Significant BASIC differences in updated 49G ROMs
kje...@gmail.com
running 1.10) and doesn't really care one way or another about doing
so. He also doesn't do much more than some simple calculations for
microwave engineering (designs the antennas on our SAR), doesn't use
RPN, doesn't have any programs installed, etc.
I can't really remember if there would be any changes that would
greatly affect someone like that in updating the ROM. Are there any
speedups that he would notice in an updated system during basic
calculations? Serious math bugs he might hit? If so, I'll find and
dig out my old wire and do it for him. . .
Any thoughts?
TW
John H Meyers
Who can know whether anything he'll ever use was either fixed
or added later than that (HP supplied official 49G updates
thru 1.18 and another 49G update is on www.hpcalc.org for 1.19-6),
or whether any library he might suddenly realize to be useful
might also be based on an updated ROM?
One can also run Emu48 with the emulator ROM for 2.09 (or later),
making sure to not apply any "patches" ("beep" or otherwise),
then ROMUPLOAD that out through a PC serial port,
directly to the 49G calculator,
which will then be as up to date as can be,
equivalent (except in speed and no support for HPGCC)
to a 50G (okay, shorter screen and no SD card, either :)
[r->] [OFF]
pobox...@gmail.com
Shouldn't the 49G be similar to an HP48 for basic calculations as far
as old math bugs go, unless any operational bugs were introduced in
revamping the 48 ROM? But for the CAS, I noticed some nice
improvements like better expression simplifications. More CAS
commands, menus and flags were also added in 1.19-6.
John H Meyers
> Shouldn't the 49G be similar to an HP48 for basic calculations as far
> as old math bugs go, unless any operational bugs were introduced in
> revamping the 48 ROM? But for the CAS, I noticed some nice
> improvements like better expression simplifications. More CAS
> commands, menus and flags were also added in 1.19-6.
TW originally said that the 49G owner "doesn't use RPN";
does this mean that he uses the "ALGebraic" mode,
which never even existed in the HP48,
and is essentially all part of the CAS?
Most basic math functions (COS, EXP) are likely unaltered,
but no one knows what the owner actually uses.
ROM 1.19-6 (made by the same folks who made "official" 1.18)
was a big improvement over 1.18 (and certainly over 1.10,
which was one of the very earliest).
If one is going to bother updating ROM at all, however,
one could do it via Emu48 with an un-patched "latest"
(2.09 or later) HP50G rom, and thus gain the advantage
of several years' more improvements (including the
optional Equation Library, Periodic Table,
and complete compatibility with all the latest applications).
By "unpatched" is meant not applying any "Emu48 patches"
(at run time of Emu48 on the computer)
which make some change(s) for making "beeps" and such,
said patches being wrong for "real" calculators
(and will also cause checksum errors in ROM).
Despite the hardware changes between 49G and later versions,
(even the couple of keyboard function swaps),
each ROM still identifies the hardware correctly,
and acts appropriately on every platform (including Emu48),
thanks to the thoughtful thoroughness of the developers.
[r->] [OFF]
pobox...@gmail.com
> one could do it via Emu48 with an un-patched "latest"
> (2.09 or later) HP50G rom, and thus gain the advantage
> of several years' more improvements (including the
> optional Equation Library, Periodic Table,
> and complete compatibility with all the latest applications).
>
> By "unpatched" is meant not applying any "Emu48 patches"
> (at run time of Emu48 on the computer)
> which make some change(s) for making "beeps" and such,
> said patches being wrong for "real" calculators
> (and will also cause checksum errors in ROM).
It's kind of hard to get a clean unpatched ROM for Emu48 because the
KML script patches the ROM everytime it's used, leaving no ROM
unpatched in the directory. Even the 2.10-7 ROM from Prof. Parisse's
website reports checksum error in the emulator. Unless there's some
way to upload the ROM from calculator, then use upd49rom.exe(in my
Emu48 directory) and the latest .flash file to create an unpatched
ROM, then remove the patch command in the KML, run the emulator, and
update the calculator using ROMUPLOAD. Or am I wrong about how hard it
is to get an unpatched ROM for the emulator?
Andreas Möller
if you have a 50G and a serial cable for it and a connector to connect
both serial cables (50G and 49G) together, it might be easier to
modify ROMUPLOAD (which checks for the machine and does not work on
the ARM based machines IIRC) and connect a 49G with a 50G directly.
Why ROMUPLOAD is supposed *not* to work an a SATURN machine can only
be something to speculate about. OTOH it can be modfied quite easily.
Maybe to convince the owners of the 49G to give their money again to
HP for a 49G+ with teething problems (keys, paint, etc...) and after
that switch to a 50G. That are two additional calcs sold for every
former owner of the (FHB) Frozen Hampster Butt :-)
The marketing department (Two drinks minimum according to Dilbert.)
might enlighten this...
Greetings,
Andreas
P.S.: Maybe someone takes the time to explain Frozen Hampster Butt to
a non-native speaker ?
John H Meyers
> It's kind of hard to get a clean unpatched ROM for Emu48
> because the KML script patches the ROM everytime it's used
Isn't the "patch" only in Emu48 memory?
I simply renamed the patch files (beep.*);
you can also comment out the "Patch" line(s) in the KML script.
I have no errors in my "7. Full ROM" test in Emu48 (as an HP49G),
and I updated my own HP49Gs from Emu48, with no problem,
at least back in 2003, after I first undid those patches,
having been kindly clued in by Christoph in this thread:
http://groups.google.com/group/comp.sys.hp48/browse_thread/thread/c2ad698ceaa230d2/
> Even the 2.10-7 ROM from Prof. Parisse's website
> reports checksum error in the emulator.
Even without patches in Emu48?
"Debug4x" at one time may have included some ROM file with changes
(pehaps not even this particular issue),
but I believe this was subsequently fixed, whatever it was.
> Unless there's some
> way to upload the ROM from calculator, then use upd49rom.exe(in my
> Emu48 directory) and the latest .flash file to create an unpatched
> ROM, then remove the patch command in the KML, run the emulator, and
> update the calculator using ROMUPLOAD. Or am I wrong about how hard it
> is to get an unpatched ROM for the emulator?
If current Emu48 happens to be zapping original ROM files with patches,
just find out the couple of original bytes which those tiny patches zap,
and change the patch file to put them _back_ to the original!
I understand that the latest Emu48 can understand both original ROM files
and "one nibble of ROM per byte on PC" files (the latter might once
have been the only type of ROM file usable by Emu48),
but I probably don't now recall anything correctly :)
[r->] [OFF]
John H Meyers
> if you have a 50G and a serial cable for it and a connector to connect
> both serial cables (50G and 49G) together, it might be easier to
> modify ROMUPLOAD (which checks for the machine and does not work on
> the ARM based machines IIRC) and connect a 49G with a 50G directly.
> Why ROMUPLOAD is supposed *not* to work an a SATURN machine can only
> be something to speculate about.
ROMUPLOAD (which _sends_ the ROM image to another calculator,
imitating exactly how a PC would send it via conn4x or "FDP")
does work on a Saturn machine (HP49G or even Emu48),
but not on an ARM machine.
Any ARM-series ROM has been altered,
so that at some places, a new "BUSCC" instruction,
interpreted properly only by the Saturn emulator in the ARM-based calculator,
replaces what was originally assembled for a real Saturn to execute.
So, even if you could bypass the check, and get an ARM machine
to send its ROM image to a "real Saturn" HP49G,
the HP49G would not be able to use it.
This seems to be why ROMUPLOAD is disabled on ARM calculators;
it would send an unusable ROM to an HP49G,
and there is no corresponding "receive ROM over serial port"
function in ARM calcs, so the latter can not even update each other,
as could the HP49G calculators.
One can, however, use Emu48 (which uses an unmodified Saturn
ROM image, and acts like a "real Saturn" CPU)
to perform ROMUPLOAD to send its ROM image to a real HP49G,
provided you do not alter the two or four bytes
which some "beep" patch likes to make in Emu48,
which may affect nothing more than "beeps"
(or perhaps also the "auto turn off" function),
but would be better left alone,
since it also causes the ROM CRC self-check to fail in some bank.
[r->] [OFF]
pobox...@gmail.com
> > because the KML script patches the ROM everytime it's used
>
> Isn't the "patch" only in Emu48 memory?
>
> I simply renamed the patch files (beep.*);
> you can also comment out the "Patch" line(s) in the KML script.
>
> I have no errors in my "7. Full ROM" test in Emu48 (as an HP49G),
> and I updated my own HP49Gs from Emu48, with no problem,
> at least back in 2003, after I first undid those patches,
> having been kindly clued in by Christoph in this thread:
>
>
> > Even the 2.10-7 ROM from Prof. Parisse's website
> > reports checksum error in the emulator.
>
> Even without patches in Emu48?
>
> "Debug4x" at one time may have included some ROM file with changes
> (pehaps not even this particular issue),
> but I believe this was subsequently fixed, whatever it was.
I mistakenly opened a different KML than the file from which I edited
out the patch line. For correction and clarity, the 2.10-7 ROM from
Prof. Parisse's website is unpatched. Debug4x (http://www.hpcalc.org/
details.php?id=5441) contains unpatched ROMs for Emu48 with ROM
version 2.09 and 1.24, for anyone interested.
pobox...@gmail.com
> a non-native speaker ?
I think I got part of it, maybe. Is it that the metallic blue color
makes it look like it's frozen?