With the Open Graphics Project advancing slowly a few students picked up the idea to do it all smaller, faster, and more importantly, cheaper. From the site: “We’d love to have an OGD1 card and are amazed by it’s specifications, but for us and a lot of people, this card is simply too much. So that’s where this card comes in, to give you all the fun of developing on programmable hardware, designing video cards and architectures, for a price that won’t hurt your wallet.”
I see a lot of potential here… while it does not compare to an entry-level Radeon… right now, the design is nice.
I have been waiting for this type of announcement for years now. Thankfully we have many devs working on software for Linux but not to many working on the hardware side with the aim of creating the hardware from the ground up to work on linux.
Many are working on the driver development side of other manufacturers hardware. Now if Conanical could pick up the ball a provide some funds and an existing manufacturer could get a vision for the production, well the future could start looking very good for linux.
penguin
Other open source and free operating systems could potentially benefit from this as well. I would love to start seeing hardware with fully open specs and drivers so that Haiku, the *BSDs and ReactOS can really get in the game. As others have pointed out, this is not much more than proof-of-concept at this early stage, but it’s a positive step towards a very bright future.
Looking at the site, it seems like he’s just trying to produce a VGA signal. This isn’t easy since you still need to produce a PCI controller and correctly interpret the VGA spec, but it’s also not going to compete with even a mid-90’s 2D card.
This guy’s taking some good initiative, and he’ll probably learn a lot, but the fact that he’s routing wires by hand is a tip-off that he’s not particularly experienced. It’s a great learning project, but I wouldn’t oversell it as a “Open Source” graphics implementation to compete with even an old S3. Just being open source does not make something automatically good.
But makes it open to improvements. That’s the point, ain’t it?
But makes it open to improvements. That’s the point, ain’t it?
Yes that is exactly the point. So many people will dismiss this as some sort of toy project. When in fact it is the beginning of a revolution. Sure the Linux kernel 0.0.1 sucked but look at it now. When someone else can pick up the specs and run with it, this is a fantastic idea.
I’m definitely not experienced, and the Eagle Autorouter only got up till 54% of the design, after which everything was such a mess that I couldn’t properly work around it anymore. That’s why I’m manually doing it.
Also, the goal of this project isn’t to compete with anything, we just want to get something /working/. After we’ve gained experience with this (and helped more people getting their hands on a usable platform) and support VGA, we’re going for driver development and 2D basics. After that, well, hopefully we can merge with OGP and go for the bigger stuff.
Again, we don’t want to compete, we want to get stuff rolling faster.
I don’t want you to get it wrong; I think your project is fantastic. I just don’t think it should be billed incorrectly as something that it’s not (not by you, but by others).
Sticking “Copyright” on the schematics is interesting. Yes, they are publicly available, but no, you can’t redistribute them with any of your own improvements. Unless I’m missing something else on this page, the copyright would seem to stifle innovation that other people could add and distribute.
What has copyright got to do with it? Even BSD code is copyrighted. The important bit is the license the copyright holder releases the work under, and right from the first paragraph of the front page of the project website:
I wrote it, so I have the copyrights on it. If I understand copyright law (IANAL). I also stated that I’m releasing it under the “Creative Commons Attribution-Noncommercial-Share Alike 3.0 Unported License” on the main page, but for clarity, I’ll add it to the schematic soon (don’t have Eagle here right now).
Team
I do not understand hardware & Most of you blog was barely comprehensible to me….
Despite that, I can recognize initiative and development when I see it.
I don’t know and don’t really care if this will succeed or not, I just know that your making this attempt is improving the industry and producing better tools/products for the rest of us.
Congratulation on you initiative, let me know if there is anything us (less experienced) types can do that will help.
Ideas:
1. Donations to help with costs.
2. Translations into other languages of documentation (Sorry I only speak English).
3. Purchase of available prototypes.
4. Telling anyone else who will listen how cool this project is.
Regards
Darren
I was disappointed by the price tag of the OGD1 card. I am heartened to read someone else is looking to give this a go. I think the final price will make or break these projects.
I would look at picking one up if it was available at a reasonable price. I am not a developer but think its neat to support something like this.
1) You can get a PCI interface wire-wrap card with most of the PCI interface logic included.
2) Get a standard SVGA graphics frame buffer chip (ET4000 from mid-90 should work.)
3) Get a low-power PPC or 68000 cpu.
You could then use a open source OpenGL library to implement a graphics language on the card. You could also move all or part of the X-Server to the card.
I’m not sure why everyone wants to create their own SVGA interface; thats the hardest part here. Also, they insist on creating some sort of custom graphics engine; when an off the shelf RISC type cpu will work just as good.
Can you tell me where to get these PCI ready cards if they don’t cost a arm and leg like the few I have found?
Personally I just want a huge frame buffer that looks like a single block of memory to my OS, the difference is that there would be multiple outputs each feeding only a quadrant of the frame buffer out.
The move from one to two monitors was great, three might be better but I think I need 4 or 6 monitors acting as a big screen would be seeing a real improvement.
It would be nice to have a single card giving support for 4 or 6 outputs.
Funnily enough thats all I really want from a graphics card too and feel blessed I can even use a basic $50 twinhead card at all on BeOS or W2000 for that matter.
Even nicer if the video ram was partitioned into smallish tiles with a decently fast and open spec processor per tile and software configurable as to tiles per monitor head. Still the perforamce bottleneck would still be the PCI unless one steps up to PCIe hardware interface or keeps all the rendering in the card, buts thats another project……
Even at the low end of FPGA cards, some of the ready made edu Spartan cards have a VGA output but no DAC so they are stuck with 640.480 with 8 or 16 colors, way too retro for me. As soon as these development boards get even remotely interesting, their prices go through the roof and they pick up features that are of no use to me.
I hate to break it to you.. but it’s not entirely complicated to design a VGA-compatible video card.
In high school, I designed my own 8bit ISA card.. is this group planning on supporting 3D acceleration of any kind?
Please.. how is this at all more impressive then an old 2MB PCI video card?
In high school, I designed my own 8bit ISA card.. is this group planning on supporting 3D acceleration of any kind?
In high school I wrote my own visual programming language. But where is it now? Nowhere because it was worthless and I kept no record of it. If I had of kept a record of it, it would have been closed source anyway and what would that have done? Nothing.
Can you share the specifications of your 8bit ISA card? Can people improve upon your work to make something worthwhile of it?
That is what is so special about this project compared to yours and mine. It is open for others to work on.
Well, he intends to have something to show for it by December, so I’d hope it wouldn’t be THAT complicated.
I read this as an “The OGD1 is great, but here’s a quick something for now while we wait, because I think it’s cool”
Wow… Extraordinary. It seems that in the Netherlands there is lots of people talented with hardware and so open minded to release their products with Open Source licenses.
Good!
Take a look at Minimig, the Amiga 500 Clone that it is also based on FPGA chip which emulates original Amiga Multimedia Chipset, from mr. Dennis Van Weeren:
http://home.hetnet.nl/~weeren001/minimig.html
http://en.wikipedia.org/wiki/Minimig
Hope that both projects (ProjectVGA and Mimimg) could interact and enhance the capabilitiy of this little beautiful mini-computer…
See also Open Source re-implementation of Minimig with Mini-ITX form factor:
http://web.comhem.se/illuwatar/project_pages/minimig/minimig.htm
I really hope you can make this work it is really needed and I appreiciate the effort.
Looks like a very nice project, and the intention does seem to be to create an open source platform. I just see a little problem with the licensing terms. Let’s see, according to http://creativecommons.org/about/licenses
“Noncommercial. You let others copy, distribute, display, and perform your work — and derivative works based upon it — but for noncommercial purposes only
Examples: Gus publishes his photograph on his website with a Noncommercial license. Camille prints Gus’ photograph. Camille is not allowed to sell the print photograph without Gus’s permission. ”
And then, according to the Open Source Definition, found in http://www.opensource.org/docs/definition.html
“1. Free Redistribution
The license shall not restrict any party from selling or giving away the software as a component of an aggregate software distribution containing programs from several different sources. The license shall not require a royalty or other fee for such sale. ”
So, it seems that the “non-commercial” condition prevents the project from being open source according to the OSD definition, which is by far the most widely used definition. Have you considered dropping this requirement? If what you want is no PROPRIETARY derivatives, the GPL (like any copyleft license) does just that. Maybe you were confused by the corporate propaganda of using “COMMERCIAL” as an euphemism for “PROPRIETARY”?
[edit: I’m an idiot] So all that is needed is a small, specific exception to the CC license to make it technically compliant.
Edited 2007-11-12 12:48 UTC
No, all it takes it not using the non-commercial variant. Creative Commons is a suite of licenses where a set of conditions can be chosen by the creator. Noncommercial is one of them.
The noncommercial restriction is a huge problem for a hardware specification. It means that a company can’t produce the hardware and sell it.
If I understand correctly, in the united states, technical specifications are exempt from copyright.
(not trying to troll here, just pointing out the obvious)
That effectively makes any hand-wringing about the licence effectively moot.
Thant said, The reason this is significant is that most hardware dev is done behind closed doors and NDA’s so the project has the potential to be very valuable, even if the initial resulting hardware is not.
Kudos, and good luck!
Due to popularity, I’ve run out of bandwidth (didn’t have much to begin with) so I made a last-minute move to here; http://oege.ie.hva.nl/~meeuwi10/pVGA/projectvga.php
I can’t access the other server at all at the moment so no redirs either.