@cgutjahr

I believe you when you say that. As I answered here I think it is time to free the platform of all that legal stuff. Elbox might have the right to act that way but on the mid- and longrun their "asset" (and that of other companies and devlopers) will have "zero" value because it will be replaced by opensource. That is what I try to help.

@OlafS25

there is a statement of elbox on amiga.org:
http://www.amiga.org/forums/showthread.php?p=717937

the driver was obviously not legal. But that does not change anything for me... basic components like drivers must be free and not "closed source"

@OlafS25

Quote:

As I answered here I think it is time to free the platform of all that legal stuff.

It's more complicated than that.

Example:If someone could have produced a new RoHS accelerator at faster speed and less cost than existing CSPPC/BPPC the following interests would rear their ugly head:

Do you make it backward compatible so that all classic users could take advantage of it or do you lock it to use with OS4.x?

Would it interfere with other plans deemed to have higher priority for long term growth?

If it used different methodology than what was currently in the pipeline by one group or another would it be opposed? (technical side)

Would personal egos cause those involved to think that their own contribution to a project was being snubbed in favor of another ego? (similar to the above but the personal side)

I could write more, but you get the picture. Depending on whom needed to be involved a perfectly sane money making project could die on the vine.

Quote:

There were (and are) too many big EGOs in the community and I have the impression that (in too many cases) people were more eager to make lawsuits and attack each other than to talk with each other and work together.

Yes. Add to that that most don't even know -how- to talk to one another.

#6

Last edited by number6 on 30-Mar-2013 at 12:55 PM.

_________________
This posting, in its entirety, represents solely the perspective of the author.
*Secrecy has served us so well*

@number6

Quote:

It's more complicated than that.

no, it isnt.

Quote:

Example:If someone could have produced a new RoHS accelerator at faster speed and less cost than existing CSPPC/BPPC the following interests would rear their ugly head:

first of all, dont develop a ppc accelerator. contrary to popular belief its completely pointless. the gain from that for a genuine 68k system is marginal and for os4 there is already a choice of faster machines. if people are hammering on and buying these accels for insane prices its just because they have once been told something and dont think again.

Quote:

Do you make it backward compatible so that all classic users could take advantage of it or do you lock it to use with OS4.x?

as i said above the gain for "classic users" (sigh!) is marginal and for os4 there is faster gear, so this argument is nil.

Quote:

Would it interfere with other plans deemed to have higher priority for long term growth?

what plans? are there any plans? and if there are any why should it be of concern. everyone is free to do his share. as on aros: one develops for arm/pi, the other for x86, yet another 68k, all development accompanies and helps common cause and each other by the way. even if someone of own initiative develops hardware aros may run on its not a problem (see fpga-arcade). no effort gets lost.

Quote:

If it used different methodology than what was currently in the pipeline by one group or another would it be opposed? (technical side)

we have been too much indoctrinated into that thinking. in this respect any progress is impossible because whoever developed a device and then developes one cheaper and even better, would not market it because it would hurt himself. think about x1k (hint).

Quote:

Would personal egos cause those involved to think that their own contribution to a project was being snubbed in favor of another ego? (similar to the above but the personal side)

might happen. but would not matter if they were dedicated to oppennes and cooperation in first place. again, think of aros. there are egos too, as strong as on o4 side. it is all possible to handle.

Quote:

I could write more, but you get the picture. Depending on whom needed to be involved a perfectly sane money making project could die on the vine.

it could, but as we see on example of natami, closed source, ego trip projects are rather expected to die. i dont point here to x1k, as this was backed up by an amount of money each else of us can only dream of investing in anything "amiga".

hope, i answered your considerations.

@wawa

I think you missed the point.

Olaf said:

Quote:

As I answered here I think it is time to free the platform of all that legal stuff.

I gave examples of impediments other than the legal side.

You seemed to agree with some of those.

Yet you said:

Quote:

no, it isnt.

As if there was nothing BUT the legal side to consider.

I could not disagree more.

Oh, and you seem to have missed "example" and assumed I'm arguing a point about a particular product.

#6

_________________
This posting, in its entirety, represents solely the perspective of the author.
*Secrecy has served us so well*

@wawa

Quote:

first of all, dont develop a ppc accelerator. contrary to popular belief its completely pointless.

I think the reason a lot of people sympathize with that approach is that they wouldn't have to set up yet another computer - they could simply expand their existing Amiga so it could run OS4.

That, plus you'd be using 'the real thing' and not some generic PPC motherboard with no real Amiga connection. Some people do care about that.

@OlafS25

Quote:

Elbox might have the right to act that way but on the mid- and longrun their "asset" (and that of other companies and devlopers) will have "zero" value because it will be replaced by opensource.

Again, I'm not defending Elbox. But as you say - it's their right to act this way. I would welcome a proper open source PCI system with modern drivers - I just have my doubts it's going to happen for AmigaOS - unless the AROS-68k stuff runs under OS3 without changes. There is practically no real development happening on AmigaOS 3, apart from a few assembler hackers trying to optimise some old Kickstart modules and a few AMOS or BB releases every year.

Quote:

basic components like drivers must be free and not "closed source"

I'd go one step further - if the last 20 years have teached us anything, it's that everything should be free (as in freedom, not as in free beer). Because the second someone is in charge, things start to go downhill.

Last edited by cgutjahr on 30-Mar-2013 at 01:46 PM.

@wawa

Its not issue if Radeon drivers are open or not, as long as the API standard for Exponention bus is open (thinking about AmigaOS3.x), and you can buy or get graphic drivers from some where, drivers are not so easy to make and it be better if was supported by vendor but its not the case, too small community.

Yes open source might be solution, to get people that do it because they can types can improve it, when a company does not see the value of continuing supporting a product.

Last edited by NutsAboutAmiga on 30-Mar-2013 at 01:54 PM.
Last edited by NutsAboutAmiga on 30-Mar-2013 at 01:50 PM.
Last edited by NutsAboutAmiga on 30-Mar-2013 at 01:48 PM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@cgutjahr

Quote:

I think the reason a lot of people sympathize with that approach is that they wouldn't have to set up yet another computer - they could simply expand their existing Amiga so it could run OS4. That, plus you'd be using 'the real thing' and not some generic PPC motherboard with no real Amiga connection. Some people do care about that.

Agree, and that covers the user perspective.

The business perspective in my example is supported by the statement made by Ben Hermans at AmiWest that classic OS4.x was the companies' best seller.

Any business recognizing their vehicle to increased sales and not focussing on how to exploit that (especially in light of admissions of needing cash), would be no business at all.

#6

_________________
This posting, in its entirety, represents solely the perspective of the author.
*Secrecy has served us so well*

@number6

Well you can't pay people for doing useless things, nor can it ever grow into some thing profitable.

Last edited by NutsAboutAmiga on 30-Mar-2013 at 01:53 PM.

_________________
http://lifeofliveforit.blogspot.no/
Facebook::LiveForIt Software for AmigaOS

@number6

Quote:

As if there was nothing BUT the legal side to consider.

many people chose to stay closed source or non public with their projects for personal rather than legal reasons, as in " its my project and i dont want others to mess with it". of course its their choice. but for us, as community only the legal matters, and only open approach guarantees any kind of "staying alive".

Quote:

Oh, and you seem to have missed "example" and assumed I'm arguing a point about a particular product.

neiter me, it was also an example.

@number6

the amiga-market is not "business" in a classic sense. I have heared statements like "x is the only future for amiga. If y succeeds we will leave the market". That is not a statement by business man how I understand that. In business world you are developing something for the platform that offers you the most opportunities. And if another platform replaces the platform you have developed for at the beginning you change because you would never just drop all your work. You could also say we will let our product die before someone else benefits of it. That is what I mean if I say "big Egos". And the unwillingness to work together by many developers and companies. That happened all the time and did a lot of harm. You see that as a small example with the Radeon drivers on OS4. The Acube customers had to pay for it. If the drivers would be free everyone would automatically benefit of every improvement. And what happens when the drivers are not developed anymore? Noone has the source and is allowed to do anything on them even when support has ended. I do not attack anyone here but only show why I think this way.

Last edited by OlafS25 on 30-Mar-2013 at 04:44 PM.

So we know elbox busboard can do 66Mhz,so all someone just needs the know how to write a driver for a third party accelerator card.

Waiting for something to happen for classic amiga can sometime be a little slow this is why I do something myself as I dont want to be waiting around for something to happen if at all.

Im working on 66Mhz PCI slot on Blizzard card another step-up from 37.5Mhz to 41.5Mhz

I have already posted some interesting benchmarks,but it should get quicker.

I need someone to test ragemem on classic with a busboard with screenmode set at 1024x768 32 bit.

Last edited by delshay on 31-Mar-2013 at 04:58 AM.
Last edited by delshay on 31-Mar-2013 at 04:55 AM.

_________________
The Machine: Bride Of The Pin•Bot by Williams Electronics

Quote:

delshay wrote: I need someone to test ragemem on classic with a busboard with screenmode set at 1024x768 32 bit.

You are probably more likely to get someone to help if you posted where to find ragemem, what AmigaOS (you say classic but it could be 3.9 or 4.x) you would like the test done on and which busboard you would like it done on. If the Mediator is your target, keep in mind that many users have the A1200 version which is slower.

Last edited by matthey on 01-Apr-2013 at 11:36 PM.

@OlafS25

Quote:

OlafS25 wrote: But that does not change anything for me... basic components like drivers must be free and not "closed source"

That can be tricky. If documentation is only available under NDA, then open-source may not be an option at all.

Sure, there are counter-examples, such as the reverse-engineered open Nvidia driver. How much effort goes into that? How many Amiga driver devs have that kind of resources and time? Oh, P96 is still an NDAed api, so you still cannot even make an open Amiga driver.

_________________
All glory to the Hypnotoad!

@billt

AROS uses a cybergraphx reimplementation (no documentation either) so it is solvable (of course it is harder). Your example only shows that I am right because this "it is my technology, I do not help you and it is secret" stuff has damaged the platform heavily (both involved Hyperion and MOS-Team).

@billt
Quote:

That can be tricky. If documentation is only available under NDA, then open-source may not be an option at all.

but will a serious company would even allow "amiga" developers to sing in under their nda? if there is one, that may be hans, who has been accepted, i dont know. doenst seem an option for the rest of us anyway.

Quote:

Sure, there are counter-examples, such as the reverse-engineered open Nvidia driver. How much effort goes into that? How many Amiga driver devs have that kind of resources and time?

the trick is that no "amiga" developers need to be involved. we can piggy back on work of bigger communities, as we do anyway other places.

Quote:

Oh, P96 is still an NDAed api, so you still cannot even make an open Amiga driver.

no need for cgx or p96, aros has cybergraphics, an api compatible solution. one could develop own drivers for existing zorro rtg boards, but for the time being there is just a wrapper under aros68k that takes advantage of freely available p96 dirver binaries. its working fine so far.

Last edited by wawa on 02-Apr-2013 at 12:59 PM.
Last edited by wawa on 02-Apr-2013 at 12:16 PM.

@wawa

Quote:

wawa wrote: @billt Quote: That can be tricky. If documentation is only available under NDA, then open-source may not be an option at all. but will a serious company would even allow "amiga" developers to sing in under their nda? if there is one, that may be hans, who has been accepted, i dont know. doenst seem an option for the rest of us anyway.

Sure. I'd be careful to not say "Amiga" when discussing things with them. Have genuine intentions of working with other things as well. Maybe not even talk about PowerPC. Talk about Embedded Systems. Have a genuine company or university involved. Start a small company if you need to. Sound professional. Be professional. Do not sound lik enothing more than a crazy bedroom hobbyist with 20 end-users to sell to.

I've succeeded more often than not.

Success: ATI (before AMD), AMD (gfx and embedded chipsets), Uli (m1575), Freescale (8640/41, 8610, t4240), Jmicron (Various Sata and flash memory controllers), PLX Technologies, EMS, a few PCB producers, Kontron (ComExpress, now under PicMG), Realtek (various ethernet controllers), Radisys, and half of these have had reps visiting in my dining room. (aka, home office for a startup, and I now genuinely have a corporate home office for my day job as well.)

AMD's embedded CPU/chipset stuff, and their embedded graphics are pretty easy to get NDAed into. TI has stuff on their website publically for anyone to grab.

Partial success: Via (USB3 controller, got PCB layout docs but not driver stuff, could couple with open-source driver examples), though Via has other times sent full datasheets including register defs for drivers without NDA.

Failures: Nvidia (no response of any kind from them), Intel (PCH chip, Thunderbolt), Ricoh (SD controller), NEC/Renesas (USB3).

Point is, get prepared, and ask. Be ready for some NO's, and have backup plans.

Then, if all else fails, go to Linux or other open-source for drivers, and buy some PCBs to pull stuff off of to reverse-engineer footprints and connections. IMHO, it's better to have docs from the source than something that somebody else figured out on their own. Granted, some are very good at that, and we've got some very good examples of this working very well. (Nouveau, Minimig)

But I'd still very much prefer real docs, and I will tend to support vendors that provide them in favor of those that do not (AMD vs Nvidia for example). TI has lots of great stuff freely documented, compared to NEC/Renesas. Only go down more difficult roads when something you want is not more easily available.

Last edited by billt on 02-Apr-2013 at 09:55 PM.

_________________
All glory to the Hypnotoad!

@billt

Quote:

I've succeeded more often than not. Success: ATI (before AMD), AMD (gfx and embedded chipsets), Uli (m1575), Freescale (8640/41, 8610, t4240), Jmicron (Various Sata and flash memory controllers), PLX Technologies, EMS, a few PCB producers, Kontron (ComExpress, now under PicMG), Realtek (various ethernet controllers), Radisys, and half of these have had reps visiting in my dining room. (aka, home office for a startup, and I now genuinely have a corporate home office for my day job as well.) AMD's embedded CPU/chipset stuff, and their embedded graphics are pretty easy to get NDAed into. TI has stuff on their website publically for anyone to grab. Partial success: Via (USB3 controller, got PCB layout docs but not driver stuff, could couple with open-source driver examples), though Via has other times sent full datasheets including register defs for drivers without NDA. Failures: Nvidia (no response of any kind from them), Intel (PCH chip, Thunderbolt), Ricoh (SD controller), NEC/Renesas (USB3).

yes, but in how many cases you have really succeeded (and i dont tallk about having representatives in your dining room) and in how many cases you have really had to sign that nda, because you have drawn many examples, where it was exactly *not necessary*!

Quote:

Then, if all else fails, go to Linux or other open-source for drivers, and buy some PCBs to pull stuff off of to reverse-engineer footprints and connections. IMHO, it's better to have docs from the source than something that somebody else figured out on their own. Granted, some are very good at that, and we've got some very good examples of this working very well. (Nouveau, Minimig)

but this is what you actually argument against?! open source is our best opportunity, its exactly what i say, dont switch the sits in the heat of discussion ;).

Quote:

But I'd still very much prefer real docs, and I will tend to support vendors that provide them in favor of those that do not (AMD vs Nvidia for example). TI has lots of great stuff freely documented, compared to NEC/Renesas. Only go down more difficult roads when something you want is not more easily available.

ok, thats a point. still it might be rather a general attitude that matters. even if you are officially under nda, serious partners, known to your conterpart might still become more attention when it comes to support.

im not saying, that what you say its wrong, however in our situation it rather doesnt matter that much, becuase, realistically, who in the entire "amiga" community has time and motivation left to dependably maintain such huge and demanding projects.

@matthey
Quote:

matthey wrote: Quote: delshay wrote: I need someone to test ragemem on classic with a busboard with screenmode set at 1024x768 32 bit. You are probably more likely to get someone to help if you posted where to find ragemem, what AmigaOS (you say classic but it could be 3.9 or 4.x) you would like the test done on and which busboard you would like it done on. If the Mediator is your target, keep in mind that many users have the A1200 version which is slower.

I posted in the wrong thread,there's another thread with almost similar heading.

Ragemem can be downloaded from OS4depot you also need OS4.x.

A1200 version maybe slower. part of this maybe down to memory access/bus speed (including PCI),latency and some other factors ie software or even firmware.

If users can address all the above factors & any I may of missed the A1200 can close the gap to other classic amiga(s).

Here is ragemem video bus benchmark with Blizzard PPC card PCI slot clocked at AGP speed

Read 8
Write 26

Screenmode 1024x768 32 bit. I needed to know how this benchmark compares to busboard(s)

Last edited by delshay on 03-Apr-2013 at 09:33 AM.
Last edited by delshay on 03-Apr-2013 at 07:20 AM.

_________________
The Machine: Bride Of The Pin•Bot by Williams Electronics