Given x1000 news is everywhere.....but nonchalance prevails

[Fab]

Reason always loses in amiga land. MorphOS developers tend to belittle AROS, claim that it is messy, that it has no apps, that it is mostly broken etc. Yet, where did they go when they needed to get rid of the need for OS3.1 binaries?

Well, there are still not many apps compared to 3.x, MorphOS or OS4, even if the situation has quite improved in the last months (which might even allow someone to use it for "real", with a few key apps like a browser, a video player, ...).

Obviously they can't be blamed for that. The developers have clearly focused on writing drivers and making the OS run on as much hardware as possible, which is a tedious task, leaving little time to write actual applications, or even polish existing codebase (hence numerous bugs and basic functionality in most components, like console, preferences, desktop, ...).

The situation is the opposite on MorphOS (or even OS4 to some extent), where the supported hardware is quite limited, which gives more time to polish the system, where components have more functionality and where most bugs could be ironed out. And obviously, the fact they had transparent 68k emulation since the start made the initial lack of applications much more bearable (i still don't see AROS janus uae as a real solution for 68k integration, except for games).

This is where the latter approach has an advantage... It gives an usable system (you could always argue on it, but i don't care for daily use, even if it's on outdated hardware. Having a half-finished system on modern hardware can't attract many end users. There was probably noone using exclusively AROS when there was neither a browser nor a video player, for instance.

And of course, you're right that mac hardware for MorphOS or x1000/SAM for OS4 will only help for a couple years. Then drastic changes will have to be made anyway, if they are to survive on modern hardware.

Regarding the comparison with UAE, the answer is easy: OS3.x is dead, and very few (real) applications are still written for it. The development is clearly targetted for the new OSes.

[Fab]

Janus isn't the only option available for integration within AROS. See here for more details.

Well, i thought this was what janus uae was based on. Anyway, same thing. This attempt to integrate UAE transparently seems like a very difficult task, and i doubt it will ever reach the level of integration MorphOS or OS4 can achieve.

That'd be fine if it wasn't for the fact that the vast, vast majority of the Amigas software catalogue wasn't made specifically for the 3.x series. Over and above this, none of the three follow-ons have much by way of software.

I said 3.x, but MorphOS is compatible with previous versions too, as long as they don't do evil stuff.


@RuneRequester

Isn't OWB the same browser used for Morph ?

The name and the engine (WebKit) are the same, but that's it. The UI and features of OWB are totally different on AROS, MorphOS and OS4, some versions being more advanced than others. See http://fabportnawak.free.fr/owb/owb-morphos-1.8.readme feature list: many of the listed features aren't available in the AROS or OS4 versions.

[Fab]

Any particular reason this would be the case? PowerPC isn't any closer to Turrican 2 than x86 is.

First, MorphOS and OS4 don't emulate the chipset, so Turrican would only run through UAE anyway.

Now, for the applications/games that don't rely on amiga chipset (and there are a lot of them), they can run seamlessly on MorphOS, through the integrated 68k emulator and the system's compatible API and ABI (the latter being something AROS doesn't have, and can't easily have, since the x86 endianess isn't the same as the 68k one, which would be a major issue when dealing with system structures, which are unfortunately accessed directly by almost all applications).

[Fab]

Firstly, sorry this is the link I was aiming for originally. My bad.

I imagine it is, but no more so than writing a from scratch emulation layer for MorphOS or OS4 I would imagine. The engine at least is already done. Setting up a layer to pass between native and emulated is underway now as can be seen from the screenshots.

As far as I know neither MorphOS or OS4's emulation layers offer chipset support. However as with that link above shows there is more than one way to skin a cat.

We'll see how far this project goes, but let me give a few issues that this UAE implementation currently has (and which might be difficult to resolve):
- different UI style between the native OS and the emulated applications (3.x vs AROS). More generally, the emulated applications don't benefit of any AROS enhancement.
- No transparent drag'n'drop between AROS and UAE.
- All kinds of notifications (DOS or anything else) won't be signaled between AROS and UAE.
- What if you want to run some emulated MUI application on another screen, how will UAE handle that? What will happen to the other applications that should stay on wanderer screen?
- To be seen if the actual implementation shares filesystem, clipboard and so on correctly too.
- Is network support even enabled in this UAE version?

I could probably think of hundreds of other scenarios where you could have similar glitches. Some can probably solved, but you see the kind of problem this solution has to face with.

You're right about MorphOS and OS4 not emulating chipset, but almost all applications requiring it are either games or trackers or paint apps. Most of these would run fullscreen anyway, and they can be integrated "transparently" (according to AROS terminology) by running (or reusing) an UAE instance when doubleclicking them from the host OS.

[EDIT]

I just checked that other link. Need to read about it first before commenting.

[Fab]

Note that the second link (which was the one I should have posted first) is much more in line with the emulation layer offered by both MorphOS and OS4.

Yes, i read a bit about it. It's still very preliminary, but it's more the kind of approach i would expect to achieve a decent integration, indeed. So we'll see how far this project goes, and it will be a huge work too, but definitely interesting.

[Fab]

Every Amiga "alternative" has a right to exist and it's own merits on why it should. My biggest pet peeve with the entire community, however, is the fact that there's no camaraderie. There's so many applications for A that aren't on B or C. Sure people are all, "You're welcome to port my code, if you'd like", but there also saying "but I'm not going to do it". I understand why they say, do and feel like that, but the "camps" refuse to just say "hey, we're different, but that just encourages evolution, let's work together and add compatibility between the three to allow even faster development". They would rather pick the other apart, like bitter siblings.

Since you're explicitely referring to one of my previous answers in some other thread, may i remind you once again that I gave the source code of MPlayer MorphOS for AROS, and helped Deadwood to port it (who was much better placed than me to port it, since he can actually run AROS, and also has better knowledge about the OS and its specificities)? I also gave MAME MorphOS sources to another AROS developer, who managed to port it successfully. I can also "call out for bullshit as soon as i see it".

I would do the same for OWB on AROS, if someone motivated enough volunteered to port it. It's not a question of refusing to port it... It's a question of doing it properly.

On the other hand, i would have a serious problem passing my code to people that would get donations (or even commercial outcome) issued from my work (especially since i refuse them).

[Fab]

Fab, that wasn't directed solely at you and you certainly weren't the first person to say that, either.

So it was *also* directed at me, which justifies my answer.

Now, let's get on to subject matter. He can "he's in a better position, 'cos he can run AROS and you can't (sic)". Cop-out. You can't go get a low-end, second-hand PC to run AROS? You're telling me you don't have a machine that could run AROS already? I find that hard to believe, since 98.9% of the world pretty much have a PC compatible machine in their house. Since you wish to go back to making up excuses, might I remind you that BTBuilder was linux only and was ported to Windows. Dennis uses Linux exclusively, but his wife has a Windows box. So, he used her machine to port his code, even though he doesn't (have) "better knowledge about the OS and its specificities (sic)". I believe that is called Pot, Kettle, Black.

The only PCs I use are dedicated to work, and i don't own any PC. My pegasos and mac mini are my main machines at home and my basement doesn't have room for more. I still have access to a distant linux server which allows me crosscompiling stuff for MorphOS (compiling webkit takes ages on a G4), but it wouldn't allow me to run AROS anyway. So there's no way for me to crosscompile for AROS or OS4 (68k would still be possible, though, but tedious, given the SDK state).

Anyway, the point isn't even there. It's already more than enough to have amigaos/morphos/aros code ready to be compiled for each other flavour (with very minor modifications, most of the time). It's not like porting between different amiga flavours requires starting the port from scratch. So, "use the good tool for the job", or rather, ask someone familiar with the target system to complete the port for his preferred system, it's much more efficient that way. I appreciate some people can afford compiling their programs for the 4 amiga flavours: jahc does that with wookiechat (using a crosscompiler on windows, i believe), and let me tell you it's a tedious task, from what he told me. If his source code was opened, i'm sure he'd rather like my approach, which is anyway the same used for most large opensource projects like mame, mplayer, webkit, wesnoth, scummvm, and many many others... Some programers develop the core, and then you have porters for each supported platform. Would you expect the main programmers to deal with the tedious task of porting themselves their baby to every OS out there?
Besides, delegating work is a nice way to create your so-called "camaradery", instead of keeping all the "fame" for yourself.

Finally, never once did I mention your name. Never once did I quote you verbatim. Never once did I imply anything about you, but here you come running, the charge of the Light Brigade, to  "defend" yourself, to your own chagrin against a "shame" to your "honor" that had nothing to do with you, nor was it a direct attack at your person. You made a volatile assumption, biased I might add, since from your previous post, it's obvious you have personal issues with me, blindly missing the point I was trying to make.

This, people, is exactly what I'm talking about.

I just love your dramatic style.

[Fab]

This part of your post is the only one I'm going to touch. Originally, I thought that it was the fact that English wasn't your primary language. Then, I realized that you, are in fact, just trying to goad me into a fight, which just isn't going to happen. After this, I'm going to simply ignore you and your posts.

Since you're supposed to ignore my post, this is probably useless, but there's no reason I would let you believe you made a point like that, anyway.
 
You're just totally dishonest by ignoring explicitly (sorry for committing such an atrocious spelling mistake before, I hope I didn't cause a permanent damage to your brain) the part of my message where I actually explained why I can't cross-compile for all the other platforms, and why most other projects solve this issue by delegating ports to... porters. You're also completely ignoring any intermediate effort between no cooperation at all and an actual port. And unlike you, I was honest by taking jahc's example, and also explained supporting all flavours can be unpractical for him as well, especially in the testing area (which means some bugs can often slip in the platforms he can't directly test).

And yes, I perfectly understood you didn't only refer to me, but since we had this argument a couple weeks ago, it made great sense to discuss it again, since you blindly refused to understand last time. I was somehow hoping you would try this time.

And to end this useless discussion, I really feel in your message a vast complex of superiority and scorn towards non-native English writers (or maybe everyone, actually), but I guess this is a good way for you to make an easy point when you don't have any valid argument.

I believe that this is game, set and match. I'd say it's been a pleasure talking to you, but not really. :laughing:

If that can make you feel better. This is a cheap medication, at least.

Oh, I almost forgot to finish my post the same arrogant way you did: LOL.

[Fab]

To be more accurate some people are upset at the dubious timing of some MorphOS news items.  I have been forwarded IRC chat transcripts which (if genuine) show that as sad as this is, there is at least occasional coordination to sink OS4 news items down the list.

Ahum. You mean like OS4.1 announced just a few days after MorphOS 2.0 release, or the settlement between A-inc and Hyperion announced just a few days after MorphOS 2.4 release for Mac mini? I have other examples in stock if needed.

The coincidences are quite interesting... Always announcements just a few days after MorphOS releases. Well, it's probably just paranoia from my side, obviously.

[Fab]

It would not surprise me if some of that was calculated.  It would be just as stupid and petty.

Either way as a site AW.net has published all those news items no matter the "camp" and the re-ordering of a news item to the top hardly ever happens.  Where is the problem really?

No problem to me. Just wanted to give another version of the rumour you were referring to.

[Fab]

Yeah..... that is kinda the pot calling the kettle black. :roflmao:

With the subtle difference jorkany doesn't use MorphOS (nor AROS, i believe).