The OpenAmiga Standard

[bloodline]

mrsad wrote:
The OpenAmiga platform is a great idea, but it probably will not work. Simply because we are not using an 'open' OS in the same way as the BSD or Linux people are.

AROS anyone? :-)

-Edit- Your comments are valid, but only if there was no open source alternative! But there is (AROS), and one would hope that Competition with a Free alternative will make Amiga Inc. and the MOS team put in more work and try to make better systems.

[bloodline]

Crumpster wrote:
Great thinking everyone.

It makes a refreshing change to see us pulling in the same direction for a change!!

Hell, it's great to see a thread going over 60 posts without a flame in sight!!  :-D

This is a really good idea, nice one guys.

Graham C

Well, it's about time!!!  :-D

I hope this will allow the Community to tell the companies in power what they really want.

We all want a standard, we know that we need a standard, developers want standards, users need stardards...

I'm really pleased, I'm sure it can work.  :-o

[bloodline]

Dagon wrote:

We all want a standard, we know that we need a standard, developers want standards, users need stardards...

= More apps for everyone :-) You will not have to make specific versions (only when it is very necessary) for each OS.

AmigaOS 4.x, MorphOS and Aros will be mutually benefited. For example a game that makes i.e. Epic will run or will be very easily ported in all of them. ;-)

That's the key, easy portability. If all a developer has to do is change the compiler includes directory or the CPU type and then just recompile his app/game. No modifications to his source needed, then he is more inclined to port it!  :-)

[bloodline]

downix wrote:
For the mascot, ya forgot the butterfly wings and fur!

And the pussy cat!!!! :-D


-Edit- Well, this has certainly come quite a way since Mdma and myself started this little "rant" :-D

The website actually looks cool, and people are sitting up and taking notice!?!?! All in the space of a few hours...

I think many people can see the advantages to more users, more developers and more software :-D

[bloodline]

Well POSIX needs to be thought out... as to the best solution so not yet...


But MUI is definately yes :-)

[bloodline]

mdma wrote:
I don't think we need to specify what compiler a developer should use.

It's a personal thing, just like what aftershave you wear. ;-)

What if someone wants to use a lnaguage other than C, but they write their app to be 100% OpenAmiga compliant in everyother respect?

Sure but we need to define a standard, that will work across theboard... such is the way of standards :-)

If some one write using say VBCC, ten that might not exist on another plat form, remdering the whole point of this excercise uselss :-(

[bloodline]

Floid wrote:

downix wrote:
For the mascot, ya forgot the butterfly wings and fur!

Some sort of chimera would probably be apropos...

Sod it, Someone just render a Black and white Checkmark tick that look like pressed Aluminium... sort of like the new Apple logo...

[bloodline]

greenboy wrote:
Read THS and THIS TOO and tell me if it is possible for developers to all embrace the same Initiative.

Actually, about a week ago I had talked to a developer who is both in Phoenix and on the Open AWeb team, and he came back to me with a No Go on my query of whether the AWeb team would like to present themselves as a participant. I queried because I wanted to see them involved so that I could offer up a matching donation to Amizilla with more confidence that a browser would come out the other end.

Nobody else from that team even bothered write me. I dunno, maybe I was never mentioned as the originator of the query - or maybe I WAS, and that made it impossible to talk about it ; } ...Either way, it doesn't make the picture look rosy for getting all qualified parties speaking WITH instead of TO each other, and dicsussing issues and finding common solutions.

Lets keep this in perspective... All we are doing is defining a set of minimum standards  which if develoeprs follow (and most already do wirthout realising it, due to the standards we have chosen) then it will be easy to port their apps from one platform to another... we don;t really ask anything else of them.

[bloodline]

uncharted wrote:
Ok so this seems to be more than a nice idea.  In theory it should be able to work.  The only problem I can see is that some of the APIs are for commerical products, but there are always ways around that.  Now what is needed is for someone to explain exactly how it WILL work.   We need someone to come up with a detailed plan and proposal to put forward to the current players.

The next stage after that will be to present the whole OpenAmiga initative to the AROS team, Genesi/Pheonix and Hyperion/Amiga and get them to sign up to it.

@Greenboy
This would be perfect for that progamming inititive I discussed with you.

Well, AROS is the reason we can do this. It provides all the whole standard (it has free, opensource alternatives to the comercial APIs covered in the openamiga spec) in one free package.


-Edit- A detailed plan would be in essence: make sure MOS and Hyperion keep AOS3.1 compatibility and introduce extentions in a system friendly and API documented (ie, they can tell us what they are and how they should work, but not tell us how they did it)

The Other APIs need to have OpenSource alternatives only if they do not already exist on one of the platforms (they all do, and AROS has opensource version of those that are closed).

The Openamiga group needs to be able to make sure that all the system can provide a base standard functionality...

So we can only raise the base standard when all system are able to meet it, understand... Openamiga needs to do nothing outher than define what id the minimum that all thesystems can do.

The APIs I have proposed cover that minimum already, well almost  (ie in the important areas) :-D

[bloodline]

greenboy wrote:
...But don't take it wrong, everything I've said as devil's advocate: I applaud this effort.

IMHO, the hardest thing will not be developers... since they will be using these APIs anyway... the hard thing will be getting Hyperion and MOS to make sure they stick to the script and don't try any dirty tricks to introduce incompatibilities... one would hope the communities desire for standards would force them to compliance :-D

[bloodline]

MDMA, you might want to make it clear on the website, that the choice of CGX was not in any way against P96... but simply that P96 uses the CGX API!!!

It seems many people are having a hard time grasping this concept even now... after 7 years!!!  :-x

[bloodline]

GadgetMaster wrote:

filson wrote:
Maybe it would be better to create a seperate section for ppl to look at if they were in doubt of the meaning of the specification.
It can easily become very messy if all things have to be explained all together with the specification.  :-)

and it would make my life easier  :-D  :-D

I smell a FAQ coming up

This sure is a strange community, one minute we're all fighting and the next we're all trying to work together. ;-)

I hope something positive comes out of all this.

It's been a log time coming!!!

But it is great to see so many people who generally moan and fight, come together and agree on something.
It seems we all want something that is so simple, but will ultimately benefit everyone.  :-)

What we really need is some heavy weight behind the idea, to show just how important this is.

[bloodline]

mdwh2 wrote:

Rodney wrote:
I like that idea. Having more then one library to do the same thing is a bit silly.  I'd think the OS3.1 APIs should be the path to stick to... Adding more libraries just adds fractures to the OS/community.

I think we should keep it simple. Both OSs will prolly support the base install of OS3.1  (through emulation or whatever) so i think OS3.1 is the place to start.

AIUI, the point about this is that it's a set of libraries/features which are available across the various AmigaOS clones. So MUI is included because it's available on AmigaOS and Morphos, and has a clone on AROS - not because they're trying to say it's "better" than Reaction, or are trying to split the community.

The problem with only supporting the basic OS3.1 API is that it's getting rather dated. Trying to program gadtools used to drive me insane.. later toolkits such as MUI are much more powerful.

While it is true the 3.1 kernel is old, it should be noted that it is tried, tested, it works and it is well known.

Openamiga is not trying to hold back developments, but give developers an idea of what the minimum API is supported across the Amiga Clone platforms.

I would hope that improvemetns to the 3.1 Kernel can be suggested via openamiga and then implemented on all the clones in an API compatibile way.

If we don't do this now, the three systems will probably drift apart, splitting the community irreversibly.
This, as they say,  is our last best chance for peace :-D

[bloodline]

NewRevolution wrote:
May I come with a suggestion? For the website - openamiga.tk, for each "standard" that is to be made, could it be possible to describe it?

E.g. MUI, CGX, OpenGL - they all need to be defined. By that I mean there must be a common cencus / agreement of what these standards are, and then maybe have som more technical specs. about them - as they aredescribed  now.

I don't know much about these standards, but I'll bet different ppl have different views on these "artefacts".

And just a quick Q about the network part. Wouldn't there be a good idea to develop APIs for 802  LAN/MAN Standards? That also includes 802.11 (a and b) wireless. IIRC the A1200 had a PCMCIA port? Excellent for wireless! You can read more about these standards on www.ieee802.org

It's important to remember we are not talking about drivers of the underlying systems, that is up to the OS teams.

What we are talking about is the the interface between the OS and the apps...

[bloodline]

Good point raised by Dan Cox on Ann...
We really should include AREXX compatibility into the openamiga spec.
Though I don't know if MOS has it? If not then they can use AROS's Regina system I guess.

Side note:
It would be ince for all programs to at least support Arexx in some way.

[bloodline]

Rodney wrote:
Well, as i see it, AmigaOS and MorphOS are supporting 3.1 APIs or something similar. Once they start to get a few native apps, they'll drop that emulation quicker then you can say, "drop that emulation" .

So keeping up with the times is a good thing. I think what woudl also be a good thing is to try add third party apps rather than official MorphOS or AmigaOS things to the set of features/requirements.

MUI is a good example. MUI will be included in AOS4 for some time to come. It will also be avaliable in MOS and in a compatible form for AROS.

But, as i've said in the news item discussion. In stead of trying to get users/developers to use these sets of tool, i think we should push AmigaOS/MorphOS/AROS developers to work in POSIX compliant. That way, we'll benifit the community on all platforms but we'll also have a wide range of software.

I know this will be a lot more difficult, because this will require some development time, but the result i think would be much worth it.

Not that im saying this isnt a goood idea. I needs to be done. Just correctly.

As I've tried to point out, POSIX isn't really an issues here.. we are talking about things that POSIX does not cover at all!!!

In fact choosing a POSIX compatibile API will be an integral part of the openamiga standard.


-Edit- one of the reasons why I'm a keen on using AROS (or any AmigaOSoid for that matter) other than, say the rather fine, AtheOS is because I don't want just another POSIX OS... sure have support for POSIX, but don't let it rule you... we have our own standards lets look after them. :-)