From: Han on
I have finished porting Jazz over to the HP49G+ and HP50G (ROM 2.15).
The latest version (version 1.1) includes support for FPTR2 and ZINT
tokens, as well as all new opcodes. The SysRPL debugger (SDB) is also
now available. One caveat is that Jazz has its own entries table
(which can now be installed in any port without speed limitations nor
RAM requirements to copy the tables from covered ports) whereas SDB
uses extable. This choice was due to the fact that there are users who
prefer the internal MASD compiler (which uses extable) and only want
to use SDB.

The files have been uploaded to hpcalc.org, but until the site is
refreshed, you can download the new versions via:

Jazz: http://www.math.uiuc.edu/~handuong/hp48/jazz50g.zip
SDB: http://www.math.uiuc.edu/~handuong/hp48/sdb.zip

Cheers!
Han
From: Han on
Here's a quick update to Jazz 50G. Minor fixes to label generation
during disassembly. Major fixes in SDB, including updates to
CK&DISPATCH1 emulation to handle updated sym/arry classes, conversion
of integers to reals, and a patch to work around the 2 bugs in ROM
2.15.

Until hpcalc.org does another refresh of files, you can get the
updates at:

Jazz 50G: http://www.math.uiuc.edu/~handuong/hp48/jazzv111.zip
Jazz 50G Source: http://www.math.uiuc.edu/~handuong/hp48/jazzv111src.zip
SDB (standalone library): http://www.math.uiuc.edu/~handuong/hp48/sdbv11.zip

This update of the standalone version of SDB also includes the source
code.
From: John H Meyers on
On 6/26/2010 10:39 PM, Han wrote:

> Here's a quick update to Jazz 50G. Minor fixes to label generation
> during disassembly. Major fixes in SDB, including updates to
> CK&DISPATCH1 emulation to handle updated sym/arry classes, conversion
> of integers to reals, and a patch to work around the 2 bugs in ROM
> 2.15

About this last patch -- is it a patch to the actual ROM,
or a patch for which you have to run all programs under Jazz?
(the latter does not seem very practical).

If there is a bug, what if it's in _all_ ROM versions?

If Jazz can correct a ROM bug, can it correct _all_ of them?

Say, why can't Emu48 correct ROM bugs, too?

That way, everyone using it, instead of a real calculator,
can have a perfect HP49/50 series calculator ;-)

[r->] [OFF]

From: Han on
On Jun 28, 4:51 pm, John H Meyers <jhmey...(a)nomail.invalid> wrote:
> On 6/26/2010 10:39 PM, Han wrote:
>
> > Here's a quick update to Jazz 50G. Minor fixes to label generation
> > during disassembly. Major fixes in SDB, including updates to
> > CK&DISPATCH1 emulation to handle updated sym/arry classes, conversion
> > of integers to reals, and a patch to work around the 2 bugs in ROM
> > 2.15
>
> About this last patch -- is it a patch to the actual ROM,
> or a patch for which you have to run all programs under Jazz?
> (the latter does not seem very practical).
>
> If there is a bug, what if it's in _all_ ROM versions?
>
> If Jazz can correct a ROM bug, can it correct _all_ of them?
>
> Say, why can't Emu48 correct ROM bugs, too?
>
> That way, everyone using it, instead of a real calculator,
> can have a perfect HP49/50 series calculator ;-)
>
> [r->] [OFF]

Sorry if there was any confusion. It is a patch of SDB. SDB was
designed to emulate CK&DISPATCH1 exactly as it is in ROM. I had
planned for the updated SDB to continue this -- emulating even the
bugs -- but then decided against this. I called it a patch because by
excluding the 2 bugs, SDB was no longer emulating the entry as it is
in ROM (i.e. bugs and all). It has nothing to do with patching the
actual ROM.
From: Andreas Möller on
Hello John,

as you might know, patching the ROM from the outside is quite
difficult.

The O.S. provides some "hooks" for this (e.g. the ROMPART) but not for
all (e.g. PTRs, FLASHPTRs).

One way to be able to patch the ROM from the outside is to install a
new SOL (SystemOuterLoop) and an additional RAM area where the
information for the patches is stored/handled. To my knowledge I am
the only one who has done this with the MLPs (Watch the video:
http://www.youtube.com/watch?v=Pj72miclisM).

Another way would be to patch the binary of the ROM *before*
downloading it into the calculator by exchanging the desired parts and
recalculating the needed checksums (of course while keeping supported
entries stable). IIRC this was an approach by the never released and
now abandoned HPGCC 3.0.

And a third way would be to program your own ROM :-)

> Say, why can't Emu48 correct ROM bugs, too?
> That way, everyone using it, instead of a real
> calculator, can have a perfect HP49/50 series
> calculator ;-)
But then, why should one buy a 50G with *bugs* from HP when there
would be a *bug-free* version available for free ?
What would be the reaction of HP´s legal department and the bean
counters ?
And what would that mean for the calculator division ?

Regards,
Andreas
www.software49g.gmxhome.de