23-Apr-2024 07:45 GMT.		 UNDER CONSTRUCTION
Anonymous, there are 146 items in your selection [1 - 50] [51 - 100] [101 - 146]
[Web] Newly released CAM article on OS 4ANN.lu
Posted on 09-Aug-2003 09:30 GMT by Christian Kemp (Edited on 2003-08-09 11:35:17 GMT by Christian Kemp)146 comments
View flat
View list
SimplePPC submitted news about a newly released CAM article on OS 4, where "AmigaOS 4's powerful shared library concept [is] explained by Hans-Joerg Frieden of Hyperion Entertainment."
Newly released CAM article on OS 4 : Comment 1 of 146ANN.lu
Posted by pentrite on 09-Aug-2003 11:43 GMT
nice read...
Newly released CAM article on OS 4 : Comment 2 of 146ANN.lu
Posted by KenH on 09-Aug-2003 11:55 GMT
I must confess that I'm very excited about these new OS's. They have the potential get it right now at the very begining. This is contrary to Windows which has many inherent problems that cannot be fixed without jopardizing its compatibility with programs that use it.
I would much prefer to see a streamlined OS at the expense of legacy support than a bloated OS that runs everything from Hollywood to Deluxpaint 0.2 (ahhh...that name brings back happy memories :)
Newly released CAM article on OS 4 : Comment 3 of 146ANN.lu
Posted by John Block on 09-Aug-2003 12:12 GMT
Initially running on a low cost G4 expansion board to go into pc's.
Newly released CAM article on OS 4 : Comment 4 of 146ANN.lu
Posted by Don Cox on 09-Aug-2003 12:15 GMT
In reply to Comment 2 (KenH):
"I would much prefer to see a streamlined OS at the expense of legacy support than a bloated OS that runs everything from Hollywood to Deluxpaint 0.2 (ahhh...that name brings back happy memories :)"

It's a matter of getting the balance right. No good having a streamlined OS with no programs to run on it.

IMO programs that run legally on an A1200 ought to run on future Amigas. Programs that stopped working when OS 2 or OS 3 came out could be dropped.

I do think that programs that use double screns (one for control, one for imagery) such as DPaint, Scala, Imagemaster, etc ought to run. They are too useful to throw away. For instance, Imagemaster is by far the best morphing program on Amiga - the alternative is the Windows version, really.
Newly released CAM article on OS 4 : Comment 5 of 146ANN.lu
Posted by KenH on 09-Aug-2003 12:26 GMT
In reply to Comment 4 (Don Cox):
All true. But how many people still use these programs? No offence Don, but if you're the only one that uses Dpaint still, should it hold the OS design back? :) As you say, it's a matter of balance.

Maybe Hyperion/Genesis should have some sort of united poll to see what software is still in use? Could be a good gesture that would benefit both parties.

But if a program is really too old yet still popular, it could be ported to the new OS. Yes balance is the key, but without the proper information, it's hard to get it right.


>I do think that programs that use double screns (one for control, one for imagery) such as DPaint, Scala, Imagemaster, etc ought to run. They are too useful to throw away. For instance, Imagemaster is by far the best morphing program on Amiga - the alternative is the Windows version, really.
Newly released CAM article on OS 4 : Comment 6 of 146ANN.lu
Posted by SlimJim on 09-Aug-2003 12:48 GMT
In reply to Comment 5 (KenH):
Well I still use DPaint V (less and less though, as my Amiga 4000 is dying on me).
It would be excellent to be able to run it on fast hardware - capable of higher
resolution, larger pages and 24bit colour without it being sluggish.
Together with some other similar programs on the Amiga (e.g. Brilliance), it can
do things I cannot do in programs on other platforms (even though it lacks many
handy features as well of course). But DPaint tend to be quite quirky when it
comes to running on newer hardware, if I recall correctly. Time will tell.
.
SlimJim
Newly released CAM article on OS 4 : Comment 7 of 146ANN.lu
Posted by Don Cox on 09-Aug-2003 13:32 GMT
In reply to Comment 5 (KenH):
"All true. But how many people still use these programs? No offence Don, but if you're the only one that uses Dpaint still, should it hold the OS design back? :) As you say, it's a matter of balance."

Why should it "hold the OS design back" though? On the contrary, adding a powerful and useful GUI feature like having a control screen in front of the main screen would give AmigaOS an advantage over other platforms. There are ways to get the function without using the original hardware.

I think your basic assumption that having the flexibilty to run more programs will be a liability is wrong. Consider ModePro, for instance . It allows programs to run on graphics cards that were written before they were heard of. Is that a bad thing?

The ideal OS would run every program ever written for any platform.
Newly released CAM article on OS 4 : Comment 8 of 146ANN.lu
Posted by Anonymous on 09-Aug-2003 13:45 GMT
hip teens don't wear blue jeans
Newly released CAM article on OS 4 : Comment 9 of 146ANN.lu
Posted by Kronos on 09-Aug-2003 14:37 GMT
In reply to Comment 7 (Don Cox):
>On the contrary, adding a powerful and useful GUI feature like having a
>control screen in front of the main screen would give AmigaOS an advantage
>over other platforms.

And what would that be ? (other than forcing something onto current gfx-chips
they were never made for).

Back than it was used because you could either have hicolor (if you call 32/64
or 4096 colors that) on LowRes, or high res (well 640 pixels per line) with
limited colors. But current chips will easiliy crank out 1600x1280 (or more) at
24 bit ....

A backdrop and a borderless window will just do the same, without playind such
dirty tricks.

About old SW:
You either put restrictions on how much you emulate, or you go the full UAE, and
this will mean some goods apps "falling of" the plattform, but is is absolutly
needed to built a system fit for the future. And you can allways run UAE on any
Amiga(like) OS (which should even be fast enough once Bernie releases his work
on the JIT for PPC).
Newly released CAM article on OS 4 : Comment 10 of 146ANN.lu
Posted by 3seas on 09-Aug-2003 15:20 GMT
I wonder if the now public information article might help to port AROS to PPC.
Newly released CAM article on OS 4 : Comment 11 of 146ANN.lu
Posted by BrianK on 09-Aug-2003 16:17 GMT
In reply to Comment 5 (KenH):
Maybe Hyperion/Genesis should have some sort of united poll to see what software is still in use? Could be a good gesture that would benefit both parties

----
Maybe they should take a poll of vendors that still exist and make sure that currently available products will work. This will encourage Amiga users to buy those programs and continue to support the current software companies within the Amiga environment.

I'd HATE to see OS4 not run ImageFX. OS4 would be a no go for me then!
Newly released CAM article on OS 4 : Comment 12 of 146ANN.lu
Posted by anonymous on 09-Aug-2003 17:03 GMT
In reply to Comment 7 (Don Cox):
'The ideal OS would run every program ever written for any platform.'

Good grief, the ideal *application* should run on every operating system.

Application code is reasonably easy to make portable -- the system-specific functions can be abstracted. But crippling an operating system to accommodate 10-year-old software is patently silly.

Software that is a decade old clearly needs to be updated or replaced. Users that absolutely cannot live without these applications can find inexpensive classic hardware to run them on.

Hyperion is basically borrowing a page from Apple's book. Port the OS, provide an interim solution for compatibility and then transition to the 21st century. We'll hold a vigil for what gets left behind.
Newly released CAM article on OS 4 : Comment 13 of 146ANN.lu
Posted by Don Cox on 09-Aug-2003 17:56 GMT
In reply to Comment 12 (anonymous):
"Software that is a decade old clearly needs to be updated or replaced."

If it has the functions that somebody wants, why change it?
Newly released CAM article on OS 4 : Comment 14 of 146ANN.lu
Posted by KenH on 09-Aug-2003 20:13 GMT
In reply to Comment 13 (Don Cox):
Because it won't be compatible with the new OS standards! ;)


>If it has the functions that somebody wants, why change it?
Newly released CAM article on OS 4 : Comment 15 of 146ANN.lu
Posted by EyeAm on 09-Aug-2003 20:15 GMT
That was a good article; and an indication of good things to come.

I see no reason why a new 64-Bit Amiga OS cannot run all previous Amiga software prior to it (i.e., everything from the Classic Amiga) on the same machine, in the same OS. I see no reason why it also cannot have every previous OS installed on it, if a user wanted to.

Check this out (and don't overlook their very long and informative FAQ page, linked at the top), http://www.hyperos2002.com/

I like the idea of the new interfaces or libraries that are being worked on, or have been written, for the new Amiga OS. And, in a way, an analogy can be made to programs that I even had on one of my past Amigas: While, at the time, my Amiga ran an '030, there were '040-specific libraries that existed on my machine, which could be used if I had upgraded. The point being, individual libraries, or even the LIBS directory, can hold the code or files or libraries there for access, without any damage or interruption to the over-all system; and they're called when needed by programs that use them. It's nothing to have something like that sitting there--text, ascii, whatever--as long as the program searching for it can get to it in time and make use of it.

Good article on how things are changing. While I don't care about PPC so much, I understand what is going on there. It's the same thing I want for Amiga with the x86-64 chips (now called the AMD64 Platform, at least with the AMD side of things).

It's like the old libraries were either dropped into a newly-written PPC library that does the equivalent, or like a PPC 'header' or router was tacked on, with the new features.

By the way, for those who do check out that HyperOS, while it is neat and interesting for Windows, when reading the FAQ try mentally replacing where it talks about the old non-NT Windows with 'Classic Amiga' and consider any talk of the newer Windows (NT, 2000, XP..) to be like the new Amiga OS, and it will give you some interesting ideas or new perspective.

The new Amiga OS, 64-Bit, won't have many programs for it when it first comes out. That's a disadvantage. To counter that, if it allowed older Amiga OSes to be installed on it (NOT like AmigaForever, but to a partition, and from CD or disk just like you would normally), all the old libraries and such could still be placed in the same System drawers if things haven't changed all that much. And if they had changed, the old drawer structure could just be created and wouldn't bother the new one. The question becomes 'on what machine would they be installed?'. I'm not buying an AmigaOne. I've soured on any kind of specific Amiga machine now--the new OSes need to be on existing, readily-available PCs, which you can built yourself or buy at almost any computer store, customize, and so on. It would be the most logical act of self-preservation for Amiga to do that.

I don't believe the new OS should have to emulate what the company owns. Especially when it isn't necessary.

There's an old program on this PC I'm using here, which technically should not or would not run if it did not have a software 'wrapper' put around it, making it equal to the other programs in terms of the 64-Bit environment. That's the concept I think could be applied to the classic software on the new OS. Not .adf files, either. I just think the new OS is imminently able to recognize (like CrossDOS recognized a PC disk/file) a classic Amiga file or disk when seen, and then it regards it for what it is.

Those who say it can't be done are almost saying "The new Amiga OS isn't good enough to do what the old Amiga OSes did." :-) It's bull. The old programs can run twice as fast under the new one.

Good article, again, though. It's good to see these things. It's almost like the libraries have headers to accomodate or place the context for the software programs seeking them. Or better yet, it's like a movie theater where people could sit anywhere they wanted once, because it was all the same, suddenly having an usher directing you to a particular section because the theater changed its policy and if you're rowdy (i.e., Classic Amiga), you have to sit in special area of seats.

Software segregation. ;-P
Newly released CAM article on OS 4 : Comment 16 of 146ANN.lu
Posted by Troels Ersking on 09-Aug-2003 20:17 GMT
In reply to Comment 13 (Don Cox):
Reasons could be that it holds back further development of the OS, or that it doesn't utilise new features in the OS.

Even though people can live with the program only containing the old features doesn't mean they don't want new stuff added. IBrowse 2.3 is also good enough for a lot of people but everybody still want's a v3.0 :)
Newly released CAM article on OS 4 : Comment 17 of 146ANN.lu
Posted by EyeAm on 09-Aug-2003 21:06 GMT
If the old software is placed in the new arena, the authors of the old software can continue development while their once almost-obsolete software program gets new life breathed into it. The code which will optimize their product for the new OS can be worked on while they some small profit from sales or notoriety from still being relevant or viable, until their new version ships for the same arena (the new OS).

--EyeAm
http://eyeam.darrelltown.com
Newly released CAM article on OS 4 : Comment 18 of 146ANN.lu
Posted by smithy on 09-Aug-2003 21:08 GMT
I don't like this method of providing compatibility with 68k programs. The problem is that you have legacy stuff (a 68k jumptable) in the modern OS4 PPC, and supposedly microkernel. As time goes on, this legacy stuff will get used less and less, and will simply exist to consume space. The whole point of a microkernel is to have only the stuff in there that will be used all of the time, and everything else in loadable modules.

A better solution would have been to create a proxy that would sit between legacy programs and the OS. In effect, a virtual machine, which would run programs in the same address space as the rest of the OS. The kernel wouldn't be compromised by having unnecessary legacy code in it - it would just be a PPC kernel.

I think this solution is inelegant and wastes resources.
Newly released CAM article on OS 4 : Comment 19 of 146ANN.lu
Posted by EyeAm on 09-Aug-2003 21:34 GMT
In reply to Comment 18 (smithy):
Another solution would be something like this:

1. System drawer for the new OS existing,
2. inside of that is placed drawers SystemOS1, System OS2, SystemOS3
3. The System contains all relevant new libraries, utilities, etc. for the new OS.
4. The other drawers, SystemOS1, etc., contain all relevant libraries from the classic past--for each version of the OS that you're going to run with the new system.
5. Some kind of interface, "door manager" or pointer should exist that is almost synonymous with the first System drawer, so the OS or System can know instantly that an incoming call requires to be pointed to the right set of libraries, utilities, and so on.

All of those System drawers for the classic would really be a 32-bit sandbox, if you didn't want to mix all that with the new system. It wouldn't *hurt* the new system if you did that, but for simplicity sake (or if you just wante to toss the classic pieces from your new OS later...), you could keep them separate and trim down your OS to only the new things.

There are several ways to reach the same goal. This is just one way.

If a classic Amiga program could be installed in the new OS, and upon insertion of the CD or disk the new OS could basically say 'aha, incoming classic program--everything gets filed over here', then the new OS could continue treating it similar to how CrossDOS treated a PC disk when I put it into my Amiga. Some kind of software prejudice where it would always be known to be the lesser, 32-bit organism :) (granted, whose code could be implemented at twice the rate)

There doesn't even have to be a System drawer or set of libs for *each* classic OS version, there could just be one for that. SystemClassic.

The magic of *aliases* could come in here, if the proper interface design were to be created for directing the older programs to their drawer. While their paths would be looking for "System/", if it were known by the interface to alias that for the 32-bit programs, then "System/SystemClassic" or "System/SystemOS2" could equal "System/" for that instance; and they would be allowed to pass through and ignore the first set of LIBS, UTILS, and so on.

EyeAm
http://eyeam.darrelltown.com
Newly released CAM article on OS 4 : Comment 20 of 146ANN.lu
Posted by Ben Hermans/Hyperion on 09-Aug-2003 21:51 GMT
In reply to Comment 18 (smithy):
>I think this solution is inelegant and wastes resources.

I think you don't understand it, sorry.

That conclusion is plainly wrong.

It does exactly the opposite.
Newly released CAM article on OS 4 : Comment 21 of 146ANN.lu
Posted by smithy on 09-Aug-2003 22:51 GMT
In reply to Comment 20 (Ben Hermans/Hyperion):
>>I think this solution is inelegant and wastes resources.
>
>I think you don't understand it, sorry.
>
>That conclusion is plainly wrong.
>
>It does exactly the opposite

No it doesn't. You have code in the kernel that may never be used (a 68k jumptable), but is loaded into memory everytime.

The core of the system (exec) is about scheduling, handling memory, signals, etc.... facilitating 68k emulation should not be in the system's core, yet there is code to deliver 68k jumptables in exec.library.

Imagine buying a new spiffy car that ran on natural gas. Now, imagine if it had a seperate fuel tank for 4-star leaded petrol, on the offchance someone might find some of this type of petrol. It's a waste of space!

Or a more appropriate example - imagine if Microsoft moved some vital support for running dos programs into kernel32.dll - so that everything in the system got it, whether it needed it or not!
Newly released CAM article on OS 4 : Comment 22 of 146ANN.lu
Posted by EyeAm on 09-Aug-2003 23:09 GMT
In reply to Comment 21 (smithy):
:-) Hey, that's not so far-fetched. They did, after all, move their OS into the market whether we wanted it or not.

--EyeAm
http://eyeam.darrelltown.com
Newly released CAM article on OS 4 : Comment 23 of 146ANN.lu
Posted by CodeSmith on 10-Aug-2003 00:55 GMT
In reply to Comment 21 (smithy):
> No it doesn't. You have code in the kernel that may never be used (a 68k
> jumptable), but is loaded into memory everytime.

Actually, once demand-paged executables are implemented (IIRC this is a planned feature) the memory for the jumptable will be mapped as in-use, but only when some process tries to use it will it actually get loaded into memory.

> The core of the system (exec) is about scheduling, handling memory, signals,
> etc.... facilitating 68k emulation should not be in the system's core, yet
> there is code to deliver 68k jumptables in exec.library.

No-one said that it's going to be there for ever... remember, OS4.0 is the transitional stage. If Hyperion removed all 3.x support right now, there would be zero software for it. I can live with some inefficiency for a year or two, if it means I don't have to wait a year for the first programs to appear.

> Or a more appropriate example - imagine if Microsoft moved some vital
> support for running dos programs into kernel32.dll - so that everything in
> the system got it, whether it needed it or not!

Actually, a lot of the code needed to handle v86 tasks (dos and win3.x/9x programs) is in kernel32.dll and hal.dll (I assume you mean in Windows NT/XP, in Win9x all the 16 bit stuff is needed by the OS itself).
Newly released CAM article on OS 4 : Comment 24 of 146ANN.lu
Posted by Jonny Johansson on 10-Aug-2003 05:55 GMT
In reply to Comment 4 (Don Cox):
Don Cox wrote:

> ...For instance, Imagemaster is by far the best morphing program on Amiga -
> the alternative is the Windows version, really.

Sorry for going off topic folks... :7

Oh, how does the GUI work for this one and what is the output like?
I've got a soft spot for Morph+ myself, since i find the interface perfect,
unlike those unwieldy two-separate-images-and-a-mesh all other programs I've
seen use and speed and end result give me nothing to complain about. Now, if
there's still something better, I'm damned curious. :)
Newly released CAM article on OS 4 : Comment 25 of 146ANN.lu
Posted by Hans-Joerg Frieden on 10-Aug-2003 06:10 GMT
In reply to Comment 18 (smithy):
> I think this solution is inelegant and wastes resources.

What makes you think that a "virtual machine" (however that would look like, I invite you to specify this a bit more precicsely) would use fewer resources?

Besides, if you don't want to have the library callable from 68k, you drop the legacy interface. The jump table needn't be there. But we are talking about a few hundred bytes for each library, summing up to about fifty kilobytes in the whole system. True-Color Icons tend to waste more space than that.

So I do consider this an elegeant solution. It works completely transparent for old programs, and it doesn't slow down new programs. It also allows us to weed out old functions from the system interfaces, and it allows us to extend the API without hampering compatibility.
Newly released CAM article on OS 4 : Comment 26 of 146ANN.lu
Posted by Hans-Joerg Frieden on 10-Aug-2003 06:15 GMT
In reply to Comment 21 (smithy):
> facilitating 68k emulation should not be in the system's core,

How are you going to run 68k legacy drivers when there is no 68k emulator in the system core?

> Imagine buying a new spiffy car that ran on natural gas. Now, imagine if it
> had a seperate fuel tank for 4-star leaded petrol, on the offchance someone
> might find some of this type of petrol. It's a waste of space!

No, because natural gas fuel stations aren't available around the country yet, so sometimes you will have to get regular unleaded fuel. You can't just load and unload the tank. As soon as natural gas is available around the globe, you can remove the other tank, and use that space for different stuff.

The presense of a 68k emulator and glue code doesn't make the system anything but a PowerPC system.
Newly released CAM article on OS 4 : Comment 27 of 146ANN.lu
Posted by Anonymous on 10-Aug-2003 07:45 GMT
In reply to Comment 26 (Hans-Joerg Frieden):
A bit like pc motherboards and bundled low cost systems being available everywhere.

Meanwhile why can't we have something which simply slots in to these systems

Rather than forcing people to buy harddrive, graphics board, keyboard, etc twice over and crazy money for the motherboard.
Newly released CAM article on OS 4 : Comment 28 of 146ANN.lu
Posted by Olegil on 10-Aug-2003 08:23 GMT
In reply to Comment 27 (Anonymous):
Well, why don't you fund the development and distribution, then? You're obviously certain it will be a success, so I expect you won't have ANY problems getting investors to back you up with funding.

And tell us when you're ready to ship, I'll consider one ;-)
Newly released CAM article on OS 4 : Comment 29 of 146ANN.lu
Posted by Olegil on 10-Aug-2003 08:55 GMT
In reply to Comment 26 (Hans-Joerg Frieden):
Heh, he gave two examples and both turned out to support your point of view instead of his. :-)
Newly released CAM article on OS 4 : Comment 30 of 146ANN.lu
Posted by Raffaele on 10-Aug-2003 09:33 GMT
In reply to Comment 14 (KenH):
Here follows some answers of mine to KenH msg.14 and Troels Ersking msg.16


1)
---
KenH wrote in answer to a statement of don Cox:

DC > >If it has the functions that somebody wants, why change it?

KH > Because it won't be compatible with the new OS standards! ;)

May all owners of MS-DOS will had followed your right statement at those time Windows didn't still exists...

[Oh Lord... If only they did it!]

...Now we could have had a 32/64 bit MS-DOS Micro-kernelized with built-in Internet facilities and an optional WIMP GUI such as Amiga provided with WB.

And we could had regret nothing, because the developing should had followed regular and equal-stepped improvements.

Unfortunately due to the inertia caused worldwide by millions clerks and secretaries who insisted in using the same software (i.e. for example, Autocad, Wordperfect, Quattro Pro, Norton Utilities, Dosshell, etc., etc., etc., etc.)...

...without changing it at every release of new OSes (because this will mean other expenses in buying new releases and loss of time in renewing Know-How learning), then M$ was encouraged, at early times of MS-DOS developing not to change anything and making only minimal improvements until not forced to!

Infact All versions of Windows until XP allowed STILL eldest software to run...

If we have had Windows is due to the STRONG and NEW request of the market who wanted WIMP GUI interfaces and a standard "menu and buttons" interface for applications such those Macintosh supplied.

If we have had multitasking onto WIN environment is due to the bare necessity of to beat Unix/Linux competitors.


2)
---
From another point of view it is necessary not to MIX POTATOES with NAILS into OUR SOUP as Troels Ersking did:

> Even though people can live with the program
> only containing the old features
> doesn't mean they don't want new stuff added.
> IBrowse 2.3 is also good enough for a lot of people
> but everybody still want's a v3.0 :)

IBrowse and Dpaint are mathematically speaking "quantities not comparable".

because the first one is a tool of communications and presentation of informations, and is subject to the evolution of communications protocols and to the necessity to embrace new functions to easily present the requested informations.

The second, Dpaint, is an artistic tool, and is subject to the necessity of create easily and quickly any sort of drawing.

As any tool that outputs an "object work" that is subject to manipulation (such as a grapchical file), Dpaint could became an actor in a chain of various graphical tools...
...in which the artist choose from time to time what tools he need to complete the work...

And this path requires also that the artist (or the professional graphician) could decide to transfer his file FROM and TO various programs and to step back to another program in order to RETOUCH the work he is doing...

So is not impossible to find an artist who needs his graphical file to be realized on the easy-to-use Amiga DPpaint, then to be ported to Windows Photoshop to add some filters, and then switched to a MAC to be outputted on an AVID environment... ...only to found at the end of the process that he needs again Dpaint to add some necessary retouches before to finally outputting it all...


3)
---
Talking about the evolution of the OS and of the program itself, at the end of all the squisite statement i just made...

...I can affirm (except of strategically choices by OS firm and the program firm itself) that:

-A good program tool must be used also with new OSes of new releases of it...

BUT until it causes some instabilities to the OS itself.


-AT LEAST, then, it can be used into emulated enviroment,

BUT only if emulated environment returns same output.


And is amazing to discover that this is quite easy to reproduce with programs such IBrowse, that is only a presentation tool...

...but only until are not changed communications protocols (such as introduction of XML and CSS, new SSL and IVP6) then it become unusable...

And on the opposite, it is difficult to reproduce an usuable environment for programs such as graphical tools like Dpaint, which need CONCRETE FEEDBACK!!!

Thingies such mouse reliability, correct pixel aspect-ratio, fill-flood parameters, and geometries of curves fixed by same mathematics, are not easy to be reproduced by emulators.

Faults made in reproducing correctly these facilities, then give the operator (artist or whatsoever) a sensible feeling that the graphical tools is not performing anymore as it did on the original enviroment...
...and make the program no more useful for the task requested!!!
Newly released CAM article on OS 4 : Comment 31 of 146ANN.lu
Posted by Don Cox on 10-Aug-2003 10:06 GMT
In reply to Comment 14 (KenH):
">If it has the functions that somebody wants, why change it?"

"Because it won't be compatible with the new OS standards! ;)"


In that case the standards are wrong. The purpose of a computer is to get things done, not to prevent things from working.
Newly released CAM article on OS 4 : Comment 32 of 146ANN.lu
Posted by Don Cox on 10-Aug-2003 10:19 GMT
In reply to Comment 24 (Jonny Johansson):
Morphing.

In Imagemaster and WinImages you have two images side by side. These could be two copies of the same image if you just want to distort, but usually you want to morph one image into another (and maybe then that into another, usw.)

Basically you mark out points or lines on the start image, they appear also on the end image. On the end image, you shift the points to where you want them to end up. For example, there might be a set of points around an eye which have to move to wherever the eye is in the end image. Then if you want to, you can adjust the fade rate to be non-linear (a morph is distortion+fading).

You set the size you want the output images to be and the number of images. Run tests without saving until it looks right. Then save. The scaling is very good quality. The two original images can be any size.

All the project data (coordinates of points etc) can be saved as ASCII files.

It is also possible (though tricky) to morph between two sequences of moving images, so for example one character can turn into another as it walks.
Newly released CAM article on OS 4 : Comment 33 of 146ANN.lu
Posted by rbl on 10-Aug-2003 10:41 GMT
In reply to Comment 20 (Ben Hermans/Hyperion):
Sounds pretty elegent and efficient to me. Considering the alternative was to have multiple versions of every library, one 68k, one PPC.
Newly released CAM article on OS 4 : Comment 34 of 146ANN.lu
Posted by Jonny Johansson on 10-Aug-2003 11:12 GMT
In reply to Comment 32 (Don Cox):
Doc Nox wrote:

> Morphing.

Hehe, I absolutely love the way you tend to turn an answer into a lecture
on the obvious, or disagree with somebody, only to then write something
that is in complete accord with what you just purported to disagree with. ;)


> In Imagemaster and WinImages...

Ok, so basically they use freely positionable vectors, like Morph+ and
not that ghastly mesh solution, but they do display the images side by side,
rather than superimposed. Should be good for overview, but I'd have to give
it a try to be sure. :7


>The scaling is very good quality. The two original images can be any size.

Good, convenient.

>All the project data (coordinates of points etc) can be saved as ASCII files.

This is very neat!

>It is also possible (though tricky) to morph between two sequences of moving >images, so for example one character can turn into another as it walks.

Lots of work to set up, although I tend to use morphers mostly to
tween video, faking smooth slow motion and so on, so that's a familiar
chore, although in the slomo case a little bit easier, since you don't have
the same considerations while recording and can reuse the previous end points
as start points for the next morph. :7

Thank you for your answers, Sir.
Newly released CAM article on OS 4 : Comment 35 of 146ANN.lu
Posted by Bert Dorhout on 10-Aug-2003 11:40 GMT
In reply to Comment 26 (Hans-Joerg Frieden):
H-J Frieden wrote:

>> Imagine buying a new spiffy car that ran on natural gas. Now, imagine if it
>> had a seperate fuel tank for 4-star leaded petrol, on the offchance someone
>> might find some of this type of petrol. It's a waste of space!

>No, because natural gas fuel stations aren't available around the country yet, >so sometimes you will have to get regular unleaded fuel. You can't just load >and unload the tank. As soon as natural gas is available around the globe, you >can remove the other tank, and use that space for different stuff.

Ah. I actually came home from a car trip the day before yesterday. We drove through The Netherlands, Germany, Poland, Lithuania, Latvia, Estonia, Finland, Sweden, back to Poland, Czech Replublic, Germany and back to The Netherlands.
The car has both an LPG (liquid car gas) tank and a 'normal' gasoline tank and we were very happy it had both! LPG is much! less expensive than normal gasoline, but in Germany, Finland and Sweden it was somewhat hard or impossible to find. So we drove mainly on LPG, but without the gasoline tank we would have had to make several detours in Germany and would have ended up pushing the car for hundreds of kilometers in Finland, I think.

I do think the new solution for libraries in this version of AmigaOS is elegant, and, as someone else has said, later versions might even see it dropped, but not before err.... LPG/liquid gas stations can be found in most/all countries. :-)
The little (?) resources/code AOS4 uses to make this possible seems worth it to me.

Bert
Newly released CAM article on OS 4 : Comment 36 of 146ANN.lu
Posted by Joe "Floid" Kanowitz on 10-Aug-2003 11:54 GMT
In reply to Comment 25 (Hans-Joerg Frieden):
Hans-Joerg said,

> So I do consider this an elegeant solution. It works completely transparent
> for old programs, and it doesn't slow down new programs. It also allows us to
> weed out old functions from the system interfaces, and it allows us to extend
> the API without hampering compatibility.

*Philosophically,* this does seem close to the Win9x mindset - a lot of code (measured by file) is equally responsible for legacy and 'modern' support, with tradeoffs for each. Some of us can (rightly?) imagine greater compatibility through the fully sandboxed approach -- hard to emulate the chipset without it, after all -- but that's just emulation, and this should help keep further development true to its roots, at least for as long as 68k support is a necessity. (For that matter, think of the early Win95 authors who took it upon themselves to backport to at least Win32s?)

...but it really *is* just a philosophical issue, when discussed in these terms. Some people ("wrongly") thought chunks of DOS, Win16, and who knows what else were worth saving, even while *knowing* they'd be a hindrance (and indeed, maintaining the sandboxed approach seen in OS/2 in NT, only to finally 'merge' things for XP... though there's more history to it than that, of course) ... On the other hand, the 'red team' rightly or wrongly believes there are things to be saved from AmigaOS -- you'd kinda hope so, because as everyone keeps asking, "What's left of Amiga in these designs?" -- but are aware of the tradeoffs, feel it's a solid enough foundation (vs. DOS/Win3.x ;)), and are willing to address the issues and try to move it forward with a little more order than, say, Microsoft or even Apple ever did. Sort of like if '95 were actually produced with an eye to 'becoming' NT, instead of practically being replaced by it.

Meanwhile, in some ways, MorphOS is looking like XP from the get-go. With so many things seeming tangled within the A/Box, there are people like me trying to figure out what the system would look like were it 'unencumbered' from legacy support. (Much as some of the interesting design decisions of NT/XP - I've gotten curious about how the whole filesystem tagging/mounting thing is working there, since it seems they've snuck some interesting stuff in 'under the hood' - are masked by the layers of cruft, marketing, and nagging Messenger popups up front.)

In short, an OS built on the OS/2, OS X, or AmigaOS XL? mindset could doubtless be pretty cool (though whether we'd allow it to be called "Amiga" would be yet *another* run of the old argument...). What 4 has evolved into (from initial declaration as 'the compatible version,' with the A1200-attaching hardware, on through to 'the stable version,' as perceived with the introduction of the VM system, on through to what we can now recognize it as: a series of tradeoffs, managed well, that should certainly take the old girl kicking and screaming into the 21st century, while keeping each major surgery 'evolutionary' enough that nobody ceases to recognize her) is equally cool... and we already *have* emulators/compatibility systems that run atop Linux, QNX, and other admirable systems, so let's see how far that wonderful and crazy hybrid of TripOS, Amiga and Commodore know-how can really be stretched!
Newly released CAM article on OS 4 : Comment 37 of 146ANN.lu
Posted by Joe "Floid" Kanowitz on 10-Aug-2003 12:13 GMT
In reply to Comment 36 (Joe "Floid" Kanowitz):
Oh, and for EyeAm... You're aware that that's basically the approach DragonFly hopes to take, right? Wonder how ridiculous it would be to apply a PEACE-like (http://chiharu.haun.org/peace/) API layer and appropriate emu-libs (Reaction or MUI -> Qt/GTK/???) when that project's ready. :)

[Very, of course, but one of the objectives is to make such things easier, even if only as a side effect of 'cleaning' up the overall kernel design. Cripes, I can't wait until it's finished enough for a luser like me to benefit from it... They even have Billsey types showing up and thanking them for losing the daemon iconography, promising to evangelize it to all the institutions who snubbed other BSDs over such concerns!]

It's a strange, strange planet, and I've had too much coffee again.
Newly released CAM article on OS 4 : Comment 38 of 146ANN.lu
Posted by Hans-Joerg Frieden on 10-Aug-2003 12:29 GMT
In reply to Comment 36 (Joe "Floid" Kanowitz):
> *Philosophically,* this does seem close to the Win9x mindset

What does this have to do with Win9x philosophy? Either you make a clean cut, and don't have any compatibility, or you must be dragging old stuff around. Some programs *will* continue to call OldOpenLibrary, because their binary says so, this means you cannot remove this from exec's 68k jump table without impairing compatibility.

There is a distinct line in AmigaOS 4's design, where the part on one side of the line is 68k/legacy and the other part is PowerPC/New stuff. The 68k side may vanish. In an instant. Without a trace. But for now, you gotta stick with it because it would otherwise leave you with very little programs to run.

OTOH, the way that it is implemented in AmigaOS 4 makes it possible for "old" programs to co-exist in the new environment without needing extra handling; they even benefit from some of the achievements. You *can* run FinalWriter on OS 4 and edit a document that doesn't fit in your main memory because the virtual memory applies equally to 68k and PowerPC programs.

Like I said, I consider this an elegant solution.
Newly released CAM article on OS 4 : Comment 39 of 146ANN.lu
Posted by Don Cox on 10-Aug-2003 12:38 GMT
In reply to Comment 34 (Jonny Johansson):
Well you said "Oh, how does the GUI work for this one and what is the output like?" so I told you. I assumed you wanted an answer to your question.
Newly released CAM article on OS 4 : Comment 40 of 146ANN.lu
Posted by Don Cox on 10-Aug-2003 12:49 GMT
In reply to Comment 38 (Hans-Joerg Frieden):
"There is a distinct line in AmigaOS 4's design, where the part on one side of the line is 68k/legacy and the other part is PowerPC/New stuff. The 68k side may vanish. In an instant. Without a trace."

That's the point at which I stop buying updates. If my programs don't work any more, I may as well go over to Mac. A better aim would be to try to increase the number of Amiga programs that work.

That word "legacy" is pejorative. It assumes there is something "wrong" with a fully legal Amiga program that doesn't work on OS 4 or MorphOS.


"Like I said, I consider this an elegant solution."

I think so. I look forward to reading the next part of the article.
Newly released CAM article on OS 4 : Comment 41 of 146ANN.lu
Posted by Joe "Floid" Kanowitz on 10-Aug-2003 13:36 GMT
In reply to Comment 38 (Hans-Joerg Frieden):
Hans-Joerg said,

> What does this have to do with Win9x philosophy? Either you make a clean cut,
> and don't have any compatibility, or you must be dragging old stuff around.
> Some programs *will* continue to call OldOpenLibrary, because their binary
> says so, this means you cannot remove this from exec's 68k jump table without
> impairing compatibility.

Y'know, I *knew* I should've qualified that I wasn't disagreeing at all - I was just presenting my "farsighted" view, in an attempt to rationalize it to everyone screaming 'this is awful!' for no good reason. ;)

If you sandbox everything in an emulation,

> There is a distinct line in AmigaOS 4's design, where the part on one side of
> the line is 68k/legacy and the other part is PowerPC/New stuff. The 68k side
> may vanish. In an instant. Without a trace. But for now, you gotta stick with
> it because it would otherwise leave you with very little programs to run.

I oversimplified; obviously, you've already said that. But for now, it gives you added impetus to keep the 'spirit' (versus just completely doubling implementations), and that's a good thing, right?! :)

> OTOH, the way that it is implemented in AmigaOS 4 makes it possible for "old"
> programs to co-exist in the new environment without needing extra handling;
> they even benefit from some of the achievements. You *can* run FinalWriter on
> OS 4 and edit a document that doesn't fit in your main memory because the
> virtual memory applies equally to 68k and PowerPC programs.

There you go - instead of "legacy support" dragging things down (as noticed in Win9x), the older programs have a chance to be "buoyed up." Like if the authors of '9x had paid any attention to what they were doing, instead of grafting on and on and on... (But then, as I said, the DOS foundation wasn't good enough to allow for much else. Amiga has mechanisms worth salvaging and updating!)

> Like I said, I consider this an elegant solution.

I'm agreeing. Throwing the child out with the bathwater can sometimes seem equally elegant (OS/2 was fun to use, some people like OS X), but it wouldn't make sense for OS4.
Newly released CAM article on OS 4 : Comment 42 of 146ANN.lu
Posted by Joe "Floid" Kanowitz on 10-Aug-2003 13:45 GMT
In reply to Comment 41 (Joe "Floid" Kanowitz):
Ugh, like I said, too much coffee.

The dangling thought was:
If you sandbox everything in an emulation, of course it works, but it's just an emulation... If the cut was made all at once, you might as well be running anything else (as Don said), but you guys *aren't* doing that just yet, you're giving it time to evolve, and time for serious thought and 'evolution' as you consider what is/isn't possible to implement... If worst comes to worst, we can always *fall back on* a 'native' port of UAE or whatnot, but you're doing your best to keep us from having to just yet. If the... what is it, 'component architecture' or 'services architecture?'... comes together, maybe we won't need to worry about it by then, because any such support could be implemented transparently enough that there wouldn't be a *reason* to bitch and moan.

...and so, through it all, it remains *Amiga* OS 4. :)
Newly released CAM article on OS 4 : Comment 43 of 146ANN.lu
Posted by Kjetil on 10-Aug-2003 14:00 GMT
In reply to Comment 41 (Joe "Floid" Kanowitz):
The word box gives a impression that environment boxed inn, can't relay say it boxed in quite open if you tell me,

and for the emulation part I call that API integration and 68k Jit translation, it's all about playing with words :-)

what I think integration is a better word in this case is that there is no translation phase whit the JMP_TABEL and the ppc library function,

And I think "Jit translation" is better words for "JIT emulation" first off all JIT is not trying to be some thing else it becomes some thing else.

To emulate means "trying to act as if it where some thing else", it's on going thing (Interpreted emulation).

Box/environment emulation: act as an otter environment
api emulation: make an otter set of API look like some thing else.

Apparently there are tow sets of API's in this design, how ever none of them are dependent of the otter directly, so there is no API emulation, it a dual headed library structure.
Newly released CAM article on OS 4 : Comment 44 of 146ANN.lu
Posted by Anonymous on 10-Aug-2003 14:33 GMT
A platform is not just about the software you can use(it's obviously important), but also the quality of experience while using the OS. Macs and Windows offer the former, but can they offer the latter? This is where new OS's have the opportunity to excel. In doing that, the former will follow. From what I read, they seem to realize this.

Signed...KenH on another machine :)
Newly released CAM article on OS 4 : Comment 45 of 146ANN.lu
Posted by Kjetil on 10-Aug-2003 15:30 GMT
In reply to Comment 44 (Anonymous):
I most disagree with you here a bit, yes new features are driven by crating some new things, just Se how Windows has changed and over time and you understand that they have Lott's of resources to make the system work, from the tiny detail up to the grater detail, competing can be extremely hard, when you have 100 of full time developers vs a few part time developers, and you can be shore that M$ are looking for new concepts in there competes products, they will just in implement them in the same way if it will give the competes way inn.

The last years I looked at Linux/Gnome/Kde and otter OS just basically duplication all the work Microsoft have done with there OS, the Gnome/Kde clones are near the quality of Windows desktop environment, the only OS i know of that is effected allot by Microsoft's features are MacOS, This is the OS that Windows has copied in meany aspects.
Newly released CAM article on OS 4 : Comment 46 of 146ANN.lu
Posted by anonymous on 10-Aug-2003 15:59 GMT
In reply to Comment 31 (Don Cox):
Don Cox writes:

'The purpose of a computer is to get things done, not to prevent things from working.'

And the purpose of any technology-driven business is ultimately to make money. I think we all understand the value of a transitional 68k solution and that has already been secured. However, moving forward we need to decide whether a few unsupported applications should hamstring the entire platform.

OS4 will provide a lifeline to some of this software and allow the original developers to update or license the rights to their solutions so that others can continue the development. As soon as this retroactive support becomes a liability all bets are off.

New hardware needs new users which in turn need new software solutions. Lots of opportunities are going to open up with OS4. It's all good.
Newly released CAM article on OS 4 : Comment 47 of 146ANN.lu
Posted by smp266 on 10-Aug-2003 16:22 GMT
In reply to Comment 27 (Anonymous):
If you plug a PPC 1.xGhz system into a 33mhz PCI bus what do you get?

Most of the components can be transferred to the Amiga 1 and back again. Thanks to Serial ATA this will become a lot less painful.
Plus who is going to write all the drivers for the cards etc. so everyones current system is fully compatible. You could be waiting awhile for that.

It doesn't seem like that much trouble to me. I'm more worried about whether my multifunction will work or not.
Newly released CAM article on OS 4 : Comment 48 of 146ANN.lu
Posted by Raffaele on 10-Aug-2003 16:24 GMT
In reply to Comment 45 (Kjetil):
Mr. Kjetil wrote:

>The last years I looked at Linux/Gnome/Kde and otter OS just basically duplication all the work Microsoft have done with there OS,
>the Gnome/Kde clones are near the quality of Windows desktop environment,
>the only OS i know of that is effected allot by Microsoft's features are MacOS,
>This is the OS that Windows has copied in meany aspects.


Yes MAC is the OS that Windows had copied in many aspects, because:

1) It has a standard menu interface programmers should follow to maintain same menu and command structure...

2) it is almost "cretin proof" because tends to advice the user on any choice he made...

BUT

M$ copied also from Acorn Archimedes (the TRAY-BAR)

MS copied IBM OS2 structure (kicking in the ass their old IBM partners) to obtain a real multitasking into WIN 95 and its sequel...

(You know it. They partecipate the developing of OS2 only to achieve as many know-how they can)

M$ copied also some amigaish structures...

THEN

because they have a lots of money...

...they hire a joint team of GUI programmers experts, a team of experts in semiotics and communication, and a lots of experienced users...

Finally M$ team collect all the suggestions of the experts and built the graphical interface for win95 as the easiest to implement versus usability, complexity and cost requests...

It is not so strange that competitors should implement something similar...

And mainly into Linux environment, because then you can allow users to migrate without necessity for experienced users and operators to come back to training again and learn another GUI usage and modify the "usage habit" they reached.

GNOME/KDE environment being similar to Windows95/98/ME/XP

MEANS

No need for training

...so no less of money & no less of time to firms who preferred Linux.
Newly released CAM article on OS 4 : Comment 49 of 146ANN.lu
Posted by Anonymous on 10-Aug-2003 16:33 GMT
In reply to Comment 38 (Hans-Joerg Frieden):
"You *can* run FinalWriter on OS 4 and edit a document that doesn't fit in your main memory because the virtual memory applies equally to 68k and PowerPC programs."

Is that actually true RIGHT NOW, or is it something still to be finally integrated?
Newly released CAM article on OS 4 : Comment 50 of 146ANN.lu
Posted by Amon_Re on 10-Aug-2003 17:03 GMT
In reply to Comment 49 (Anonymous):
IIRC it's already done, yea

Cheers
Anonymous, there are 146 items in your selection [1 - 50] [51 - 100] [101 - 146]
Back to Top