OS2World OLD-STATIC-BACKUP Forum

WebSite Information => Article Discussions => Topic started by: Saijin_Naib on 2008.04.09, 04:12:48

Title: OpenSNAP geplant/OpenSNAP planned
Post by: Saijin_Naib on 2008.04.09, 04:12:48
BWAHHHHHHHHH!  ;D ;D ;D ;D

Can this be real? Will OS/2 really have hardware overlay again?!?

Yesss.
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: Shai on 2008.04.09, 08:34:47
BWAHHHHHHHHH!  ;D ;D ;D ;D

Can this be real? Will OS/2 really have hardware overlay again?!?

Yesss.
Wasn't that hard to get the latest SNAP legally. It is shipped with eCS and there was still the Personal Edition.
It's nice they don't just let it die quietly like so many other abandoned products - seems that they still care for OS/2.
Therefore I think it's obvious: For this to be beneficial, someone has (or better a group of people have) to maintain and enhance the driver.
btw: hardware overlays are only one feature (doesn't Panorama support this, too?). Hardware accelerated display drivers are simply faster. Also there's multihead support in SNAP - this would be especially important for guys like me who want to continue working with 2 monitors.

greetings,
Robin
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: Saijin_Naib on 2008.04.09, 08:44:05
SNAP does not support my current chipset (7600GT) and as far as I know, Panorama cant do hardware overlay because its a VESA only driver, therefore, no hardware acceleration at all.

I am most excited for the hardware overlay so video playback can not eat my CPU, and I can benefit from all the post-proc features my chipset supports (hello Desktop Painter, how I have missed you!)

I also would like to be able to scroll large windows without lag and 99.9% CPU usage, more or less most of my excitement over this news :)

Also, this gives me hope for a port of OpenGL to the OS/2 platform again, something we desperately need.
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: magog on 2008.04.09, 16:36:24
SNAP does not support my current chipset (7600GT) and as far as I know, Panorama cant do hardware overlay because its a VESA only driver, therefore, no hardware acceleration at all.

Panorama can't do overlay and Dual/Multi-Head but it's much faster than SNAPs GenGRADD.
As nVidia hasn't started to open their specs you won't see much nVidia GeForce support in OpenSNAP I guess.

I am most excited for the hardware overlay so video playback can not eat my CPU, and I can benefit from all the post-proc features my chipset supports (hello Desktop Painter, how I have missed you!)

I also would like to be able to scroll large windows without lag and 99.9% CPU usage, more or less most of my excitement over this news :)

As I said scolling without lag for unsupported cards use: Panorama

Also, this gives me hope for a port of OpenGL to the OS/2 platform again, something we desperately need.
I still have the copy of the Hardware Accelerated OpenGL Toolkit here. ;)
Using this is not really legal but since it's abandonware I would actually ignore this obstacle.

For details look in this thread and this message as starting point:
http://www.os2world.com/component/option,com_smf/Itemid,63/topic,515.msg2485/#msg2485
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: Saijin_Naib on 2008.04.09, 18:47:23
Hey Magog, I am using PANORAMA Vesa with Shadowbuffer on, and it still lags and eats 100% of my cpu when I scroll any window. Its readily apparent with larger things like OO.org windows or Firefox, but even extends down to small plain-text windows like MrMessage chat windows. I've discussed it with Eugene, and he is aware, and intends somewhere along the line to improve the scrolling performance of Panorama.

As for no support for my Nvidia chipset, I said that SNAP doesnt support it, but that is only the case for the OS/2 driver because they abandoned it long before the Windows driver. My card is fully supported under SNAP (windows) and works beautifully, so I assume that since it is in the SNAP codebase and functional, it would be relatively easy to take this (already established) code and work it into the OS/2 release, or am I wrong?
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: BigWarpGuy on 2008.04.09, 20:21:40
http://list.scitechsoft.com/mailman/listinfo/opensnap (http://list.scitechsoft.com/mailman/listinfo/opensnap)
This site might have more information?
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: Paul Smedley on 2008.04.10, 05:15:17
As for no support for my Nvidia chipset, I said that SNAP doesnt support it, but that is only the case for the OS/2 driver because they abandoned it long before the Windows driver. My card is fully supported under SNAP (windows) and works beautifully, so I assume that since it is in the SNAP codebase and functional, it would be relatively easy to take this (already established) code and work it into the OS/2 release, or am I wrong?

Well it depends what exactly they end up releasing in the opensource release.  Some of the code for particular chipsets could be covered by NDA so cannot be released under any opensource license.
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: Saijin_Naib on 2008.04.10, 05:26:24
Damn :C Is there any particular factor that would determine this, or is this all an internal (to Scitech) process and we just have to stand by and see what the result is?

Even if it turns out it doesnt support my chipset, I'll still be incredibly happy to see it become a reality!
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: jep on 2008.04.10, 10:57:39
Hello,

SNAP from ScitechSoft may not have support for all the latest chipset (ATI and NVidia) but contain what I would say is the more stable and well designed product.

It would be most interesting if "all" developers that write drivers for graphics card could unite and continue to improve OpenSNAP, add parts from Panorama (as I asked from the start 05...) as it does have at least some support for ATI chipset and get HW OpenGL at some point in the future. My GeCube Radeon HD 3870 does work with Panorama at least  ;D

Piggy backing on the Linux community? If we had the soources to GRADD then we'd have something to bargain with so that we could share technology between platforms. We could offer something that could be used by many, one wheel design to improve.

The Linux way: One developer, one driver, one system that it works on, develop the wheel over and over again.

//Jan-Erik
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: Paul Smedley on 2008.04.10, 11:17:01
Damn :C Is there any particular factor that would determine this, or is this all an internal (to Scitech) process and we just have to stand by and see what the result is?

Even if it turns out it doesnt support my chipset, I'll still be incredibly happy to see it become a reality!

It would depend on how Scitech obtained the information they needed to be able to support the chipset.  Some manufacturers make this publically available, others require a NDA.... IIRC in some cases, Scitech used information from opensource linux drivers also...
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: magog on 2008.04.10, 20:31:23
As Steve already wrote on the normal snap os2 newsgroup, they can't release the Matrox driver.
As SciTech as been ATI's development partner, they had an NDA for the ATI driver stuff.
The question is now, can they release the ATI stuff now as AMD/ATI has opened the access to the chipset documentation?
Intel has also some open documents of their integrated graphics engine, so it might be possible to add them as well.

As SciTech had always problems with nVidia the non-team player, I would guess that they can release those drivers. The information needed came from X-driver sources as far as I know.
Then there is the old stuff, that was already part of the open SNAP DDK in April 2004 (SNAP 2.3r14).

A quick look into that package tells me the following driver where included in that release:
-3dfx (TDFX_BANSHEE, TDFX_VOODOO3, TDFX_VOODOO4, TDFX_VOODOO5)
-VESA
-IBM VGA

Some MESA-3D (v3.5) stuff is also included.
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: miturbide on 2009.08.27, 19:13:00

Does Anybody know anything about this ? was any source code released ?
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: warpcafe on 2009.08.28, 09:23:08
Heya,

did someone already mention that the scitech site no longer exists but redirects you to...
http://www.altrichmond.ca/ (http://www.altrichmond.ca/) ?

Quote
December 2008
Alt Richmond closes acquisition of SciTech Software’s SNAP technology

October 2008
Alt Richmond agrees to acquire SciTech Software’s SNAP technology

I never heard of that up to now... did you?? Wondering what that means for the future of OpenSnap (if any)

Cheers,
Thomas
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: Joachim on 2009.08.28, 14:08:57
did someone already mention that the scitech site no longer exists but redirects you to...
http://www.altrichmond.ca/ (http://www.altrichmond.ca/) ?

We're already in touch with them, don't worry.

Regards,

Joachim - Mensys BV
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: warpcafe on 2009.08.28, 17:23:09
We're already in touch with them, don't worry.

coooool. ;)
(And thanks for the info!)
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: Saijin_Naib on 2009.08.28, 18:49:12
Not to pry, but what does that mean for us? Do you have access to the Windows Scitech info? They supported much newer chipsets with 3D accel and that could be potentially useful for us.
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: alex on 2009.08.31, 09:42:02
Joachim,

Will Alt Richmond make the OS/2 SNAP technology open source as promised by SciTech before their acquisition, or do you have to make a deal with them ?

Regards,
Alex
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: madcrow on 2009.12.08, 15:32:46
Sorry for the bump, but after several months, I was wondering if anything more had been heard on this front? Will SNAP's new owners keep Scitech's promise to open the source or have they decided to be jerks and break the promise?
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: rwklein on 2009.12.08, 18:37:41
Sorry for the bump, but after several months, I was wondering if anything more had been heard on this front? Will SNAP's new owners keep Scitech's promise to open the source or have they decided to be jerks and break the promise?

I would need to check again with Altrichmont where we stand (Mensys).

BTW I could be wrong but was it not Scitech that would open source SNAP. I don't know what the statement was from Altrichmond. And in this type of cases its better to have a name of somebody that stated there are willing to open source.

Mensys was in touch with them and we still needed to work out details how to deal with the code. But lets just say it will not be done free or charge. From what I remember a developer would need to split off the code thats OS/2 specific.  That would come with a price tage.

Roderick Klein
Mensys
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: miturbide on 2009.12.08, 18:46:45
Roderick

Here it is the link.

http://www.os2world.com/content/view/17898/2/ (http://www.os2world.com/content/view/17898/2/)

The source of that news was Steve Wendt.
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: rwklein on 2009.12.08, 19:08:14
Roderick

Here it is the link.

http://www.os2world.com/content/view/17898/2/ (http://www.os2world.com/content/view/17898/2/)

The source of that news was Steve Wendt.

Thanks but now comes the nit picking. Alt richmond never released such statement that I can remember.
But I could be wrong...
I need to keep better track of a lot of the stuff I work on but some much stuff goes over my desk.
In other words I don't know if Alt richmond released ever such statement as Scitech has done.

Roderick Klein
Mensys
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: miturbide on 2009.12.08, 19:18:48
uhmm..
where is Steve Wendt now?
where does he works now?

Is there any way to contact him and clarify his statement?
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: herwigb on 2009.12.08, 22:57:01
Steve Wendt still hangs around quite frequently in the mozilla.os2 newsgroup in usenet.
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: madcrow on 2009.12.09, 04:38:51
Sorry for the bump, but after several months, I was wondering if anything more had been heard on this front? Will SNAP's new owners keep Scitech's promise to open the source or have they decided to be jerks and break the promise?

I would need to check again with Altrichmont where we stand (Mensys).

BTW I could be wrong but was it not Scitech that would open source SNAP. I don't know what the statement was from Altrichmond. And in this type of cases its better to have a name of somebody that stated there are willing to open source.

Mensys was in touch with them and we still needed to work out details how to deal with the code. But lets just say it will not be done free or charge. From what I remember a developer would need to split off the code thats OS/2 specific.  That would come with a price tage.

Roderick Klein
Mensys
So OpenSNAP is dead before it was even born? If Mensys is planning on taking up the code and developing it as either payware or as an eCs-only closed source driver kit, it would certainly seem that way... C'est la vie, I guess.
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: rwklein on 2009.12.09, 14:51:51
Sorry for the bump, but after several months, I was wondering if anything more had been heard on this front? Will SNAP's new owners keep Scitech's promise to open the source or have they decided to be jerks and break the promise?

I would need to check again with Altrichmont where we stand (Mensys).

BTW I could be wrong but was it not Scitech that would open source SNAP. I don't know what the statement was from Altrichmond. And in this type of cases its better to have a name of somebody that stated there are willing to open source.

Mensys was in touch with them and we still needed to work out details how to deal with the code. But lets just say it will not be done free or charge. From what I remember a developer would need to split off the code thats OS/2 specific.  That would come with a price tage.

Roderick Klein
Mensys
So OpenSNAP is dead before it was even born? If Mensys is planning on taking up the code and developing it as either payware or as an eCs-only closed source driver kit, it would certainly seem that way... C'est la vie, I guess.

Mensys needs to go back to Altrichmond and see what we can do. If Altrichmond would stick to the statement Scitech made about opensnap, don't know. But Altrichmond is a different company and Scitech basically had to close down because of lack of business from what I remember reading on statements on the website.

Roderick Klein
Mensys
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: magog on 2009.12.17, 02:24:03
uhmm..
where is Steve Wendt now?
where does he works now?

Is there any way to contact him and clarify his statement?

Steve is working for Kendall Bennetts other company since SciTech closed business...as far as I know no hardware stuff but business applications (I don't have up-to-date information so I might be wrong).
Michal Necasek moved on to Sun.
If you need to contact Steve look here (the dark gray line at the top): http://www.os2bbs.com/os2news/

Altrichmond(.ca) never released anything about SNAP beside that they bought the code. OpenSNAP was only something that Steve released as selling SNAP wasn't going forward at that time. When selling the tech was going forward again OpenSNAP was dead (before it even started).
So now it only depends on Altrichmond if it's possible to continue developing drivers. As Mensys already talked to them, they have to push it.
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: abwillis on 2009.12.17, 05:05:04
I had also spoken to Altrichmond about the same time as Mensys did.  They were not opposed to opening the source (at first it seemed as if they might just open it) but decided that the cost to separate the code would be cost prohibitive.  About the time that they started talking cost I suggested we should bring in Mensys as they had the best infrastructure to raise money and found out they were already talking to them.  I don't know how many man-hours of work there is to split the source out but what they were asking for was the cost to be covered.  It may be possible for someone to sign an NDA and do some of the work to lessen the cost but I hadn't asked that at the time.
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: magog on 2009.12.17, 20:47:19
I don't know how many man-hours of work there is to split the source out but what they were asking for was the cost to be covered.  It may be possible for someone to sign an NDA and do some of the work to lessen the cost but I hadn't asked that at the time.

I don't know anything about the cost as I have no idea what the "split the code" is all about. It's not possible or at least not very well-thought-out to only release the OS/2 part. Driver developed should be running on all platforms otherwise SNAP doesn't make sense at all. If someone is fiddling around with the OS/2 SNAP backend modifying it would break using drivers from other platforms. Is this something they want? I don't think so.

What Altrichmond has should be only the part that can be open sources. I don't think it is possible they also got the drivers in source SciTech developed under NDA (ATI and Matrox drivers).
When it comes down to the source the best options to pick up the work are Kendall, Steve and Michal....supported by other people (whoever that might be).
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: RobertM on 2009.12.17, 23:08:48
I kinda suspect that it is the GRADD components that may have to be stripped out before SNAP is open sourced... meaning they'd have to be recreated or re-added via one of the DDTKs for an OS/2 release.
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: Saijin_Naib on 2009.12.17, 23:13:42
Why would the GRADD stuff have to be removed? Isn't it available in the DDK and other documentation?
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: RobertM on 2009.12.17, 23:35:13
Why would the GRADD stuff have to be removed? Isn't it available in the DDK and other documentation?

It's just an uneducated guess. But as far as the law goes, it doesn't matter if it's obtainable via other means if the license SciTech was given indicated they could not release the source to it.
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: rwklein on 2009.12.18, 00:38:56
Why would the GRADD stuff have to be removed? Isn't it available in the DDK and other documentation?

Here is the exact thing! Not everything of GRADD is in de the DDK. I know from the guys at Scitech that they did not have access to the GRADD stuff other then in the DDK. For fixes to GRADD they have to count on IBM.
That is not to say it can't fixed/updated. GRADD.SYS for Panorama was partly updated by the eCS development team :-)

Roderick Klein
Mensys
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: mobybrick on 2009.12.19, 12:56:19
AFAICR, the part of the code that would have to be removed wouldn't be GRADD (as I don't think it contains anything more than was available on the DDK) but... SciTech I believe had an NDA with ATI. I suspect that the ATI specific code would have to be removed... and as that is more than half of the supported chipsets, I'm afraid that renders most of the product far less useful.

MenSys, or whoever, would need to take on SNAP on a commercial basis - buy the IP - and ask ATI to be successors under the NDA. It would then need to remain closed-source.

So, IMHO, asking the new owners of SNAP to open-source it is counter-productive. Not only would it be expensive to get the code reworked to get the ATI stuff removed, it might actually be cheaper just to buy the rights to the code and IP - assuming MenSys would be intrested in doing so.

Moby.
Title: Re: OpenSNAP geplant/OpenSNAP planned
Post by: magog on 2009.12.19, 19:35:41
Forget about GRADD that's not an issue. GRADD was IBMs attempt to get more video drivers by reducing the efford needed to get a working driver. It allows to implement hardware functionality in parts. Parts not covered are done in software (--> softdraw.dll).
Best example is GenGRADD which is doing everything in software using VESA functions. GRADD allows direct hardware access which SciTech used. I don't think IBM added special functions not available in the GRADD DDK.

As I said before I'm not sure Altrichmond has access to the source code of the closed source drivers as SciTech developed them in Partnership with ATI (plus some stuff under NDA with Matrox). If they (Altrichmond) have the full code, I'm sure at least ATI had to agree (most likely also Matrox).
If there are parts in the SNAP base that might be a problem but I'm not sure that's the case.
To get more information I strongly suggest to get in touch with Kendall Bennett (directly or over Steve Wendt).
Both should be able to clear the basic questions currently discussed here (are NDAs and code developed using NDAs part of the SNAP package Altrichmond got,...). After that you have to deal with Altrichmond directly.