Scorp and Kai

Here in our network, SPBZXNET, a
letter has appeared, in my opinion very interesting:

==============================================================
>> area | HARDWARE.ZX
>> from | Dmitry Lomov
>> to | Kirill Chubenko & ALL
>> subj | Scorp and Kai
>> date | 01-02-98
>> time | 01:40

>> Hi, Kirill & ALL!

>> KC> Dim, anything that will improve the performance of the spec at least
>> KC> to some extent should be implemented
>> KC> unless of course it costs more
>> KC> PC... :)

>> Hehe...

>> Then I have a proposal for ALL:
>> But share
>> opinions/considerations/suggestions/wishes etc
>> on the topic:

>> Is there a need for a device called “hellish HARD disk controller”
>> with the following capabilities:
>> (note: read to the very end.
>> Think about it. Read more. Then judge.)

>> mandatory-

>> 1) full VG emulation (absolutely everything goes, emulated
>> flop drive speed, index holes, write protection,
>> inter-sector delays, i.e. ABSOLUTELY everything)
>> 2) point 1), but without emulation of speeds/delays;
>> pumping speed about 1.75 MB/s asby #3D13,
>> and through the fast-loader, everything goes except for the bootloaders
>> time-bound (EYEACHE_2, POWER_UP etc)
>> 3) partitioning into tr-dos disks - the same as
>> in SMUC; own shadow program for MAGIC without damaging the stack or
>> anything (all this for any machine,
>> and not just for Scorpio)
>> 4) work in IS-DOS - similar to SMUC; 16 meters,
>> HARD-boot, 1.75 mb/sek
>> 5) internal cache of 1 MB.
>> 6) download speed from cache - 1.75 MB/s
>> (always, excluding mode 1; for is-dos and tr-dos,
>> regardless of the type of bootloader (#3d13, #3d2f, whatever)
>> 7) the speed of suction from the screw is equal to the transfer of the screw (but not more than 1.75
mb/sec),
>> 8) the device is stuck into the slot; no soldering iron needed.
>> 9) saving all setups somewhere.

>> if you get perverted, you can also -

>> 1) creating images of floppy disks with any format, emulation
>> of the process of inserting a nightmare floppy disk into the drive.
>> 2) cache up to 4 MB, rubber RAM disk (it’s not clear how,
>> but oh well), transparent to the user caching
>> all four pseudo-trdos disks in the frame (after selecting a disk
>> it is transferred to the cache within 1 second)
>> analogue of SMARTDRV on a PC(record after) etc.
>> (EVERYTHING works both via #3d13 and via fast;
>> everything is absolutely transparent to the user)
>> 3) using part of the internal cache for the RAM disk
>> in IS-DOS (it’s not clear why, it won’t be faster anyway)
>> 4) using the internal DMA controller for your needs
>> (switching between the internal cache and the 48th memory
>> at a speed of 1.75 MB/sec)
>> 5) adapt the STS to the built-in shadow server (MAGIC can be pressed
>> always, auto-detection of the contents of #7ffd and #1ffd, the stack does not
>> spoil... complete crap!)
>> 6) Reset supervisor (press RESET, machine actions -
>> in accordance with what you specified plan)

>> if you get too perverted, you can make -

>> 1) images of absolutely any diskettes, including scratched ones,
>> with two index holes, with any garbage
>> between sectors (very difficult, everything depends on the software;
>> but - install the CV, run the installer,
>> throw away the floppy disks)
>> (2VM - don’t kick it; I thought for a very long time - this
>> is possible)
>> 2) working with MSDOS, TRDOS, ISDOS, AMIGAOS partitions :),
>> the ability to project any partition onto any
>> system, for example: youyou work in IS-DOS with a device
>> with a volume of 16 meters, which is actually a directory
>> in the MS-DOS partition, and when you insert a screw into the PC you work
>> with the same data, but in MS-DOS; or working in tr-dos
>> with the is-dos subdirectory.
>> (I may be talking heresy, but, judge, who is stopping me
>> from broadcasting systems at the physical level -
>> roughly speaking, with this work I create a pseudo-trdos or
>> pseudo-isdos disk and work with it, the reverse broadcast occurs
>> in the background, the user does not see it)
>> 3) working with XTR in the background, roughly speaking, modem
>> screw controller using the bus and Spectrum modem.
>> Spectrum processor regularly gives the bus for analysis
>> lines to the screw controller processor. Naturally,
>> there will be a slowdown, but you can play dynamic games via a
>> modem - the Spectrum processor is relieved of the need to count
>> time, this function is taken over by the controller processor.
>> 4) pseudo-multitasking (tasks sit in the screw cache, you can
>> quickly switch from one task to another using
>> MAGIC, simultaneous operation of tasks is impossible; although, if you think about it -
>> it’s possible, only it will beblink the screen :) and swap 128 memory
>> will take a lot of time)
>> 5) the ability to control the task.
>> once, say, a minute, the screw controller
>> interrupts the task and analyzes its work according to some
>> signs (the BBC is frozen - restart for the hell of it :))

>> Well, how? If you think carefully, the volume
>> of possibilities will double... Let's say, with minimal
>> additions (physical), you can also turn
>> into a controller for a PC keyboard (not slow!) and a mouse, as well as
>> into a graphics accelerator (by modifying DMA)...

>> I give you a certificate - I am of sound mind and sober memory...

>> Now the comments:

>> 1.75 MB/sec - transfer speed directly
>> sectors into memory. It takes time to move to the next sector/track
>> error analysis (don’t forget, VG is emulated)
>>. For a fast loader, it’s estimated -
>> the speed will be 1.73 MB/sec :)

>> I repeat, 1.75 MB/sec in _all_ modes
>> for _all_ loaders for all systems,
>> excluding the mode of forced _special_ brake set by the user
>> for the operation of all kinds of music loaders.

>> About perverse points - everything depends
>> on the software, and only on the software(meaning the internal
>> software controller).

>> Don’t ask me _how_ all this can work;
>> I won’t tell. Don’t say that this is impossible;
>> I say that all this is possible, both at the
>> level of hardware and at the software level. The whole question is
>> how long it will take. At the level of circuit design
>> solutions, everything is thought out (in theory), there are no barriers
>> no, everything depends on the cost of my time.
>> I’ll say right away - I really want all this, but
>> I need to know if anyone else needs it.
>> And will this someone be able to adequately pay
>> my sorrowful work...

>> Note. I still
>> will do it for myself, only it turns out to be a non-commercial
>> project, therefore - creativity, which means
>> I have to open the window wider and wait for the arrival of the
>> MUSE, which will inspire me and give me strength
>> and patience for the high art of circuit design
>> and coding. For me it will be a pleasure,
>> and it is desirable to prolong the pleasure; I’m in a hurry
>> there’s nowhere to go, I know that the Spectrum is forever for me
>> a machine for creativity, which is what I do,
>> and all my aspirations are based
>> hedonistic orientation.
>> If anyoneMy sublime
>> words are incomprehensible, don’t pay attention. Just do
>> mental skip.

>> But, if I know that my work, in addition to
>> a feeling of deep satisfaction from a job well
>> done and a feeling of well-spent
>> time, together with the pleasure I experienced,
>> will also bring me some money, it will turn out
>> that I am combining business with pleasure; in this case
>> the work will go much faster.

>> ---SO---, I want to know your opinion on
>> everything I said. Say what
>> you think (this is not a hint..), I will be glad
>> any questions/wishes/subscriptions/offers/
>> crap pouring down on me in buckets :)
>> Just, please, less emotions... Substantively,
>> only substantively...
>> I want to know how much money people
>> are willing to pay for such a device.
>> I I assume its cost (manufacturer price)
>> is no less than 150 tons. Think.
>> Offer your prices, we'll see...

>> And here's another topic for discussion.
>> The fact is that... Sprinter is very cool,
>> it's just best (without irony), it's amazing
>> flexibility and good capabilities for an 8-bit
>> computer, but...no one will buy it (it seems to me so
>> I won’t talk about the reasons); GMX - a lotion for texts +
>> large memory, all for very good money...
>> see the commentary on the Sprinter!
>> So, comrade DANGEROUS has an idea (so far only an
>> idea) to make such a thing as a kind of axel with a concept
>>: on a Spekov screen, but it _FLYS_!
>> without exaggeration... Z180 (21 MHZ) inside + DMA.

>> hehe, this is his project, I’ll refrain from commenting,
>> I’ll just say that I like it.
>> I love it when it flies! And the first software, the creation of which
>>, by the way, he is ready to pay for (coders, gee-gee, we already have),
>> is DOOM. In a frame (maximum of two), pixel size 1*1.
>> And I’ll try to persuade him to write not DOOM, but HERETIC,
>> because I’m obsessed with it, but, unfortunately, I’ve already
>> passed it (at the most evil level). For me, Duke is worse and
>> Toad is worse. There, you know, there is nothing there that
>> I specifically need. I love flights of fancy and really
>> appreciate design. HERETIC is what you need...

>> P.S. And yet I will do it. Someday...

>> P.P.S. May the moderator forgive me! The thing is
>> that Echo GEN is a garbage dump; I don’t want to join a bunch of crap.
>> I know these are all people I respectread HARD.
>> About price conferences... This has already happened somewhere.
>> So, I’m ready to pay, say, 30 tons to a person
>> who will write something like SMM, only taking into account the requirements:
>> 42 characters, automatic text wrapping (disabled), editor no worse
>> ZXASM, fast procedures (disabled) for the disk, and more something.
>> 2DP: do you need money? I am sure that I will have like-minded people
>> in terms of requirements, and some of them are capable
>> of unfastening 10..30 tons to the author of the program. Ten people
>> will chip in - the author will receive money for his work.
>> From the world one by one - monopolies shirt :)
>> But I don’t know in what echo to organize such a conference?
>> So, do you agree with my requirements for the mail
>> editor? Speak up, let’s persuade Denis and give him money
>>? Or is everyone satisfied with the existing SMM? Then excuse me...

>> I'm ready to give money also for TR-DOS pkUnZip, a packer of code
>> blocks (better than MS-PACK and without glitches), tr-DOS commander according to
>> my requirements (I'm sure, not only mine...) and much more
>> for what. Join us. 2KIR: I’m ready to give money for yours too
>> INET. Or did the train leave and everything broke off?
>> I would have written all this myself, but it’s not interesting to me from the point of view
>> from the point of view of coding, II just want to use it
>> and not give a damn about it.

>> (hehe, in terms of coding, I’m interested in something else -
>> at the moment we are making STORM as impossible as possible,
>> and then we write an OS on it.
>> Don’t laugh, the solution is quite
>> solid. We are interested. When - I don’t know, and it doesn’t matter.
>> We’re just interested! If you’re interested too - to CODE. I will answer/discuss any questions you like.)

>> P.P.P.S. Once again, may the moderator forgive me.
>> I don’t know where this is, but GEN doesn’t raise a hand.
>> It’s the right thing. Or am I the only one who is so wrong?
>> I have long been accustomed to the fact that no one cares, and, in general,
>> I don’t expect much activity... We’ll wait and see!

>> Thank you all for your attention.

>> All the best.

>> Dmitry aka -=LD=- / X-TRADE GROUP.

>> -+- ZXASM 3.0
>> + Origin: Write letters... (500:812/08.14)
============================================================

This is the letter.

Sorry, Boris.

-+- iS-EDIT 6.06+

Whoa, finally someone figured it out, it’s time for Speck to have an axel. And then the memory
was expanded, the turbo was made, the screw was tightened :), the music was cranked up beyond belief,
but the screen remained unchanged as it was. IMHO it’s time to do it.

With maximum respect, Andy Harchenko.

Vot, skazhem, kak u menya seichas...;)