AmigaOS 3.1.4 successor wishlist

[wiser3]

OS wishes
(i previously posted in some other non OS related thread, don't remember where)

My 2 cents on how to move AmigaOS forward:

Regardless of who has/gets the rights to AmigaOS this is the approach i'd like to see:

1. AmigaOS 3.1.4 is a move in the right direction. Create and release version 3.2 that continues the approach of bug fixes and adding minor but important features. For example 3.1.4 allows >4BG drives but requires too much RAM and time to validate large drives during bootup.

2. Low end Amiga's are too slow and limited to properly handle some features. I'd like an enhanced AmigaOS version for enhanced Amiga's. For example, a 68000 CPU isn't fast enough to anti-alias text to make it more readable. An 030 CPU with fast RAM would enable that and much more. Lets put our accelerators to use.

3. Money from every Vampire sale should be put aside to help fund a Vampire optimized AmigaOS. Optimizing the AmigaOS and adding features for the Vampire to really shine is a huge project. This work would have to be spread over several versions. Releases need to be far enough apart to include noticeable improvements but not so far apart that users feel abandoned as they wait for the next release.

Finally, as a programmer, i want to stress the importance of including features in the OS as opposed to having them available through a hack on Aminet.

AmigaOS really needs a modern and complete system for developers to easily make a GUI. Not gadtools or MUI or some other addon. But something integrated into the OS that i know will be available on every system of at least a certain version. Right now applications require you to install some third party package for the software to make a GUI. Or, like in the case of PageStream, they developed their own GUI system.

Having hacks to support TrueType fonts, or UTF-8 encoding or other features on Aminet doesn't properly support developers. Features need to be properly integrated into the OS and available to every program.

Thank you for your time... now i'm sure the roasters will find a way to roast me while most people give it some thought without saying anything.

[madlax]

Originally Posted by E-Penguin

The next version is 3.1.41, no?

For the changes the OP is suggesting this would be a version 3.2. If it were a big fix release with no new features it would be 3.1.5. How do you get from 4 to 41?

[Thomas Richter]

Originally Posted by wiser3

1. AmigaOS 3.1.4 is a move in the right direction. Create and release version 3.2 that continues the approach of bug fixes and adding minor but important features. For example 3.1.4 allows >4BG drives but requires too much RAM and time to validate large drives during bootup.

This is entirely up to you to change. The default block size of 512 bytes creates a bitmap that is just huge because it reserves a single bit per block. RAM requirements go down significantly if you use larger block sizes. The HDToolBox of 3.1.4.1 will already take care of this automatically, so you will not have to wait for long, but you can change the block size already. This requires, unfortunately, to copy the data once as you cannot change the block size of a life file system.

Originally Posted by wiser3

2. Low end Amiga's are too slow and limited to properly handle some features. I'd like an enhanced AmigaOS version for enhanced Amiga's. For example, a 68000 CPU isn't fast enough to anti-alias text to make it more readable. An 030 CPU with fast RAM would enable that and much more. Lets put our accelerators to use.

Text anti-aliasing is not going to happen any time soon. This is due to the way how Amiga graphics work, and how Amiga fonts work. To make this an effective technology, you need true-color displays, which only exist on Amigas equipped with RTG cards.

Originally Posted by wiser3

3. Money from every Vampire sale should be put aside to help fund a Vampire optimized AmigaOS.

Third party vendor support is up to the third-party vendor. AmigaOs neither contains third-party extensions to support GVP cards or P5 cards either. All they do is that they announce themselves properly to the Os, and supply the firmware they need to the Os by the means of AutoConf or an F-Space boot ROM.

If Gunnar wants to take advantage of the Vampire, it is up to him to follow the existing Os design that allows already transparent extension. It does not require drilling up the Os for this. So, for example, if a DMA-enabled scsi.device is something that should be supplied, then this can be done already today. Just that Gunnar does not want to supply AutoConfig (his choice) and believes that the F-Space is better put aside for Atari support than to support AmigaOs.

If you ask me, both choices are strange, if not foolish, but it's his work, not mine.

Originally Posted by wiser3

AmigaOS really needs a modern and complete system for developers to easily make a GUI. Not gadtools or MUI or some other addon. But something integrated into the OS that i know will be available on every system of at least a certain version. Right now applications require you to install some third party package for the software to make a GUI. Or, like in the case of PageStream, they developed their own GUI system.

Well, there is Reaction, but I'm myself not yet certain whether this is really appropriate. It is rather slow on low-end machines, and given that we have a system that is defined by legacy hardware, I'm not sure whether depending on more CPU horsepower is the right direction.

Originally Posted by wiser3

Having hacks to support TrueType fonts, or UTF-8 encoding or other features on Aminet doesn't properly support developers. Features need to be properly integrated into the OS and available to every program.

I don't think that true type support requires any hacking. I do not know how it is implemented, but all it needs in reality is just to announce itself as a font engine to diskfont.

UTF-8 support is again something that will not come anytime soon. It requires a much more sophisticated font system than what we have, and extending this would break a lot of things. Codepage support is probably the best we could go for, and even that is a lot of work (but at least doable).

[Bruce Abbott]

Originally Posted by wiser3

a 68000 CPU isn't fast enough to anti-alias text to make it more readable. An 030 CPU with fast RAM would enable that and much more. Lets put our accelerators to use.

Anti-aliasing will even slow down an 030. I would rather put my accelerator to good use.

Money from every Vampire sale should be put aside to help fund a Vampire optimized AmigaOS. Optimizing the AmigaOS and adding features for the Vampire to really shine

I want an OS that is compatible between machines. OS 'features' is just another word for 'incompatibilities'.

My Vampire shines pretty well already, but I wonder if it still will when bogged down with an 'optimized' OS? The machines that really need optimizing are those that don't have the processing power of a Vampire. Optimize the OS to work well on them, and the Vampire will shine even more!

AmigaOS really needs a modern and complete system for developers to easily make a GUI. Not gadtools or MUI or some other addon. But something integrated into the OS that i know will be available on every system of at least a certain version.

Yet another 'easy' GUI building system? I don't want that. What I want is something I can put in my code that works on all Amigas, not just those of a certain version and above (getting sick of having to work around OS differences). And I don't care how 'hard' it is to make the GUI, only that it works properly.

Having hacks to support TrueType fonts, or UTF-8 encoding or other features on Aminet doesn't properly support developers.

TrueType fonts and UTF-8 encoding are Windows crap. I don't want it on my Amigas, and as a developer I don't want to be bothered with it (I made one app for Windows Xp, and man was that Unicode a pain!). Bitmap fonts are a better match to the Amiga's capabilities, and do we really need 1,112,064 character codes?

[Locutus]

Originally Posted by bruce abbott

do we really need 1,112,064 character codes?

???? ?? ?? ?? ???? ????

???? ?? ?? ?? ???? ????

[Locutus]

LOL, forum code isn't utf8 safe?

[kolla]

Originally Posted by Bruce Abbott

I want an OS that is compatible between machines.

Then Amiga OS is not for you, and you should rather look at Linux and/or NetBSD.

[Bruce Abbott]

Originally Posted by kolla

Then Amiga OS is not for you, and you should rather look at Linux and/or NetBSD.

What a childish comment. I couldn't install Linux on most of my Amigas even if I wanted to. Yet every one has a compatible version of AmigaOS on it. I just want to keep it that way.

AmigaOS most definitely is for me. I won't tell you what you should get though - if you want a version of AmigaOS with programs specially written for it that won't work on most Amigas then go for it. I don't think that would be good for the community, but who knows? Perhaps we could join the ranks of deluded Linux users who think they are 'compatible' (oh look, I managed to get a Linux68k kernel running on my A1200. Now I can install Firefox!).

[scu98rkr]

I'd like to see bootable compact flash drives from the pcmcia slot.

Is there somewhere to vote / donate to help the outcome you want ?

[ExiE]

Originally Posted by Bruce Abbott

Anti-aliasing will even slow down an 030. I would rather put my accelerator to good use.

I am really curious what is good use for 030 accelerator in Amiga these days. Because for me good antialiased fonts sounds much better than crappy bitmap fonts. But the reason could be that my Amiga got 060 and RTG.

Originally Posted by Bruce Abbott

I want an OS that is compatible between machines. OS 'features' is just another word for 'incompatibilities'.

So if I extend your wish, we should ignore AGA to stay compatible with OCS machines, do we?

Originally Posted by Bruce Abbott

Yet another 'easy' GUI building system? I don't want that. What I want is something I can put in my code that works on all Amigas, not just those of a certain version and above (getting sick of having to work around OS differences). And I don't care how 'hard' it is to make the GUI, only that it works properly.

Exactly what I wrote above. Maybe you should add "512 kB of chip ram is enough for everybody" and get famous.

Originally Posted by Bruce Abbott

TrueType fonts and UTF-8 encoding are Windows crap.

What a childish comment. Unicode is used in many systems for a good reason - whole world is not just plain english. And btw if you talk about Windows, you should say UTF-16.

[EzdineG]

I would like to see a command line switch that prevents the new setpatch from scanning the IDE bus mid boot on every boot.

This has doubled the boot time of any machine I’ve installed 3.1.4 to, and I understand this is to address an LED issue that doesn’t even apply to most machines anyway.

setpatch /noscan

Thanks!

[Thomas Richter]

Originally Posted by ExiE

So if I extend your wish, we should ignore AGA to stay compatible with OCS machines, do we?

The problem is a bit how to position the system. For 3.9, it was pretty clear that everyone had the perspective that the lower end machines do not matter that much anymore, anyone should upgrade, and 3.9 should offer a transition path for newer, more powerful machines, for a bright future of the Os.

For 3.1.4, the perspective was a somewhat different one. Instead, the Amiga system was taken as a retro machine, and the approach was that nobody should be left behind, allowing the system to be used even on weaker machines without too much hassle. So the product addressed a somewhat different market.

I personally believe we should continue in this tradition. If we "lock out" lower powered machines, then the success and acceptance of the product might be limited.

[Thomas Richter]

Originally Posted by ExiE

I am really curious what is good use for 030 accelerator in Amiga these days. Because for me good antialiased fonts sounds much better than crappy bitmap fonts. But the reason could be that my Amiga got 060 and RTG.

The current font system of graphics does not allow for that. Graphics Text() has no idea about the colors of the screen it renders on, leave alone its subpixel arrangement. This would require a completely new, and completely different, and certainly less primitive rendering system. Given the amount of changes this requires, and the amount of compatibility risks, I consider this quite unlikely to be possible in the near future.

[Thomas Richter]

Originally Posted by scu98rkr

I'd like to see bootable compact flash drives from the pcmcia slot.

Is there somewhere to vote / donate to help the outcome you want ?

The problem is not lack of money - none of the developers is earning anything, anyhow. The problem is the lack of manpower. The above would require drilling up the scsi.device, essentially - this is a more realistic approach than re-implementing another device just for the purpose of booting from pcmcia, or extending carddisk.device, which can only handle memory-type pcmcia cards.

Given the current development pace, this would delay any release by a couple of months. Not because nobody wants to work, but simply because we don't have so many developers, and each of us has a day job as well.

So while this is not out of scope in any particular way, and neither unrealistic, I afraid that there are currently a couple of more important issues that need to be tackled.

[digiflip]

well Hyperion bought reaction, so id think that be in next update

[ExiE]

Originally Posted by Thomas Richter

The problem is a bit how to position the system. For 3.9, it was pretty clear that everyone had the perspective that the lower end machines do not matter that much anymore, anyone should upgrade, and 3.9 should offer a transition path for newer, more powerful machines, for a bright future of the Os.

For 3.1.4, the perspective was a somewhat different one. Instead, the Amiga system was taken as a retro machine, and the approach was that nobody should be left behind, allowing the system to be used even on weaker machines without too much hassle. So the product addressed a somewhat different market.

I personally believe we should continue in this tradition. If we "lock out" lower powered machines, then the success and acceptance of the product might be limited.

I do not agree. OS 3.1.4 is targeted on people with harddisk and some more ram as minimum, not plain A500. There are so many different accelerator projects running atm like Furia, ACA, Vampire, HC, Witcher, TerribleFire so there is no reason not to push the limits a bit more.

OS 3.1.4 is good base OS with working support for big discs, but people who just play old games with WHDL will probably not buy every next OS update coz they will simply not need it.

[gulliver]

The requirements have been seriously lowered for 3.1.4 comparing it with 3.9 (68000 2MB or even less, versus 68020 + 6 MB). And this means broadening the user base. It was the right call given the circumstances.

You can even run 3.1.4 on an A500 with 3.1.4 roms and only 1MB ram and a harddisk.

It is difficult to find a balance between features and hardware requirements looking at the small user base we currently have. And more if you take into consideration the amount of developers actively involved.

[Bruce Abbott]

Originally Posted by ExiE

What a childish comment. Unicode is used in many systems for a good reason - whole world is not just plain english.

Amiga already supports languages other than English. As a retro computer platform I think it is better to maintain its character than try to make it all things to all people.

The 'good reason' for UT-8 is that English-speaking people can have the most efficient encoding and don't have to be bothered by text written in foreign languages. But some others aren't too happy about it.

Comparison with single-byte encodings

UTF-8 encoded text is larger than specialized single-byte encodings except for plain ASCII characters. In the case of scripts which used 8-bit character sets with non-Latin characters encoded in the upper half (such as most Cyrillic and Greek alphabet code pages), characters in UTF-8 will be double the size. For some scripts, such as Thai and Devanagari (which is used by various South Asian languages), characters will triple in size. There are even examples where a single byte turns into a composite character in Unicode and is thus six times larger in UTF-8. This has caused objections in India and other countries.

But hey, if wanting to maintain traditional ASCII encoding is 'childish' perhaps we should be looking at other areas where foreign languages aren't treated with proper respect. Take programming languages for example. It's bad enough that US versions of BASIC misspell COLOUR as COLOR, but using English keywords at all is an insult to other cultures. One of the most popular programming languages in use today is Python, and all its keywords are English! Disgusting.

[systmcrsh]

Originally Posted by gulliver

It is difficult to find a balance between features and hardware requirements looking at the small user base we currently have. And more if you take into consideration the amount of developers actively involved.

there will always be haters, keep doing what you're doing

[ExiE]

Originally Posted by systmcrsh

there will always be haters, keep doing what you're doing

People with different opinions are not haters!
I greatly appreciate the work of Olaf and Thomas and and I think discussion about future development is quite important...