I have a 128K HP RAM card which became "confused" during a transfer 
between my HP48GX calcs.
On startup the calc displays 'Invalid Card Data' warning.

I downloaded the program P0Fix but it reports that there is no 
corruption on the card.

Problem: When I run PVARS on port 0, it returns an empty list and 
reports that the memory is just 75,000 bytes.

Now that figure would be correct if the previously stored programs 
visibly existed. And I assume that they still do.

Question:  How do I recover them?

Geoff 

Sometime I can get an "Invalid Card Data" when I plug the card from one
calc to another.
Try to replug the card on the source HP-48GX, maybe it can fix the
problem.
If this problem continues, I guess you have lost the data.
It happened recently when I changed the battery of my 32KB RAM card
where I stored a copy of the HP-41CV emulator: when I switched the
machine on after replug the card, I got an "Invalid Card Data" error.
Everything was lost. To avoid the message, just store something on the
card and press ON-C. The message should disapear.

fred.

On Tue, 24 May 2005 14:47:20 -0500, R. Geoff Newbury wrote:

> I have a 128K HP RAM card which became "confused"
> during a transfer between my HP48GX calcs.

It could be worse -- imagine how Star Trek's "beam me up/down"
procedures must scramle a little DNA every time :)

> On startup the calc displays 'Invalid Card Data' warning.

> I downloaded the program P0Fix but it reports that there is no
> corruption on the card.

> Problem: When I run PVARS on port 0, it returns an empty list and
> reports that the memory is just 75,000 bytes.

> Now that figure would be correct if the previously stored programs
> visibly existed. And I assume that they still do.

> Question:  How do I recover them?

Some old postings on this very subject:

Using MetaKernel:
http://groups-beta.google.com/group/comp.sys.hp48/msg/4d073d4bc16b57c6
http://groups-beta.google.com/group/comp.sys.hp48/msg/92301ec9c04d6ed0

About memory chips and card socket shorts:
http://groups-beta.google.com/group/comp.sys.hp48/browse_frm/thread/923fdaa74b25bd6d/1de927180df3e792

Programs with "recover" in their name, for HP48GX:
http://www.hpcalc.org/search.php?query=recover&hp48=1

You already tried P0FIX, but for anyone else:
http://www.hpcalc.org/details.php?id=2454
http://www.hpcalc.org/hp48/utils/memory/p0fix.zip

And finally, you could try ALT+A+F -> YES
(try to recover lost objects, works with "merged" RAM)

If your card has a low battery (startup warning),
keep it FULLY PLUGGED IN with FRESH AAA BATTS IN CALC
and CALC TURNED ON as you change the card battery
(it can be tricky not to pull out or wiggle the card
in its socket as you pull out the coin cell; it's good
to "break in" new cards by sliding cell in/out several times,
until it slides more easily, before you start to use the card, or
apply very tiny amount of naphtha/lighter fluid/elect. cleaner/WD40).

[r->] [OFF]

On Tue, 24 May 2005 21:50:50 UTC, fbarbaise@yahoo.fr wrote:

> Sometime I can get an "Invalid Card Data" when I plug the card from one
> calc to another.
> Try to replug the card on the source HP-48GX, maybe it can fix the
> problem.
> If this problem continues, I guess you have lost the data.
> It happened recently when I changed the battery of my 32KB RAM card
> where I stored a copy of the HP-41CV emulator: when I switched the
> machine on after replug the card, I got an "Invalid Card Data" error.
> Everything was lost. To avoid the message, just store something on the
> card and press ON-C. The message should disapear.
> 
Replugging in the source 48GX does not work.
Nor interestingly, does the warning disappear when I store something 
on the card.
It still shows only 74k of memory available, when there should be 
about 127k after storing something small.

The fact that there is about 50K unaccounted for leads me to believe 
that the original programs may be recoverable.

But it appears I now need another 128K card on which to load 
Metakernel!

Anyone in the Toronto/Mississauga area with a spare card for sale, or 
who is willing to lend me a card for a day or two?

Geoff

"R. G. Newbury" <newburynospam@mandamus.org> wrote in message 
news:JcldVHe8EppF-pn2-NRMGBbPxtqdr@Tor2/Mandamus2...
>
> The fact that there is about 50K unaccounted for leads me to believe
> that the original programs may be recoverable.
>
> But it appears I now need another 128K card on which to load
> Metakernel!
>
> Anyone in the Toronto/Mississauga area with a spare card for sale, or
> who is willing to lend me a card for a day or two?
>


Maybe you could try to copy it to emu48?
That way you work on a copy and not on your only flaky original.

-- 
Steve Sousa 

R. G. Newbury wrote:
> 
> Replugging in the source 48GX does not work.
> Nor interestingly, does the warning disappear when I store something 
> on the card.
> It still shows only 74k of memory available, when there should be 
> about 127k after storing something small.

If you run PINIT or store an object on the card it will fix the card at 
hte last objects recognised valid.
So if you had any corrupted data *after* then it will be unaccounted for 
in the available space.

> 
> The fact that there is about 50K unaccounted for leads me to believe 
> that the original programs may be recoverable.
Sure. but it does indicate that there have been data corruption. How 
much and how serious: nobody can tell and IMO not worth it unless you 
must retrieve it and can't get it somewhere else.

JY

On Wed, 25 May 2005 13:33:22 -0500,
R. Geoff Newbury wrote:

> The fact that there is about 50K unaccounted for leads me
> to believe that the original programs may be recoverable.
> But it appears I now need another 128K card
> on which to load Metakernel!
> Anyone in the Toronto/Mississauga area with a spare card for sale,or who is willing to lend me a card for a day or two?

If any of the following programs work by mapping the entire
card space to transfer it to a computer, then there might be
some hope for preserving the entire binary card content,
after which you might be able to snip parts of it
using a hex editor; I have not looked any more closely
to see how each works, but you could try them all
on Emu48 first, or read any docs first:

http://www.hpcalc.org/search.php?query=backup&hp48=1

If you get hold of a spare 128K card (or buy a spare),
you could also copy your entire existing card before
playing further with it:

http://www.hpcalc.org/details.php?id=2394
http://www.hpcalc.org/hp48/utils/memory/copy.zip

Wait a minute -- this one might do the trick
(copies any whole card to a computer, it says):

http://www.hpcalc.org/details.php?id=2397
http://www.hpcalc.org/hp48/utils/memory/copyrom.zip

[r->] [OFF]

On Wed, 25 May 2005 19:25:39 UTC, Jean-Yves Avenard <me@privacy.net> 
wrote:

> R. G. Newbury wrote:
> > 
> > Replugging in the source 48GX does not work.
> > Nor interestingly, does the warning disappear when I store something 
> > on the card.
> > It still shows only 74k of memory available, when there should be 
> > about 127k after storing something small.
> 
> If you run PINIT or store an object on the card it will fix the card at 
> hte last objects recognised valid.
> So if you had any corrupted data *after* then it will be unaccounted for 
> in the available space.


Running PINIT does not change the state or status. Still the same 
(incomplete) memory reported.

Unfortunately.

 
> > The fact that there is about 50K unaccounted for leads me to believe 
> > that the original programs may be recoverable.
> Sure. but it does indicate that there have been data corruption. How 
> much and how serious: nobody can tell and IMO not worth it unless you 
> must retrieve it and can't get it somewhere else.

Well its not as if I must as in 'I must file my tax return', but I 
really would like to get it all back, and it does not exist elsewhere.
The coding represents quite a few hours of work, so I would consider 
spending $100 to get it all back. It does not exist anywhere else. 
This code is the result of hours of fiddling with the program while 
watching TV or in bed before turning out the light.

Geoff

R. G. Newbury wrote:
> I have a 128K HP RAM card which became "confused" during a transfer
> between my HP48GX calcs.
> On startup the calc displays 'Invalid Card Data' warning.
>
> I downloaded the program P0Fix but it reports that there is no
> corruption on the card.
>
> Problem: When I run PVARS on port 0, it returns an empty list and
> reports that the memory is just 75,000 bytes.
>
> Now that figure would be correct if the previously stored programs
> visibly existed. And I assume that they still do.
>
> Question:  How do I recover them?

But the card would be either port 1 or port 2, depending on which
slot you plug it in to. Port 0 is shared with "System RAM", so
0 PVARS should show you approximately the same amount of free
memory as MEM, with any difference being due to the last stack,
last arguments, and last command line saves. Try 1 PVARS or
2 PVARS, as appropriate.

I take it that you've tried a fresh battery in the card? Note that
the calculator should be turned on while changing the card
battery.

Do you have the same problem with the card in either slot of
either 48GX?

Manually slide back the connector cover on the card; do the metal
contact "fingers" look okay? No corrosion, gouges, worn through
metal, or foreign material?

Check the links that John posted.

Since PINIT didn't fix the problem, there's always the possibility
of a genuine hardware failure. See
http://stores.ebay.de/Klotz-electronic for a replacement. His
4x128kb card should work in either slot of either the 48SX or the
48GX, but only lets you access one 128KB bank without "freeing"
the port (if it was "merged"), turning off the calculator,
changing the switches, and turning the calculator back on. His
512kbyte, 1Mbyte, 2Mbyte, and 4Mbyte cards should work correctly
only in slot 2 of a 48GX (and can't be "merged" with system
memory), but except for the 4Mbyte card, don't require any
switching or rebooting to access all banks (each 128KB bank is a
separate port). For his 4Mbyte card, which two banks the 48GX see
as ports 30 and 31 depend on the switch setting, and I expect that
switching requires turning the calculator off, changing the
switch, and turning the calculator back on. At least, that's my
understanding; I don't actually have any of his cards. Anyone who
does have one, feel free to correct me.

By the way, it's a good idea to at least occasionally back up your
work to your PC, at least if you have anything on the calculator
or card that you really don't want to lose....

-- 
Regards,
James

On Mon, 30 May 2005 19:56:20 UTC, "James M. Prange" 
<jmprange@i-is.com> wrote:

> R. G. Newbury wrote:
> > I have a 128K HP RAM card which became "confused" during a transfer
> > between my HP48GX calcs.
> > On startup the calc displays 'Invalid Card Data' warning.
> >
> > I downloaded the program P0Fix but it reports that there is no
> > corruption on the card.
> >
> > Problem: When I run PVARS on port 0, it returns an empty list and
> > reports that the memory is just 75,000 bytes.
> >
> > Now that figure would be correct if the previously stored programs
> > visibly existed. And I assume that they still do.
> >
> > Question:  How do I recover them?
> 
> But the card would be either port 1 or port 2, depending on which
> slot you plug it in to. Port 0 is shared with "System RAM", so
> 0 PVARS should show you approximately the same amount of free
> memory as MEM, with any difference being due to the last stack,
> last arguments, and last command line saves. Try 1 PVARS or
> 2 PVARS, as appropriate.

The card is in slot 1. It is a 128K HP card.

You are right. When I do 0 PVARS, it shows me the P0FIX program (in a 
list) and gives a 74K memory figure. That number is within 30 bytes of
the number provided by MEM.

1 PVARS returns an empty list and "ROM" as its return objects.
Sorry to have mislead everyone on this.

I stored the missing programs in port 1. I know I did that because I 
could and did transfer them to the other calc by moving the card.

So... different question: why does it show an empty list and "ROM" for
port 1...?

(Looks like I still need to play with the same programs which have 
been suggested to look at the card.)

Geoff






> I take it that you've tried a fresh battery in the card? Note that
> the calculator should be turned on while changing the card
> battery.
> 
> Do you have the same problem with the card in either slot of
> either 48GX?
> 
> Manually slide back the connector cover on the card; do the metal
> contact "fingers" look okay? No corrosion, gouges, worn through
> metal, or foreign material?
> 
> Check the links that John posted.
> 
> Since PINIT didn't fix the problem, there's always the possibility
> of a genuine hardware failure. See
> http://stores.ebay.de/Klotz-electronic for a replacement. His
> 4x128kb card should work in either slot of either the 48SX or the
> 48GX, but only lets you access one 128KB bank without "freeing"
> the port (if it was "merged"), turning off the calculator,
> changing the switches, and turning the calculator back on. His
> 512kbyte, 1Mbyte, 2Mbyte, and 4Mbyte cards should work correctly
> only in slot 2 of a 48GX (and can't be "merged" with system
> memory), but except for the 4Mbyte card, don't require any
> switching or rebooting to access all banks (each 128KB bank is a
> separate port). For his 4Mbyte card, which two banks the 48GX see
> as ports 30 and 31 depend on the switch setting, and I expect that
> switching requires turning the calculator off, changing the
> switch, and turning the calculator back on. At least, that's my
> understanding; I don't actually have any of his cards. Anyone who
> does have one, feel free to correct me.
> 
> By the way, it's a good idea to at least occasionally back up your
> work to your PC, at least if you have anything on the calculator
> or card that you really don't want to lose....
> 

R. G. Newbury wrote:
> On Mon, 30 May 2005 19:56:20 UTC, "James M. Prange"
> <jmprange@i-is.com> wrote:
>
>
>>R. G. Newbury wrote:
>>
>>>I have a 128K HP RAM card which became "confused" during a transfer
>>>between my HP48GX calcs.
>>>On startup the calc displays 'Invalid Card Data' warning.
>>>
>>>I downloaded the program P0Fix but it reports that there is no
>>>corruption on the card.
>>>
>>>Problem: When I run PVARS on port 0, it returns an empty list and
>>>reports that the memory is just 75,000 bytes.
>>>
>>>Now that figure would be correct if the previously stored programs
>>>visibly existed. And I assume that they still do.
>>>
>>>Question:  How do I recover them?
>>
>>But the card would be either port 1 or port 2, depending on which
>>slot you plug it in to. Port 0 is shared with "System RAM", so
>>0 PVARS should show you approximately the same amount of free
>>memory as MEM, with any difference being due to the last stack,
>>last arguments, and last command line saves. Try 1 PVARS or
>>2 PVARS, as appropriate.
>
>
> The card is in slot 1. It is a 128K HP card.
>
> You are right. When I do 0 PVARS, it shows me the P0FIX program (in a
> list) and gives a 74K memory figure. That number is within 30 bytes of
> the number provided by MEM.
>
> 1 PVARS returns an empty list and "ROM" as its return objects.
> Sorry to have mislead everyone on this.
>
> I stored the missing programs in port 1. I know I did that because I
> could and did transfer them to the other calc by moving the card.

So, do you now have the programs on the other calculator?

Does the card work in one calculator but not the other?

> So... different question: why does it show an empty list and "ROM" for
> port 1...?

Normally, the empty list means that no objects are stored on the
card, and "ROM" means that the switch (on the top edge of the
card) is in the "write protect" position (slid toward the edge of
the card).

I think that "ROM" explains why PINIT has no effect; PINIT
normally stores (and removes) an object in each free RAM port, but
it would have to skip ROM ports. If there's nothing (corrupted or
not) on the card that you want to try to salvage, then slide the
switch toward the center of the card and run PINIT again. On the
other hand, if you think that there's still something on the card
that you want to salvage, having it write-protected may not be
such a bad thing; at least you won't make the situation worse by
overwriting whatever is there.

<snip>

-- 
Regards,
James

On Tue, 31 May 2005 23:44:48 UTC, "James M. Prange" 
<jmprange@i-is.com> wrote:

> > I stored the missing programs in port 1. I know I did that because I
> > could and did transfer them to the other calc by moving the card.
> 
> So, do you now have the programs on the other calculator?

NO
 
> Does the card work in one calculator but not the other?

Previously (3 months ago) itworked in both. Then it glitched during 
transfer and gave an Invalid error whichever calc it was in. 

Now I wonder whether in fact I got an old battery when I changed the 
card battery about 9 months ago.

> > So... different question: why does it show an empty list and "ROM" for
> > port 1...?
> 
> Normally, the empty list means that no objects are stored on the
> card, and "ROM" means that the switch (on the top edge of the
> card) is in the "write protect" position (slid toward the edge of
> the card).

Well, I've been misreading that funny little graphic under the switch.
I thought that was the R/W position.

Now  1 PVARS give { } and 131072.


Looks like I may be entirely out of luck. Looks like I have lost 
everything which was on the card. AND because of the glitch betweeen 
the card and calc, I've lost the programs from the calc too...


> I think that "ROM" explains why PINIT has no effect; PINIT
> normally stores (and removes) an object in each free RAM port, but
> it would have to skip ROM ports. If there's nothing (corrupted or
> not) on the card that you want to try to salvage, then slide the
> switch toward the center of the card and run PINIT again. On the
> other hand, if you think that there's still something on the card
> that you want to salvage, having it write-protected may not be
> such a bad thing; at least you won't make the situation worse by
> overwriting whatever is there.

I'm toast...
Thanks for the help everyone. I've re-read the manual a few times now 
and have a little better understanding of things..
Not that helps a lot..

Geoff

R. G. Newbury wrote:
> On Tue, 31 May 2005 23:44:48 UTC, "James M. Prange"
> <jmprange@i-is.com> wrote:
>
>
>>>I stored the missing programs in port 1. I know I did that because I
>>>could and did transfer them to the other calc by moving the card.
>>
>>So, do you now have the programs on the other calculator?
>
>
> NO
>
>
>>Does the card work in one calculator but not the other?
>
>
> Previously (3 months ago) itworked in both. Then it glitched during
> transfer and gave an Invalid error whichever calc it was in.
>
> Now I wonder whether in fact I got an old battery when I changed the
> card battery about 9 months ago.

That reminds me of an unfortunate event the first time that I
replaced "low" N cells in my 28S with brand new fresh off of the
shelf batteries. Those batteries must've been in storage for a
very long time. I did have "backups" for the 28S, but they were
printouts. Since then, I'm in the the habit of testing every
battery before I put it in a calculator.

But actually, your calculator should've given you a warning if the
battery on the card were low.

>>>So... different question: why does it show an empty list and "ROM" for
>>>port 1...?
>>
>>Normally, the empty list means that no objects are stored on the
>>card, and "ROM" means that the switch (on the top edge of the
>>card) is in the "write protect" position (slid toward the edge of
>>the card).
>
>
> Well, I've been misreading that funny little graphic under the switch.
> I thought that was the R/W position.

I have the same problem with it; I have to try it to see which way
gives me "ROM", sort of like the "bar" and "circle" for on/off
switches; couldn't they simply print "on" and "off" in plain
English? Silly little icons give me some idea of how Egyptologists
must've felt before the Rosetta stone was discovered.

> Now  1 PVARS give { } and 131072.

And if you put the card in slot 2, then 2 PVARS should return the
same, and it should be the same on both calculators.

> Looks like I may be entirely out of luck. Looks like I have lost
> everything which was on the card. AND because of the glitch betweeen
> the card and calc, I've lost the programs from the calc too...

It looks that way to me too.

>>I think that "ROM" explains why PINIT has no effect; PINIT
>>normally stores (and removes) an object in each free RAM port, but
>>it would have to skip ROM ports. If there's nothing (corrupted or
>>not) on the card that you want to try to salvage, then slide the
>>switch toward the center of the card and run PINIT again. On the
>>other hand, if you think that there's still something on the card
>>that you want to salvage, having it write-protected may not be
>>such a bad thing; at least you won't make the situation worse by
>>overwriting whatever is there.
>
>
> I'm toast...
> Thanks for the help everyone. I've re-read the manual a few times now
> and have a little better understanding of things..
> Not that helps a lot..

Look at the bright side of it. Now you get to make a fresh start
with a clean slate and all that....

-- 
Regards,
James

On Thu, 2 Jun 2005 07:23:17 UTC, "James M. Prange" <jmprange@i-is.com>
wrote:

> R. G. Newbury wrote:
> > On Tue, 31 May 2005 23:44:48 UTC, "James M. Prange"
> > <jmprange@i-is.com> wrote:
>> >
> > Now I wonder whether in fact I got an old battery when I changed the
> > card battery about 9 months ago.
> 
> That reminds me of an unfortunate event the first time that I
> replaced "low" N cells in my 28S with brand new fresh off of the
> shelf batteries. Those batteries must've been in storage for a
> very long time. I did have "backups" for the 28S, but they were
> printouts. Since then, I'm in the the habit of testing every
> battery before I put it in a calculator.
> 
> But actually, your calculator should've given you a warning if the
> battery on the card were low.

There was no warning and it was probably the battery. I put a program 
on the card, and pulled it from the calc for 24 hours.
The program was not there when re-inserted.
The card still gives the 'Invalid Data' warning on startup of the 
calc, even though the card is now entirely empty...I guess because 
there is now no marker for the end of used memory...

So that is the actual explanation: low battery....Not quite low enough
to trigger the calc to give a warning, but low enough to lose the 
stored data when removed from the calc.... Worst of all possible modes
of failure..Had the batt warning come on, then of course I would have 
never moved the card.

> > Well, I've been misreading that funny little graphic under the switch.
> > I thought that was the R/W position.
> 
> I have the same problem with it; I have to try it to see which way
> gives me "ROM", sort of like the "bar" and "circle" for on/off
> switches; couldn't they simply print "on" and "off" in plain
> English? Silly little icons give me some idea of how Egyptologists
> must've felt before the Rosetta stone was discovered.

Well even when I know what its supposed to mean, I have to stretch to 
read that icon that way.  <Comments regarding the mental acuity, 
heritage, marital status of parents, etc. excised...>

> > I'm toast...
> > Thanks for the help everyone. I've re-read the manual a few times now
> > and have a little better understanding of things..
> > Not that helps a lot..
> 
> Look at the bright side of it. Now you get to make a fresh start
> with a clean slate and all that....

I believe you are misrepresenting yourself on this news list. Surely 
you are a 12 year old girl named Pollyanna!


Geoff

 

"R. G. Newbury" <newburynospam@mandamus.org> wrote in message 
news:JcldVHe8EppF-pn2-F44vKj2UsxLO@Tor2/Mandamus2...
> So that is the actual explanation: low battery....Not quite low enough
> to trigger the calc to give a warning, but low enough to lose the
> stored data when removed from the calc.... Worst of all possible modes
> of failure...

That's not the explanation, the explanation is: you have a defective or 
badly designed card.

>> > Well, I've been misreading that funny little graphic under the switch.
>> > I thought that was the R/W position.

Original cards have a very clear simbol, with a single pointed or double 
pointed arrow, others i've seen have a padlock, what symbol does your card 
have?


Best Regards.

-- 
Steve Sousa 

On Thu, 02 Jun 2005 02:23:17 -0500:

> your calculator should've given you a warning
> if the battery on the card were low.

On the 48SX/GX you need to plug in the card and turn the calc ON
before changing the card battery (at which time the calc
can supply power to the card, even while the card battery
is removed).

However, the calc only warns of low card battery
when the calc is OFF and then is turned back ON
(unless an ON+C warmstart can also do it).

Therefore, if you insert a dead battery into the card,
you won't get a "low card battery" warning until
after you have first turned the calc OFF,
at which time you might already lose your card contents,
unless ON+C also tests the card -- does anyone
recall whether it does? -- if so, then be sure
to do ON+C immediately after inserting a new card battery,
while the calc is still on, as it should have been
while changing the battery.


Certain RAM cards which were not engineered especially
for 48SX/GX may also short out their inputs if the
calc batteries run low, and might then fry
the calc's own power supply -- that was the incidental
result of adapting memory cards originally designed
for *typewriters* for use in HP calculators.

HP had its own supplier (Epson) modify the HP brand cards
to match the lower voltage of the calc power supplies,
but when some users obtained "plug compatible" typewriter
or computer ("Grid") cards and used them in their calcs,
they occasionally destroyed their calc
when the "calc vs. card" voltages got out of balance.

When I was a kid, I once also noticed that my fingers
woud easily fit into the light bulb socket of a lamp;
that was my earliest lesson in watching out
what I plugged into what, just because it fit :)

[r->] [OFF]

Steve Sousa wrote:

<snip>

> Original cards have a very clear simbol, with a single pointed or double
> pointed arrow, others i've seen have a padlock, what symbol does your card
> have?

Yes, my "original" cards (marketed for the 48SX) are clear enough.
Trying a little ASCII art, the symbols below the switch look sort
of like this:

        | |
        | |
        | |
        | |
    /\         /\
   /  \       /  \
/___o  \   /___\  \
\      /   \   /  /
   \  /       \  /
    \/         \/

It's not too difficult to realize that they're supposed to mean
that with the switch to the left, communication with the card is
one-way (read-only), and with the switch to the right,
communication is two-way (read/write).

But when they redesigned the artwork about the time that they
released the 48GX, the symbols were changed to something like:

    /\
   /  \
/___o  \
\      /
   \  /
    \/

/____
\

Not only are both arrows single-headed now, but they're arranged
one over the other, instead of side-by-side. What in the world is
that supposed to mean? I can only guess that whoever designed the
new artwork had no clue as to what the symbols were on the card
for.

Of course, with both styles, the symbols are on the side of the
card that can't be seen with the card in the calculator, but in
that case, you can always use the PVARS command to see whether a
RAM card is write-protected.

Of course non-HP cards may well use a variety of different
symbols.

Personally, I rather like the open and closed padlocks as symbols
for read/write and write-protected.

-- 
Regards,
James

P.S.:

The "ASCII art" is better viewed with a "fixed" font; it looks
even worse with a "proportional" font.

-- 
Regards,
James

James M. Prange <jmprange@i-is.com> wrote:
> 
> Of course, with both styles, the symbols are on the side of the
> card that can't be seen with the card in the calculator, but in
> that case, you can always use the PVARS command to see whether a
> RAM card is write-protected.
> 
> Of course non-HP cards may well use a variety of different
> symbols.

One of the things I like about the Klotz cards is that the switch has one
position marked ROM and the other marked RAM, and it's on the side of the
card that can be read while it's in the calculator.  MetaKernel also has
a nice on-screen indicator that shows the status of the card in slot 2.
(Slot 1, of course, contains the MetaKernel card itself, which normally
would be a ROM card, so I guess it was felt that there was no need to
have an on-screen indicator for that slot.)

-- 
Wayne Brown  (HPCC #1104)  | "When your tail's in a crack, you improvise
fwbrown@bellsouth.net      |  if you're good enough.  Otherwise you give
                           |  your pelt to the trapper."
"e^(i*pi) = -1"  -- Euler  |           -- John Myers Myers, "Silverlock"

John H Meyers <jhmeyers@miu.edu> wrote:
> 
> However, the calc only warns of low card battery
> when the calc is OFF and then is turned back ON
> (unless an ON+C warmstart can also do it).
> 
> Therefore, if you insert a dead battery into the card,
> you won't get a "low card battery" warning until
> after you have first turned the calc OFF,
> at which time you might already lose your card contents,
> unless ON+C also tests the card -- does anyone
> recall whether it does? -- if so, then be sure
> to do ON+C immediately after inserting a new card battery,
> while the calc is still on, as it should have been
> while changing the battery.

I just removed the battery from one of my cards (with the calc on) and
did an ON+C warmstart.  It definitely *does* give the "LowBat" warning.

-- 
Wayne Brown  (HPCC #1104)  | "When your tail's in a crack, you improvise
fwbrown@bellsouth.net      |  if you're good enough.  Otherwise you give
                           |  your pelt to the trapper."
"e^(i*pi) = -1"  -- Euler  |           -- John Myers Myers, "Silverlock"

If you can't decipher the card switch settings:

On a 48SX/GX, to determine whether a card in port 1 is writeable,
just do 1 PVARS; if the card is either ROM or write-protected RAM,
the calc will return the string "ROM" on level-1;
otherwise it will return the amount of writeable free space.

[r->][OFF]

Hi,

using SysRPL, you can also use GETBVARS,
and in ML, you can read out the port status (merged, readable, writeable,
number of banks) directly.
For this the three entries starting at
=CONFTAB (PORT0EOS, PORT1EOS, PORT2EOS)
are the relevant ones.


Raymond

"John H Meyers" <jhmeyers@miu.edu> schrieb im Newsbeitrag
news:op.srwzwyw9wmwqgg@news.cis.dfn.de...
> If you can't decipher the card switch settings:
>
> On a 48SX/GX, to determine whether a card in port 1 is writeable,
> just do 1 PVARS; if the card is either ROM or write-protected RAM,
> the calc will return the string "ROM" on level-1;
> otherwise it will return the amount of writeable free space.
>
> [r->][OFF]

On Thu, 2 Jun 2005 21:15:43 UTC, "Steve Sousa" <etsteve@yahoo.com> 
wrote:

> "R. G. Newbury" <newburynospam@mandamus.org> wrote in message 
> news:JcldVHe8EppF-pn2-F44vKj2UsxLO@Tor2/Mandamus2...
> > So that is the actual explanation: low battery....Not quite low enough
> > to trigger the calc to give a warning, but low enough to lose the
> > stored data when removed from the calc.... Worst of all possible modes
> > of failure...
> 
> That's not the explanation, the explanation is: you have a defective or 
> badly designed card.

The card is an HP card.... its about 12-13 years old now. And I have 
changed the battery only about 3 times since it lives in the calc.

> >> > Well, I've been misreading that funny little graphic under the switch.
> >> > I thought that was the R/W position.
> 
> Original cards have a very clear simbol, with a single pointed or double 
> pointed arrow, others i've seen have a padlock, what symbol does your card 
> have?

It's one of the early ones with the not very clear single/double 
pointed arrow. In retrospect, there was no reason that some actual 
explanation in English could not have been printed on the card...

Geoff

On Fri, 3 Jun 2005 10:22:45 UTC, "James M. Prange" <jmprange@i-is.com>
wrote:

> > Original cards have a very clear simbol, with a single pointed or double
> > pointed arrow, others i've seen have a padlock, what symbol does your card
> > have?

> It's not too difficult to realize that they're supposed to mean
> that with the switch to the left, communication with the card is
> one-way (read-only), and with the switch to the right,
> communication is two-way (read/write).

Well I completely misread that.

 
> But when they redesigned the artwork about the time that they
> released the 48GX, the symbols were changed to something like:
> 
>     /\
>    /  \
> /___o  \
> \      /
>    \  /
>     \/
> 
> /____
> \
> 
> Not only are both arrows single-headed now, but they're arranged
> one over the other, instead of side-by-side. What in the world is
> that supposed to mean? I can only guess that whoever designed the
> new artwork had no clue as to what the symbols were on the card
> for.

That sounds likely.... They turned the outside printing over to the 
graphic artists.


> Of course, with both styles, the symbols are on the side of the
> card that can't be seen with the card in the calculator, but in
> that case, you can always use the PVARS command to see whether a
> RAM card is write-protected.

Well of course! the symbols are printed on the side you cannot see! To
do otherwise would be.....too obvious???

> Personally, I rather like the open and closed padlocks as symbols
> for read/write and write-protected.

Much more understandable...

Geoff
 

On Sun, 5 Jun 2005 20:44:48 UTC, "John H Meyers" <jhmeyers@miu.edu> 
wrote:

> If you can't decipher the card switch settings:
> 
> On a 48SX/GX, to determine whether a card in port 1 is writeable,
> just do 1 PVARS; if the card is either ROM or write-protected RAM,
> the calc will return the string "ROM" on level-1;
> otherwise it will return the amount of writeable free space.

In retrospect, this *also* is a failure of imagination/understanding 
on the part of HP's engineers. I saw that display. It did NOT tell me 
that the RAM was write protected. It actually made no sense to me. 
What would have been wrong with reporting 'Write-protected' and 
reporting the *remaining, not at this time available RAM*?

Geoff

 

On Thu, 2 Jun 2005 23:43:47 UTC, "John H Meyers" <jhmeyers@miu.edu> 
wrote:

> On Thu, 02 Jun 2005 02:23:17 -0500:
> 
> > your calculator should've given you a warning
> > if the battery on the card were low.
> 
> On the 48SX/GX you need to plug in the card and turn the calc ON
> before changing the card battery (at which time the calc
> can supply power to the card, even while the card battery
> is removed).
 
> However, the calc only warns of low card battery
> when the calc is OFF and then is turned back ON
> (unless an ON+C warmstart can also do it).

Yes, but it never did so. The night before all this happened I had 
been fiddling with a program while in bed. I decided that I would make
a backup on the old calculator. (As noted, I have been having problems
getting the calcs to talk to my laptop: it will only do so under DOS 
6.22....no idea why, although I suspect that THinkpad configuration 
program sets things in the BIOS that operating system level programs 
do not reach...The symptoms seem to imply a mismatch of parity or flow
control settings....)

When I got to work, I pulled the old calc out of the drawer, put 
batteries in it and turned it on. I pulled the new one out of my 
briefcase and out of its case and turned it on. NO messages.

I made a backup of the program ( 'prog', RCL, ':1:prog', STO). 
Shutdown, pulled the card. Turned off the old calc, inserted the card,
turned on the old calc... 'Invalid Card Data' error.
Shutdown, transferred the card back to the new calc.
Turned on new calc....... 'Memory Lost...Recover Memory?'....

'Yes' brought back some programs, but of course (Murphy's Second Law 
in action) not the one I really wanted to have....

So the exact mechanism of what happened can be the item of some 
conjecture...
But the result is unfortunately all too clear.

Geoff

R. G. Newbury wrote:
> On Sun, 5 Jun 2005 20:44:48 UTC, "John H Meyers" <jhmeyers@miu.edu>
> wrote:
>
>
>>If you can't decipher the card switch settings:
>>
>>On a 48SX/GX, to determine whether a card in port 1 is writeable,
>>just do 1 PVARS; if the card is either ROM or write-protected RAM,
>>the calc will return the string "ROM" on level-1;
>>otherwise it will return the amount of writeable free space.
>
>
> In retrospect, this *also* is a failure of imagination/understanding
> on the part of HP's engineers.

Maybe so.

>  I saw that display. It did NOT tell me
> that the RAM was write protected. It actually made no sense to me.

It seems to me like a failure of imagination/understanding on your
part.

To be sure, "RAM" stands for "Random-Access Memory" (which could
be either read-write or read-only) and "ROM" stand for "Read-Only
Memory" (which could be either serial-access or random-access),
but as usually used, both RAM and ROM are random-access, with the
differences being that RAM is read-write, volatile, and usually
faster, and ROM is always read-only, and usually non-volatile.
Surely "write-protected RAM" is effectively ROM, at least as long
as the switch is in the write-protect position and the battery is
supplying power?

> What would have been wrong with reporting 'Write-protected' and
> reporting the *remaining, not at this time available RAM*?

Perhaps it has something to do with that famous "Life is short and
ROM is full." answer from Bill Wickes. Or maybe they just didn't
realize that someone wouldn't understand it, particularly after
reading the fine manual.

-- 
Regards,
James

Do remember that the HP48SX/GX adopted Epson plug-ins
originally created for electronic *typewriter* memory,
so don't expect miracles from them!

For further details from HP maestro Dave Arnett, see:
"6.25 How does RAM card locking work?"
http://www.hpcalc.org/hp48/docs/faq/48faq-6.html

"The HP CPU chip detects the write protect indicator line
  of the card.  If it is reporting as unwriteable,
  the hardware locks out write functionality from the CPU chip.
  If this Card Detect line reports as writeable,
  then write transactions are permitted in hardware.
  If the Card Detect line floats,
  the hardware reports that no card is installed."

NOTE: there is only ONE signal line to check,
but it serves to report THREE possible states!
(writeble, not writeable, or no card present at all).

Sorry, there is evidently no possible way to distinguish
true ROM from write-protected RAM, so don't blame the calculator
for not being able to tell you anything more about the internal
hardware which it can not see into!

Interpret "ROM" on the stack as meaning literally "Read-Only Memory,"
with write-protection indeed creating that same condition, at least
as far as anything that the calculator itself can do about it.

[r->] [OFF]