Hi,

you might have stated your keyboard driver first ...

but as you later state that the problem occurs with MKEYB as well,
here's my educated guess what happens:

the UP, DOWN, LEFT, RIGHT keys in the cursor block are similar to the
UP, DOWN, LEFT, RIGHT in the numeric key pad as they send similar
scancodes, but send an additional 'E0' scancode

now it seems this additional scancode is lost, and instead of
   'cursor block UP'  E0 48
only
   'numeric key pad UP'  48
is received (and processed).

since this has not been reported by anybody else so far, and only
if your machines are slow, and only sometimes:

something in your machines blocks interrupts for too long, and keys
get lost - most likely some TSR or driver, ...

please try running ONLY  HIMEM and MKEYB (or the other keyboard
driver), and see if the problem vanishes.

add more software until the problem reappears

Tom



am 9. Mai 2014 um 00:19 schrieben Sie:

> Hi,

> A couple of years ago I noticed that, in some computers, when
> the Up, Down, Left or Right keys were held pressed for several
> seconds, sometimes they acted as numeric keypad keys, i.e.,
> instead of just moving the cursor in these directions, they
> actually *wrote* the characters 8, 2, 4 or 6 respectively.

> These are *not* the numeric keypad keys, but the four-key
> cluster between the alpha and the numeric keypads.

>     Let me state the whole thing once more, to avoid
>     misunderstanding.

>     The keys involved are not the numeric keypad keys. However,
>     the effect is *as if* we were using the numeric keypad arrow
>     keys with the NumLock feature off, and these keys, when held
>     pressed, would act most of the time as per NumLock off (move
>     the cursor), but sometimes -- say, one in 30 iterations --
>     would act as if NumLock was on (write digits 2, 4, 6, 8).

> This is of course serious, as you may inadvertently write a
> digit in your text when you merely wanted to move the cursor.

> This usually happened in old computers (early Pentiums) donated
> for use in my Health Center volunteer work. At the time I
> thought it was some hardware fault because the machines were old
> and presumably worn out.

> Actually I used this as a test to evaluate if the machines were
> reliable. I saw this happen in several computers, and still have
> one of those. All tests were done with pure FreeDOS.

> One trick that apparently eliminates the problem completely is
> keeping NumLock off. (This may sound odd, because NumLock is
> supposed to change the behaviour of the numeric keypad keys,
> whereas we are talking about the other arrow keys here.)

> Because this is a keyboard issue, I (rather vaguely) remember
> trying MKEYB.EXE by Tom Ehlert instead of the standard FreeDOS
> KEYB.EXE, but the problem remained.

> Today, however, for the first time, the same problem happened
> with my own computer -- a machine I have been using intensively
> for over a year, and which so far has been completely reliable.
> An old text editor was running which requires reducing processor
> speed because of the Borland runtime 200 bug. I use Eric's FDAPM
> to reduce the speed. The problem was exactly the same as
> described above, and it also disappeared when the NumLock
> feature is off. It also disappeared when the processor was
> returned to its full speed (1 GHz).

> So perhaps it is not a hardware fault of old computers, but some
> other issue. As I never had this happen with MS-DOS which I used
> from 1995 to 2007, I thought it might be a FreeDOS issue.

> On a side note, when I started running the Health Center
> database in a network 2 years ago, with all machines running
> FreeDOS, a problem arose which *also* involved keeping the Up,
> Down, Left or Right keys pressed for a few seconds in the client
> machines. The effect was that the whole network crashed.

> I mention the network problem just because it was also triggered
> by keeping the same keys pressed. I discussed the issue in our
> list, and several people responded. The solution was to run
> MS-DOS in the server. Clients continued to run FreeDOS, and the
> problem never came up again. It seemed that the problem was
> related to running FreeDOS in a network.

> In today's issue, there is no network involved. This is a
> standalone machine I use at home as my main computer, and it has
> been very reliable so far. I also run Puppy Linux and Seamonkey
> or Nightly in it to browse the web. So apparently it is not a
> hardware problem.

> Has anyone experienced anything like that? Any idea what it
> might be?

> Regards,

> Marcos



> --------------------------------------
> Marcos Fávero Florence de Barros
> Campinas, Brazil



> ------------------------------------------------------------------------------
> Is your legacy SCM system holding you back? Join Perforce May 7 to find out:
> • 3 signs your SCM is hindering your productivity
> • Requirements for releasing software faster
> • Expert tips and advice for migrating your SCM now
> http://p.sf.net/sfu/perforce
> _______________________________________________
> Freedos-user mailing list
> Freedos-user@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/freedos-user



Mit freundlichen Grüßen/Kind regards
Tom Ehlert
+49-241-79886


------------------------------------------------------------------------------
Is your legacy SCM system holding you back? Join Perforce May 7 to find out:
• 3 signs your SCM is hindering your productivity
• Requirements for releasing software faster
• Expert tips and advice for migrating your SCM now
http://p.sf.net/sfu/perforce
_______________________________________________
Freedos-user mailing list
Freedos-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freedos-user

Reply via email to