term vs dosx (is-dos r

Hi Kirill, In a message of to Michael Kondratyev (2:5030/299.18@fidonet), you wrote: KC> Here's a question: why is a term using a memory "extender" a good idea KC> under SMUC? And not for your modem connection scheme? Or according to your scheme too KC> yes, is it just that Boris is under SMUK? everything is much simpler: what I can check on, that’s what I do. and for other ports The modem is basically redone in about ten minutes. also, probably, in addition to the second Variations of ports will have an option without a cmos timer. Naturally, all changes concern only the “extender” - the program itself communicates with the outside world only through him. In general, it’s too early to do anything - the main task now is to give “therm” name. Bye, Michael.

Hi Boris, In a message of to Michael Kondratyev (2:5030/299.18@fidonet), you wrote: MK> the task now is to give the "term" a name. BS> term-mk1 excluded ;). although I was finally tired of guessing, I cast the simplest lot and determined. as it happened, so it turned out (there were about a hundred options - this means there was an element of chance). BS> And don’t stop there, but test, test and write a soaper. BS> I would like to have the simplest thread by September 1st! BS> If only you called at the appointed time on the appointed phone and picked it up BS> what lies with him. It may very well be. it is already added in the header that the mailer. ;) although you have to take some timeout with any outputs because at the moment nothing is stable there is no worker. (by the way, that version contained at least five only serious mistakes) By the way, is a device for reading the entire hard disk necessary or not? I'm on her I revived the next day (if I haven’t already forgotten where it is). Bye, Michael.

Hello Michael! MK> by the way, is a device for reading the entire hard disk needed or not? I her MK> revived the next day (if I haven't already forgotten where it is). I've been trying to call for a week - busy :( Throw a file at Gruzdev and you promised to ensure the reverse process. At the same time, explain to me how to use your zip to create an archive with the name of the underlying file in the archive without writing this name. Boris.

Hi Kirill, In a message of to Michael Kondratyev (), you wrote: MK> ports there will be an option without a cmos timer. Naturally, all changes MK> concern only the “extender” - the program itself communicates with the external one MK> the world only through him. KC> Hmm... Is there no way to work with the clock through the driver? Definitely separate KC> program with/without a clock? Consider that this small boot program is the driver. what does she really care about interferes? and so it is, a combined driver for buttons, character printing, computer and hours, plus several procedures that were useful to me. MK> It’s too early to do anything - the main task now is to give the “term” a name. KC> In the sense that the term will be called in general? Hmm... Problem... KC> "Is_Term_MK", "IsTeriMKa", "HaIsTerm" :)))))))))))))) extinguished, there will be little Richard Bye, Michael.

Once on 07/28/98 at 03:45:11 someone Michael Kondratyev wrote to Felix Knajzev regarding term vs dosx (re:is-dos r and I decided to intervene. ├┤i Michael & Felix! [...] MK> there will be 4 options (mine and moash schemes; moash schemes and no watches). I wrote it! We need a standard method for determining time in the system! You are now lame and making is-dos a typical tr-dos with pipeloaders. But writing software only for a specific option - for zonova, is strange. Now there are at least 4 (!!!) used clock connection schemes! Come up with a method for taking time out of the driver (or restarting, which is better) and write here! [...] MK> I didn’t want to upset you, but I have to: first of all, I’m interested in doing it MK> something close to a mailer. and the terminal will probably remain so poor. Well, it should be possible to launch transferring devices from it. Write in the description how - maybe someone will make a terminal. And instead of making fifty versions, you can work with a modem reduce to the resident installed before starting the program. And it will be possible to support _all_ versions then (if anyone needs it - will write the driver). Like a fossil on PeTs, although I have a hard time imagining it what he is. W!7h Be$7 W!$he$, FK0.

Hi Kirill, In a message of to Michael Kondratyev (), you wrote: MK> there will be 4 options (mine and moash schemes; moash schemes and no watches). KF> Wrote the same! We need a standard method for determining time in the system! KF> You are now lame and making is-dos a typical tr-dos with pipeloaders. KF> But writing software only for a specific option - for zonova, is strange. KF> Now there are at least 4 (!!!) clock connection schemes in use! I’ll try to separate the flies from the cutlets: I need a standard method, but in a system. in native system. for native programs. I actually leave for my .prg from isdosa large file system driver. there was an option to take everything out required in separate _isodn_ drivers, but it disappeared after basic calculations of required memory. therefore, we can assume that without taking into account disk operations is generally its own small system, one big driver (serial port, keyboard, ticker, clock, text screen, simplest memory manager). Is it bad that you get a lot of different ones? so I don't yet hard to assemble any. and when it gets hard, I’ll just announce it as a public domain with the ability for the user to compile any of the provided options himself, oradd according to your requirements. we get a simple Unix one technique (assemble your own kernel). KF> Come up with a method for taking time out of the driver (or restarting, which is better) KF> and write here! I've already figured it out for myself. It sounds something like this in my header: ========= dosxhDosX_Misc.inc ========= [...] Rtc_GetDateTime macro call ApiEntry_Misc db ORD_RTCGETDATETIME endm Rtc_GetUnixTime macro call ApiEntry_Misc db ORD_RTCGETUNIXTIME endm Rtc_GetDosTime macro call ApiEntry_Misc db ORD_RTCGETDOSTIME endm Rtc_SetDateTime macro call ApiEntry_Misc db ORD_RTCSETDATETIME endm Rtc_CorrectTime macro call ApiEntry_Misc db ORD_RTCCORRECTTIME endm [...] ======================================================= As you can see, there is no information about the clock for the program, but everything is removed from one restart ;) and access to other devices is also the same (except for one places where screen attributes are directly set, but this place is also easydone, and it is even possible (and necessary) to add a start-up recalculation of all used screen coordinates depending on those received from the driver sizes. KF> Well, there should be an opportunity to launch transferring devices from it. KF> Write in the description how - maybe someone will make a terminal. already in this form there is no way (there is a lot tied to each other). although there is nothing impossible (if necessary / if someone convinces me that there is no life - it’s possible diversify the command line). there will also be an opportunity to work with the code return (or rather, the return is already there, all that remains is to write an extended exebat for its processing) - it will be possible to “connect” external applications. only because There is no exec/spawn device, there are no child programs, and most importantly, I and I already have memory is not enough (it has already exceeded 32k), such a batch method will be the only one possible for their connections. KF> And instead of creating fifty versions, you can work with a modem KF> reduce to the resident installed before starting the program. KF> And it will be possible to support _all_ versions then (if anyone needs it - KF> will write the driver). Like a fossil on PeTs, although I have a hard time imagining it KF> what he is.Well, that’s what he is. All that remains is to approve the assignment of entry points. it Initially, I intended it to be a general program and a private “extender”. Bye, Michael.

Once on 08/19/98 at 05:13:39 someone Michael Kondratyev wrote to Kirill Frolov regarding term vs dosx (re:is-dos r and I decided to intervene. ├┤i Michael & Kirill! [...] MK> well, some little things should be taken into account when writing, for example, nizzya MK> use the stack for other purposes (nmi will not approve of this), you should not MK> disable interruptions, etc. Well, in general, it’s okay if 80 letters in a line are possible. But here's the problem.... THE EXTENDED SCREEN IS THE WHOLE CAN!!! Moreover, the 7th or 5th. Better make a buffer in another bank, huh? And such an incomprehensible thing. When programming a momed with interrupts sometimes an interrupt comes and the no interrupt pending bit is set... You have in the ancient term, in this case, the output from the interrupt howled, and the modem believed that the interrupt was not processed. The INT4 level was log.1. and therefore /NMI is constantly 0. Everything hung until reinitialization modem :-( Why is there such a ^^^? W!7h Be$7 W!$he$, FK0.