27-Apr-2024 03:47 GMT.		 UNDER CONSTRUCTION
Anonymous, there are 277 items in your selection (but only 77 shown due to limitation) [1 - 50] [51 - 100] [101 - 150] [151 - 200] [201 - 250] [251 - 277]
[Web] AmigaShare Software Library expands its Platform support!ANN.lu
Posted on 10-May-2004 10:53 GMT by choochy277 comments
View flat
View list
Today, AmigaShare announces that it has included additional platform support that relates to uploaded software. That is, the Software Library now allows software items to show additional emulation compatibility. These include the following platforms:

* MorphOS v1.4
* AROS
* UAE and E-UAE
* WinUAE
* MacUAE

"It is important that people understand that we are first and foremost an Amiga site; our focus is AmigaOS. We want to encourage everyone to help grow the community to its full potential. That is, we are mindful that Amiga software does run, in emulated mode, on other hosts, so we have allowed for this.", said Chris Nillissen of AmigaShare.

"We also see this may be a perfect opportunity to expose users that are emulating Amiga Software, like UAE users, that aren’t up to date with the latest progress relating to the Amiga Platform!"

Visit www.amigashare.com
AmigaShare Software Library expands its Platform support! : Comment 201 of 277ANN.lu
Posted by Kjetil on 12-May-2004 04:32 GMT
In reply to Comment 199 (Kronos):
I agree whit you, 99.999% of all classic programs, and all AOS4.0 programs that do not follow the guide lines of AmigaOS40 SDK, will break, when Isolated memory space is added,
I just hope this will be feature you can turn on and off. As they say the feature exits inactive, I think that feature is going to stay inactive for some time, until there is large number of applications that can run native
AmigaShare Software Library expands its Platform support! : Comment 202 of 277ANN.lu
Posted by Sammy Nordström on 12-May-2004 05:43 GMT
In reply to Comment 201 (Kjetil):
Well, I don't think classic programs will be affected since they run within the emulators own address space or something along those lines. Please forgive me if I'm wrong, I'm not an expert when it comes to operating system memory managment. :-P
AmigaShare Software Library expands its Platform support! : Comment 203 of 277ANN.lu
Posted by Ole-Egil on 12-May-2004 06:09 GMT
In reply to Comment 133 (Amon_Re):
And for some reason you seem to reply to every single post Eva makes. What does that make you?
AmigaShare Software Library expands its Platform support! : Comment 204 of 277ANN.lu
Posted by Ole-Egil on 12-May-2004 06:10 GMT
In reply to Comment 140 (Jupp3):
Come on, you're not going to see a public beta of OS4 at ANY time. Arguing like that just makes you look stoopid.

A LOT of applications come out every year that has NOT seen a "public beta" period. A public beta isn't a holy grail...
AmigaShare Software Library expands its Platform support! : Comment 205 of 277ANN.lu
Posted by Amon_Re on 12-May-2004 06:38 GMT
In reply to Comment 203 (Ole-Egil):
Bored? :P
AmigaShare Software Library expands its Platform support! : Comment 206 of 277ANN.lu
Posted by Fabio Alemagna on 12-May-2004 07:25 GMT
In reply to Comment 195 (Choochy):
> How could I have worded it correctly what do i call this type of execution if
> its not "Emulated or VM'd".

"Compatible"?
AmigaShare Software Library expands its Platform support! : Comment 207 of 277ANN.lu
Posted by Fabio Alemagna on 12-May-2004 07:27 GMT
In reply to Comment 195 (Choochy):
> Secondly I dont understand how 68k code can run on PPC with some form of
> emulated execution. JIT is a form of emulation.

Sure, but the same applies to AmigaOS4, so by your logic AmigaOS4 is a VM or an emulator (take your pick).
AmigaShare Software Library expands its Platform support! : Comment 208 of 277ANN.lu
Posted by Fabio Alemagna on 12-May-2004 07:27 GMT
In reply to Comment 195 (Choochy):
> Secondly I dont understand how 68k code can run on PPC with some form of
> emulated execution. JIT is a form of emulation.

Sure, but the same applies to AmigaOS4, so by your logic AmigaOS4 is a VM or an emulator (take your pick).
AmigaShare Software Library expands its Platform support! : Comment 209 of 277ANN.lu
Posted by Ole-Egil on 12-May-2004 08:19 GMT
In reply to Comment 208 (Fabio Alemagna):
Help stamp out redundancy
AmigaShare Software Library expands its Platform support! : Comment 210 of 277ANN.lu
Posted by Ole-Egil on 12-May-2004 08:19 GMT
In reply to Comment 208 (Fabio Alemagna):
Help stamp out redundancy
AmigaShare Software Library expands its Platform support! : Comment 211 of 277ANN.lu
Posted by Amon_Re on 12-May-2004 08:22 GMT
In reply to Comment 210 (Ole-Egil):
This is redundant
AmigaShare Software Library expands its Platform support! : Comment 212 of 277ANN.lu
Posted by Amon_Re on 12-May-2004 08:22 GMT
In reply to Comment 210 (Ole-Egil):
This is redundant
AmigaShare Software Library expands its Platform support! : Comment 213 of 277ANN.lu
Posted by takemehomegrandma on 12-May-2004 08:43 GMT
In reply to Comment 174 (Kronos):
> Descent SMP can't work with current AOS-structures, so yes it will either be
> badly implemented or the have to break lots of old stuff.

I have tried on some occasions to discuss an idea about how this perhaps could be achieved; with an analogy to the way QNX handles it.

Please read this link:

http://www.ann.lu/detail.cgi?category=forum&file=1045927640.msg

(Read the above link before continuing)

My "thesis" was that since the Abox in practice is a multithreaded Quark process(or could be made that way?), the Abox itself (and all OS components and legacy apps running "inside" the Abox) could benefit from SMP once that is implemented in Quark, provided that quark is designed the "QNX way" in this aspect.

This would then work completely transparantly, without any need for hardcoding SMP support into old legacy apps, OS components etc. The "threads" inside the Quark process Abox (the Amiga "processes"?) would be dynamically spread out on different processors and scheduled by quark itself, *outside* the Abox, meaning that Abox and its running applications inside it (as well as any quark "native" OS components, drivers or *new* applications that in the future has been moved out from Abox into it's own quark memory protected space) would get the benefits from SMP without even knowing it, without even having the need to care about it.

Perhaps I am really clueless about this, but I have not yet heard any comments about this that has been able to convince me of this being impossible. This is how it's done in QNX, so the theories could evidently be turned into practice; the question is if it could be done in MorphOS.
AmigaShare Software Library expands its Platform support! : Comment 214 of 277ANN.lu
Posted by Anonymous on 12-May-2004 08:46 GMT
In reply to Comment 202 (Sammy Nordström):
> Well, I don't think classic programs will be affected since they run within the emulators own address space or something along those lines

Separate address spaces are not possible with AmigaOS, applications are exchanging address pointers all the time with each other and with the OS: messages, handles for windows, screens, files, tasks, functions etc are all address pointers and instead of well-defined interfaces that provide read/write access to such internals, it's typical to just wander through live list and structures. Bad bad bad design. What was good for a slow 68000, a lack of abstraction and encapsulation, is now a major design flaw.
AmigaShare Software Library expands its Platform support! : Comment 215 of 277ANN.lu
Posted by takemehomegrandma on 12-May-2004 08:49 GMT
In reply to Comment 195 (Choochy):
> I think what i am trying to get at is, if you were to write my announcement,
> how would you have written it?

I would have used the word "compatible" instead of "emulation". I would probably also have listed the OS's and Amiga emulators somewhat separately, but that is a minor thing IMHO.
AmigaShare Software Library expands its Platform support! : Comment 216 of 277ANN.lu
Posted by Sammy Nordström on 12-May-2004 08:52 GMT
In reply to Comment 213 (takemehomegrandma):
Well, I think you would have to make SMP support in the 68k emulator of the A/Box since it is the emulator that is providing the processor(s) with the instructions to execute when you're running 68k applications, right? Again, excuse me again if I'm wrong.
AmigaShare Software Library expands its Platform support! : Comment 217 of 277ANN.lu
Posted by Amon_Re on 12-May-2004 08:53 GMT
In reply to Comment 213 (takemehomegrandma):
I think that, for MOS, the problem comes from the ABox itself, i don't know how it handles different programs running in the ABox, but my understanding was that the Quark kernel only sees the Abox, not what's running inside it.

Also, iirc, the JIT compiler in the ABox runs all apps in the same thread, thus this won't benefit neather.

To make a good guess i'd need more info tho.

Cheers
AmigaShare Software Library expands its Platform support! : Comment 218 of 277ANN.lu
Posted by Sammy Nordström on 12-May-2004 08:56 GMT
In reply to Comment 214 (Anonymous):
Think; how is UAE able to execute classic Amiga applications ontop of certain "modern" Operating Systems with memory protection and resource tracking?
AmigaShare Software Library expands its Platform support! : Comment 219 of 277ANN.lu
Posted by Amon_Re on 12-May-2004 08:56 GMT
In reply to Comment 216 (Sammy Nordström):
It's a little bit more complicated than that, you have the Abox, wich is a Quark box, the Quark kernel only sees the Abox.

The Abox has an exec replacement, apps running in the Abox only see the exec replacement, not the Quark kernel.

For SMP to work, you'd have to find a way so that either every single thing in the ABox is run as a thread or proccess, or a way to make the Quark kernel manage the tasks run by the exec replacement.

Also, IIRC the JIT compiler runs every app in the same instance
AmigaShare Software Library expands its Platform support! : Comment 220 of 277ANN.lu
Posted by Amon_Re on 12-May-2004 08:57 GMT
In reply to Comment 218 (Sammy Nordström):
UAE however, doesn't provide these features to the apps running within the emulation, from an application viewpoint, there's no difference to it from running in UAE, the ABox or a real amiga.
AmigaShare Software Library expands its Platform support! : Comment 221 of 277ANN.lu
Posted by takemehomegrandma on 12-May-2004 09:12 GMT
In reply to Comment 216 (Sammy Nordström):
In QNX, no applications need to have "SMP support", they (the processes) only need to have enough threads to be spread out evenly on the processors, thus balancing the CPU load between the processors.

http://www.qnx.com/products/rtos/smp.html
AmigaShare Software Library expands its Platform support! : Comment 222 of 277ANN.lu
Posted by Sammy Nordström on 12-May-2004 09:20 GMT
In reply to Comment 220 (Amon_Re):
Sure. However, if I understood things correctly about how ExecSG runs native and emulated software, the emulator runs pretty much like an application ontop of ExecSG. All ExecSG does is identifying the application as either PPC or 68k and if 68k, then launches the emulator as any other ExecSG task while the 68k application itself is executed within this emulator task. This means that while the emulator task itself runs as an ExecSG task and is able to benefit from ExecSG features such as resource tracking and memory protection, the emulated 68k application is not. This is all just AFAIK and I may be wrong.
AmigaShare Software Library expands its Platform support! : Comment 223 of 277ANN.lu
Posted by Anonymous on 12-May-2004 09:24 GMT
In reply to Comment 222 (Sammy Nordström):
you are wrong
AmigaShare Software Library expands its Platform support! : Comment 224 of 277ANN.lu
Posted by Amon_Re on 12-May-2004 09:30 GMT
In reply to Comment 222 (Sammy Nordström):
You are partially correct, every task that's running in emulation is a seperate task asfar as ExecSG is concerned, the emulation is transparent (just like in MOS), if for instance, every task would be tracked by exec, tasks running in the emu would also be tracked.

Unless i'm wrong that is
AmigaShare Software Library expands its Platform support! : Comment 225 of 277ANN.lu
Posted by Alkis Tsapanidis on 12-May-2004 09:35 GMT
In reply to Comment 222 (Sammy Nordström):
The problem is that while the task is run into a seperate emulation process,
it still has to access the OS structures, so either you let it access everything
or you have a copy of the whole OS for use in a shared box of all 68k tasks.
AmigaShare Software Library expands its Platform support! : Comment 226 of 277ANN.lu
Posted by takemehomegrandma on 12-May-2004 09:46 GMT
In reply to Comment 219 (Amon_Re):
If the Abox process is multithreaded, then the quark kernel doesn't have to see the Amiga apps inside, and the Amiga apps does not have to see the quark, it would be enough if the exec processes in practice would be handled by their own respective Abox thread. But I am probably too clueless about this to see the real problems in this ...
AmigaShare Software Library expands its Platform support! : Comment 227 of 277ANN.lu
Posted by Anonymous on 12-May-2004 09:49 GMT
In reply to Comment 218 (Sammy Nordström):
> Think; how is UAE able to execute classic Amiga applications ontop of certain "modern" Operating Systems with memory protection and resource tracking?

Dude, rent a brain. Of course you can run any OS in an emulator that simulates a complete computer. Great, that gets you separate address spaces. That's not the point though. We are talking about MorphOS and AmigaOS4 and those do not emulate a computer that boots its own OS on top of a host OS. These operating systems just emulate a virtual CPU, to run under the same OS as the PPC CPU. You have one OS and one address space.
AmigaShare Software Library expands its Platform support! : Comment 228 of 277ANN.lu
Posted by itix on 12-May-2004 09:52 GMT
In reply to Comment 201 (Kjetil):
OS X does it nicely. You run old Mac OS9 software in a box and if it locks up you only lose the box, not an entire system. If you want isolated address space and proper MP and resource tracking this is way to go...
AmigaShare Software Library expands its Platform support! : Comment 229 of 277ANN.lu
Posted by Anonymous on 12-May-2004 09:57 GMT
In reply to Comment 222 (Sammy Nordström):
> All ExecSG does is identifying the application as either PPC or 68k

I wonder how you think that can work? On AmigaOS, an application (in the most simple form) is a task. That task constantly enters into third party code: the OS code, libraries that are called, callback hooks etc. On a system with truly transparent emulation, there is no way to say an application "is" PPC or 68k: It can be PPC one millisecond and m68k the next. I would assume that the decision if a task "is" PPC or m68k at a specific point in time is made on basis of the code page that is executed, not on basis of the task?
AmigaShare Software Library expands its Platform support! : Comment 230 of 277ANN.lu
Posted by Anonymous on 12-May-2004 10:06 GMT
In reply to Comment 228 (itix):
> OS X does it nicely. You run old Mac OS9 software in a box

Yawn, that's called UAE. Old OS in a box on top of a new OS. The inflexible, design of AmigaOS does not allow to reach a point of acceptable, seamless integration: instead of old applications that appear to run under the new OS, the best we can do is to run them on a virtual Workench screen on top of the real workbench screen. Not good enough.
AmigaShare Software Library expands its Platform support! : Comment 231 of 277ANN.lu
Posted by Amon_Re on 12-May-2004 10:19 GMT
In reply to Comment 226 (takemehomegrandma):
No, you are right, but i think that's only part of the problem, the OS functions rely heavilly on the use of registers, (eg D0 for passing pointers to functions iirc), in an SMP environment, you have 2 CPU's, and 2 sets of registers (or 3, or 4 or... you get the picture), part of the problem would be to make sure everything ends up in the correct registers, and there are more issue's wich i'm equally clueless about :P

In theory, it all sounds simple enough, in reality however, it's a whole different ballgame

Cheers
AmigaShare Software Library expands its Platform support! : Comment 232 of 277ANN.lu
Posted by Ole-Egil on 12-May-2004 10:21 GMT
In reply to Comment 229 (Anonymous):
Oh for heavens sakes, you know what he meant. Even if an m68k application calls a ppc library the application is still m68k. However, the TASK, that is, the ACTUAL RUNNING version in memory will at various points in time be ppc even if it is an m68k application that was loaded from disk.

And how do you distinguish between them? Simple, they are separated in space (memory space, that is).

Now all we need to do is to warp the space-time continuum...
AmigaShare Software Library expands its Platform support! : Comment 233 of 277ANN.lu
Posted by Choochy on 12-May-2004 10:22 GMT
In reply to Comment 206 (Fabio Alemagna):
"Compatible"?

LOL!!! Nice one!!!! It must have been late when I wrote that Announcement. Mind you that word was supposed to sum up all of the listed platforms!

Sure, but the same applies to AmigaOS4, so by your logic AmigaOS4 is a VM or an emulator (take your pick).

Hey, I never said that AmigaOS4 didnt do the same. One of my replies was quite to the contrary.
AmigaShare Software Library expands its Platform support! : Comment 234 of 277ANN.lu
Posted by Amon_Re on 12-May-2004 10:22 GMT
In reply to Comment 230 (Anonymous):
wrong, OSX hooks into the classic OS9 sandbox, a closer example would be Amithlon (replaces certain things with *nix things)
AmigaShare Software Library expands its Platform support! : Comment 235 of 277ANN.lu
Posted by Sammy Nordström on 12-May-2004 11:01 GMT
In reply to Comment 229 (Anonymous):
>On AmigaOS, an application (in the most simple form) is a task.

Well, it *becomes* a task when it's executed. Until then, it's simply an executable binary compiled for a specific processor. What happens when it's executed depends entirely on the operating system and what it decides to do with it. Again, I'm not an expert in this area and I may be wrong, but I do believe what Ole-Egil just wrote, ie that AmigaOS4 seperates 68k tasks from PPC tasks by using seperate address spaces.
AmigaShare Software Library expands its Platform support! : Comment 236 of 277ANN.lu
Posted by ms on 12-May-2004 11:33 GMT
In reply to Comment 235 (Sammy Nordström):
In AmigaOS everything must be in the same address space including PPC code. The difference in OS4 is that memory area reserved for PPC code is marked executable and rest is non-executable.

You can have isolated address spaces in limited use if memory is marked private. But you can't isolate PPC and 68k programs from each other, really. Read exec.library/PutMsg() doc and you understand it.
AmigaShare Software Library expands its Platform support! : Comment 237 of 277ANN.lu
Posted by Kronos on 12-May-2004 12:08 GMT
In reply to Comment 201 (Kjetil):
SW for OS4.000 still uses MSG-ports and the like, so you just can't put them in isolated address-spaces.

The problem with OS4 is that it runs all SW on the same kernel, utilizing the old structures, which makes it absolutly impossible to protect that kernel. Even if they add new APIs, one old app would still be enough to make MP for the kernel impossible. And even SW compiled for OS4.0000 falls under "old" in this context. The only way out of this is a sandbox, enclosing 68k/OS4.0 SW and a adapted ExecSG running on top of another kernel (which might be called ExecTG, but won't have much in common with Exec as we know it).


@Sammy
UAE is one big overgrown sandbox (AOS-tasks are not directly running under the host kernel, and neither can they directly access any resource).
AmigaShare Software Library expands its Platform support! : Comment 238 of 277ANN.lu
Posted by Alkis Tsapanidis on 12-May-2004 12:27 GMT
In reply to Comment 237 (Kronos):
It might as well be ExecSG with memory protection enabled running a stripped
version of itself with MP disabled in a private space.
That would actually be a boxed approach, all legacy apps would run in an
isolated environment. IMHO that would be a nightmare, as they would have to clone all libs, resources, etc. in the shared environment as well.
AmigaShare Software Library expands its Platform support! : Comment 239 of 277ANN.lu
Posted by Sammy Nordström on 12-May-2004 12:30 GMT
In reply to Comment 236 (ms):
Well, again I'm not an expert. All I know is this:

--- 8< ---

"AmigaOS 4 uses a task-based emulation which is directly integrated into ExecSG to emulate legacy 68k code. Each emulated task is scheduled independently and has its own emulation process and interpretation/translation stage independently from other emulated tasks."

--- >8 ---


--- 8< ---

"While this is in theory less compatible than a sandbox or "virtual machine" type approach, it offers a few key advantages:

More responsive; such an approach does not break time-critical/real-time behavior.

Clarification: With a sandbox approach, a separate scheduler runs inside the sandbox to schedule 68k tasks, which increases delay times substantially. A task-based scheme does not separate 68k and PPC tasks. They are treated equally."

--- >8 ---

--- 8< ---

"Legacy applications may benefit (to some extent) from additional functionality introduced by ExecSG (resource tracking and memory protection, for instance), albeit in a limited way."

--- >8 ---

Reference: http://os.amiga.com/os4/OS4FeatureSet.php

Exactly how this is accomplished is for Hyperion to know but for me and you to speculate about. :-P
AmigaShare Software Library expands its Platform support! : Comment 240 of 277ANN.lu
Posted by Ole-Egil on 12-May-2004 12:47 GMT
In reply to Comment 237 (Kronos):
What you're not thinking about is that OS4 could be used as a bridge here. If you have enough software that handles the newer API fine, you can enable MP. If not, it'll break your software. So yes, it will break old software. But at some point that will be considered "fine". For instance, once there is enough new software to make it acceptable to run all the old things in UAE along with all the machine-dependant programs out there.

Only time will tell, and for some odd reason I'm not willing to take your gloomy forecast as the ultimate truth ;-)
AmigaShare Software Library expands its Platform support! : Comment 241 of 277ANN.lu
Posted by Alkis Tsapanidis on 12-May-2004 12:49 GMT
In reply to Comment 239 (Sammy Nordström):
Clarification: With a sandbox approach, a separate scheduler runs inside the sandbox to schedule 68k tasks, which increases delay times substantially. A task-based scheme does not separate 68k and PPC tasks. They are treated equally."
--

Depends on what you mean.
In MorphOS, the whole A/Box is a sandbox, there is no seperate scheduler for
68k or PPC tasks. They are all scheduled using ExecPPC. For exec, all tasks
are PPC, there is NO distinction between them.
AmigaShare Software Library expands its Platform support! : Comment 242 of 277ANN.lu
Posted by Ole-Egil on 12-May-2004 12:51 GMT
In reply to Comment 240 (Ole-Egil):
To rephrase that a bit:

Don't see OS4 as a better way to run your old programs. See it as a way forward. SDKs in the hands of everyone will hopefully create gazillions of bedroom coders, just like it did once before.

It's just like with MOS. Some people can discuss the m68k emulation all day long, while others are busy making new software instead. If all you want is to run DPaint quicker then WinUAE is the right solution. If you want to be part of something new and creative, OS4 or MOS would be a smarter place to start.

Can we at least agree on that one? ;-)
AmigaShare Software Library expands its Platform support! : Comment 243 of 277ANN.lu
Posted by Sammy Nordström on 12-May-2004 12:59 GMT
In reply to Comment 242 (Ole-Egil):
I second that! =)
AmigaShare Software Library expands its Platform support! : Comment 244 of 277ANN.lu
Posted by brotheris on 12-May-2004 13:02 GMT
In reply to Comment 239 (Sammy Nordström):
You need to spill scheisse on competitor product in your featurelist. Nothing can beat that! ;-)
AmigaShare Software Library expands its Platform support! : Comment 245 of 277ANN.lu
Posted by ms on 12-May-2004 13:16 GMT
In reply to Comment 239 (Sammy Nordström):
"Legacy applications may benefit (to some extent) from additional functionality introduced by ExecSG (resource tracking and memory protection, for instance), albeit in a limited way."

Yawn. We had MP in AmigaOS 2.0 already. Do you remember Enforcer and MuGuardianAngel? AmigaOS 4.0 does some, this time it is integrated into OS rather than 3rd party software.
AmigaShare Software Library expands its Platform support! : Comment 246 of 277ANN.lu
Posted by ms on 12-May-2004 13:17 GMT
In reply to Comment 240 (Ole-Egil):
"What you're not thinking about is that OS4 could be used as a bridge here. If you have enough software that handles the newer API fine, you can enable MP. If not, it'll break your software. So yes, it will break old software."

Sounds like MEMF_PUBLIC failure.
AmigaShare Software Library expands its Platform support! : Comment 247 of 277ANN.lu
Posted by takemehomegrandma on 12-May-2004 13:24 GMT
In reply to Comment 242 (Ole-Egil):
Since "enabling" certain functions will brake everything we today (and probably quite a considerable amount of the future) consider to be "Amiga" (tools, applications, "how things are done", even current OS components(?)) with a great crash-boom-bang, I see a risc of this never happen. The "bridge" as you say, might be endless, and thus a poor "way forward".

A better (the only?) option I see would have been a forced crash-boom-bang approach without delay, a completely new OS all the sudden, that brakes everything, and starts on a "square one" and try to build a future from there. Kind of like Amiga Inc's original strategy, when they announced that "the Amiga is dead, long live the Amiga" and introduced the Amiga DE as the next AmigaOS. But the question is, who wants that (I'm not talking about AmigaDE, I'm talking about braking what we know as "Amiga")?
AmigaShare Software Library expands its Platform support! : Comment 248 of 277ANN.lu
Posted by Don Cox on 12-May-2004 13:45 GMT
In reply to Comment 242 (Ole-Egil):
"Don't see OS4 as a better way to run your old programs. See it as a way forward."

Problem is, I have plenty of programs. I just want to run them on faster hardware with more memory. AOS4 might be a way to do this.

Particularly ImageMaster. It doesn't work on Amithlon, and I havent managed to get UAE/AF to see an Amithlon partition yet.
AmigaShare Software Library expands its Platform support! : Comment 249 of 277ANN.lu
Posted by Anonymous on 12-May-2004 14:33 GMT
In reply to Comment 242 (Ole-Egil):
> Don't see OS4 as a better way to run your old programs. See it as a way forward

If it was meant as a way forward, it would have a completely different design. As it is, neither MorphOS nor AmigaOS4 nor AROS are much of a way forward. They all re-invent an obsolete OS, this time on PPC and x86, with no substantial changes. To be fair, it's not as if you have much choice, there is only one fundamental decision to be made: should it be compatible with AmigaOS3? If that is answered "yes", all important decisions are made for you. You can still make decisions about the color of icons but not about the fundamental issues of the OS. You will never gradually get from there to a modern OS. Since the answer to OS3 compatibility was "yes", your comment above, that you shouldn't think of OS4 as an OS for old software, makes no sense: that is exactly what OS4 is for. If it was not, it would have a smarter design. There wouldn't be 108 character file name limits, five different gadget toolkits (intuition, boopsi, gadtools, reaction, MUI), exposed system structures without tracking, unaccounted passign around of pointers, heavy-handed arbitration with Forbid() etc. etc. I don't know where to start, there is so much wrong with AmigaOS. My personal hope is that OS4 and MorphOS allow these companies to make enough money to stay alive long enough to make a real OS.
AmigaShare Software Library expands its Platform support! : Comment 250 of 277ANN.lu
Posted by Alkis Tsapanidis on 12-May-2004 14:41 GMT
In reply to Comment 249 (Anonymous):
The whole point in the design of MorphOS is that it is ready to accept a new
OS structure in the Q/Box and yet keep compatability using the A/Box.
Of course this is a *LONG* term solution:-)
Anonymous, there are 277 items in your selection (but only 77 shown due to limitation) [1 - 50] [51 - 100] [101 - 150] [151 - 200] [201 - 250] [251 - 277]
Back to Top