What should be the principles of the new OS?
ZXNet echo conference «zxnet.soft»
Hello SonOD
Here the question was once raised that there is supposedly no timer on NMI, for a full
writing a full-fledged axis and so on, why not come up with some kind of scarf
which is inserted into the slot, on this board there will be a Flash-ROM where there will be
BIOS of our super OS, which will be used instead of the main ROM when
when turning on the computer, also place a programmable timer on this board,
maybe some kind of coprocessor, somewhere on the forum the idea of creating
something like a coprocessor for the TCP/IP stack. Since the size of the scarf can be
if you have more, you can immediately provide a DMA controller, USB or something else,
controller manager for executing programs in RAM for example, and then some
the questions raised above will disappear by themselves. And with this approach the minimum
alterations to the main board, almost the only modification is shading
system ROM. When using Flash-Rom, it becomes possible to quickly
change the BIOS version and do not need to unsolder and run for the programmer.
Hello,geners
One thousand five hundred times right. Shit is not done for everyone. Do it for yourself or for
cart.
Hello ws_mason
ws_> One thousand five hundred times right. Shit is not done for everyone. They do it for themselves
ws_> or add to cart.
So to speak, between the lines I remind you that the TASiS release was released only under
Turbo2+, in fact, only two people were engaged in Yura Korsunin and Maksagor.
Well, of course, one cannot help but notice the virtual TRDOS for Turbo2+ - which allows
load and save TRD images and work with them from any OS on Turbo2+.
Which ensures the requirement of OS compatibility with TRDOS. This extension was written by
In general, one person Yura UKMSz.
So everything is done normally when there is a normal leader who leads
a project (in our case Maxim Timonin) and a desire to do something useful.
Hello, CHRV
CHR> Well, of course, one cannot help but notice the virtualTRDOS for Turbo2+ - which
CHR> allows you to load and save TRD images and work with them from under
CHR> any OS on Turbo2+. WHAT ensures the OS compatibility requirement with
CHR> TRDOS. This extension was written by one person, Yura UKMSz.
CHR>
Well, in the case of xBIOS/vTR-DOS, I also contributed as an encoder. Although
The share of my code, of course, compared to Yura’s is extremely minimal. But the concept
what should be implemented, put forward by me and Yura, have been implemented
in approximately the same quantity.
Hello, Maksagor
gentlemen and again tasis - this is for Atmka but what about other cars
it turns out - if you want to work in a new axis, buy an ATM
some bad advertising 6)
Hello, CHRV
Well, as far as I remember, Scorpio is also quite expensive
well, then it’s not an option and it’s morally outdated
Damn it, I bought it from you
Hello,geners
then you need a new format for floppers - a sector system :(
but the system needs to be developed not as it goes with you
in the system requirements column -ATM 2+
but seriously as in pussies
what frequency, what periphery. memory size
in fact, all the problems of the spec are due to the fact that everyone refuses with a show-off
achievements of third-party platforms.
and then they grab it - damn... it would be nice to connect a hard drive, then let’s do it like this
so that there is nothing incompatible white except ours!!!!!!
and it’s also necessary that all the old standards were upheld - which have accumulated...
knows how much
Dc., your energy spent on the development of all sorts of hardware emulators for hbios
and finalize the BASIC - direct it to the development of a new axis and a converter
the same programs from tyrdos to the new axis.
since the coders are so cool, then write a program that replaces the appeal to
ports for accessing points on the system is a piece of cake :)
and besides, who canceled the SRM
Hello,geners
gen> then you need a new format for floppers - a sector system :(
The reproach is incomprehensible. Like the essence of your phrase.
gen> and the system needs to be developed not as you have in the system column
gen> requirements -ATM 2+
gen> and in a serious way as in pussies
gen> what frequency, what peripherals. memory size
gen> actually all the problems of the spec are due to the fact that everything is a show-off
gen> refuse the achievements of third-party platforms.
The problem of the Russian spec, as you correctly noted, is the divergence of standards.
The same “standard” Pentagon is the crookedest machine, but widespread and
that’s why everyone suffers from this (especially the organizers of various parties when recording
received on video).
Well, it’s impossible to make a system so that it works equally well on the Pentagon
and Turbo used up all the resources. After all, WinHR doesn’t work on a 386 machine, well
at least kill yourself :).
gen> and then they grab it - damn, it would be nice to connect a hard drive, then we’ll do it
gen> it's so incompatible white nothing but ours !!!!!!
And what do you suggest should be compatible with: Kai, Scorpio, ZXNEXT.. By the way
Turbo was one of the first to have IDE. It makes no sense to drag other people's standards,Moreover, IzDOS has a DRIVER for this.
gen> and it is also necessary that all the old standards were maintained - which
gen> has accumulated x.. knows how much
gen> Dc., your energy spent on the development of all kinds of hardware
gen> hBIOS emulators and BASIC improvements - direct to the development of a new one
gen> axis and converter
gen> the same programs from Tyrdos to the new axis.
As practice has shown, few people need the new AXIS because there is nothing underneath it
coming. And we went the other way, we made a layer that allows you to run
a lot on any axis (for Turbo of course) everything that is already there and most importantly WITHOUT
ALTERATIONS. And now that we have this foundation, we can improve
the AXIS itself.
gen> since the coders are such cool, then write a program that replaces
gen> accessing ports and accessing points in the system is a piece of cake :)
gen> and besides, who canceled the SRM
But no one canceled it, it works great together with Tasis. I have it on hard
disk and both.
As for accessing ports, everything works through drivers and input points
systems. It’s probably just that you’re not very familiar with the internal kitchen (and even
SPM everything works through the table of entry points).
Hello,geners
gen> well, as far as I remember, Scorpio is also quite expensive
gen> well, then the isdos is not an option and it’s morally outdated
gen> ktati pozhklsta I bought from you
Our topic was moved to hardware, where I again said goodbye to Penatgon :).
It's good that you bought it. I respect you!
Hello, Dima_ZX
Dim> Here the question was once raised that there is supposedly no timer on NMI...
The question is not about boards at all, look in the “hardware” section - Caro offered a lot
options - both for the TCP stack and for the PS/2 mouse and much more, however
each of these little things puts a strain on the pocket and leads to an increase in the price of the already not
such and such a cheap computer (just ask how many people bought
Pen-1024 or any other Spectrum at least for a year - there are very few of them and the price
is still a significant factor, because you will compare the same 2000 rubles with
a similar computer for the same money, but IBM-compatible), if all
add bells and whistles and the price will be budget but MODERN
IBM compatible computer. And of course there are people who won't care
Solder the glands to your Speck, but it won’t work to graft this as a standard
(just healthy skepticism) - we’ve tried more than once and to no avail
led this.
gen> I look at you and understand why you will never be satisfied
gen> more than one axis for spec - YOU JUST DON'T WANT IT, you don't need it
No, it’s just that everyone wants to see their own set of functions, and therefore in different ways
understands how they should be implemented. You can ignore the question
compatibility, however, it is the old programs that still give life to the Speck (mostlyto the extent that they are the stimulus that brings people nostalgic and
just interested) and those new ones who are still (I’m just extremely happy
this fact) appear on the platform at least periodically.
Therefore, at least to some extent, the system must work with old programs.
and then look below:
ws_> One thousand five hundred times right. Shit is not done for everyone. They do it for themselves
ws_> or add to cart.
here it is, what I wrote about in the paragraph above - add to trash - this is the same refusal
from all compatibility - for yourself - similarly. Any OS must be supported
- that is why the “lucky” owners of IBM-compatible computers bet on
it is Windows and not Unix and its clones - there are many times more commercial programs,
There is more pirated software, and there is simply a lot of exclusive (non-)pirated gaming software. Well
the author wrote a super-mega-unique system - so what? - again the main thing -
support. From this point of view, the best solution is iS-DOS (the most
old OS on Speck which actually has drivers and all the equipment, although
there are also many disadvantages), the most practical - TR-DOS - its functions in general
essentially trimming, but they work from time to time - as in the joke about
programmer - that’s why it is “where they were and where they stayed.”
I started writing a memory manager myself, but again I don’t have time, I want tospeck and do something else, etc., I’m not the only one, that’s why the project is moving like this
"quickly".
Hello, Umka
2 umka
Yes, no one will write a new axis - this is a fact, they will mutate too isdos
cpm
and no one will be involved in porting, everyone wants everything ready,
and with a bunch of application programs that they themselves don’t want to write
Hello,GriV
forgive me dear comrade, but over is not a stump :)
this is 486 with an extended set of functions
but pentium overdrive = pentium 1
486over doesn’t even have overdriva rows involved
Hello,geners
and by the way, I didn’t see the card that was mentioned at the beginning
pages
and in general the launch of Sandra is highly doubtful
One user's ana on 166 mx slowed down when I tested the system!!!!
and if anyone doesn’t believe my arguments, I can repeat this operation as on the website
described - there is iron
Hello, Umka
Umk> Also, IMHO, you should not try to support Spectrum 48
Umk> This is the same as if Microsoft tried to add it to Windows XP
Umk> support for 386 machines....
A little offtopic, on a 486overdrive (=pentium) with 16 MHz and 32 meters of memory
launched WinXP link http://winhistory.de/more/386/xpmini_eng.htm
I checked it works
Hello, Umka
off: I had win95 rus oem (lic.:)) running on my 386sx-33/ram-4mb
Once again we slipped into flame... it's a pity...
ws_mason, you and I also worked hard on the ballot box... project thunder moon
has sunk into oblivion. You should post here your developments on drivers... I also based on the archives
I'll dig around. after all, we were thinking about the axis :)
2 all. I support GriV's candidacy as a coordinator.
I can provide all possible assistance in writing modules/subroutines, since I have
quite rich prof. programmer experience.
I also support the minimum configuration: ZX-16 with a tape recorder (may be,
hardtaper?), although kay1024 himself will soon have a screw...
Hello jdigreze
I posted “Thunder Moon”, only in the topic about graphics, about the tutorial on
drawing one-bit graphics. At the moment I am not directly involved with Spectrum,
I'm just following the news.
Hello dwt
As far as I understand, the main technical issues have been addressed. Currently
all interested parties were divided into several groups. Every group
develops its own OS in accordance with its ideas. There is no single OS
it will work out.
Hello,geners
2Umka flea market, where for 50-100 rubles. you can buy Spectrum 48
Isn’t it fate for the 128th Spec to be switched to mode 48?
:(
Both from the 48th and from the 128th Spec, the memory area 49152-65535 is clearly visible.
In addition, there is the area 32768-49152 as a kind of “window”. Thus
"system" (the simplest multitasking control shell) will work on both
cars. The advantage of the 128 machine is the additional memory banks that
will be marked by stage, probably so.
CHRV once said that rather than figuring out how to dodge the codes, it’s easier to do
professional program loader (boot manager). Worked in a graphic editor -
saved the picture, went to the manager, launched the HTML editor - inserted the picture and
etc. Or in general, on command - bypassing the manager, he left one program for another.
2geners yes, no one will write a new axis - this is a fact, they will mutate too
isdos cpm
Why is this so? However, it may be possible to write, but surely the concept can be laid out? A
whoever likes it will implement it. So, besides Doors and Beeos there is nothing, no
docks... almost. I'm busy with the PC "stupidity" right now, after running it in I'll get to work on Speck.
Much has been forgotten, so the mechanism itself will be debugged at the PC, and then,
little by little something similar will be written for the Spectrum (possibly 48/128 for TR-DOS) not
necessarily by me, but by a different coder. In any case, documentation is 40%prepared and will be posted online, and then it’s an amateur’s game. Let's be brave
look ahead! Apparently some forum members are sitting and waiting for some thread
raw dema, in order to bark at her later, isn’t that what happened with the beautiful Doors?
PS: regarding the “stupidity”, I can send you a demo by email, but it’s unlikely to be of interest,
these are just parts of the whole.
Hello Vovoi
I already wrote somewhere... If anyone wants to advance their axle, then they will have to go to it
also write an assembler + debugger, so that it’s cooler, or at least not
worse than the best available. The best option, I came to him myself,
write a program, for example a game, as if it were running under control
your not yet written OS. Accordingly, some parts of the program
for example, working with a disk, writing it as part of the OS (so to speak, a foundation for the future).
But most likely it will just be a set of more or less universal libraries.
Hence the conclusion -> you need to write an assembler with a debugger: we will get a new tool
and the beginnings of the OS.
Hello, CHRV
CHR> There must be a working group that deals with the Axis
And I suggested making a separate invisible section for such a group!
Hello, CityAceE
Cit> And I suggested making a separate invisible section for such a group!
Why is it invisible, readonly is enough for everyone, and whoever submits the application is the one
hits.
Breeze, have you thought about taking over leadership yet? ;)
Hello ws_mason
ws_> I already wrote somewhere... If anyone wants to promote their axis, then they should
ws_> you will also have to write an assembler + debugger for it, but to make it cooler
ws_> or, at least, no worse than the best available. Optimal
ws_> option, I came to it myself, to write a program, for example a game,
ws_> since it seems to be running under your control, which has not yet been written
ws_> OSes. Accordingly, some parts of the program, for example working with
ws_> disk, write as parts of the OS (so to speak, a foundation for the future). Only
ws_> most likely it will just be a set of more or less universal ones
ws_> libraries. Hence the conclusion -> you need to write an assembler with a debugger: we get
ws_> and a new tool and the beginnings of the OS.
As I already wrote above. There must be a working group that deals with the Axis,
Moreover, there must be a fairly strict leadership of the group by a person who
clearly sees what needs to be done, and can write some technical specifications (from
I have already named those present who can be trusted with this).
If this does not happen, then naturally nothing will happen.
Hello, CHRV
2VOVOI
Please send me the demo
Hello, CHRV
CHR> Breeze, have you thought about taking up management yet? ;)
hmm... I don’t even know what to say :( my head is a little different now
problem, I borrowed a large sum and got screwed... now I’m thinking about how to repay...
:o
Hello, breeze
bre> hmmm... I don’t even know what to say :( my head is a little full right now
bre> another problem, I borrowed a large amount and got screwed... now here
bre> I’m thinking about how to give... :o
I played it on the slot machines :). Kidding.
Well, everyone has such problems.
Hello, CHRV
CHR> Somehow I scrolled through the slot machines :). Just kidding.
CHR> Well, everyone has such problems.
Well, I don’t think that everyone tried to open their own company, and their money was stolen
:mad:
hmm... that's the golden rule! trust - but verify! and preferably more often :(
PS: okay! let's go! I will contribute as much as possible! but on team leader
I obviously can’t handle it :( sorry!
Hello ws_mason
ws_mason> I already wrote somewhere... If anyone wants to promote their axis, then they should
ws_mason> you will also have to write an assembler + debugger for it
If this is not done, then no one will need such an OS. That. you create
a simple boot manager and attach to it a type of microasm, tailored for boot.
Anyone who wants to find fault, then wants to do something and when the program
will compile and work in tandem with the boot, then the cell will begin to think further, and
What else could I screw in there? what to improve?
I don’t know much about Linux, but I heard somewhere that there was a C compiler and therefore Linux
I didn't die. Those. there were people who immediately realized that for the new system,
You CAN easily write YOUR OWN programs. Without the compiler, the project would probably have died out long ago
or would be the plaything of the chosen ones. (IMHO)
For example, Spec's "system" Windows 1.4, well at least the author described HOW
make programs and they actually work out (though sometimes with swear words), if he
I screwed the asm there, then both for iS-DOS (with is-assemler) and for Win1.4
It would be possible to create applications. And so, only the entry points are described, and
implement it as you know, this, IMHO, should not happen. That is, if in the system
says "don't use #3D13!", then you shouldn't break the rules, no matter how
I wanted the same to be true for displaying texts on the screen and using RAM.
The best option, I came to it myself, is to write a program, for example a game,since it seems to be running under the control of your OS, which has not yet been written.
I, and probably many others, came to this conclusion. And (if I'm not mistaken) in DomenOS the author
said, if coders left at least 50 bytes of free space, then in these
sections of the program could be pushed into cutting down the process, etc., for organizing
at least plug-in operation of the system.
Accordingly, some parts of the program, for example working with a disk, should be written as
parts of the OS (so to speak, a foundation for the future). But most likely it will be simple
a set of more or less universal libraries.
You will get a set of plugins, apparently this is one of the outputs on Spec.
:smile:
It’s even easier to create basic libraries for loading and unloading data, etc.
operations (which are typically performed by many programs, even output
text). These libraries are loaded at the end of the asm text, well:
ORG PROGA
LOAD EQU 10000; I forgot how this is formatted, but the general meaning, I think, will be clear
LD HL,Address
LD BC, Lenght
CALL LOAD; something like this
RET
LOAD ORG; the library itself, the procedure for loading data into RAM
RET
After creating the system and publishing entry points, the source codes
undergo minimal editing, and all changes to the system concept (for example,
I implemented the loading this way, but in the system it’s slightly different) will concern
modifications to the LOAD procedure so that the data received fromsystems (loading) were brought into a form understandable to the program, but for SAVE
the order given by the program was formed in the format accepted in the system and so
further.
Hello Vovoi
Vov> For example, Spec's "system" Windows 1.4, well at least the author
Vov> described HOW to make programs and they actually work (though sometimes with
Vov> obscenities), if he had screwed the asm there, then the same as for iS-DOS (with
Vov> is-assemler), and for Win1.4 it would be possible to create
Vov> applications. And so, only the entry points are described, but implement it as you know
Vov> this, IMHO, should not happen. That is, if the system says "not
Vov> use #3D13!", then you should not break the rules, no matter how
Vov> I wanted, the same goes for displaying texts on the screen and using RAM.
Your information is outdated: a more advanced one has existed for a long time
version ZX-Windows v1.6
Among the "goodies" - you can finally freely move icons around the desktop
table as you please, you can have more icons at the same time, and more
little things...
Hello, Maksagor
> Your information is outdated: there have been more for a long time
> advanced version of ZX-Windows v1.6
> Among the "goodies" - you can finally move icons freely
> desktop as you please, you can have the icons themselves at the same time
> more, and some other little things...
Sorry if I offended you in any way, but it seemed to me that the changes were more
significant.
:o
Hello dwt
dwt> I have thought about this question many times - is it enough
dwt> Spectrum's computing power for full implementation on it
dwt> multitasking OS?
dwt>
How many angels can fit in a beer bottle?
> And in general, if we consider the problem more broadly, is a new OS needed
> Spectrum?
>
What processor command does the kolobok represent the devil?
> The purpose of creating this topic is to define the basic principles, so to speak
> "ideology" of the OS. In my opinion, there is no need for
> initial installation of a graphical interface, etc. Necessary
> create an OS whose kernel will be minimal. The core that
>
The kernel belongs in the same place as graphical interfaces.
Kernels, GUI, mice, start button - these are the main attributes of any OS.
Of course... And we will write the software. Nothing to do. Let's stop carrying
all sorts of nonsense in the forums and let's go write software. What if
If only the software could write at such a speed, windows would have already been rewritten
would. Twice.
> What do you think about this?
This is absolute nonsense. OS on Spectrum...
Which makes sense. Take for example a Spectrum like Pentagon-512.Or KAY-1204. And TR-DOS. At least Kaevsky, at least 5.13, at least 6.xx.
What is memory above 128 kilobytes used for? For RAM disk.
And if you have one drive 3.5 and another 5.25, or in general, horror,
one disk drive. Or even if there are 2 disk drives - a RAM disk is the most useful thing. Here
why: when did you lose track of Lara Croft?
another warez that fell from ZX.SPECTRUM you need:
1. run unzip and unpack.
2. launch, I’ve already forgotten what it’s called, to get it out of SCL.
3. look, erase, copy the rest somewhere.
Drive B: take Lara and mail base, drive A: Melon
and MMD. Without a ramdisk you will be damned, these disks are unreadable,
reformat, paste here and there. Everything is slow. And with
With a ramdisk everything is done in just 10 minutes. That's what it's for
RAM disk. IT IS CONVENIENT TO WORK WITH. It's comfortable to hold
temporary files. Files transferred from program to program.
It MOVEs in a minute. And not for 10. There are no failures on it. Except
power failures.
But there is one problem. There are people like coders who write
programs that corrupt this ram disk. I even had a switch for
launching such programs. So that the top
turn off memory. BUT THEN THE RAM DISK IS NOT AVAILABLE!
This is the problem.
Why does this happen? Take, for example, ALASM. He needs
Well, a can, well, two or three. Store the sources. And because of this he is a ram-diskruins. Because he doesn't know that there is a RAM disk. And even if he knows, he doesn’t
knows how to use it. And all that would be needed is
create a file in this ram disk, find out which jar it went into
and use that jar. There is a plus: the state of the source is reliable
saved between Alasma runs. Because other programs
if they also work correctly with the ram disk, they will not spoil the file,
which was occupied by alasm.
Moreover, the same analysis can be temporarily interrupted, recorded
like @ file to ram disk, work with another program, and
return back to alasm. Why not multitasking? And no one
this memory-resident alasm will not spoil it. Even if there are two alasmas in parallel
run. Because they know -
This is my file, but this is someone else’s. And where is which bank?
More to come. Let's introduce the concept of "virtual disk" instead of
ram disk With custom driver. Of course for tr-dos.
What will we get? The user *without ROM firmware* has the ability
use different disk virtualization programs.
For some on a hard drive, for others on a RAM disk, for others on a CD-ROM...
Where is the driver located? In the same RAM disk. It is used
simply as a means of allocating memory.
It’s not at all difficult to write a program that would virtualize a TRD file
recorded on CD-ROM...
And this will make it possible to fully use the collections
software on Spectrum. Not like now:
1. CD-walk.2. Unzip
3. dehobeta or unscl
THREE DIFFERENT PROGRAMS. And it’s good if this is not a collection of warblers
Kalinina, where everything is exclusively in *.RAR.
Multitasking again. This does not mean PARALLEL
EXECUTIONS. Magic button. Firstly, to reach out to her, and secondly
it doesn't always work correctly. This might be good
(I mean NMI) for programs with protection. But bad for
the usual programs in which the stack gets corrupted. There is a solution.
PATCH ROM, function RST #38. So that, for example, pressing
SS+CS+Space (SS+BREAK) worked as SAFE MAGIC.
With user defined function. From the same RAM disk. Pressed SS+BREAK, from
ram disk launched a special
program. She can, for example, write to the same ram disk
compressed memory dump and call boot. From a ramdisk, of course.
THE USER LOADS THE MOST INTO THE RAM DISK ONCE
NECESSARY PROGRAMS AND THEN WORKS FAST!.
Adaptation of existing programs that spoil any
memory. Yes, this is a problem. Fortunately, such programs are relatively
not enough. And they either run in 128 mode. Or they can be patched to highlight
cans from ram disk.
But the most important thing I wanted to say is every time the conversation
it comes to all sorts of operating systems, they start inventing all sorts of crap. ZX-Windows, ZX-Linux,
windows, mice, kernels, task managers. What the hell
Is there any point in them if I want to write code in ASAM under all this?But the elementary thing is a ram disk. This is a ready-made manager
memory. And in general a very convenient thing. Without it, use the Spectrum at all
unrealistic. Only if there is no SMUC.
However, what is there to write about? The pussies-emulators are nit-picking.
Let's sit for a week in real life...
Hello Vovoi
Vov> Sorry if I offended you in some way, but it seemed to me that the changes
Vov> were more significant.
Vov> :o
What abides, daragoy, slobber! With grievances - to the author of the subject. What other ones were there?
I’ve already forgotten the changes myself, since I haven’t used them for a thousand years. Although there were
times... :)
Hello, CityAceE
Cit> Unfortunately, I have practically no free time - I can only
Cit> cut it out a little on the weekends. But I undertake to take part in
Cit> of this idea. I can write some procedures, work on
Cit> interface, submit some ideas, etc. The main thing would be over what and
Cit> with whom to work.
Right now we are working on TASiS 6.0, but it is only for Turbo2+ and higher.
Hello jdigreze
Explain why cp/m is bad? The fact that it can’t work on classic 128k?
Or something that was not originally for the spectrum?
In my opinion, this is a completely professional system, there is software for it, too
professional. Well, there’s a lot of commercial stuff, really... But you can also write under cp/m
ok! A normal API containing file operations and any other I/O.
You just need to add some of the current chips there (before
total, FS with catalogs). Again, you don’t need to invent anything - at least do it
like in MS-DOS.
And multitasking, java - this, IMHO, is clearly not for the Spectrum... Is there anyone here?
imagine the complexity of the java virtual machine?
Hello, maximk
max> Or something that was not originally for spectrum?
Each platform has its own axis... And more than one... Why would that be?
Why does x86 have at least “Vyn” and “Tux”?
At the moment, many OSes and OSecs have been written for the Spectrum, which is also
for a reason...
The problem lies in using a computer to solve a problem.
Let's take existing examples:
TRDOS, with some exceptions, is suitable for any job, but to a limited extent
using floppy disks...
ISDOS is good enough for typing large texts and transferring data from one
disk format to another, but it’s difficult to write toys for it...
I haven’t seen the others, so I won’t write about them.
The only thing I want to note is that virtually everyone, to a sufficient extent,
a quality product (eg: ZX-Format, CV, UFO, etc.), had its own shortcomings,
tailored to the needs of a specific product...
And sometimes you want to “stretch” your brains ;)
Hello jdigreze
jdi> Decision on the implementation of one or another rational proposal in the project
jdi> is accepted ONLY by the project coordinator (manager).
Are you ready to take on the role of project coordinator?
Hello, Maksagor
> Returning to you about axes in general - I think that the OS should be written
> immediately with an emphasis on HDD (at least).
Any normal system should be able to do without a hard drive. This is mine
conviction. But, of course, you need to make sure that the presence of a HDD provides
complete comfort, and not like SMUC works in TR-DOS mode.
Hello, CityAceE
At the moment I am the coordinator of the OS TASiS project. The results, I think, are everything
anyone who wanted it has already seen it - either on a real ATM or in an emulator. When determined
conditions, I could take on this project. HO! So far I think that the most
it is rational to support existing systems and, above all, iS-DOS (at all
not necessarily in its TASiS modification), which, in my personal deep opinion
I am convinced that it has not yet exhausted its potential. For example, according to vague rumors,
who occasionally reach me from Ilyasov, I can say that work is underway on
creating a graphical shell for iS-DOS in addition to the standard SHELL.
Another branch of the system - OS TASiS v6.xx will also be equipped with loadable
graphic libraries, on the basis of which a GUI shell will also be created,
and we won’t deprive the system of pseudo-multitasking either - in principle, it’s possible
implemented within the framework of the most common iS-DOS already now, but, of course, in
normal "text" mode of the SHELL shell.
What I need to improve in this system is the file system - it
in itself is good, but it’s painfully straining the maximum volume of the device in
16MB. We need to increase it, and increase it seriously! Therefore, as a coordinator
project TASiS I announce a competition for rational proposals for IMPROVEMENT (or
painless replacement) while maintaining compatibility with the main quantitysoftware, file system of this OS. Please write to people, NOT by hearsay
familiar with iS-DOS and its FS, and above all from the point of view of knowledge about its structure,
ideology, restarts. People who have started the system from a floppy a couple of times,
poked at the options and went through the subdirectories and dropped this one after
the floppy disk is collecting dust on the mezzanine - please do not disturb it. I'm not interested
flame about the advantages and disadvantages of iS-DOS, its necessity/uselessness, speed or
slowness, etc. I need specific improvement proposals. Preferably with
examples in code.
Returning to you about axes in general, I think that the OS should be written right away with
focusing on HDD (at least). I am deeply convinced that for Spectrums with a minimum,
the “classic” configuration 128+AY+Beta Disk doesn’t even need it - there’s only one
TR-DOSina is enough for the eyes. And the new OS will only irritate, because it will have to
intermediately load, and only then launch software from floppy disks from under it. Why,
When can I do this directly through TR-DOS?
Hello, CityAceE
Cit> Any normal system should be able to do without a hard drive.
Cit> This is my belief. But, of course, you need to make sure that the availability
Cit> HDD would provide complete comfort, and not the way SMUC works in
Cit> TR-DOS mode.
I agree, I SHOULD. But the question is that without a screw almost the whole meaning is lost, because
then compared to TR-DOS we lose:
a) In the speed of working with the flop - since we are working with the logical structure
(subdirectories, segments, etc.), rather than direct reading/writing of physical sectors.
b) In the speed of working with applications - first you need to boot the system, then
change the disk with the toy (since the system floppy disk is probably filled almost to capacity
start with applications and docks) and only then launch it. immediately arises
thought - maybe it can do without the “extra” link in the form of the OS? I don't think so
Let's say that other users will think...
c) In the possibilities of writing programs: since, unlike TR-DOS, it is different
the system will occupy part of the RAM and, accordingly, impose its own
own rules and restrictions on programs. The screw allows you to bypass this
firstly, by overlay loading of parts of the program (which can be
difficult if you change the system disk all the time), and secondly,organization of swapping of "virtual" RAM pages, as compensation for
occupation of several “real” pages by the system itself.
So without a screw, TR-DOS actually has advantages over others
systems and is inferior only to those that cannot be developed and are inflexible.
If the Spectrum initially had only a BIOS with a loader, then these moments
would not be fundamental, since there would be nothing to compare with. Exactly the same
the situation would have been if a whole culture had not already developed on the platform
programming under TR-DOS, with its traditions of working with disk, using
address space, handling any ports. Then it becomes difficult for yourself
force retraining, and impose certain restrictions (for example,
be formed into memory exclusively through firewood). This slows down the spread
new axes. But connecting the screw radically changes things - TR-DOS monopoly for us
magically instantly loses all its power. Therefore, it is necessary first of all
focus on the screw, otherwise on computers without a screw in most cases this
The OS will be launched only to demonstrate to the writers, “what, supposedly, is on Spec
there is such a thing"...
Hello, Maksagor
So, who has the desire and feels the strength to participate in the project on
creating a new operating system for the ZX Spectrum, welcome to the closed
section for developers. Who plans to actually do something in this
direction, and don’t just scratch your tongue, contact me for the password to this
section.
Hello, CityAceE
Cit> Are you ready to take on the role of project coordinator?
Me? No! There are many objective and not so good reasons for this ;)
But Maxagor is right! Why is IS-DOS bad? Because she was not supported, because
de facto "TR-DOS" when working with a floppy drive.
By the way, we are not the first to step on this rake...
Maksagor, there are ideas to overcome the “16MB” barrier, but so far this is only at the level
ideas, now I’m considering one of them for compatibility with the current version
IS-DOS. If I don’t find any conceptual contradictions, I’ll post...
Hello dwt
In connection with this topic. Related question.
What is the situation with the legality of ZX firmware?
The circuit solutions seem to be fine, the software is free....
Hello ASDT
ASD> In connection with this topic. Related question.
ASD> What is the situation with the legality of ZX firmware?
ASD> ....
ASD> Circuit solutions seem to be ok, software is free....
According to Amstrad (as the copyright holder of the ZX Spectrum, ZX Spectrum+, ZX
Spectrum+ 128, ZX Spectrum+ 2(A/B) and ZX Spectrum+ 3) firmware available for
use in emulators (somewhere here I gave a link to this statement, or
go to worldofspectrum.org, there is a link on the title page). About
Amstrad has his own opinion about the Soviet clones :), but I still don’t understand
let's say how they feel about the fact that some Englishman (or maybe
Spaniard, I don’t remember) picked and shoveled the ROM from +3. Actually, if you do it like this
in Sprinter (this is not a clone, but a ZX Spectrum emulator) - then there should be no problems,
because it is not declared how it can be emulated, but how it cannot be emulated (software or hardware).
Hello icebear
Well, if we look at it from the angle of “iron” emulation, then all Spectrum-compatible
clones fall under this category... After all, a clone does not use the original
circuit design neither Sinclair Research nor Amstrad, but only “emulates” on
“loose”, where it is more accurate and where it is less, the physical and logical model
original. However, the attitude of the brand owner is in any case negative...
However, getting off topic...
Hello dwt
On the hardware side, everything seems to be ok.
Schemes and documentation for the original have been published,
and clone implementations have their own circuitry.
Regarding “going away from the topic” - this is the topic!
Initial data:
1. There is a fleet of hardware that is relatively compatible with ZX.
2. The OS must support the existing hardware (from 48k to 1M).
3. However, there are questions about ROM 48/128, similarly -
how to treat TR-DOS...
That is you need to properly position the OS according to these
questions.
Hello jdigreze
jdi> Well, if we look at it from the angle of “iron” emulation, then that’s it
jdi> Spectrum-compatible clones fall under this category... After all, a clone
jdi> does not use original circuitry from either Sinclair Research or
jdi> Amstrad, but only “emulates” in “loose” form, where it is more precise, and where
jdi> less, the physical and logical model of the original. However, the attitude
jdi> the copyright holder of the brand is in any case negative...
jdi>
jdi> However, we are moving away from the topic...
I don't know if Amstrad has the rights to the circuit design. However, any Pentagon
without ROM - an empty piece of hardware. And if you are using ROM not for emulation, what is it?
then? Another thing is with Quorum and Profi, for example - they work with SR/M, although I don’t
I know how much they can do without Spectrum ROM.
Hello caro
car> In Quorum, as in Pro, when CP/M is running, the ROM is completely disabled.
So they won’t be able to stress :)
Hello icebear
ice> Another thing is with Quorum and Pro for example - they work with SR/M, although I
ice> I don’t know how much they can do without Spectrum ROM.
In Quorum, as in Pro, when CP/M is running, the ROM is completely disabled.
Hello jdigreze
jdi> 1. Memory manager level
jdi> 2. I/O level
jdi> 3. SHELL level
jdi>
2. and 3. are already in ROM, why reinvent the wheel? Look here
http://www.zxplus3e.plus.com/ - GREAT IMHO that such an implementation would be
“more dear” than all sorts of booties and shells from Unix.
Hello jdigreze
jdi> icebear, I apologize for the flame... Apparently I got excited
jdi> holiday... ;)
Come on, what kind of flame? Everyone has their own opinion. That was my IMHO. On the other hand
on the other hand, it’s interesting to know why people come up with all sorts of “bells and whistles” for 8
bit computer, which then in most cases will lie dead
cargo?
Hello jdigreze
jdi> Great solution! But...
jdi> ...what should an ordinary user do, let’s say me, with mine
jdi> kay1024+nemo_ide? Flash ROM? With adjustment for Nemo-IDE ports?
jdi> And if I want to make it work with FAT16(32)/NTFS/HPFS/ReiserFS...
jdi> not as with the main file system, but as with an additional one
jdi> say, mounted?
Why do you need all this? Do you really need it on the Spectrum? Do you want to hook up?
hard drive with NTFS to Spectrum? What is the use of FATs (and everything else) on
Spectrum? For one-time use of these systems (read copying from
media with such FS) one program is enough. For a possible question about
placing everything in ROM - do you often copy files from NTFS to TR-DOS?
jdi> But there are many real problems, as well as possible solutions...
jdi>
jdi> Maybe let everyone express their wishes here?
jdi> I'll start with myself:
jdi> - I want the disks to be like in Tux, so without letters
jdi> - I want it to be possible to switch the “shell” to another, or remove it
jdi> at alljdi> - I want to boot from a screw or CD
jdi> - I want to be able to work with a screen other than 256x192
jdi> - I want, I want, I want!
jdi> - and the last “I want”: I want a new program written in any language
jdi> clone/original, you could run it on your own without having to go into every time
jdi> mouse/memory/sound settings...
jdi>
jdi> How much I want... And how little I have... And apparently I’m not the only one like this
jdi> "unique"... Otherwise people wouldn't write here...
There are more other computers for all this. Why pull on the Spectrum
face of other platforms? Why deprive him of look & feel & usage ZX
Spectrum?
Hello ASDT
1. Memory manager level
2. I/O level
3. SHELL level
The third level must be able to “take off” in order to make room for creativity
"boot" writers ;)
The second level is the level of working with blocks and files, as well as with channels and
streams...
The first level, in fact, is the core, which should work minimally
starting with "48K + tape recorder"
Hello icebear
icebear, I apologize for the flame... Apparently I got excited about the holiday... ;)
Total:
We settled on the memory manager...
Does anyone have any thoughts on this? For me personally, it’s still a “mess”
head...
Hello icebear
Great solution! But...
Flash ROM? With adjustment for Nemo-IDE ports?
And if I want to make it work with FAT16(32)/NTFS/HPFS/ReiserFS...
not as with the main file system, but as with an additional one, for example,
mounted?
But there are many real problems, as well as possible solutions...
Maybe let everyone here express their wishes?
I'll start with myself:
- I want the disks to be like in a Tux, without letters
- I want it to be possible to switch the shell to another one, or remove it completely
- I want to boot from a screw or CD
- I want to be able to work with a screen other than 256x192
- I want, I want, I want!
- and the last “I want”: I want a new program written in any
clone/original, you could run it on your own without having to go through the settings every time
mouse/memory/sound...
How much I want... And how little I have... And apparently I’m not the only one like this
“unique”... Otherwise people wouldn’t write here...
Hello dwt
After some thought...
This is the model, for starters.
ZX+boot and PC as "remote"
disk (possibly terminal).
Communication - som-port.
If this is interesting...
Hello jdigreze
jdi> It may not be in demand, and will lie as a dead weight, or maybe
jdi> be and will not be implemented at all... Here, I still want to do
jdi> emphasis on the modularity of the system so that possible future ones
jdi> developers did not encounter the notorious “16MB problems”... But, IMHO...
Modularity is certainly good. Only my remark should be regarded as “not
It’s worth burdening the system with unnecessary components.”
Hello icebear
Maybe it won’t be in demand, and will lie like a dead weight, or maybe not
will be implemented at all... Here, I still want to focus on
modularity of the system so that possible future developers do not have problems
the notorious “16MB problems”... But, IMHO...
A typical, one might say historical, fact: due to the heterogeneity of iron,
In the West, the de facto standard remained the tape recorder, but in our country, due to the specifics
initial market, BetaDisk became the standard, and it took about 10 years to break through
We can’t, because there is no flexibility.
This “gap” is partially closed by is-dos, but initially the “rigid” organization
kernels, today creates some problems in the use of modern hardware...
Hello ASDT
icebear, well, it seems like we got the average opinion ;)
ASDT, and we are just thinking about the boot... As if with minimal resources
make maximum scalability ;)
What minimum configuration will we start from?
I’m still inclined to think that 48K with a tape recorder.
The basic configuration, with a minimum set of "drivers", should be no problem
“fit” into 48K and not occupy more than 16K in them.
Why a tape recorder? Because it is actually ready “to eat” hardtaper by
DVS, plus “Westerners” still don’t shy away from tapes...
IMHO, it's debatable. Again, most "drivers" support the base model
“48K”, “hardwired” into ROM, i.e. no confusion about “how does it work”
should be (but it will be... ;))
And also, let's decide what should be included in the basic configuration
AXIS...
Let's say:
- support for 32x24 text screen (let's say in ascii-866 encoding, I
fundamentally against koi8r)
- support for standard keyboard
- support for files on tape
All this should be enough to write at least a text
editor and any development environment.
Hello dwt
The tape recorder is a utopia...
Well, if only... There was one option in the iron theme.
Direct access to files is required.
And why is somport bad? Send files and cables
and terminal...
I have a ROM emulator attached to the 48, there is
where to upload the bootloader? Add another catfish,
more software...
Hello dwt
“that the device is quite unique” is bad, the thing is extremely useful.
"This is HARDTAPER" - but it must be done...
And the catfish itself is needed, and can be used for OS purposes...
“It’s not there yet, but it will be soon” - what about standardization?
"start from the minimum" - for a real tape recorder no OS is required, anyway...
Hello ASDT
ASD> A tape recorder is a utopia...
It may be a utopia, but the “Iron” section confirms the opposite... 48K +
The tape recorder is still a very common variation...
ASD> Well, if only... There was one option in the iron theme.
This is HARDTAPER ;)
ASD> Direct access to files is required.
ASD> And why is the com port bad? Send files and cables
ASD> and terminal ...
ASD> I have a ROM emulator attached to the 48, yes
ASD> where to upload the bootloader. Add some more catfish
ASD> more software ...
Direct access to files is needed, and it will be.
About the som-port I can only say that the device is quite unique, and far from
there is everyone... I don't have it yet, but I will soon ;)
And yet you need to start with the minimum, and then add “uniques”...
Hello jdigreze
jdi> I am still inclined to think that 48K with a tape recorder.
jdi> The basic configuration, with a minimum set of "drivers", should be without
jdi> problems “fitting” into 48K and not occupying more than 16K.
jdi> ........
jdi> And also, let's decide what should be included in the basic
jdi> OS configuration...
jdi> Let's say:
jdi> - support for 32x24 text screen (acceptable in ascii-866 encoding,
jdi> I am fundamentally against koi8r)
jdi> - support for standard keyboard
jdi> - support for files on tape
jdi>
jdi> All this should be enough to write at least
jdi> text editor and any development environment.
And then ShaOS appears [http://www.nedopc.org/nedopc/shaos/shaos_r.shtml]...
:v2_smoke:
Hello ASDT
Com-port, at the moment, although not a standard, is not a problem either. B
At present, de facto, the “Kondratiev scheme” is used, which
supported by controller Kamil Karimov (caro). That's where we'll go
navigate. The controller can and should be issued by mail.
A real tape recorder does not require a driver, and OS support would be very
by the way. For example, when working with a virtual disk, for example, as done in ArtStudio
(for 128).
Hello,GriV
Gri> is this a set of subtitles? or is it the OS?
Gri> In any case, the sources are in the studio.
ShaOS? This is quite an axis with a set of subroutines and a command line, as well as
some unified file formats and data transfer protocols via
Several specially designed hardware interfaces.
Hello Shaos
Okay, I still don’t understand - where is the finished code? I just followed the link there
function descriptors, or did I miss something? And yet, we’re still waiting, from the sources
(-: :v2_tong2:
Hello,GriV
Gri> Okay, I still don’t understand - where is the finished code? I followed the link
Gri> are there only function descriptors, or did I miss something? And yet, still
Gri> we are waiting, from the sources (-: :v2_tong2:
And the finished code is on my screw and several copies on disks ;)
ShaOS sources consist of 2600 kernel ASMO lines and several
hundreds - utilities and tests (about 3 thousand lines in total). To post them, you need
convert from mnemonics of my cross assembler RASM (Robo Assembler) to ZMAC
(so that spektrumists can read without stuttering on every line) well, not bad
I would like to ensure the compilability and launchability of the converted sources :)
Hello Shaos
Sha> And the finished code is on my screw and several copies on disks ;)
Sha> ShaOS sources are 2600 ASMO lines of the kernel and
Sha> several hundred - utilities and tests (about 3 thousand lines in total). To
Sha> lay them out - we need to convert them from the mnemonics of my cross-assembler
Sha> RASM (Robo Assembler) in ZMAC (so that spektrumists can not read
Sha> stuttering on every line) well, it would be nice to collect and
Sha> ensure the launchability of the converted sources :)
Automatically converted from RASM to ZMAC what was left from 1998 and posted
read:
http://www.nedopc.org/nedopc/shaos/shaos_zx0_.shtml
I also compiled and made a small BASIC loader for this case, putting everything in
TAP [http://www.nedopc.org/nedopc/shaos/s004_v0_.zip] (thanks2glukalka) -
The command line automatically starts, which makes it possible to do something
edit, but does not launch anything (except that by command # you can launch
subroutine test_em).
The system was designed for electronic quasi-disks - to bring it to mind, I
you either need to find a way to attach quasi-discs to your ZX Spectrum +, or
adapt the system to TR-DOS and debug it in the emulator. Or maybe a virtual one?make a disk in memory - but 48K memory is not rubber - all supposed utilities
the systems won't fit.
Hello jdigreze
> 1. I am still inclined to believe that 48K with a tape recorder
> 2. Still, you need to start with the minimum, and then add more
> "uniqueness"...
> 3. All this should be enough to write at least
> text editor and any development environment.
>
1. The tape recorder is already supported in ROM, and what is missing can be organized in
applications, this is an extra burden in the system.
2. It won’t be possible to add more weight to 16k, unless you remove one and install another
:(
3. Yes, the development environment is one of the ways (IMHO) to
get coders interested in just looking at the system.
It is not easy to integrate a language compiler into the environment and organize it correctly
interaction of applications and their fragments with RAM, that is the task. Demos
many turn out excellent, but draw and save a 256*192 picture in this
demo, not to mention writing at least a simple calculator and
immediately run in the environment.
Hello Vovoi
Vov> 1. The tape recorder is already supported in ROM, and what is missing can be organized
Vov> in applications, in the system this is an extra burden.
The tape recorder in ROM is supported in an outdated style. For example, we download for five minutes
48K block of code and we get an error at the end. It's a shame? It's a shame. Moreover, it's stupid.
You need to read from a tape recorder in pieces of 1 or 2 kilos (and write accordingly
also). In case of incorrect reading of the next block, do not upset
user with a message - like everything is gone, press reset and let’s do it all over again, and
politely ask to rewind a little and try from this last one
unread block. In general, you can build a device that can be poked
Spectrum via standard MIC and EAR holes through which the OS itself will speak
the block number (standard bit parcel with piloton) that it needs, and
the device will buzz in response with the desired piece (again in standard
Specov format) of a certain not very large size.
Hello Shaos
Sha> The tape recorder in ROM is supported in an outdated style. <...> Must be read from
Sha> tape recorder in pieces of 1 or 2 kilos (well, write accordingly
Sha> also). <...> And in general, you can build a device that can be poked
Sha> Spectrum via standard MIC and EAR holes through which the OS itself will be
Sha> say block number.
Well, yes, I heard that there are drivers for 4x speed, but the tape must be of high quality
probably super :)
Regarding the pieces, I also remember that Art Studio once finished it off, it took place in my
three pieces: Katinka-6912, main block ~ 41k, hacking block ~ 2k, which then
and started.
The refrigerator will turn on and go crazy, but I wanted to draw :) I found it by chance
control unit from the keyboard (I pushed the RST8 output somewhere and the address was found),
Immediately, on the way, I changed QAOPM to 67890, like “Sinclair-2” :) dumped it onto a cassette
48k (or so) cut into 2k and made a “loader” with a strip (I saw it once
on the PC in DosNavy and copied it, but we didn’t have such a thing, except digital
gave a countdown), so, well, the raw materials are still lying around on GENS4 and
the program means it shows a progress-bar from left to right and if
tape_loading_error then it will neatly highlight the message “Searching for BLOCK 4 (from 6)Please Rewind". This was already easier, and then GENS4 itself also loaded, with
stripe. I have a Legend-404 tape recorder, it has double the speed, but the driver is too slow
I didn’t want to load it, but if the speed switch was carefully moved to
position "middle" (well, you have to search for a long time, experimentally :))), then the speed will be
just between 4.75 and 2.23 (I wrote games at 2.23, I don’t remember the exact numbers,
Vertak's passport must be read) and the entire GENS in less than half a minute, all 10k
rushed into RAM. But it was simply amazing to work, and at the same time the joystick
connected and he began to work for the benefit of the Art Studio and they lived happily ever after
until the joystick ended up in some toy.
:v2_happy:
Actually, what was meant was the possibility of converting cassette programs and data in the OS
to disks, but it doesn’t work in principle. I have one of these lying around at work
cool device, a tape recorder with cables and it says Commodore
unit. This thing can fit a whole cassette and can probably work from signals from
computer, in theory even the OS will work on this computer, maybe.
:)
Hello Vovoi
I apologize, I deceived you a little. The name of the thing is 1530 DATASSETTE UNIT Model
C2N, further noted - Suitable For Use With Commodore Computers. From
The device comes with a cable with a flat connector, where four contacts are located
together, then on the same block a little further away there are two and separately hanging earth. For
Saving and load are enough one at a time (the land is common), even if they had two lands,
then there are two more left, perhaps to power the device itself. No passports
Obviously not, but according to the instructions to the commodore you can find out how much he gives "for
side" (for example, Specky gave +5, and later developments also +12, for
floppick). We also see that there are buttons on the “object”, which means
There is no automatic cassette control (otherwise there would be only one “eject” button
cassettes"). The tape recorder is similar to a portable player (with a recording head),
there is a counter.
:)
For the OS you will have to press a lot of buttons to load applications, but in the mid-90s
I would dream of such a contraption, instead of a standard, huge turntable.
:rolleyes:
File: Datassette.JPG http://zx.pk.ru/attachment.php?attachmentid=3018
Hello Vovoi
Vov> We also see that there are buttons on the “object”, which means
Vov> there is no automatic cassette control (otherwise there would be one button
Vov> "cassette ejection"). The tape recorder is similar to a portable player (s
Vov> recording head), there is a counter.
It’s not a fact that it was without auto control - I had the same one on an Atari 65x with
auto-control and at the same time there were fur. buttons. After all, you can just eat
disable =). And I think that Atari, as a competitor to the commodore, stole this idea
precisely at the Commoduri's...
Hello jdigreze
jdi> And also, let's decide what should be included in the basic
jdi> OS configuration...
jdi> Let's say:
jdi> - support for 32x24 text screen (acceptable in ascii-866 encoding,
jdi> I am fundamentally against koi8r)
jdi> - support for standard keyboard
jdi> - support for files on tape
Come on, text screen? Why not screen support via graphics lib?
Yes, the speed is lower, but it can be converted to other screens without rewriting
kernels, what if someone finally figures out not only the spec for fpg, but also the graph.
do axel. then it’s generally a win in everything.
Tape.... No, that's not it at all. Of course we can’t rule it out, but also about tyrdos
Don't forget to do at least 128k support, or even better to do liblu so
it was possible to work for each specific clone. That's all
, what is less - why is Vasik not satisfied?
Hello fk0
fk0> "The main problem when developing a new hard drive is in the brain
fk0> user" (C) mute. Very accurately said. It’s the same here. When
fk0> connecting an 86x24 screen to CP/M
fk0> there were no problems and all the software started working right away. Feel
fk0> difference.
fk0> There is generally a level of drivers, modularity and other horrors -
fk0> scanty.
CP/M is a STANDARD that is already about 30 years old. And it is accepted as a standard there
use a text screen, the parameters of which are determined through
OS tools. I hope I'm right here? ;)
fk0> Aftar is burning. If only because it works both from the ram disk and from
fk0> HDD.
Can you tell me how to access the FAT partition of my "screw"? Just don't
say that “write support and you will be happy!”... I know what I need to write,
and there is little free time...
fk0> Because demos and toys need to be written under LOAD "" and not TR-DOS.
That's what I'm talking about...
fk0> The standard Spectrum is Scorpio and Pentagon. And not discussed
fk0> even.
Uh-huh. I won't. Because I have KAY... :v2_tong:fk0> I agree to add standard and ATM-turbo to the list if he learns
fk0> switch memory like the Pentagon.
Is this important for the OS?
I had a Pentagon... But this is a topic for another conversation.
fk0> Technology Research Ltd - Soviet standard?
TRL is not a standard! The Soviet "standard" was the use of the TRL product
everywhere, anywhere... But that was a different time, and at that moment with
there were few alternatives. So to speak, “reaping the benefits” :v2_cheer:
Hello jdigreze
jdi> The same goes for everything else... Typical example: iS-DOS
jdi> is designed to work only with BetaDisk! Try connecting
jdi> something like Disciple - you'll probably have to rewrite about 70%
jdi> code, including the principles of loading the OS itself and so on... And such
jdi> there are many examples... Including not only OSes, but also software (any),
jdi> adapted for TR-DOS, reverse deadaptation is "fraught with side effects", and
jdi> I want to keep Dizzy and run it from a screw or even SD/MMC... ;)
jdi>
Well, you're a little wrong!
iSDOS later versions work not with BetaDisk, but with block devices!
Which may be segmented or not.
That is, having written a block device driver, you can, in principle, use
any medium.
Hello jdigreze
jdi> Roman, I give up! Gave a blunder... ;)
jdi> However, you must agree that loading from a “non-standard” procedure is quite
jdi> non-trivial... Moreover, for each, for example, HDD interface, its own, and
jdi> is quite unique.
jdi> I have one interesting idea about bringing all HDDs under one
jdi> "comb", but I suspect that it "eats up" several resources
jdi> more than the traditional way... In general, I think... :)
Well, on some OSes, adjustment to the machine is done by the so-called “compilation”
kernels", i.e. essentially assembling only the necessary drivers. For spec this is optimal
option.
To be honest, I’m not a very big expert on iSDOS, but it seems that preparations are also underway there
kernels to download.
Hello jdigreze
jdi> Yes, in iS-DOS the kernel is assembled from modules, and then, if any
jdi> desire, then this “assembly” can be made as a single whole
jdi> of the bootable kernel. Which greatly increases the download speed
jdi> rather than through loading following the configuration file. But it is fraught with glitches
jdi> when running a unique build on a clone with excellent hardware
jdi> configuration.
For this purpose there is a distribution kit, in which the most minimal
The configuration is available for download. And then the work of reading begins
manual and "user's hands". Those. I want to say that glitches for the most part
are a consequence of user laziness.
BUT making a kernel that includes everything and everyone - it seems to me that this is extremely wrong, given
our resources.
Hello fk0
fk0> I agree to add standard and ATM-turbo to the list if he learns
fk0> switch memory like the Pentagon.
Already knows how ;) : http://www.nedopc.com/ATMZAK/atm710re.htm
Improvement called "Adjustment that eliminates taking into account the A9 line when sampling
port #7FFD".
All the new cars I have assembled have this modification.
Hello, acidrain
The main emphasis is on modularity! ;)
And the text screen is 32x24 - the support module turns out to be very simple, but if
there is something cooler, then take the flag into your hands - write support, or, at least,
specification for modes. Otherwise it turns out to adapt iS-DOS to 80x25
the guys had to rewrite the whole WIND level (and most likely not only
it), instead of just writing and connecting the driver...
The same goes for everything else... A typical example: iS-DOS is designed for
work only with BetaDisk! Try connecting something like Disciple -
you will probably have to rewrite about 70% of the code, including the loading principles
The OS itself and so on... And there are many such examples... Including not only the OS, but
and software (any) adapted for TR-DOS, reverse deadaptation is fraught
sideways", and I want to keep Dizzy and run it from a screw or even SD/MMC... ;)
Standard Spectrum is 48K + m/f! Why? Because there are clones
using #7FFD in a somewhat non-standard scheme, which “knocks out” the 128K model
from the standard... I’m generally silent about BetaDisk... This Russian-Soviet
"standard" ruined a lot of interesting ideas, besides being forced to be unchanged
for compatibility, and what kind of twists those who try have to go to
modernize it without losing compatibility... It's a shame! :(
Well... Something got me into a flame... In an attempt to explain the reasons for my"Die-hard" ;)
Hello, CHRV
Roman, I give up! Gave a blunder... ;)
However, you must agree that downloading from a “non-standard” procedure is quite
non-trivial... Moreover, each, for example, has its own HDD interface, and
quite unique.
I have one interesting idea about bringing all HDDs under one
"comb", but I suspect that it "eats" a little more resources than
traditional way... In general I think... :)
Hello, CHRV
CHR> Well, on some OSes the adjustment to the machine is so-called
CHR> "kernel compilation", i.e. essentially assembling only the necessary drivers. For
CHR> specka is the best option.
CHR> To be honest, I’m not a very big expert on iSDOS, but it seems to work there too
CHR> preparing the kernel for loading.
Yes, in iS-DOS the kernel is assembled from modules, and then, if there is a desire, then
this "assembly" can be made as a single whole, bootable kernel. What
quite significantly increases the download speed than through downloading following
configuration file. But it is fraught with glitches when running a unique assembly on a clone with
excellent hardware configuration.
Hello jdigreze
jdi> The main emphasis is on modularity! ;)
jdi> And the text screen is 32x24 - the support module turns out well
jdi> simple, and if there is something cooler, then take the flag - write support,
jdi> or, at least, a specification for the modes. Otherwise it turns out that
jdi> the guys had to rewrite the whole thing to adapt iS-DOS to 80x25
jdi> WIND level (and most likely not only it), instead of
jdi> just write and connect the driver...
jdi>
"The main problem when developing a new hard drive is in the user's brain" (C)
nemo. Very accurately said. It’s the same here. When connecting an 86x24 screen to CP/M
There were no problems and all the software worked immediately. Feel the difference.
In general, the level of drivers, modularity and other horrors there is miserable.
> The same goes for everything else... Typical example: iS-DOS
> is designed to work only with BetaDisk! Try connecting
> something like Disciple - you'll probably have to rewrite about 70%
> code, including the principles of loading the OS itself and
> After all, it's burning. If only because it works from both the RAM disk and the HDD.
> And there are many such examples... Including not only OSes, but also software
> (any), adapted for TR-DOS, reverse deadaptation is "fraught
> sideways", and I want to hold Dizzy and run it from the "screw" or even
> SD/MMC... ;)
>
Because demos and toys need to be written under LOAD "" and not TR-DOS.
> Standard Spectrum is 48K + m/f! Why? Because there are clones,
The standard Spectrum is Scorpio and Pentagon. And it's not even discussed.
I agree to add standard and atm-turbo to the list if he learns
switch memory like a Pentagon.
> I’m generally silent about BetaDisk... This Russian-Soviet “standard”
> ruined a bunch
>
Technology Research Ltd - Soviet standard?
Hello, CHRV
CHR> Offtopic!
:v2_huh::v2_conf3: Apparently you didn’t understand the meaning of my phrase?
CHR> Actually, I remembered a good phrase:
CHR> "Don't tell me what to do, and I won't tell you where to go" :D .
Well, this looks like you :v2_confu:
Hello fk0
fk0> The standard Spectrum is Scorpio and Pentagon. And not discussed
fk0> even.
That’s why you’re marking time, unlike the bourgeoisie (just don’t write right away
boiling water, keywords Spectrum SE, Spectrum +3E, etc. or also YABUS (or
what is it called there?)). Find a way to “bring together” the Pentagon and Scorpio -
If you're lucky, if you don't find it, you'll continue to write at 128K and BDI.
Hello icebear
ice> That's why you're marking time, unlike the bourgeoisie (just not
ice> write immediately with boiling water, keywords Spectrum SE, Spectrum +3E, etc.
ice> or also YABUS (or whatever it's called?)). Find a way to "reduce
ice> together" Pentagon and Scorpio - you will be happy, if you don’t find it - you will be
ice> write further under 128K and BDI.
Offtopic!
I actually remember a good phrase:
“Don’t tell me what to do, and I won’t tell you where to go” :D .
Hello, CHRV
CHRV
Why the question [censored] did you delete my message? Felt the power
or how?
Hello icebear
ice> CHRV
ice> Why [censored] did you delete my message? Power
ice> did you feel it or how?
I deleted mine too! And the reason is "offtopic".
If you don’t like it, complain about me; if I’m wrong, I’ll be punished.
Regarding power, no one is stopping you from becoming a moderator ;)
Hello jdigreze
jdi> Can you tell me how to access the FAT partition of my "screw"? Only
jdi> there is no need to say that “write support and you will be happy!”... I know
jdi> what needs to be written, but there is little free time...
An offline comrade, specifically ZET-9, asked to answer this question. In general,
use DNA OS, it can do it. Currently works with FAT16 and CDFS.
Work is underway in the direction of FAT32, but it’s not working quickly :( At the same time
DVDFS is being studied. If anyone needs it, the distribution contains the source code, as well as
a description that is highly recommended to read.
An outdated version can be found in the appendix to the ZX Info Guide. Fresh is sought from
author: zet9_zx(et)mail.ru. If necessary, you can post it on the forum.