@xispo
>Traditional Amigans can continue with their OS4, MOS, AROS, Natami if they want >to. Their numbers will continue to dwindle, slowly fading away, until the >inescapable conclusion: Total disappearance.
yes thats right for many, but i see it this way.
some people code in C, some in visual basic, some in Java, some in .net. somein windows some in Linux.
I like code in amiga os and especial in amiblitz.I have also old programs that i develop long years and fit mostly my needs.if i need new feature i add it in short time.I like not only code, i like use my new features.i do many imageprocessing, i make musik and i want use programs not only code.I have only limit time and if i must choose much work or by for a few euro a faster X86 i buy a faster X86.
with winuae i and all that have a rom can use this programs without much work.
.net or java programs are too emulate, and here can see, it is possible to integrate this better.
using emulation is not bad, and help to use more programs without reboot as can see the many users that use vmware and co.
what i not understand is, that many devs on amiga dont like code on a emulate system, many of them give up amiga complete.
And as far i see very few learn then code in windows Linux or something else.they code then nothing in their free time.
But i loose not motivation to code on amiga, i like it and hostet i can run forever my programs.
sure as a native system i dont use AOS, i like news so i want test windows or linux programs too.
>My point is that, in my view, Anubis tries to appeal to people outside the Amiga >community. It's not about bringing more power to the Amigans. It's about exporting >Amiga-inspired ideas to a new public. If it fails in that regard, then bad luck. At least >they tried.
now i understand, lets see what happen.But i have less hope here, because there are many OS out with memory protection, but only very few get a user base of more than 10 000 Users.but there is still hope, maybe a embedded developer use Anubus because of their lightness and cheap.But they must hurry up, because i think in 5 years handies or netbook have powerful CPU and much ram to run too bigger OS.
>The problem is that Amigans don't want new ideas.
thats not only the amigans because it cost time to learn new and work to change.and if the advantage is not much noticiable wy do much work.
X86 is for asm coders a bad CPU, but asm is selden used and X86 have all features as other CPU and was first availible.when PPC is introduce in 199x then PPC was in compare to X86 faster, and price was ok.A 604 outperform a X86 Pentium pro or pentium2 at same clockrate.and was cheaper.That was the reason wy i buy a cyberstorm PPC 233 cybervison PPC as soon i can get.
but then newer X86 get later better and cheaper, but thats a result that not all change to PPC because it give better price power.they stay on X86 because a change need more work as the advantages can get.
I like new ideas, so i do AFA.it have in some points more in some points less than OS4 MOS features.
It depend, of my matter of taste what i want.better look GUI i see all time, but memory protect i notice only in few situations.also a crashing program is always bad, so i like more write or use programs that dont crash.also my windows box and amiga box crash very very selden.when i develop a reboot is most time faster.
But i also add a safer system mode to AFA, that do parameter check of OS calls.for example it is impossible to crash AFA due to wrong msg address, port address or, devide address.
It do it by checking for node id and cookie.
most programs work without problem and it increase stability a lot.i know this from blitzbasic.when runtime check is on then it is near impossible that a program crash whole system.when you call a blit command with wrong rastport, then bring an error, wrong rastport and stop the program or jump in debugger.normaly AOS crash then without runtime check
I dont see that MOS OS4 support that features.and thats a first important step and is also possible on AOS without new kernel as see in AFA.I begin since years with that, but AROS and zune is still very sensible and totalcrash with wrong parameters.MUI and AOS 68k are better resistence again wrong OS parameters, so i add at least the checks AOS do in AFA.for example AOS dont crash if a freemem is called with 0 pointer.AROS code do
memory protect did not help to keep the system alive, when a OS function can call as in AOS with bad values.if a OS function can do illegal mem access and write wrong intern structures (for example add a bad message port address in the window ports)
when whole AOS is protect by paramtercheck it is near impossible that it crash.programs that overwrite mem are rare and can also easy find when use memtracker.
some programs access uninialized pointers (access adresses in range 0 -$800) they also can easy detect and avoid.
So i think the NEW what is offer by OS4 is not so important, that it break compatibility much, it is only a excuse of the lazyness of the developers to fix Problems.
It is much more easy to write a OS for that programmers must change their programs, but it is hard to write a OS that do not break
I aggre that for step to full memory protect compatibility give off, or run old OS in a box, but that what MOS OS4 currently offer as memory protect i think need not loose much compatibilty.
and i dont like to change code, pay much money when OS developers are lazy and add features that can do without much compatibility loss in very few time.
also when i activate safer system mode in AFA and use tlsf mem, there run many 68k programs.
Last edited by bernd_afa on 30-Mar-2009 at 09:46 AM.
Last edited by bernd_afa on 30-Mar-2009 at 09:45 AM.
Last edited by bernd_afa on 30-Mar-2009 at 09:43 AM.
Last edited by bernd_afa on 30-Mar-2009 at 09:42 AM.
Last edited by bernd_afa on 30-Mar-2009 at 09:41 AM.
|
22 crawler(s) on-line.
105 guest(s) on-line.
1 member(s) on-line.
Leo 
