emulsion for Linux... yeah...
ZXNet echo conference «zxnet.pc»
Hello, boo_boo
There is no bullshit in UNRIL, there is no need to slander SMT. I can suggest an option:
see how the music and video output are made in the fuse, and replace this matter in
unrile. Unril should be able to exercise quite normally (just a little bit of VC++
may interfere, GNU C++ is more common on Linux, there is still some kind of
there is a small difference).
Hello, Vladimir Kladov
There's a lot of changes that need to be made in the world.
there is a lot tied up in API functions (CreateFile with all sorts of HANDLE and
etc.)
There is an idea, of course, to write a wrapper that will “emulate” the necessary APIs
functions (not all of them, of course, only those that are used a lot).
and there are no problems with video output, because in enril all these video filters are like
how easy it is to write.
You can mess around with the sound ;)
in principle, there is a way to remove direct access to the hard drive (in Linux it’s even easier
done than in Windows, but in the emulator you will have to change a lot), leave one
video filter (which should be written under SDL), sound also modified under SDL
and the guy (which still has one window) for now, remove it for now, then it’s fine
you can port it ;)
That's exactly what I'm doing ;)
but I have very few completed projects (5 out of 100) ;)
It’s easier at work - they pay money for it, but for themselves... they quickly forget about it.
so if anyone seriously ports it (and not like me ;) then I’ll help in any way
I can, fortunately I have experience in programming for anything (zx, ammy, pc-windows,
windows-mobile, pc-linux, and other things not related to emulators)
Hello, boo_boo
you read the documents - you whistle
Hello Sinus
I read, I whistle. kind of like a maz... essentially the same garbage as wine (same code),
however, you can rewrite parts that are buggy/not working - at least under SDL, at least
under what. Tomorrow I'll try to put it together, throwing out outright leftism like bass --
I wonder if it will even start? ,)
Hello Sinus
hmm, maybe it’s really annoying... yeah, there’s bullshit in the form of a settings dialog, but oh well
with her. What’s worse is that calls to vinapi functions are scattered everywhere. heh, don't try it
Is it possible to build it with winelib? %)))
2 sinus - oh, yeah, it would be such a job to port the emuls... darling =)
*went off to think and read the winelib description*
Hello, Electrician
> By the way, who has VC++, under which it is assembled (on mine
> cut 6.2 from a disk for 70 rubles, it will not compile)
It’s better to buy a disk with vc7.0/7.1 for another 70 rubles. the thing is that I started
rewrite MMX/SSE inserts to a higher level, via intrinsics
> at least fix it so it works under Cedega
will not work, Cedega does not implement functions for reading from ini files, and without them
emulsion does not initialize
Hello, boo_boo
well try it ^_~
although for me if it’s “the same garbage as wine” then it’s too cumbersome and
slow for the emulator port.
Hello SMT
And, that is, everything is because of the ini... Well, this can be completely corrected. If I find it
correct VC++...
Hello Sinus
By the way, who has VC++, under which it is assembled (on my chopped 6.2 with
disk for 70 rubles it does not compile) at least correct it so that it is compatible with Cedega
it worked. There you just need to remove the check for the Windows version.
Hello, Electrician
The problem with wine is not the general slowness, but the slow and unfinished implementation
separate winapi-fiy and subsystems. that is, if pieces of emulsion using these same
functions, rewrite them so that they use native Linux interfaces, problems
There shouldn’t be... like sound, video. and something from which speed is not required
(creating threads/opening files/bullshit... - what is most difficult to port
due to the fact that it is in small scattering throughout all varieties) let it slow down to its heart’s content
)
however, life will tell...
Hello, boo_boo
> gentlemen, emulator writers! and tell me, plz, how to do autorun
> of drive A, similar to the type in US or Pentagon48 using the reset_trdos button
After reset, you need to set the variable responsible for the “TR-DOS trigger”
> Isn’t there a lot of asma there? Rewrite in tse
In general, I have already rewritten a lot. only CPUID and RDTSC timing remain (if
works without sound) and a couple of filters (text64, bilinear). so it's okay
just delete
Hello, boo_boo
boo> b) assembler inserts have an Intel syntax, but AT&T is needed
boo> (or you will have to use an external assembler). plus, they will probably come out
boo> hard-to-find glitches related to data type dimensions
boo> alignment of structure fields and other such nonsense
boo>
Isn't there a lot of asma there? Rewrite it in tse.
boo> maybe it will be easier to polish glukalka and screw it to it
boo> little by little features from the US?...
Then FUSE is better. The glitch was not working when I last saw it.
in general. Well, that is. gives a gray background, you can hear the cursor clicking on the menu and
that's all.
Hello, Electrician
> Then FUSE is better. Glyukalka when I last saw her not
> worked at all. Well, that is. produces a gray background, you can hear the cursor on
> clicks the menu and that's it.
I just have the latest (0.90) glitch that works better than fuse.
maybe you assembled it with a left motif/lesstiff?
I rummaged around in the glitch, corrected the Kempston mouse emulation there, made a download
disks, snapshots and tape from the command line...
gentlemen emulator writers! plz tell me how to autorun disk A,
like in US or Pentagon48 using the reset_trdos button?
Hello, Electrician
heh, not everything is so simple - when I tried to compile US with winelib, the following came out
dirty tricks:
a) incompatibility of the syntax of C compilers, well, that’s bullshit.
b) assembler inserts have an Intel syntax, but AT&T is needed (or you will have to
use external assembler). plus, hard-to-find glitches will probably come out,
related to dimensions of data types, alignment of structure fields, etc.
so stupid
ve) the code is unstructured, you’ll get tired of debugging
maybe it would be easier to polish up the glukalka and add it to it little by little
features from US?...