in different ways to see how things progress. The other is a process wiki and we use that to manage multiple wikis on it. We want to manage the process of installing extensions across our wiki parts and staging extensions through the testing process to our production wiki parts and we use the process wiki which is also a part of this we can close it down using saved forms to manage that process. So it's very easy and we're able to create each of these wikis within a few hours and then of course they evolve over time so it's great to add additional fields to the page, to the form to our data model and we can survive at this point that that is very easy for us to create within a few hours and are enough to be able to really manage our workflow. Can you explain a little bit the wikis you are using these are enterprise which is like probably separate or which wikis are they? So we have our work and we also are developing wikis for customers in the government so we have a variety so we have different platforms where we host different wikis to talk to the security question some of our wikis are closed wikis that we need to authenticate to in order to be able to view the content of all of our wikis within an enterprise the benefits of attribution are much higher and the reputational aspects of it. So we have a variety of different types of wikis for ourselves as well as wikis we've developed for customers and the security topic that you mentioned before that's really completely orthogonal to you know whether you're using Samantha, Makita, Carter or whatever in an enterprise you know either you're okay with people seeing the content or not and if you're not okay with everybody then you restrict your set of users that can read it and you have to log in in order to be able to read but having per page access control really is contrary to the law did that answer your question? I guess we're mostly out of time in fact I'm just the final point is I don't think any of these extensions are really usable on wikipedia for workflow management or anything else with possible exception of external data but it just seems to me like there's a you know there's this bad setups right now in place on mewiki.org and others and this would be a way to add structure to all this data in a way that would be easy to remove at a later time if it ever came to that so it's a lot lower risk I think than any other solution to this problem of data entry and findability