USAG v1.4 for 49 series - Hewlett Packard

This is a discussion on USAG v1.4 for 49 series - Hewlett Packard ; Version 1.4 [28.07.2008] ------------------------ * improved: "PICT" for common commands (GOR, REPL, STO, SIZE,...). * improved: extended for commands like "FOR", "THEN, "WHILE",... * added: input supports command, {command}, "command", {"command"}. * added: support to ALG mode. * added: key ...

+ Reply to Thread
Results 1 to 13 of 13

Thread: USAG v1.4 for 49 series

  1. USAG v1.4 for 49 series

    Version 1.4 [28.07.2008]
    ------------------------
    * improved: "PICT" for common commands (GOR, REPL, STO, SIZE,...).
    * improved: extended for commands like "FOR", "THEN, "WHILE",...
    * added: input supports command, {command}, "command", {"command"}.
    * added: support to ALG mode.
    * added: key [+/-] and menu [LANG+].
    * added: support to 48gII.
    * removed: support to current font.
    * removed: freeze for some commands.

    http://www.gaak.org/hp/usag/

    - Gaak -

  2. Re: USAG v1.4 for 49 series

    Gaak kirjutas:
    > Version 1.4 [28.07.2008]
    > ------------------------
    > * improved: "PICT" for common commands (GOR, REPL, STO, SIZE,...).
    > * improved: extended for commands like "FOR", "THEN, "WHILE",...
    > * added: input supports command, {command}, "command", {"command"}.
    > * added: support to ALG mode.
    > * added: key [+/-] and menu [LANG+].
    > * added: support to 48gII.
    > * removed: support to current font.
    > * removed: freeze for some commands.
    >
    > http://www.gaak.org/hp/usag/
    >
    > - Gaak -


    Thanks for the great little utility!

    This last one (1.4) has on my v.2.10-7 HP50G an unfortunate side effect
    When using it on single type argument command (like CLKADJ or
    ->KEYTIME), and pressing afterwards any key, the calculator
    warmstarts... I downloaded the 1.3 from hpcalc.org for checking. 1.3
    runs flawlessly.

    Robert Tiismus
    ZZ

  3. Re: USAG v1.4 for 49 series

    On Aug 7, 2:54*pm, Robert Tiismus wrote:
    > Gaak kirjutas:
    >
    > > Version 1.4 [28.07.2008]
    > > ------------------------
    > > * improved: "PICT" for common commands (GOR, REPL, STO, SIZE,...).
    > > * improved: extended for commands like "FOR", "THEN, "WHILE",...
    > > * added: input supports command, {command}, "command", {"command"}.
    > > * added: support to ALG mode.
    > > * added: key [+/-] and menu [LANG+].
    > > * added: support to 48gII.
    > > * removed: support to current font.
    > > * removed: freeze for some commands.

    >
    > >http://www.gaak.org/hp/usag/

    >
    > > - Gaak -

    >
    > Thanks for the great little utility!
    >
    > This last one (1.4) has on my v.2.10-7 HP50G an unfortunate side effect
    > When *using it on single type argument command (like CLKADJ or
    > ->KEYTIME), and pressing afterwards any key, the calculator
    > warmstarts... I downloaded the 1.3 from hpcalc.org for checking. 1.3
    > runs flawlessly.
    >
    > Robert Tiismus
    > ZZ


    Uhm?, not for me. :-(
    Testing on ROM 2.09 (50g) and ROM 2.10-7 (Emu-49G) runs flawlessly.
    Commands like CLKADJ and ->KEYTIME are commands with common
    structure...

    More details?:
    - Input mode (remember that ALG and RPN are supported now).
    - Command mode (lists, strings, command).

    Thanks for report bugs ;-)

    - Gaak -

  4. Re: USAG v1.4 for 49 series

    Gaak kirjutas:
    > On Aug 7, 2:54 pm, Robert Tiismus wrote:
    >> Gaak kirjutas:
    >>
    >>> Version 1.4 [28.07.2008]
    >>> ------------------------
    >>> * improved: "PICT" for common commands (GOR, REPL, STO, SIZE,...).
    >>> * improved: extended for commands like "FOR", "THEN, "WHILE",...
    >>> * added: input supports command, {command}, "command", {"command"}.
    >>> * added: support to ALG mode.
    >>> * added: key [+/-] and menu [LANG+].
    >>> * added: support to 48gII.
    >>> * removed: support to current font.
    >>> * removed: freeze for some commands.
    >>> http://www.gaak.org/hp/usag/
    >>> - Gaak -

    >> Thanks for the great little utility!
    >>
    >> This last one (1.4) has on my v.2.10-7 HP50G an unfortunate side effect
    >> When using it on single type argument command (like CLKADJ or
    >> ->KEYTIME), and pressing afterwards any key, the calculator
    >> warmstarts... I downloaded the 1.3 from hpcalc.org for checking. 1.3
    >> runs flawlessly.
    >>
    >> Robert Tiismus
    >> ZZ

    >
    > Uhm?, not for me. :-(
    > Testing on ROM 2.09 (50g) and ROM 2.10-7 (Emu-49G) runs flawlessly.
    > Commands like CLKADJ and ->KEYTIME are commands with common
    > structure...
    >
    > More details?:
    > - Input mode (remember that ALG and RPN are supported now).
    > - Command mode (lists, strings, command).
    >
    > Thanks for report bugs ;-)
    >
    > - Gaak -


    Thank you for asking! I tried again. I did "ON-A-F NO", then turned on
    the RPN mode, downloaded the program and tried again:
    1: { CLKADJ }
    usag.hp

    It displayed the correct information, but when I pressed ANY KEY, I saw
    very spectacular random pattern on the screen and "Try To Recover
    Memory?" afterwards. "No" was unresponsible but "Yes" functioned.

    Is it possible that my unit is faulty? I got it only half a year ago and
    have not had much time to use it.

    Robert Tiismus

  5. Re: USAG v1.4 for 49 series

    On Aug 14, 10:45*am, Robert Tiismus wrote:
    > Thank you for asking! I tried again. I did "ON-A-F NO", then turned on
    > the RPN mode, downloaded the program and tried again:
    > 1: { CLKADJ }
    > usag.hp
    >
    > It displayed the correct information, but when I pressed ANY KEY, I saw
    > very spectacular random pattern on the screen and "Try To Recover
    > Memory?" afterwards. "No" was unresponsible but "Yes" functioned.
    >
    > Is it possible that my unit is faulty? I got it only half a year ago and
    > have not had much time to use it.
    >
    > Robert Tiismus


    Maybe another buggy object?... libraries with bad $CONFIG?...
    Please, delete your 'STARTUP' variable from HOME and after Warmstart
    press and hold [BS]... and retry again USAG.

    Sorry for the inconvenience :-(

    - Gaak -

  6. Re: USAG v1.4 for 49 series

    Gaak kirjutas:
    > On Aug 14, 10:45 am, Robert Tiismus wrote:
    >> Thank you for asking! I tried again. I did "ON-A-F NO", then turned on
    >> the RPN mode, downloaded the program and tried again:
    >> 1: { CLKADJ }
    >> usag.hp
    >>
    >> It displayed the correct information, but when I pressed ANY KEY, I saw
    >> very spectacular random pattern on the screen and "Try To Recover
    >> Memory?" afterwards. "No" was unresponsible but "Yes" functioned.
    >>
    >> Is it possible that my unit is faulty? I got it only half a year ago and
    >> have not had much time to use it.
    >>
    >> Robert Tiismus

    >
    > Maybe another buggy object?... libraries with bad $CONFIG?...
    > Please, delete your 'STARTUP' variable from HOME and after Warmstart
    > press and hold [BS]... and retry again USAG.
    >
    > Sorry for the inconvenience :-(
    >
    > - Gaak -


    I erased all libraries, did again "ON-A-F NO". Bu still the same
    behavior... Then I downgraded ROM to version 2.09 (from HP site), and it
    behaved! The bug was back after upgrading to 2.10-7. I think it is
    somehow related to LANGUAGE selection, because after every crash, the
    "LANGUAGE->" command returns "5.". I tried to look into program, to try
    to debug it, but it was inside CODE object. Is it this way for
    aesthetical reasons? That this way the stack is not cluttered with
    EXTERNAL's. Anyhow, I got it disassembled by this:
    << ->H DUP SIZE 33 SWAP SUB "D9D20D9D20" SWAP + H-> ->s2 >>
    But could not find debugger like Jazz library had for HP48 series. So
    again no avail

    Robert Tiismus

  7. Re: USAG v1.4 for 49 series

    On Aug 15, 4:45 am, Robert Tiismus wrote:
    > Gaak kirjutas:
    >
    >
    >
    > > On Aug 14, 10:45 am, Robert Tiismus wrote:
    > >> Thank you for asking! I tried again. I did "ON-A-F NO", then turned on
    > >> the RPN mode, downloaded the program and tried again:
    > >> 1: { CLKADJ }
    > >> usag.hp

    >
    > >> It displayed the correct information, but when I pressed ANY KEY, I saw
    > >> very spectacular random pattern on the screen and "Try To Recover
    > >> Memory?" afterwards. "No" was unresponsible but "Yes" functioned.

    >
    > >> Is it possible that my unit is faulty? I got it only half a year ago and
    > >> have not had much time to use it.

    >
    > >> Robert Tiismus

    >
    > > Maybe another buggy object?... libraries with bad $CONFIG?...
    > > Please, delete your 'STARTUP' variable from HOME and after Warmstart
    > > press and hold [BS]... and retry again USAG.

    >
    > > Sorry for the inconvenience :-(

    >
    > > - Gaak -

    >
    > I erased all libraries, did again "ON-A-F NO". Bu still the same
    > behavior... Then I downgraded ROM to version 2.09 (from HP site), and it
    > behaved! The bug was back after upgrading to 2.10-7. I think it is
    > somehow related to LANGUAGE selection, because after every crash, the
    > "LANGUAGE->" command returns "5.". I tried to look into program, to try
    > to debug it, but it was inside CODE object. Is it this way for
    > aesthetical reasons? That this way the stack is not cluttered with
    > EXTERNAL's. Anyhow, I got it disassembled by this:
    > << ->H DUP SIZE 33 SWAP SUB "D9D20D9D20" SWAP + H-> ->s2 >>
    > But could not find debugger like Jazz library had for HP48 series. So
    > again no avail
    >
    > Robert Tiismus


    Good news for you... your calculator is ok, don't worry ;-)

    Upgrading my 50g to 2.10-7 and debugging (for you), I found the bug,
    many thanks.
    The Flash Pointer 7 8E (aka ^WRITEMENU) available on "all" ROM
    versions, is wrong on 2.10-7.

    ^WRITEMENU (2.10-7)
    ::
    DUPTYPEREAL?
    case
    ( ... )
    ;

    ^WRITEMENU (another version like 1.18, 1.19-6, .. 2.09)
    ::
    ZERO SIX
    ZERO_DO
    ( ... )
    ;

    ^WRITEMENU is called for commands that requires "One Argument" (-
    >KEYTIME), "No Arguments" (MEM), and metas (->LIST).

    Obviously, ROM 2.10-7 is beta :-S

    Sorry for the inconvenience, again... :-(

    - Gaak -

  8. Re: USAG v1.4 for 49 series

    Gaak kirjutas:
    >
    > Good news for you... your calculator is ok, don't worry ;-)
    >
    > Upgrading my 50g to 2.10-7 and debugging (for you), I found the bug,
    > many thanks.
    > The Flash Pointer 7 8E (aka ^WRITEMENU) available on "all" ROM
    > versions, is wrong on 2.10-7.
    >
    > ^WRITEMENU (2.10-7)
    > ::
    > DUPTYPEREAL?
    > case
    > ( ... )
    > ;
    >
    > ^WRITEMENU (another version like 1.18, 1.19-6, .. 2.09)
    > ::
    > ZERO SIX
    > ZERO_DO
    > ( ... )
    > ;
    >
    > ^WRITEMENU is called for commands that requires "One Argument" (-
    >> KEYTIME), "No Arguments" (MEM), and metas (->LIST).

    > Obviously, ROM 2.10-7 is beta :-S
    >
    > Sorry for the inconvenience, again... :-(
    >
    > - Gaak -


    I already realized that 2.09 is the way to go and downgraded the ROM
    version. Thank you! And don't be sorry, Mea Culpa for use of the
    uncertified ROM version!

    Robert Tiismus

  9. Re: USAG v1.4 for 49 series

    but now that the problems is found
    is there a real+beta ROM version compatible
    1,4,1 USAG available somewhere?

    thanks Gaak

    "Robert Tiismus" wrote in message
    news:g84oqi$q5e$1@aioe.org...
    > Gaak kirjutas:
    >>
    >> Good news for you... your calculator is ok, don't worry ;-)
    >>
    >> Upgrading my 50g to 2.10-7 and debugging (for you), I found the bug,
    >> many thanks.
    >> The Flash Pointer 7 8E (aka ^WRITEMENU) available on "all" ROM
    >> versions, is wrong on 2.10-7.
    >>
    >> ^WRITEMENU (2.10-7)
    >> ::
    >> DUPTYPEREAL?
    >> case
    >> ( ... )
    >> ;
    >>
    >> ^WRITEMENU (another version like 1.18, 1.19-6, .. 2.09)
    >> ::
    >> ZERO SIX
    >> ZERO_DO
    >> ( ... )
    >> ;
    >>
    >> ^WRITEMENU is called for commands that requires "One Argument" (-
    >>> KEYTIME), "No Arguments" (MEM), and metas (->LIST).

    >> Obviously, ROM 2.10-7 is beta :-S
    >>
    >> Sorry for the inconvenience, again... :-(
    >>
    >> - Gaak -

    >
    > I already realized that 2.09 is the way to go and downgraded the ROM
    > version. Thank you! And don't be sorry, Mea Culpa for use of the
    > uncertified ROM version!
    >
    > Robert Tiismus




  10. Re: USAG v1.4 for 49 series

    On Fri, 15 Aug 2008 13:03:32 -0500, Gaak wrote:

    > The Flash Pointer 7 8E (aka ^WRITEMENU)
    > available on "all" ROM versions,
    > is wrong on 2.10-7


    There's more to it, because there was not just one "2.10";
    there must have been a "regular" ROM (never released by HP),
    and there was an experimental "Geometry/Spreadsheet" ROM
    (made for math teachers in France?)
    which was apparently derived from that,
    and you are probably using the latter.

    The problem appears to be that some of the FPTR "slots"
    in bank 7 seem to have been "borrowed" (or re-arranged)
    in the "Geometry" ROM, no longer matching their original functions
    (or at least not matching what "extable" says),
    while the "regular" orignal 2.10 ROM was no doubt actually fine.

    I think you would have to ask BP about how that ROM was created,
    to find out how these substitutions occurred,
    and what else might be affected by them;
    for example, was there not enough room to add Geometry+Spreadsheet,
    and were other original functions (or just their bank table pointers)
    replaced or re-arranged instead?


    > NOT_WRITEMENU (2.10-7G? wrong function in bank table?)

    (at 7:42797 ? in Geometry ROM)

    > ::
    > DUPTYPEREAL?
    > case
    > (...)

    ( clearly CAS symbolic code here, called from "symbolic" functions )


    > ^WRITEMENU (another version like 1.18, 1.19-6, .. 2.09)

    (at 7:461C0 ? in "regular" 2.10, called from ^MENUext and other sensible places)

    > ::
    > ZERO SIX
    > ZERO_DO

    ( this is correct code )
    > ;



    > ^WRITEMENU is called for commands that requires "One Argument"
    > (->KEYTIME), "No Arguments" (MEM), and metas (->LIST).
    > Obviously, ROM 2.10-7 is beta :-S


    It would seem that only the altered "Geometry" version has been rearranged,
    evidently so obscurely that it doesn't seem to have been noticed before.

    Evidently _you_ call ^WRITEMENU (using its original FPTR 7 8E)
    from USAG, although the commands themselves (in ROM) do not;
    perhaps you could choose another function
    (or copy the original code) to become free of this dependence,
    and thus to be compatible even with the "Geometry" ROM?

    It would be nice if the very caring HP (new?) management
    would care to either post the improved "regular" 2.10 ROM themselves,
    or let someone else post it "unofficially,"
    like 1.19-6 for old 49G, which was never acknowledged by HP,
    but which was clearly improved over HP's "official" 1.18

    Otherwise the altered "Geometry" version of 2.10
    might remain the only one found,
    and might prove incompatible with some "bank 7 dependent" SysRPL/ML,
    just as apparently in this case.

    [r->] [OFF]

  11. Re: USAG v1.4 for 49 series

    to my unhumble opinion:
    HP (Cyrille) just steel one more bank to the system
    and/or
    even another one for the extable 2

    50g+ could have twice the flash and 90 MHz speed
    and -5V/+5V real RS-232C
    plus full HPGCC compatibility

    "we can patch it up, baby"
    VPN

    "John H Meyers" wrote in message
    newsp.ufyc14dann735j@miu.edu...
    > On Fri, 15 Aug 2008 13:03:32 -0500, Gaak wrote:
    >
    >> The Flash Pointer 7 8E (aka ^WRITEMENU)
    >> available on "all" ROM versions,
    >> is wrong on 2.10-7

    >
    > There's more to it, because there was not just one "2.10";
    > there must have been a "regular" ROM (never released by HP),
    > and there was an experimental "Geometry/Spreadsheet" ROM
    > (made for math teachers in France?)
    > which was apparently derived from that,
    > and you are probably using the latter.
    >
    > The problem appears to be that some of the FPTR "slots"
    > in bank 7 seem to have been "borrowed" (or re-arranged)
    > in the "Geometry" ROM, no longer matching their original functions
    > (or at least not matching what "extable" says),
    > while the "regular" orignal 2.10 ROM was no doubt actually fine.
    >
    > I think you would have to ask BP about how that ROM was created,
    > to find out how these substitutions occurred,
    > and what else might be affected by them;
    > for example, was there not enough room to add Geometry+Spreadsheet,
    > and were other original functions (or just their bank table pointers)
    > replaced or re-arranged instead?
    >
    >
    >> NOT_WRITEMENU (2.10-7G? wrong function in bank table?)

    > (at 7:42797 ? in Geometry ROM)
    >
    >> ::
    >> DUPTYPEREAL?
    >> case
    >> (...)

    > ( clearly CAS symbolic code here, called from "symbolic" functions )
    >
    >
    >> ^WRITEMENU (another version like 1.18, 1.19-6, .. 2.09)

    > (at 7:461C0 ? in "regular" 2.10, called from ^MENUext and other sensible
    > places)
    >
    >> ::
    >> ZERO SIX
    >> ZERO_DO

    > ( this is correct code )
    >> ;

    >
    >
    >> ^WRITEMENU is called for commands that requires "One Argument"
    >> (->KEYTIME), "No Arguments" (MEM), and metas (->LIST).
    >> Obviously, ROM 2.10-7 is beta :-S

    >
    > It would seem that only the altered "Geometry" version has been
    > rearranged,
    > evidently so obscurely that it doesn't seem to have been noticed before.
    >
    > Evidently _you_ call ^WRITEMENU (using its original FPTR 7 8E)
    > from USAG, although the commands themselves (in ROM) do not;
    > perhaps you could choose another function
    > (or copy the original code) to become free of this dependence,
    > and thus to be compatible even with the "Geometry" ROM?
    >
    > It would be nice if the very caring HP (new?) management
    > would care to either post the improved "regular" 2.10 ROM themselves,
    > or let someone else post it "unofficially,"
    > like 1.19-6 for old 49G, which was never acknowledged by HP,
    > but which was clearly improved over HP's "official" 1.18
    >
    > Otherwise the altered "Geometry" version of 2.10
    > might remain the only one found,
    > and might prove incompatible with some "bank 7 dependent" SysRPL/ML,
    > just as apparently in this case.
    >
    > [r->] [OFF]




  12. BP, major bug in geometry rom (WAS: Re: USAG v1.4 for 49 series)


    On Aug 15, 8:02 pm, "John H Meyers" wrote:
    > On Fri, 15 Aug 2008 13:03:32 -0500, Gaak wrote:
    > > The Flash Pointer 7 8E (aka ^WRITEMENU)
    > > available on "all" ROM versions,
    > > is wrong on 2.10-7

    >
    > There's more to it, because there was not just one "2.10";
    > there must have been a "regular" ROM (never released by HP),
    > and there was an experimental "Geometry/Spreadsheet" ROM
    > (made for math teachers in France?)
    > which was apparently derived from that,
    > and you are probably using the latter.
    >
    > The problem appears to be that some of the FPTR "slots"
    > in bank 7 seem to have been "borrowed" (or re-arranged)
    > in the "Geometry" ROM, no longer matching their original functions
    > (or at least not matching what "extable" says),
    > while the "regular" orignal 2.10 ROM was no doubt actually fine.
    >
    > I think you would have to ask BP about how that ROM was created,
    > to find out how these substitutions occurred,
    > and what else might be affected by them;
    > for example, was there not enough room to add Geometry+Spreadsheet,
    > and were other original functions (or just their bank table pointers)
    > replaced or re-arranged instead?



    I just checked.
    Bank 7 contains 13 commands, and then the contents of library 222.

    In the geometry rom, somehow after the FTR 7 C (the 13th command)
    apparently 22 garbage entries got inserted into the the bank table.
    This messed up all later bank 7 flash pointers (so FPTR 7 8E became
    FPTR 7 A4, etc) BP should probably be able to fix this.

    The issue did not impact any code compiled into bank 7 because the
    compiler/linker replaced the symbols with the flashpointers that had a
    slot incremented by 22 compared to the normal location.

    Unfortunately, the extable/extable2 use the original flashpointers, as
    do any existing programs that used those flashpointers.

    BP, could you please generate a new rom that does not have those bogus
    entries (or if they are not completely bogus that have those entries
    in slots at the end of the banks' flashpointer table)? Thanks.

    Oh, BP, I also have a few unrelated things to ask.

    I presume you must have had access to what would be the official 2.10
    if HP had released it (if you did not have access to it, that would
    make basing your geometry rom on it a bit difficult i would imagine) .
    Any chance you you leak it? No? Ok. Just wanted to ask. :-D

    (BTW, the posted cas2user.s file for the geometry app was clearly not
    used to compile the posted rom, as CASRESERVED62 and CASRESERVED63 in
    the posted rom are RCLVX and STOVX. That means that two of those
    pseudo-userrpl commands found in CASRESERVED64-99 must have been
    removed. Which two where they? (Just curious).)

  13. Re: USAG v1.4 for 49 series

    On Aug 7, 2:54*pm, Robert Tiismus wrote:
    > This last one (1.4) has on my v.2.10-7 HP50G an unfortunate side effect


    Version 1.5 [18.08.2008]
    ------------------------
    * improved: language correction and optimization.
    * improved: " ?" -> "Undefined Name" (with language support).
    * fixed: bug ROM beta 2.10-7 with ^WRITEMENU.

    http://www.gaak.org/hp/usag/

    Cheers.

    - Gaak -

+ Reply to Thread