[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [MiNT] minimal support base



On Fri, 3 Dec 1999 01:44:59 +0200, Martin-Eric Racine wrote:

>On 2.12.1999, Lonny Pursell <atari@bright.net> wrote:
>
>> if you have too little memory to run MiNT properly then it's
>> your bad luck.  Why hold back the entire platform?
>
>This has been brought up before, but not in this perspective of
>holding down the platform evolution to fit the ST's limitations.
>This is a quite novel and interresting perspective.

I don't see it as novel, unless I miss-interpret the meaning
of the word "novel".  I see it as very important.

The /kern issue is a good example, since I have 72meg of ram I am
not concerned one bit (bad pun there) if the human readable aspect
uses up an extra 32k bytes.  That is not to say I'd be happy with a
bloated kernal, but I think that 32k or so is not a bad trade off
for output that is easily dealt with.

>Recently, Konrad also asked wether we should continue adding more
>features and optimization options, despite the increased memory
>requirements, or if maybe it would be better to slip development
>into separate kernels for 68000 and 68k.
>
>This leads to one question that hasn't been agreed before:
>What should our minimum support level be?
>
>68030?  With or without assuming a 68881 FPU is present?
>Or more generic, like 68020-40, perhaps?
>
>And how much ram should we expect?
>
>Any specific TOS version for each hardware type?
>
>Or should we keep on assuming 68000 with 4 MB and unknown TOS?
>
>Assuming 68000 support is still desirable, should we fork this
>into a separate kernel, optimized for 68000 and minimal ram?
>
>Do we assume an AES will be used and how much ram do we set
>assides for it and its applications?
>
>There are probably other issues I might have forgotten, but this
>should provide a good base for discussion.

I honestly dunno how to vote on those issues.

>PS:  it would be nice if as many people as possible expressed
>     themselves on this issue, as it might determine a lot of
>     things on how future kernel development will procede.
>
>     Mere users are particularly welcome to express themselves
>     on this issue, since they know better what hardware they
>     need MiNT to perform.

All I can say is I've went after the upgrades I needed in order
to get my machine to what I feel is a robust setup, so I don't 
have such limitations.  Granted it wasn't cheap, but it is my
main hobby this old Atari.  If others want to be stuck in a timewarp
with limited machines and not upgrade at all, then how can the
software improve?  I think sometimes people expect to much from
an old ST.  ;-)

__________________________________________________________________
Atari Computer User: Lonny Pursell        E-Mail: atari@bright.net
WWW: http://www.bright.net/~atari/       IRC: lp @ #Atari @ IRCnet