Any plans to increased the AmigaNG software?

[itix]

Even back in 1992 Commodore will have found it hard to make people use the new OS 3.0 API - but of course now we take it for granted. Developing for OS 2.04 or even OS 1.3 is unheard of.

I develop for Kickstart 2.0 and I have even gone so far that i have removed Kickstart 3.0 dependencies to get there.

In some cases i have made my programs Kickstart 1 compatible if it is possible without too much effort (lack of CreateNewProc is biggest obstacle).

I see no problem there. Other my sw then might require the latest MorphOS version.

[itix]

Target 68k and you can forget about using those features.  Use those features and you can have some pretty cool software that gives a more compelling reason for people to add OS4.x hardware to their collection.

Excuse me but why would developer give a damn if his software helps to generate more HW sales?

Lack of nice API calls like alpha blending and such is good argument to not have 68k build (you have to provide your own method to do that) but idea that developers are here to promote others work (OS developers or HW developers) is daft.

Unless you are OS/HW developer of course.

Personally I'd rather see compositing used wherever and whenever possible especially if software using it attracts more users.

Are developers getting paid if they use one of new features?

MorphOS has layers3d.library which I think has a lot of the same functionality of the compositing in OS4.1.  

In fact it is built-in to layers.library.

MorphOS has Reggae which can probably best be described as datatypes on steroids.

I'm sure MorphOS users would prefer software fully exploits these powerful features too.

Actually it does not matter as long as it does its job. Reggae is more flexible and advanced but user dont give a damn. If they did, you would not use OS4 anymore

If you just target 3.x and everyone gets the same thing no matter what system they are using then there's really no little point in having MorphOS, OS4 and to a lesser extent AROS.

Wrong. Even if you are just using 68k applications in MorphOS you benefit from greater speed, better user interface, hardware support and selection of software that comes as a standard.

[itix]

..and thats probably the main reason the Amiga never became anything other than a flash in the pan.  Everything was made to be compatible with an Amiga 500.  512k ram, 1 floppy.   If a majority of people think "all new hardware is alien", then they'll be stuck with software thats for underpowered machines instead of working on NG OS software.  But different strokes for different folks.. not my deal.  I have other hobbies I spend money on.  :-)

That is just Commodore's fault they didnt make new Amigas compelling enough. Kickstart 2.0 made sw development easier and it gained popularity in productivity sw market. But to regular users, no.

[Rob]

Excuse me but why would developer give a damn if his software helps to generate more HW sales?

I'd have thought that most users and developers of NG systems would like to see an increased user base.

Lack of nice API calls like alpha blending and such is good argument to not have 68k build (you have to provide your own method to do that) but idea that developers are here to promote others work (OS developers or HW developers) is daft.

Unless you are OS/HW developer of course.

You have to reinvent the wheel for each application that way unless developers share/open source well documented routines for other to use and will the hardware be there or will it be slow software rendered stuff.

Again most people wanted an increased user base for there NG system of choice, dont' you?

Are developers getting paid if they use one of new features?

Generally they use the new features because it makes development easier or perhaps allows for features that might not be possible/as fast otherwise.

Amiboing make heavy use of compositing in their games and there's also LiveForIt-mplayer, neither are getting handed bundles of cash for using the new (a number of years old actually) OS features.  They do it because it allows them to do something better, the whole reason the OS devs put the features in in the first place.

In fact it is built-in to layers.library.

Thanks for the technical correction.

Actually it does not matter as long as it does its job. Reggae is more flexible and advanced but user dont give a damn. If they did, you would not use OS4 anymore

The user may not care what code is responsible behind the scenes but it doesn't stop them appreciating the result.  Was Reggae developed in the hope that nobody would bother to use it?

Wrong. Even if you are just using 68k applications in MorphOS you benefit from greater speed, better user interface, hardware support and selection of software that comes as a standard.

Faster than 68k software on Amithlon, for instance.  So you want a better UI and better hardware support but you don't want the software to take advantage of it, strange?


Ultimately devs will develope for 68k, multi platform or a specific NG system based on their design goals and whether they need/want to use better feature offered in the NG systems.

[OlafS3]

The compositing engine in OS4.1 brings hardware acceleration to a lot of 2D drawing functions such as alpha blending, scaling, texturing and more.

Target 68k and you can forget about using those features.  Use those features and you can have some pretty cool software that gives a more compelling reason for people to add OS4.x hardware to their collection.

Ringhio is handy for notifying the user of events.  It would be fairly easy to have OS4 version that supports this and 68k version that doesn't.

Those are just 2 things I can think of.

Personally I'd rather see compositing used wherever and whenever possible especially if software using it attracts more users.

MorphOS has layers3d.library which I think has a lot of the same functionality of the compositing in OS4.1.  

MorphOS has Reggae which can probably best be described as datatypes on steroids.

I'm sure MorphOS users would prefer software fully exploits these powerful features too.

If you just target 3.x and everyone gets the same thing no matter what system they are using then there's really no little point in having MorphOS, OS4 and to a lesser extent AROS.

From what me have been told MorphOS (AROS/3.X) and AmigaOS are developing differently so if you use specific functionality of AmigaOS you are stuck to AmigaOS user (around 1000-1500 user I would guess) or you use MorphOS specific functionality then you are tied to the MorphOS user base. 68k is by far the biggest part and (if written properly) runs on AmigaOS and MorphOS too so from a commercial point of view it makes much more sense (as long as the specific features do not give obvious big advantages). Also writing PPC versions often means rewriting big parts (f.e. look at what happened with Art Effect and StormC AmigaOS ports). Magellan is a real exception but mostly there are reasons why the software never was ported. If you stay on 68k you can take the existing sources and add new functions and remove bugs (much less time than rewriting it).

There might be projects where it makes sense to drop 68k completely but that depends on the project.

[Thorham]

but don't hold back progress on our multi-GHz NG machines

You talk about progress, but are happy to be stuck with AmigaOS derivative operating systems. If you want progress, use a peecee with a modern operating system.

If you like Amiga land, then you've got to accept that you're always going to lag FAR behind what's available now, and it's most likely NOT going to change. Doesn't matter to me. I use an A1200 and a peecee (yeah, with Winblows). Best of both worlds.

Also, why should 68k developers who enjoy 68k cater to NG users?

for the sake of a 20 year-old machine.

Those old machines are what it's all about to many people, including myself (no nostalgia).

[OlafS3]

I'd have thought that most users and developers of NG systems would like to see an increased user base.



You have to reinvent the wheel for each application that way unless developers share/open source well documented routines for other to use and will the hardware be there or will it be slow software rendered stuff.

Again most people wanted an increased user base for there NG system of choice, dont' you?



Generally they use the new features because it makes development easier or perhaps allows for features that might not be possible/as fast otherwise.

Amiboing make heavy use of compositing in their games and there's also LiveForIt-mplayer, neither are getting handed bundles of cash for using the new (a number of years old actually) OS features.  They do it because it allows them to do something better, the whole reason the OS devs put the features in in the first place.



Thanks for the technical correction.



The user may not care what code is responsible behind the scenes but it doesn't stop them appreciating the result.  Was Reggae developed in the hope that nobody would bother to use it?




Faster than 68k software on Amithlon, for instance.  So you want a better UI and better hardware support but you don't want the software to take advantage of it, strange?


Ultimately devs will develope for 68k, multi platform or a specific NG system based on their design goals and whether they need/want to use better feature offered in the NG systems.

Wrong ... that is the widespread view in the so called NG camp. If people think commercial they want to sell software. Why was MS-DOS successful? Because of its great features? It was successful because of the huge user base (=huge number of buyers). Simply as that. So if you think in economic terms (what is not widespread at amiga developers right now) you do not care if AmigaOS or MorphOS give you any advantage as long as this features do not give you advantages that are so big that they justify the small number of potential csutomers. Or you are convinced that AmigaOS will become the next big thing like A1000 and you want to jump onboard early. I do not see this here.

[itix]

I'd have thought that most users and developers of NG systems would like to see an increased user base.

Do you ask Windows developers to not support Win95 anymore because of that?

Users/developers certainly wish there is more new and neat applications coming in the future and the platform is growing. But I dont think developers supporting 68k are slowing us down. MorphOS is growing nicely and I think that only way to boost that growth is to boost OS development. Writing new software must be made easier and that is task to OS developers.

You have to reinvent the wheel for each application that way unless developers share/open source well documented routines for other to use and will the hardware be there or will it be slow software rendered stuff.

It is up to developer choose what APIs he uses, what systems he supports and how much time he can or want to invest. I write the software for MorphOS only but if it looks I can port software to AmigaOS quite easily and I have time to do that I might do that. I have developed bunch of (slow) alpha blending functions that are compatible to CGX and so on. If it is too slow, too bad, buy faster Amiga. I have even replacement memory pool or AllocVec() routines so I can go to support even 2.0 or 1.3. Usually I cant because I am using MUI in GUI applications.

So I have limits what I am not ready to trade off for backward compatibility but may occasionally build 68k binaries without using any Kickstart 3+ calls.

Again most people wanted an increased user base for there NG system of choice, dont' you?

I would not mind but I am happy with MorphOS as a user and developer.

Generally they use the new features because it makes development easier or perhaps allows for features that might not be possible/as fast otherwise.

That is what I do. Then later I can write replacement functions for missing pieces. Maybe.

Amiboing make heavy use of compositing in their games and there's also LiveForIt-mplayer, neither are getting handed bundles of cash for using the new (a number of years old actually) OS features.  They do it because it allows them to do something better, the whole reason the OS devs put the features in in the first place.

What is so special in MPlayer? I presume there is no decent build for 68k and if there is there is no GUI and it would be too slow for real 68k. 68k build tools might be horribly out of date making porting too difficult and so on.

The user may not care what code is responsible behind the scenes but it doesn't stop them appreciating the result.  Was Reggae developed in the hope that nobody would bother to use it?

Reggae is part of OS and OS components are using it. I think MUI is not using it and relies on datatypes (IIRC, I have not checked it lately) but it is just using it to load images. No really reason to invest time to rewrite working code.

Users can ask if 3rd party programs are using Reggae and if not ask to add it. Developer may agree or not, it is his choice. We provide the toolbox and developers pick their favorite tools from there.

Faster than 68k software on Amithlon, for instance.  So you want a better UI and better hardware support but you don't want the software to take advantage of it, strange?

I dont have any specific needs in my mind at the moment. There is new software in development taking advantage of new APIs, like ebook readers. I suppose they use Unicode and charset conversion APIs provided by the MorphOS and not relying on 3rd party libraries. I dont have any ebooks yet.

[wawrzon]

@itix
there is 68k build of ffmpeg and ffplay from amistuff heavily asm optimized with the help of matthey especially for 060. it is already few years old, im not sure it has been updated. the result using it was mostly a slideshow, but it was a cool experience to mes with it.

what concerns your cgx replacement for alphablending functions, is this c or (68k?) asm, for chunky or planar graphics? would you consider to contribute it to aros(68k) if it was of advantage for them? i would happily establish a contact to krzysztof or neil.

[Wilse]

maybe its just me, but imho hd-rec is great, simple and intuitive tool to make raw recording and cut/edit them on the fly....

It's not just you; I agree with all of that.

Nonetheless, and retunring to my original point, I often need to record 3 or 4 inputs simultaneously. Even if I just want to record an acoustic guitar I'll almost always have two microphones on it. There is nothing on the Amiga that can do this.

If HD-Rec/OS4 could be modernised to accommodate multiple, simultaneous inputs (again, as one example), it would be a considerable improvement.

[wawrzon]

It's not just you; I agree with all of that.

Nonetheless, and retunring to my original point, I often need to record 3 or 4 inputs simultaneously. Even if I just want to record an acoustic guitar I'll almost always have two microphones on it. There is nothing on the Amiga that can do this.

If HD-Rec/OS4 could be modernised to accommodate multiple, simultaneous inputs (again, as one example), it would be a considerable improvement.

im not sure ahi allows this. im not sure too if it could have or ever been done with any sound card on amiga using any alternative driver system. i suppose if this was doable wanderer would have implemented that.

[itix]

@itix
what concerns your cgx replacement for alphablending functions, is this c or (68k?) asm, for chunky or planar graphics? would you consider to contribute it to aros(68k) if it was of advantage for them? i would happily establish a contact to krzysztof or neil.

I dont think you want that. It is slow generic ReadPixelArray() + modify pixels + WritePixelArray() without any optimizations. It only works for ARGB. But it has properly working GlobalAlpha parameter, I think.

[wawrzon]

@itix
i see. i think this is satisfactory implemented so far. im mostly concerned about improving  planar gfx in this and every other respect if even possible. but thx for the answer.

[kamelito]

Itix
". Writing new software must be made easier and that is task to OS developers."

That is not the first time I read that from you, so we're waiting for your tools to speed up development under Morphos or AmigaOS

Kamelito

[cgutjahr]

68k is by far the biggest part

That might be true, but how many of the people still 'using' m68k-amigaos belong to the target audience of a software developer?

Software development on 68k completely died down a decade ago. What was released after that is either a port of something maintained by outspoken PPC supporters (YAM, Wookiechat...) or written in AmiBlitz - which means that due to the limitations of the compiler, there'll always be a 68k built even if the developer doesn't care much about 68k (i.e. HD-Rec). As a platform for desktop users, 68k is in much worse shape than AmigaOS 4 or MorphOS. As in: "Even more dead".

A m68k port of AROS is nice for creating free/libre emulator distributions without depending on AInc. And it's a good gimmick for those people who collect, repair and polish old hardware and then want to have a shiny Workbench environment running on it that almost looks like it belongs to this century.

For everything else, m68k-AROS simply appeared way to late - which seems to be the common theme for AROS in general. If AROS would have been there when Amithlon took shape, the Amiga scene might look very different these days. But it wasn't, and now it's 2014.

[wawrzon]

That might be true, but how many of the people still 'using' m68k-amigaos belong to the target audience of a software developer?

Software development on 68k completely died down a decade ago. What was released after that is either a port of something maintained by outspoken PPC supporters (YAM, Wookiechat...) or written in AmiBlitz - which means that due to the limitations of the compiler, there'll always be a 68k built even if the developer doesn't care much about 68k (i.e. HD-Rec). As a platform for desktop users, 68k is in much worse shape than AmigaOS 4 or MorphOS. As in: "Even more dead".

A m68k port of AROS is nice for creating free/libre emulator distributions without depending on AInc. And it's a good gimmick for those people who collect, repair and polish old hardware and then want to have a shiny Workbench environment running on it that almost looks like it belongs to this century.

For everything else, m68k-AROS simply appeared way to late - which seems to be the common theme for AROS in general. If AROS would have been there when Amithlon took shape, the Amiga scene might look very different these days. But it wasn't, and now it's 2014.

christoph..  hello?!? anybody home?

aros68k a decade too late? wow?? as if anything else amiga was so much ahead of it. we are not discussing here any plans to catch up with windows, right? we just discuss what would benefit most of the interested parties involved in projects that all are inevitably doomed to fail.