Still on schedule for DP2, Jeto.Media has been refined to a state I find suitable for release. There is currently one IMediaIoEngine implementation, which is XineMediaEngine (provided in Emotion.Backends.Xine.dll). The API covers almost all the areas I would like it to, with the noteable exceptions of chapter and caption support. A new event loop MediaEventLoop has been created (and is now useable with EmotionEventLoops constructed with EmotionModules.Media) for handling media-related events. More detailed information on this topic will come in an upcoming Knowledge Base article on the Jeto forums at http://komodoware.com/forums . If you want to check out the changes yourself, feel free to check out the new code from CVS. I'm tired, so I sleep :).

> Read More... | Digg This!

Work progresses on the new release of Emotion due out this month. The basic functionality of Jeto.Media is working great, and I shall spend my free time (that is, time not in school) expanding and improving it. The default media engine plugin will be Xine, and media format conversion will be done by SOX. After DP2 we will expand these options. My time estimate for bringing the Media module to a level ready for DP2 is about 2 or 3 days. After that there's just Hardware and Volumes to go. Hardware is mostly broken right now but Volumes just needs some design improvements (that I couldn't do the other day when I was super caffeinated by this awesome coffee)

> Read More... | Digg This!

Many things have been done to the Graphics module of Emotion to increase it's performance. First of all, a method WaitForEvent() has been added to the IEventLoop interface. When an event loop is used seperate (that is, not with EmotionEventLoop) it is expected to use the underlying event blocking mechanisms which are available. For instance GraphicsEventLoop with the X11 graphics backend will use X11.NET's X11.Application.WaitForEvent() method to do the dirty work. [...] Read more!

> Read More... | Digg This!

Finally got off my ass and adapted Pango to work with Emotion. Thus anywhere where there is layouted text in Emotion (aka the TextLayout/ParagraphLayout objects), Pango is used. This means the Pango markup language can be used with most UI controls. In order to support Pango in Emotion properly, a new Portal binding was made and added to CVS, PangoCairo.NET. It wraps the libpangocairo library for using Pango with Cairo. Beyond that, the whole Jeto.Graphics text system was revised to better support the advancedness of Pango and the custom chunks I was working on have been removed or commented out. All the changes are now in CVS, please feel free to test it and send feedback. Among the foreseeable problems: TextLayout and ParagraphLayout currently cannot be used in conjunction with virtual Artists, such as CanvasItemArtist or any artist which cannot produce a CairoContext via the new IArtist.GetCairoContext() method. The main concern is figuring out the proper behavior for when drawing is delayed (that is, the artist caches commands to apply with a real artist later). There have also been many other Jeto.Ui control-specific tuneups and bug fixes.

> Read More... | Digg This!

First of all, some updates on the status of the Emotion library. The Jeto.Graphics and Jeto.Ui code has been fully updated to use the new drawing/surface semantics from Cairo 1.0. Approximately 90% of the graphics examples work properly and probably around 50% of the Ui examples work. A new set of examples for the Jeto.Kernel namespace has helped me iron out a bunch of bugs in the UserList, GroupList and Kernel classes. UserList now uses the standard C library method for accessing /etc/passwd, which makes the code much more lightweight. GroupList still parses /etc/group on it's own though (not sure if stdlib has an interface for it). The Kernel class is seeing impressive changes at the moment, which can be seen in CVS. Soon Kernel.Filesystem will be subclassable to allow easy implementation of user space filesystems, using SULF (Stackable User-Level Filesystem) and FUSE (Filesystem in USEr space). Secondly, I think all processor's will eventually move to similiar models as those used by the Cell and Sun Rock processors. That is, they will become Massively Multi-Cored. Sure, we're used to hearing about dual-core chips with awesome performance and all that, but Rock has 16 processing units + 16 floating point / graphics units, and Cell has 8 vector processing units along it's primary processor. I think there will need to be a new, dramatically different way to program these bastards, at least to get the full performance. There has been news about IBM's new Octopiler for Cell which uses very advanced compiler technology to auto-vectorize code and in general keep all parts of Cell busy. However, to truly make use of the system in the way it was intended, you will need a _Master_ programmer. Massively Multi-Cored does make sense, as the difficulty of lowering the sizes of chip manufacturing processes increases. But, I don't know Cell in and out (it's mindboggling to know Cell in and out) and there could be a lot I'm missing. I do think it's almost entirely unreasonable for Sony to expect the PS3 to be done and released with actual TITLES for it this year (you can see video game companies scraggling to find funding for the master programmers).

> Read More... | Digg This!

Oh yeah. Apparently 4 to 5 years for a new Windows release is "bad". No my friends, the longer the wait to upgrade, the better: at least for users. Microsoft forgets our criticism of their cocaine-induced release and upgrade policy. Ballmer wants Windows and SQL Server and Office et cetera to be on more interim cycles, which I would agree with for security but ultimately we know Microsoft will consider these "interims" full releases and ask for the whole fee over again. Go ahead Ballmer, screw up again. It's tearing, being an open source advocate (who naturally dislike the Redmond beast) while still hoping that Microsoft will come to it's senses and join us once they realize we can't be fought. I would be happy with either outcome, whether Microsoft dies because consumers get fed up with the constant hassling, high prices, fast upgrades (which apparently will return) or they take the effort to be reborn into a decent company. My what fun reading the news is today :). Or at least the tech news..

> Read More... | Digg This!

ftp://komodoware.com/snapshots/komodo-dp1-i686-2005.10.02.iso.torrent The next DP1 is now out (round 2). Pick up yours at ftp://komodoware.com/snapshots/komodo-dp1-i686-2005.10.02.iso.torrent

> Read More... | Digg This!

The new DP1 snapshot is very close to being ready for seeding. Much has been changed, streamlined, and verbosed to improve the testing process. There's still some mono specifics which must be worked out for the shared cache creation in system setup. I do have to go to school tomorrow so it will have to wait until I get home. This one should be much less problematic: - Programs which reported missing libraries were converted to static executables - Most of what was done in post-install is now done in pre-install to detect errors earlier. - The initrd for the live CD has been bumped up to 5 mb from 4mb (shouldn't affect anyone) and pre-install, post-install and Pkg (the package installer) have been moved into the CD and out of the initrd. - Much more I'm not trying to rush too badly. I know a lot of people are looking forward to the next snapshot, but bandwidth can be expensive and I'd like to get a whole _new_ round of problems instead of regressions of the old ones :)

> Read More... | Digg This!

http://osnews.com/ We have been OSnews'd :). I submitted it though when I realized no one was coming from the forums I submitted the alpha testing story on OSNews. Before I move on to my replies to a LOT of OSnews comments, I should note that although 100% of the information in our About pages describes the battle plan, really only perhaps 60% of the systems are on par and working properly. But you get the goals of it. Regarding many comments on OSnews: Martin: Wow SkyOS called, they want their UI back. Is it just me, or do they look similar? Me: It's just a theme. It is, of course very themeable. I just am not going to produce twenty themes and then change some interface and change all twenty of them. Just one is good for now. I haven't seen SkyOS, I was just playing with what Cairo gave me to make it look nice, trying to utilize some ARGB colors too around. The real ARGB support both between widgets and between windows is implemented but doesn't quite work yet. When this finally works (mostly I just have to debug the X11 backend), we'll be changing things up a lot. Anonymous: More software infected with .NET Only in this, case they've taken full advantage of C#'s "portability" and created a new API so that all of your System.Windows.Forms and GTK# apps won't run. I really don't understand why these .NET fanatics just don't run Windows Vista. Me: Wrong, these applications _DO_ run and we include QT, GTK, and .NET bindings for both (weeelll really we don't have much for QT bindings yet because, well it's really hard to do). Emotion is _not_ just for graphics! It encompasses an entire object-oriented API for all the services of the OS (or at least it aims to). In this way, Emotion could theoretically be rewired to USE Qt or GTK# for it's UI operations, especially since Cairo is the core of the graphics system: it allows a lot of interoperation with other systems now that almost everyone's using it. Never once has Linux had a _complete_ orchestrated API and we give them not one, but two: one for functional languages and one for object oriented languages. This framework is flexible, so that removing one from the other is easy, thus making it more reusable. In fact, most functionality in Emotion is abstracted to backends (aka, plugins). MikeGA: um, why does it look so like Aqua OS X? So, they added some more buttons to the window title-bars. That doesn't deny the fact that the rest of the title bar scheme is identical! And I could similar things about their dock, hell they even call it a dock (which has always seemed a slightly stupid name to me!) Well at least they have a sense of humour, one of the screen-shots shows a desktop picture which fairly clearly features a stylised iMac G4. Me: Thanks, you flatter me. But honestly, guys this stuff is all themes, and the title bar is a theme for XFCE's window manager. I happen to like the simplicity. It's not even the desktop that will be used for the final release! The "iMac G4" is a background from KDE-Look which I thought looked cool for my normal desktop. In fact, none of the screenshots on that page are shots of the actual desktop. They are shots of Emotion tests. The "Dock" is really only a rather fancy test of Emotion. We've never mentioned anything about actually including a dock or calling it that in a release, and if we do we'd use it for a base and innovate on top of that design. Anonymous: Where'd these guy coming from? This looks real interesting. I haven't read everything on the site yet, but what is the emotion engne based off (Xlib), so it's more of a native toolkit? Me: Where indeed! We've been right here, waiting for you! Emotion is actually a fully-backend-capable toolkit, with backends wrapped in .NET plugins. Any graphics system that can have a cairo backend can have an Emotion backend. qlscohen: I don't talking about kill the actual desktop projects, only rethink. maybe join the efforts to One Desktop. i like the diversity but linux there is very very very very, diversity. ( on Desktop Side ) there are so much personal developer touchs on his developments. Me: I like the way your thinking there. I agree, KDE and GNOME have their own thing going and the projects are just too big to accommodate the kinds of crazy ideas we can with Emotion. Not that Emotion is just a graphics library, but it is a major part of it. A lot of times it is impossible to create something crisply designed without starting from scratch. That's my opinion, so that's what I'm pursuing. Unknown: Also, what do you mean by "fully freedesktop-compliant". fd.o is meant to be a cooperation-platform between _existing_ environments. Me: We speak more of the freedesktop _software_ than the standards, although those are included as well in our goal of freedesktop compliance. Anonymous: Well, there's the irony. People want choice and open source, but people (some) want widespread adoption of LoTD. Why would these guy join Gnome or KDE. So they can have RedHat or Suse as a boss? Me: It's more the fact that I think development with KDE and GNOME are actually not tight enough. Also, we'd prefer a completely .NET native environment over one which constantly marshals to the native OS (drops performance). It also feels more natural that way. Many: What's up with the Komodo name man? Didn't you google ActiveState? Me: The name was chosen before I was aware of it. We changed our names many times (once it was Calix, then Calyptos when Calyptos and Calix merged, then a few other temporary names, and eventually we really liked Komodo). We are not at all closed to a name change, and there's a name change sticky in our forums for you to post your ideas. It's not that easy: Very few good names are left unused in the software industry already. Anonymous: It certainly looks good but what about performance? Does anyone know how fast it runs in comparison to other distros? Oh please don't hehe. I might be embarassed. No really, it should run pretty quick as it's based on Linux 2.6 and all the newest performance goodies. My home system (which is also the main Komodo dev system) is 1.7ghz and never gives me any performance problems or freezes/crashes. Well, with the exception of when a rogue GL game goes out of control, but I'm working on those things even if all I can do is make the system more receptive to those problems. Anonymous: how they think they could be taken seriously with screenshots with words like "fucked" "ownage" "bitchin" etc..? ps im not a moron, i use this words everyday but it doesnt seem to me a great idea to have this approach promoting your own product... Me: Jeez Anonymous, you raise a good point. I wasn't aware that those were still there of course. Also those screenshots are NOT meant to be an indication of the product. Microsoft puts swears in their products though, why can't I? :) Bascule: Considering the look is blatently stolen from OS X, I don't think they've really considered the legal ramifications of anything they're doing. Me: That's insulting. I have considered every side of the legal spectrum that applies to us. You fail to realize that our name is in the process of a name CONTEST in the first place. You also fail to realize that the "look" is only the theme which is being used to develop it. You need to investigate things more thoroughly before you point fingers at me. Anonymous: Some of these comments are pretty funny..."why don't they just join Gnome or KDE"..."we already have enough desktops". The funny part of this is when you advocate a "standard" desktop some of these people will be the first to throw a hissy fit. So I guess when it's Gnome and KDE, it's all cool, but when these guys do something different (and this is very different down to the toolkit) it's not ok. But when another random distro comes out that does nothing different, but has a new name and a new repository, then "choice" is good. A bunch of hypocrites. Let's face some facts. Not having a "standard" which really means dominant desktop has hurt LotD, and since there isn't going to be one at least there are people being a little bit innovative - rather than YAD Me: Thanks for sticking up for me and my free choice to do whatever I want (within applicable OSS license guidelines) with free software. Anonymous: Yeah, and I said previously now they're defacto working for RedHat and Suse....but without the pay. But at least these guys are doing something pretty different. I have a feeling the performance isn't going to be great -especially on low-end machines because of Cairo. Me: You've been saying a lot tonight Anonymous! :) Different: yes, that's the point. The performance of Cairo _is_ actually a bit crufty but the library is very new and a lot of optimizations have yet to be implemented to allow good performance on slow machines. Even fast machines with Render accel can stumble. Performance is more limited however to .NET at the moment. This is really out of our hands at the moment but the great guys at the Mono base have been working hard to improve performance and have done so _considerably_. wazoox: I don't care, SkyOS is quite nice BUT it's 1) proprietary 2) there aren't any apps. Komodo is 1) GPL 2) compatible with Linux apps. Me: Yes, free software forever! Anonymous: Wow, this guy sure has a lot of good things to say about Komodo. Me: Haha, yes I do. It's a big part of my life and I think it's gonna put a dent into Microsoft's ass. (Yes, I use .NET, but I've been a member of the Microsoft Eradication Society for three years. Coincidence?) Roguelazer: So therefore they're proponents of the "Less is less" design philosophy? Nihilism and Realism combined as a software design philosophy? Sounds cool, but I think for all of my "futuristic, Über-advanced DE" needs, I'll stick to E17. At least the enlightenment website doesn't time out. :-) Me: Ouch! I'll communicate that to my host. You must also understand though, we got 7000 unique visits from the OSNews article, and (although I haven't confirmed with our official server stats the exact number) tens of thousands of hits in a very short period of time. We also have a lot of other stuff going on in the server, including seeding for the ISO. No surprise there were some timeouts. Anonymous: Actually, it looks a heck of a lot like one of the Aqua-esque themes that comes with XFCE4. Well, the whole thing looks vaguely like XFCE4, except for the Gaim icon in the panel. That looks more like the Enlightenment 'Engage' launcher. Me: That's because it is! It is in fact the Agualemon theme. That's not really what the screenshots were showing... bullethead: This is great. It's great to see some new ideas for the desktop (Finally!) Me: Thanks! I love seeing these comments :) Anonymous: Open Source is even catching up to Microsoft's own Windows roadmap. This should be interesting. Me: Won't it? The traditional distros seem to just change by version number and I haven't seen a single new thing that hasn't been there the last five years. The other distros are just at more of the end of an OS development cycle, where they are trying to get everything working PERFECTLY whereas we are just trying to get things to work in a new way, and the perfection comes later. jeremy: Sure Linux can do all of that. But it's more complicated and disconnected than it needs to be. I agree all system admins should know how to setup everything listed but it should be easier. Why does it need to be complicated? To limit who can do it? I think he is talking more about creating a combined system with tools that allow you to easly manage all of these things. Look at what apple has done so far with their server stuff. A large portion of it's open source but it's one contained unit when looking from the outside. They also offer you all the tools you need to complete any basic task with out much work. Me: You hit it right on the head!

> Read More... | Digg This!

If you download the Komodo snapshots, please make sure your computer can receive traffic from port 6881 to 6889 in your firewall/router/NAT settings. These ports are essential for clients to communicate with your PC to get parts of the file from you. Please also seed it (aka, leave BitTorrent open after it's done downloading) so more can download the file. Those who can provide long-term seeding with good uploading bandwidth please do so, and if you email me I'll give you credit on our website.

> Read More... | Digg This!