On Fri, 21 Apr 2000 04:21:29 +0200, Patriek Lesparre wrote:
>>But I think even if only one board is made, a few people will buy it.
>And that's exactly why only one board should me made, with the few people
>buying a new board anyway you had better make sure they all buy the same one!
It's just a matter that I'll hardly buy a faster MSX that will not run
faster my old programs. (-:
>Depends on what you think is the MAIN MSX base. There are almost
>exclusively turboR's in Japan. Lot's of turboR's in holland and the rest is
>almost all MSX2. Except some countries where people still use MSX1, but
>there seem to be a lot of them.
Yup, there is a lot of MSX1 in the world. (-:
>So does that mean we should only make MSX1 programs? Or only MSX2 programs?
>Or??? One should program a program (^_^) for the system that the program NEEDS.
There is the need for MSX1, MSX2, MSX2+, MSX TR and even more... (-;
I think Ricardo's MSX1 programs are better than lots of other software
available even for TR. See fudebrowser, in the last version... HTZ up to
8Mb! The speed is really fast...! And if you forget the Mapper, it's a
MSX1 program.
>If you're going to do a 3D-shooter, you probably won't do it on MSX1.
>Probably not on MSX2 either. You probably would go for turboR and maybe
>even Gfx9000! If there was Z380 you'd most definately use that, because you
>can create a better program!
(-: If you plan to do a 3D shooter, you will use the fastest machine,
it doesn't matter if it is Z180 or Z380... (-: And better than gfx9000 would
be a ADVRAM + Gfx9000. But you know what will happen if you make a program
for padial's Z380 board plus MoonSound plus ADVRAM plus gfx9000... ? You'll
be the only person in the world that will play it... Well, there are
one or two more people, but that's all. This is the real problem. ((((-:
In the perfect world everyone has the topmost system... unfortunately,
we do not live in the perfect world. Sometimes a MSX1 program may
me more apreciated than a TR one.
>So my AMD Athlon system is a new computer and should not be called a PC?
In several aspects, NO! (-:
> It
>has a totally different bus than my previous PC. It has many many many more
>instructions than my previous PC.
If you think, you'll see that it's NOT a PC! It's processor is not
even a x86! It's a RISC processor emulating a x86. All these new
instructions are emulated instructions. So, not only Athlon, but even
PIII are not PCs (read x86) anymore!
PC was never a standard. (-: They changed the processor. They changed
the bus. They changed the memory system... The only backward compatibility
they preserved (badly) was software.
>Emulators, especially MSX emulators, aren't perfect. If you develop on a
>emulator, you will probably have to rewrite parts of the program or do some
>heavy bughunting if you later try it on the real hardware.
>Heh, I see proof of this everyday (in the gameboy programming scene).
The difference is: when we are programming on BrMSX and we find a single
difference between the emulation and the real machine, we say it to
Ricardo and he corrects ASAP.
And MSX1 emulation of BrMSX *is* perfect. Even on timing. MSX2 is not
perfect yet, but it'll be some day. And even TR emulation will be perfect
in the future. I know Ricardo. He is capable and has the patience. (((-:
>Sure, but Z380 with ADVRAM will be faster yet.
>You're still not proving Z180 is a better choice than Z380.
Well, I said it's faster on overal and the bus will be faster using
Z180. (-: The 50% speed will do a lot difference when talking about
disk-to-memory transfer. Play an Eva will be faster on Z180 than on
Z380... but generating a fractal will be faster on Z380 than on Z180.
It's just a matter of target.
I do not like 3D games, btw.
>Heh, after debugging/programming with NO$GMB BrMSX always leaves me wanting
>more :(
Ask Ricardo. What is on BrMSX is what he needed and what he was asked to
add. There is no reason to create a debug with hundreds of info that is not
used by any one. BTW, what do you think it's missing on BrMSX FuDebug??
>No, I meant Z380 isn't much slower than Z180, so its useless to want a Z180
>for only a small speed increase while you can have all the extra processor
>power of Z380 if you just tap into it!
You only think on matemathic functions? When will we use 32 bit mult on a
MSX?
>Besides, I can't think of any program that needs speed and shouldn't need a
>rewrite...
>You say PMEXT/PMARC. It's old and stupid (rounds off files to 128 bytes).
>It's better to use LZH. And writing a LZH compressor/decompressor is as
>easy as downloading a C-sourcecode from the internet and compiling a Z380
>version. (or a legacy Z80 versions that will still run fast on Z380)
>And then you have subdirectory support!
C compiled maybe becomes slower than PMExt. BTW, someone will have to
write a C compiler, right?
>You say UZIX. UZIX is still in development and would benefit greatly (and i
>mean GREATLY) from Z380's advanced features. It would be a shame to not
>create a seperate Z380 version.
How compile the C parts optmized for Z380? Adriano uses HitechC which
generates Z80 code. There is more problems. This is why have Z80 code
running at faster speed. We have not tools to build programs on Z380,
no even on Z180 opcodes... Just ASM.
>That bullshit. If you program Z180 using MLT opcodes, no Z80 will run it!
>If you program Z380 using 4GB linear, 32 bit registers, no Z80 will run it!
>The same counts for both processors! It's just that the Z180 doesn't have
>much more to offer over a normal Z80, just like R800 did. That's why so
>many R800 programs (including my own GEM) run on a Z80.
What I'm talking is that you will run VERY FAST on Z180 code that runs
on Z80. To have VERY FAST code on Z380, you'll need to generate code
that DOESN'T run on Z80!
Did you know, btw, that Z380 does prefetch, which may increase it's
performance, but sometimes we need to byte align it's code, inserting
NOPs? So, a code that runs faster on Z380, p.e., will hardly run usable
on Z80...
>Well, ASCII doesn't seem like its upgrading the MSX standard, so why
>shouldn't we?
Because a new standard is not discussed. Someone do something and
says: This is the new standard (as Panasonic done with TR). When
a new standard is discussed, we never get to the point.
>>All programs that run on a normal
>>MSX will run with ADVRAM also. (-:
>Heh, that's a nice thing to say, but it doesn't work the other way around!
>If you want to create cool new soft using ADVRAM, people without it can't
>run it! Just like if you create cool new soft running on Z380, people
>without it can't run it. And so on...
>It's all a little contradictory, is it not?
Are you comparing a main processor with a new video system...?
I don't know why you insisit on saying I'm against Z380. NO! I had
fought in the past to use Z380 on a new MSX. I just changed my mind
because it will be faster...! (-: (you will say... but it has
dozens more opcodes, twice registers etc, etc, etc, etc... that
is a point, I agree, but - again - which everything bring to
my old programs?)
>As far as I have seen the Z380 manual, the assembly language is still 100%
>the same as Z80.
>Have you seen official R800 assembly spec's??? NOBODY uses that!
>MOVEM [.HL++,.DE++] in stead of LDIR, hah!
Well, there is only one Z380 assembler: the zilog one. If the ASM used by
it is different, we will need to adapt ourselves. I'm ready to this.
This is not a problem to me. The problem remains the same if we want
to program using Z180 specific opcodes, also.
>So upgrading to a new system hightens his productivity? At least he
>promises good things.
>I'm in the same position, although slightly different, and I think others
>are as well. I have here a turboR, Video9000, MoonSound 640kB, but no
>reaaaaallly great processor power. When I get a Z380 (and I WILL get a Z380
>for sure) I will make a Z380 GEM, and I will probably make lots of other
>stuff! Coz Z380 is cool!
Yes, it is! And I think you'll really do it. GEM is a nice program (althought
on version 0.4 it hangs badly my MSX, with RType, that was previously working...
but I had not tested it extensively yet.)
>*yawn* I'm tired, I'm probably talking nonsense again, and nobody probably
>cares about these messages... I think we're about done talking anyway,
>since we seem to agree on many things. The only difference is: I want a new
>MSX, you want a faster MSX.
Yes... (-: But it's nice to talk about this... ((((-:
AbraçOS/2, Daniel Caetano ([EMAIL PROTECTED])
...!m.tag
OS/2 Sites: http://www.quasarbbs.com/daniel/
http://www.geocities.com/SiliconValley/8752/os2hp/os2index.html
MSX Sites: http://www.fudeba.cjb.net/
Drawings: http://www.djgallery.tsx.org/
****
MSX Mailinglist. To unsubscribe, send an email to [EMAIL PROTECTED]
and put "unsubscribe msx [EMAIL PROTECTED]" (without the quotes) in
the body (not the subject) of the message.
Problems? contact [EMAIL PROTECTED]
More information on MSX can be found in the following places:
The MSX faq: http://www.faq.msxnet.org/
The MSX newsgroup: comp.sys.msx
The MSX IRC channel: #MSX on Undernet
****
