 OK, first thing, what is a Gnusda Publication Project? The Gnusda Publication Project, you see the URL at non-Gnusda.org. It's a reasonably old project, which has been around since 2002. Currently, it's under the process of being conditioned to a full Gnusda Project, after the question of Richard's climate himself. So you heard me and I'm proud of that. And we are making certain things out. Probably we will split up the project in two, and decide which part go into a full-blown Gnu project, and which will remain on Gnu, which still means it's open source project, just not under the Gnu umbrella. And probably we need to change the name. You can find us sometimes on our IFC channel, which is just because we didn't like the noise on the Gnu step channel. Now the Gnu step channel is so silent that we don't have the noise problem anymore, but it's useful because people sometimes expect to ask questions, and if they find us on how they return to the line. But in fact, the goal of the Gnu step application project is to fill the gap which Gnu step has of lack of applications. Because Gnu step has many users. So it's such an extensive framework. Some people just used to develop for it and target a specific platform. But actually, since it's an open step child, the goal of it is to have a complete test of environment with it. Since this is not the explicit goal of the Gnu step itself, which provides Gnu step project, which provides the main libraries and a couple of useful application and developer tools, the rest is to be supplied by other ones. Since picking out applications from different projects is always a mess and confusing for users, because you never know what to get things. Projects started to cut up and give up consistent environment. I don't know if we are the oldest of the background is older than us. I just know that that one is almost old, and so older, but I think it was a funny death. Yes. Officially, it's not there, if you ask them. But they didn't release something for two years or so. So essentially, we can express that the Gnu step class gap gives you our own space. A bit so amazing we don't have a window manager and some tools are missing, but they should have a usable environment. What didn't go up? The project originally started as a porting repository. So Gregory John Casamento, which is currently the Gnu step maintainer and project leader, he had the idea we would look for open step code and port it to Gnu step. Theoretically, it should actually compile. It's not that way, but what we discovered is that open source wasn't around then. It was not going down. So there is no more source of many of the three open step applications, which is sad. Some source code is lying around, but the authors are not willing to give it away. So we know a couple of applications which we do want and which we don't get, because the authors say, oh, the code is messy. I don't want to release it. No way, Gregory. Many times, just give it to us and we'll clean it up away. And finally, so this project is a bit on a dead end. So if something comes up, we will do it. But currently, it has been lying around. Of course, the application we will have, like a time monitor, which is a CPU monitor, has been ported, it's maintained. And so now we have the Macintosh code. So the applications come from that side. Sometimes they are a ban on project for Mac, not maintained anymore. We can just rescue them for them. So this is the other point here, rescued by other project. Sometimes people start a project, code a bit, and then we remain there. This is typical of open source. We have many examples. Sometimes these applications are fine, or even depend on them. And you don't have a place to download them. Nobody even patches them for, for example, trivial API changes. We have no steps. So essentially, it's a bit top, which is a pity. So we try to cut them here, and we have a couple of them. And we at least maintain them functional and improve when we can. So this is, I think, important for the end user experience. Because in one place, you cannot load them. And the applications, especially when we write an application, we will put in here. Like I said, I have not started any extra product anymore. Everything which I have to write, I put it in gap. So what's been outside for historical reason or anything there, application, utilities, very much everything which is needed. The goal is not to make something which has a lot of kits, where we depend on, so you essentially need to install the whole project, which is a bit more required. We try to be more lightweight. This is maybe not so exciting. We don't have so many extra features. But the idea is that Knustep actually has enough feature API features that you can write complete application. If we need a library, we will put in here. So complete this. Key people. Gregory Jones, as I mentioned, needs to be mentioned, because he is a founder in the project he does. So he needs to know, he is in the space and put in countries here. So tenders. And he will be disappointed with his body. Then it's me, I'm a co-leader. And I wrote most of the new applications, because Greg is more than 14. And I wrote original applications, which is like this Gucci, FTP, database, and battery monitor in other tools. Then there is Adam Fedor, which is a former Knustep project leader, and a gifted programmer. He's a five-person, but he actually maintains a picture of them, which is a very nice and cute application. Other people. Some are past committers, which disappear into nothing. Some are not very active. They can report, can attend, and they appear for six months. And of course, all the original authors, which I don't mention here, they're the authors. Not a long list of active developers. And for the first application, the Panamagica, it's an image viewer tool. It was inspired by xv. So you just open it. If you open an image, you have a list of images. You can click. It's in a single window. So if you have a list of images, make a slideshow in the window, or you can make a full screen. And it's currently, as far as I know, the only full-screen application in GNU-STEP, which is many images, and it works both on the Mac and on GNU-STEP. Because the main goal we define about one year ago is to try to port every application in both on Mac and GNU-STEP, the demonstration that we can do there. If it makes sense, it's not, of course, a system-dependent application. We won't do that. And this one walks, even the full-screen port walks on longer than this. Can I? Come in, come in. We should have a screenshot here, but since we can see some of this application live afterwards. It's also a reference for people, for example, wanting to make full-screen port and scaling port. It should be a lightweight tool, I think, especially. And there's an FTP application. The interface is pretty standard. It's a two-paying, low-climbing mode. It tries to implement the complete FTP specification, which is hardly found in modern FTP clients. It has both active-passive and default modes. If you ever hit them, they do work. And there's a special note here. It has been ported both to the Macintosh and to Windows at one point. I wrote the whole core socket code so that it can work with Windows sockets. Because I used plain, unique way to do it, and it was destined to work in hardly on Windows. Even if it should, it didn't work. No way. Now it works. It has some if-desk, but this application actually works on the Mac. On Linux, Unix, Solaris, and Windows. This is a major demonstration of the power of the core kit. It may not be perfect, but it's a useful tool. And I think they only maintain the FTP application. There were others, but where they are now, I don't know. I use it daily. So for what I use it works, it still has some shortcomings. This is a very nice application. We are playing with that before the beginning. It's a vector, it's an application. I need to credit the original author, which is Enrico Cessale. He's a very politic Moonstep author. He wrote the World War Space, System Preferences, and other utilities. And he started this application eight years ago. I think it's from 2000. It was called GDO. It was written when Moonstep was still limited. It was very open-step-specific, post-coded code. And the application just remained in beta for eight years. It wasn't even completely functional with that updated Moonstep. And completely respected code. So it has a code class here. It's clean now. You can expand it. All the comments were written using the ZEPA. So it can also work on Poco, because Poco doesn't have displayed post-coded anymore. This is now portable code. The performance is good. Nice. It's really, it's a similar tool to this medium. For example, you can't do all circles, because what's in there, it was a portable original. So it did not get extended. I made it work. What I'm doing, I made a box walk. What's was to make a box walk in a clean way, and into the store, all the editors. You can save the format. It's compatible with the one format, even if I don't think there are many ties around. Of course, you can print. It's a document-oriented application. I think it's a nice application with a reasonable potential, it still needs some work. It's not official now, so you don't have to make yet a release of the new version. This inner space, this is a regular, personal toy. This is black space implementation. So essentially, you have a screen saver running on your desk, but maybe it's not very useful, but it's nice. It was a GT open-step application, and now you can use that. You can animate back drops, and the modules are compatible with the original version, but there are a couple of modules. Not all are included, because not every module has exactly a license, because people drop them in open source without giving it an open source license. So we can't include them here. And this is one of the problems that we might be able to move part of, probably, to take them out, or leave them somewhere else. Of course, it's easy to write new ones. Pictureframe is the address, the catcher. It's called Pictureframe, because the ultimate goal is to make a digital pictureframe. So it takes a screen, uses an old laptop, and made a nice one mount to the player. And it looks very nice. On the website, there are pictures of it after you're walking. This is actually a screenshot. And she did some nice blending code. It uses alpha channel, and it can display additional information, like how time, when you shoot the picture, whether it's, of course, intended to run unattended as a calculator. It uses, of course, a vast quality notation, because this is what I use. HP calculators are just superior, and I made it one. I was totally knuckling a real scientific calculator. It's not being elated under this name, but you could look for it with an old name, which is LPN Guy or something, but it's the same code. I got a change of license from BSD to GPF, because somebody asked that. I think it was Guy, or if he has a little better memory, he's going to be GPF, and BSD is too time-considered. I like BSD, but no, they're not sure. It has a couple of shortcomings, not as much in the Mac version, but in the Newster version, but it's really, really minor thing, because it's the same. There's only all the kinds you swap. There is only one thing which bugs me about this kind of one that's pretty obvious. There is a mathematician who can have a discussion on how to implement the exponential function for non-integer numbers, but I don't think it's interesting. It's just a personal problem. This is maybe the fastest class of application lately, and probably the iPhone, because the rest of the code is still unreleased. Because it's changing, it's overalder. As we discussed for years, we like to have an active Newster browser environment for two reasons. First, because our browser is expensive to have. Second, because it acts as a testbed for the underlying HTML framework, which can be used, of course, outside of the browser, because it's everywhere when you want to display hypertext. In this case, we use simple WebKit, which is a complete implementation in pure Objective-C of the WebKit API format. So it's a complete scratch on the item, which is, of course, a major goal. But it has many advantages, so the performance is cleaner. It's not Objective-C++, so it can run from GCC to an active-fibre app, which means it can buy some embedded devices on different platforms, so we can support a much wider audience into them, that's all. So it's nice. Since it's WebKit-compatible, on the Mac, you can build the browser. Again, simple WebKit on Poco, or even WebKit inside. So this helps, of course, debugging. And we can even display much better pages. The interface is extremely simple. Currently, it just has four buttons. History, but it has bookmarks, and it can read and write Safari bookmarks on the format. It is just convenient, because you can use it as a screenshot. So instead of page each other, it's already reasonably walky, non-substantial compliant. It's coming up. Not to speak about the browser itself, we can play to present a simple WebKit. They want extra additional information. Then we have terminal original answer from Backbone. So here we actually made use of the open source concept. We can take other GPA code and put it in our project. Because essentially, Backbone wasn't maintaining it correctly. They did not accept some patches. It was not affordable. And so at the end, I just put it in. I asked the author many times, and it was just not progressing as we wish. So we made compile the latest new step to O-Version, which was a major shift in Mac-wise. We added I-Version, I added code to make it compile in the platforms in the project. The emulation is based on the Linux kind of emulation. So for Linux machines, it's very convenient, because it's extremely comfortable for us. But it uses codes which were not available outside Linux. And I wanted to place them for this course, so I put them in Solaris. I know we have a user which uses them in Solaris. It has been a long requested function. It's a multi-window 10 minutes. It's fun to 10 minutes. It's usable, and it works very well. Others framework, which is sadly to say a rescue project, because the developer was brilliant, but I think he got busy with other projects. It's an implementation of the address framework of Apple. So API-wide is the same, essentially. And it has a front end. So the main use is, of course, the address book, but it provides the other services for other applications and new menus. So there is one major application which is framework. And this is also the most important reason why it needs to be kept up to date, because other alignment, I think it's very convenient, especially on Android devices. You want to have your addresses ready. Small utility for who uses laptops, battery monitor, time monitor, remote desk. These are small applications. I'm going to check the whole page. Battery monitor just uses your SDPI to display your battery status and some battery information. Time monitor is a CPU monitor. Remote desk is a front end to a desktop. So if you need to connect to Windows camera services, what's convenient, the time it doesn't set up online. It seems better. Flexi-sheet. The support. This was an application already written for Coco. So it's open source, a very long source watch. But the project's told me on the auto. And display. We wrote him. We are able to contact him. And we have the office of permission to put him in our project with SDPI. So it's fine. It's a quantity-stack spreadsheet limitation If you know one flex, it was a special kind of spreadsheet written for OpenStep. At the time, it was a major or sort of lotus input. They had the same concept to complete the spreadsheet. There is an innovative way of looking at spreadsheets. Let's see. Yeah, I should understand that it's not that clear. It's different for myself, because you can write that set here in columns, and each row is a Bucer data set. And then you can drag and create different views of the same data. So Excel eventually copied it to the Pivot table. But this one is actually more convenient. And the formulas are not written inside a cell, but they're global for the spreadsheet. Because so when you drag it out, and you change the view, essentially, you have a queue, and you are slicing it. It's a simple way of looking at a data queue. Can I call it a hypercube, more dimensionally? Essentially, it's a hypercube. And the formulas need to work on all the faces of a queue. This is why you can put them in a static place. Currently, the application is reasonably working on the map. Unfortunately, they auto-left it like a bitters page. So what we need to do is clean up the existing code for blacks, because there are some, even on the map. Identify the faulting difference between Cocoa and Knewstep, and either work them around in the FlexiSheet code itself, if it doesn't come to panels in Cocoa, or improve Knewstep so that it's a part of Cocoa features. We have a fairly good job, because the application started working here, which was six months ago, I wasn't thinkable, because not even the window was coming up. We are using complete nip, so we did not rewrite the interface. We did not change the code. So we just alumni file, and some details, but very few. So we had yesterday a presentation about portability. And this is a good example of how you can even rewrite the interface. Eventually, we have to do it to tweak some aspects and some fine tunes, but the general approach works quite well. The code itself has been made portable, because Mac programmers are quite sloppy. I've seen them on other projects. It wasn't the Mac that's fine. The Mac has one architecture. Now, OK, it has two, because it has Intel. At that time, it was just obviously, I think. So the code is absolutely reliant on a specific way, the API works, the processor works, the operating system works. So it needs to be made portable. Because currently, for example, it works reasonably on Intel, on Linux, but we tried on other platforms. We made it tweak. And at the end, everything is done. We make a release, and we can completely miss features, because they are missing features and expand applications. I think it's a major application, which is interesting, because there are currently no other open source at China Pixel, so if somebody is interested in this kind of spreadsheet model. Another portable project is Bing. It's called a simple working process on this project is called Gregory. This tool is a co-op board. We took the last version of Bing, which runs on 10.4 map, and called the code and put it in here. And this goes in the title down the wall, because it doesn't even have a main track. We used directly a PBX bit to build the X code. So it's pretty much the extreme way we do it. Completely unmanaged. Yes. It has an if-desk in the port, currently some of them can be removed. And this complexity really prompted a lot of work. Minor and major include the 2B protocol. And it's almost usable now, so it's very nice. There are still some bugs, which prevent it from daily usage and save problems with the FTF panels if you have images and X, but it will be walked out. And this is too easy. If you don't want to use text editor here, for example. Very simple. The screenshot is the tool that I guess might be used. The main will get interrogated from the man. Of course, they are separators, which makes no sense in the rest. But it works, so it's very good. The good thing is, before you have the same application on the Mac, and on GnuStep, which is the core of the GnuStep application project, you'll be able to provide this. It's much easier to exchange data. And you may have a Mac laptop and GnuStep workstation on the other way around. We have a couple of games, which are written by a gifted programmer and maintain them in the end of the speech on the GnuStep games. You can just play with games in the end of the speech. But the future of the GnuStep application project, we have another couple of applications we want to implement. They are actually had written, currently, but not maintained, because we shoot the tool. The last project, we have, like, TaxiSheet and Nihar, and Best Putschia, are consuming all of it because it's so big. And also, because this application I'm not so blocking, but it would be nice to have a login, which would be one way to implement it would be to be compatible with XQS Player Manager, for example, so it wouldn't be mainly unique, super simple, accessible for all of the booking. We are thinking about thinking starter for the end user, so if you don't want to use your distribution packages, which is one way to do it, we want to have an installer which just plays like the other one, but a bit more advanced because it should be able to handle different architectures. So the best thing we have our draft for that is to have a single travel, which contains, for example, Solaris, Linux, BSD, PowerPC, Intel, and then an installer which is for the point one. Dependency tracking and uninstalling our plans. Apple doesn't have them, we don't have them. Yeah, we have to do the pushup, or pushup. Some people say it's not useful, because maybe you want to have an interaction with the distribution installer that depends. Either use completely that one and the using installer, or if you have a clean environment, theoretically, you can just, if you just have genus that you just can use or install, you can install the system and mean how you use it on the Mac. I think it depends on the user. If the user wants to have a GNOME environment, or maybe Ubuntu, or what kind. It depends if it's a mixed environment or not. It's like if the user decides to use some environment and some packaging tool, maybe you should provide our programs for those packages tools. And somebody who wants to use a pure, pure, twisted environment can then use GNOME. It's not a goal not to provide packages for other distributions, if you want to set a kind of environment. And there's another tool which is an app that configs. The idea would be to have a tool to configure and mount an app that's just a convenience. Other applications of the same kind flow in the air, like system administration tools and systems here. So not portable applications, but which are convenient. So a bunch of tools to administer users, administer your network cards, so things you expect from an environment. We're very tall, but first we make amounts of things about the editor, but sometimes it's handy to have a web editor for writing documentation. So something lightweight, we made it something really lightweight, where you write FTI to just write your titles. Not a complete CSS, not a real irreplacement, but a handy tool. Just to avoid the landing machine error. Yes, exactly, because often the goal is to have HTML in widespread use. It's a job effort to do, so you have documentations, program head, notes, in HTML. So that's what it is, tedious. Some people request an instant messaging application. Some people want Java, some people want Microsoft Messenger, Yaku, ICQ, and you can't name them. There are so many protocols. Even next projects have paid for that. So at the start, they work for a bit. They become uncomfortable, because they depend on some libraries, as libraries are not contained anymore. So there's a new one. It's a Moot area, and pretty people have started to work for it. More obvious of these applications. We have a big failure over Scratch it, but maybe not enough. But why do we have this with us, of course, is a big task. We have a Lighthouse application for Sun. We will never get them. We will never get them. Some people still dream of that, but that's not going to happen. So maybe that's not a passion, but if GNOME step gets more less documented, that might be a point to work on, because I think somehow people expect it. I mean, KDE has KO is in GNOME. You have GNOME office, because you have GNOME marriage, of course, there is open office, which is alien to everything. So everybody has an idea about it. And we'll see if other developers come in, they might bring your ideas on. Dictionary tool, because it's just a way to. There was one under OpenStep. We might think there is an OpenDictionary format. We were also thinking about something which integrates with Wikipedia, for example. Just something where you can get a look up in that space. The digital website application. You're thinking about the digital website application? Yes, it existed. There is already a translation dictionary by Robert Burns. So there are ideas for things around it. So of course, neither pressing, neither unessential. Nobody can do it. I mean, if you have it, you need to look up on the water. That's it, essentially, as far as gap goes. OK, you had the eyes on the list, which is an amazing location. I can present that in your presence. I have also a couple of sites on Simple Electric, which might be of more interest if you are on the web. May I ask some questions? Yes, of course. OK. How about a license? You said that you changed one project license from BST to... I can do that because I'm the original author. Yeah, of course. But why that? And is it OK to put a BST license project into the gap? Theoretically, it's OK because Sarana just requires an official open source license, which PSP is. But if we have the political change license, we prefer GPM and GPM. OK, that was decided by Gregorian. Essentially, I agree with it, because it's clear that way. And it's competent with Githniusdip itself, since we've already had license problems in the past. It's better to avoid them at the beginning. Next question. What about Knoome app? It looks like it's abandoned. Nothing changed over a long time right now. Do you know what's up with the author? The author is still alive. He says he has patches. If you report problems, he has an attitude that's saying, always a thing, no step problem, which is not always true. So it's a bit of a game. Whose fault is it? And I don't like his attitude in this case. It's a brilliant application. I think it's one of the most complete applications we have for Knoome app. I used to use it every day, all the time, under Knoome's. We don't have plans to import the app. We don't have the resources to track another big application. We have plans, actually, to rewrite the code. The whole group is hard-coded. The whole group is hard-coded. So on the Mac, the multiple users' nips and the new step fault uses hard-coding of GUI, which several applications do. Tamina is also completely hard-coded, because some author like that way. And at the end, we need one idea, gradually, of course, because it's a government-coded instance that everything should use to run. For Tamina, we already have the idea to write the interface to use to run, because it's normally never translated, because we are going to provide translations. Because some people request translations, which is of course correct. And with Go-on file, it's easy. Just provide another Go-on file, a language file. Do that with localized pain. You can do it. It's a useless pain, probably. And the same should be done for Knoome. But the idea is, if we do that, and we send him patches, what will he do with his patches? OK, he's not very... He's a positive, correct channel. He has a clear thing. He had provided 90 tablets on his website, so... Yes, they're not thought of, David, but it's much better than the last release. So it has features like better eyes. I don't know why the release never happened. Maybe you should poke them a little bit. We do that. Maybe I'd make myself two or whatever. Yes, maybe if it's always the same casting, poking in the world community is important, because I know that it has a wider usage here. The guy was contacted by Apple to pitch Apple Mail, and especially the mine thing, or Hunter Mine, to use the Apple. I don't know if that then it happened. That's a piece about the quality of his mine thing. OK. He told me that one. It's used by some colleagues of the mine, the pattern I'm playing on. Yeah, I do use it. Yes, but... Do you plan modernizing the interface of several applications, like adding the toolbars and stuff like that? Some only have the toolbars, some are incomplete. Yes, I know. It depends on applications. For some applications, definitely I'm going to... Definitely update the Grappos interface, which is updated and complete. And I want, first of all, to release like this, the original one, and then I plan to change the inspectors and update it to sort of be more coir-like in the loop, like use colorwares, liners, instead of... Do you know, in terms of the remarks, do you know Bookmarker, written by Marcus Miller? No. It's a small bookmarking application, which... What's the word in it, in English? It's heading German and... Provide and... Manage? Oh, yeah, which manages your bookmarks, and can open bookmarks from visiting the application to central place to manage all your bookmarks. And he ported it to Knuster once, but then he ended up... There's a use for it at Knuster, because there's no Knuster application who opens or sends them URLs. So maybe you should talk to him? We can... He's not around here at first, no? He was here yesterday. We can talk to him, because... I added an experimental support for services, if you know Mac, you know there are services. And on the Mac, you can already use the scoochie and say, open URL into... So it should be easily integratable. And I try to hack this on Knuster and it works. But you have a URL in mail, you can double-click on it, but you can select it and say... Which is a bookmark application. Maybe you could use... I don't know if he integrates, for example, with Safari. I don't know how he does it, but we can talk. My feeling is that it's competent with the Safari one. There's a sensor... A little bit given up on it too, because there's not much time for it, whatever. We could ask him for it. If it works on the Mac and it's... One of the key words of this integration, so the more we integrate, the more complete the user experience is. Did you look also in other projects, from the Mac side, like CyberDark and Colloclue? CyberDark is an FTP application. I don't know it. Maybe I'll show you it later. I remember... Yellow Dog. Yellow RubberDog icon. I remember CyberDog as an open-dog web browser. I mean it's like a duck. It's like the Donald Duck. No, the only FTP application I know was Wypo, which was interesting, but sometimes it was open by design and it implemented only possible FTP. So I added it up for every month. Do you know Colloclue? I don't know how it's... No, I don't. I know it exists, but I never used it. It's a chat application. I think the Atari guys use it. Yeah, they use it. It has support for that special protocol. The Atari guys use this slick protocol. But it's a Cocoa application and it could be... Maybe it could be ported. Maybe it could be ported. And then the OpenOffice guys are planning to use Cocoa for the MacPort. Actually, I know there has been talks about that, but I'm not sure. Presentation of a simple webkit. Yes, we can. It's going to be fine. This is the core of this bootchip. I said the goal is to provide a represented engine support and it's part of FunoStep. So it's not just a project. It's under the FunoStep. The main author is Nicolas Schaller. Managers, founders, you're called delicious computers. His main goal is to target 100 devices like that. He's a big business computer. He's the important of this state of computers. He works in the open mode of DAOS and all those small and nifty tools. So he needed that. It's more of an engine. He's the founder of the main programmer for the work map. It's me. I'm the tester. I'm the guy who insisted that he put his code open source and FunoStep, and I maintain the core of the main files. And since I use it in this bootchip, I also test. That's hard. I also mentioned Fred Keeper, which is a Korean bootman trainer. He did not write a single line code in simple webkit, but he has been very kind in fixing bugs and completing GUI so that it works. Because the webkit engine is really taking out the maximum FunoStep GUI. Because it uses a text-active NS text-optic thing inside the text cell to display the world web page. So it's an interesting approach. It works. It can work, of course, in the other bugs. Can it be future complete in the end? Can it display tables and stuff like that? Technically, yes. It can. It's not. Tables is the most complicated thing. But it's half-working on Mac 10.5. Mac 10.5 has a table, which can be embedded in text view. So since it can be done, it's just a matter of putting it. OK, it's not in Moostop, it's there. No, not yet. So what the main features should be our web comp, but it will drop in the placement. So the API is exactly the same. And in fact, the most basic as it's putty on the Mac can be combined either with the Coco webkit or simple webkit just by linking different target, Iran, and the same nation world. It's not the same binary for the original, how Apple manages frameworks, but it's not a big deal. Strong points. We are cross-platform, really cross-platform. Since Moostop is cross-platform, and the code is entirely based on the code libraries, we work on Moostep. It works in Coco. So the author continues to check the same code on Coco, Moostep, and Mystep. So we see the differences. What is Mystep? Mystep is like a smaller cousin. Moostep environment optimized for very small devices. So it's for handheld, hard devices. The author is here to show it. We are tested on many architectures. Since Moostep works, we have to do some Solaris on PSV, pre-PSV, open PSV, Linux, Linux PowerPC. Some platforms were not so easy. For example, we know that Spark has issues with nil structures, and how they help it when we work. I don't think that it works as well, because they don't need to. And pure Objective-C, I will say it loud again, it's no Objective-C++ search. Apple made another minion. It combines from GCC to 95 upwards. So this makes us extremely portable. And having support for all the compilers is very convenient for the angelic architectures. So on Linux, we have always four, one, or two, whatever. On other platforms, it's not always the way you're there back. So it's convenient to support any compiler. In fact, we run on Perth. We are one of the very, very few browsers that run in Google Perth and work. I hope we still work. We worked one month ago, so. And as I mentioned, so that is 2.5 on Spark. Let's put your hands. Open DSD Spark, where there are very few browsers that we run. This put print is small. So the answer to binary is between one and three megabytes depends on the processor architecture, of course. If it's a bit bigger, the size goes down, probably. This is probably the heritage from being an embedded framework. Possibly uses browsers, but still. Both on workstation and handhands. So the goal is, for example, if you need to run it. But it goes farther since the WebKit and the goals are similar to the WebKit itself. So you can have help and documentation viewers with HTML. And one of the expansion to a transit new mail is to support the viewing of HTML mail in mail applications. We have to ask Ludovic to look about it, because, currently, the framework is something simply to support, at least a simple email and HTML attachment. So the rendering of simple styles and headers is enough. So it would be already a big improvement. And another usage is like kind of, this is not implemented. It's not like to-do list, because everyone wants it and nobody wants it, so I can record it. It's in my step. So you must have been supported for sure. That's a matter of writing it a bit differently. Converting HTML to an NTR document. Since we display an NSR to route it today, you can do that conversion. And it can be very convenient sometimes. But in the future, we need to fix some essential issues like line items, tables, there are some bugs. We do have potential for JavaScript support. So, overalls are probably much wider than, for example, other embeddings, more cloud-less like dildo. Maybe you should ask, talk to David, she's now in this very advanced, but when it comes to runtime. We can learn with pictures and stuff like that. And you implemented several languages and scripting stuff. And maybe it's the right thing to talk to. We have JavaScript interpreter, so that one is pretty good. But we can at least pass the syntax much better than other simple browsers, of course. It's not Mozilla, but the little boxes of pages where we at least don't display anything because we pass code to the page. And another thing is CSS, at least inline is tiger. So complete CSS support is very difficult. But modern HTML, they call standard HTML attributes as CSS inline. And so it's still HTML as far as features go. So the dump tree you get is very similar. But it's written differently. And right now it's just ignored. So I think that should not be too difficult. And of course, tons of bug fixes. In the long-term roadmap, so where we want to go would be nice to have a complete JavaScript support, at least as far as ECMAS script specification goes, which doesn't mean you can do any JavaScript page because there are so many queries and works around. But at least you know you have some JavaScript you can use for example applications like Help. And most comfortable pages. CSS one and two is only to do this, but that will be probably the most difficult part. Speed improvement, there can be many. There should be auto optimization of the parser. And in the way we display it, because since we use a text view, we need to be a bit smart on what we are doing. Otherwise, we spread the text view. Throwing it back a bit too much. Other bug fixing, I can write it on every page bug fixing, and the goal of the author, Nicholas, is to have SVG processing because for 100 devices it can be very convenient. I know that not many web pages actually use it, but it's a nice feature, I think, because it's standard, and if you need to display graphics inside a page, it should be standard. It's not, I don't know why. That's it. That was simple web kit. And if somebody is instead interested in price, which was listed on the program, you can ask me personally, I'll go to the website. Other questions, I think I will end up on time. Yeah, Tim, so. Everybody asked, came late. We can see some applications live. What's next on the agenda? Next is Quentin in exchange for Nicholas Rott, but he did not show up yet. If he did not show up, you can look at Graphos, which is a drawing application, where you can show vector drawing using NSPCA path. These are squares, but I think they're all NSPCA path, and of course, more exciting, the CFS port. Here, we're using the art backend, which just optimizes this kind of practice. Anyone who's very moved and is very happy to zoom, I guess you should be able to zoom, maybe it works. But it's a good starting point for the application. If you are really adventurous, we can try the Rancho Spudgy, if we have network connection, because we have a project for it, because of FNF and FNM displays. What do you want more? What do you want? Music related. We can check out test files and see what we can. We also see the blocks with the shortcomings we have. You should have here. We see the title here, not quite sure they called it FNM, which is a shortcoming with NSPCA path. We have support for quite complete, special characters, like Omenar to the Gatures. The translation for drawing characters is good and the translation table is quite standard complete, which was a lot of work, but there are hundreds of simple something, called within the unique code and the display. But we see, for example, at the least, it's not complete, but it's coming out nice. This is intentional that we can say the bookmax because we don't have the right to write them. That's it, questions, otherwise we wait for a point in the house. Yeah. Thank you. Thank you.