Mail Archives: opendos/1998/10/14/19:11:05
Guti AT mail DOT qsystems DOT es writes:
> SUGGESTIONS
> þ Include in each new release a file called WHATSNEW.TXT with
> the history of improvements, bugs fixed, ... for every part
> of DR DOS (guti AT bbs-ce DOT uab DOT es).
It might also be nice (and in marked contrast to Microsoft's handling
of bug reports and suggestions) if Caldera listed a name -- *not* an
email address! -- for the source of each suggestion. "Corrects a bug
in CHKDSK which reported 0-byte files as cross-linked at cluster zero,
reported by John Doe." "This update to EDIT displays the line number
in the status line, as suggested by J. Random User."
> þ Implementation of the SERIAL command for allowing the user
I still don't see the value of this one, especially as DR DOS doesn't
display the serial number in most commands. However, all the "standard"
methods of changing it appear to be supported, so any program to do this
should work correctly under DR DOS -- it does not have to come from
Caldera.
> þ FASTOPEN and DOSKEY (HISTORY) should be out of the kernel,
> this way users that do not use it, will not have an
The way it's implemented now has a nice side effect. Any program which
uses the kernel console calls gets its own private command history.
DEBUG is one such program.
> þ Rename the programs called in DR-DOS 6 as xxxMAX to xxxMgr
>
> þ Include the 'Starting Caldera DR DOS...' message using
> color 10 (Caldera Green) like in Mathias Paul IBMBIO.COM;
>
> þ The same with NWCACHE, NWCDEX, ... should be DRCACHE, DRCDEX,
> this will increase the homogenity with the new DRMOUSE.
I would argue against changes which threaten compatibility for no better
purpose than mere cosmetics. Changing the name of a program, for
example, means that any batch file which calls the program must be
updated (or the program renamed back to its old name.) Look how
LOADFIX works under DR DOS, for example.
I haven't seen Mr. Paul's IBMBIO.COM, so I don't know how that change
is implemented. But it's probably worth pointing out that DOS may be
run on monochrome systems -- MDA, HGA. Would attribute 10 be visible
on an MDA?
> þ Have you seen PTS-DOS (www.phystechsoft.com)? It has nice features that
> could be implemented in DR DOS. For instance, pressing
> Ctrl-Alt-Del, automatically flush all the buffers, and performs
> a quick reboot (similar to those available in QEMM). It is very
This is in fact a feature of the memory manager. Does QEMM run under
DR DOS? I don't have a copy to try, but I suspect that it does.
Something similar could probably be added to EMM386.
> þ PASSWORD command should include a warning for preventing the
> protection of itself with a password (doing that, the password
> could not be removed unless you have it in a disk).
> (guti AT bbs-ce DOT uab DOT es)
DR DOS already has a certain amount of "pistol" protection, e.g.
the FORMAT /X switch. But it's easy to go overboard with this sort
of thing: should FORMAT be prevented from erasing any drive with a
copy of FORMAT on it? Should DEL be prevented from nuking UNDELETE?
In fact, the putative problem isn't really all that serious anyhow.
DR DOS file-level password protection is nice to have, but easy enough
to circumvent. Just boot up any version of DOS which doesn't provide
it: MS-DOS, PC DOS. Remove the hidden attribute from the passworded
file, copy it somewhere else, and you're set.
> þ TaskMgr should implement multitasking for 8086+ and not only
> task switching.
Think about this....
> þ Internal commands like DIR, COPY, when invoking help with
> the /? switch should include the two lines of the top like
> [command version number]
Internal commands, by definition, reside in the command shell. You can
get version info by typing VERSION COMMAND.COM -- but perhaps detailed
version info should be added to the COMMAND /? display.
raster AT highfiber DOT com
- Raw text -