A Mac, a Palm, and a suddenly-blank SD card?

Archived from groups: uk.comp.sys.mac,comp.sys.palmtops.pilot (More info?)

Please note crossposts - trim if required.

I noticed the SD card had popped out of my Palm Tungsten E while in my
pocket. I popped it back in and got the dialog asking whether I want to
format the card. Odd, it shouldn't do that.

I slotted it into an SD card reader on my Mac (running OSX 10.3.5) and
everything appeared fine - all my data was on it.

Popped it back in the Palm; this time it recognised that it was already
formatted, but a) the card name was corrupted (odd characters, etc) and
b) the launcher didn't display any of the applications that were on the
card. I didn't check any further - eg using Filez.

Back into the Mac, run Disk Utility on it. Loads and loads of errors [1]
then it said it was fine and fixed. Didn't actually check the card, just
ejected it from Disk Utility.

Back in the Palm, and the disk name was OK, but launcher showed nothing
again.

Back in the Mac, and it showed a corrupt file at the root (which
vanished when clicked) and an empty folder structure consisting of
PALM/Launcher/


So.... the info on there wasn't that important so I'm not too fussed
about recovering it (it's just a mirror backup of the Palm and a bunch
of eBooks), but what I would like to know is:

1. What was the likely cause of this? I use TealBackup to make auto
backups every night (and was awake to see the last one appear to work
OK), had a few applications on there (just copies of stuff on the Palm)
and a couple of eBooks that I was halfway through using Weasel. I can't
really see anything that may have caused it - might the physical act of
popping the card in my pocket have shorted something? Should I send the
card back to be replaced?

2. Next time this happens, what could I do to repair it? Is it
possible/likely that Disk Utility screwed things up further because it's
handling a non-HFS+ volume? I missed out a few obvious steps (partly
because I didn't expect anything to work and wasn't that bothered about
the data) but are there different tools I should have used?


Thankyou.

-zoara-





[1] snipped version - removed loads of similar entries (just different
cluster numbers):

Repairing disk for "TUESDAY"
** /dev/disk3s1
** Phase 1 - Read FAT
Cluster 53472 in FAT 0 continues with out of range cluster number 1
Truncate? yes
** Phase 2 - Check Cluster Chains
Cluster chains starting at 235 and 53471 are linked at cluster 256
Clear chain starting at 235? yes
Clear chain starting at 53471? yes
Cluster chains starting at 43587 and 53503 are linked at cluster 43605
Clear chain starting at 43587? yes
Clear chain starting at 53503? yes
Cluster chains starting at 49351 and 53477 are linked at cluster 49376
Cluster chains starting at 49351 and 53477 are linked at cluster 49376
in starting at 49351? yes
Clear chain starting at 53477? yes
Cluster chain starting at 52908 ends with cluster marked free
Clear chain starting at 52908? yes
Cluster chains starting at 53474 and 0 are linked at cluster 49376
Clear chain starting at 53474? yes
Clear chain starting at 0? yes
Cluster chain starting at 53475 ends with cluster marked free
Clear chain starting at 53475? yes
** Phase 3 - Checking Directories
/Launcher starts with cluster marked as EOF
Remove? yes
Remove? yes
ory
Truncate? yes
** Phase 4 - Checking for Lost Files
Lost cluster chain at cluster 2
1 Cluster(s) lost
Reconnect? yes
No LOST.DIR directory
Clear? yes
Lost cluster chain at cluster 4
37 Cluster(s) lost
Reconnect? yes
No LOST.DIR directory
Clear? yes
Lost cluster chain at cluster 61
174 Cluster(s) lost
Reconnect? yes
No LOST.DIR directory

[snip]

Clear? yes
Lost cluster chain at cluster 60768
162 Cluster(s) lost
Reconnect? yes
No LOST.DIR directory
Clear? yes
Clear? yes
at cluster 60942
193 Cluster(s) lost
Reconnect? yes
No LOST.DIR directory
Clear? yes
Lost cluster chain at cluster 61135
215 Cluster(s) lost
Reconnect? yes
No LOST.DIR directory
Clear? yes
Update FATs? yes
4 files, 122728 free (61364 clusters)
** Phase 1 - Read FAT
** Phase 2 - Check Cluster Chains
** Phase 3 - Checking Directories
Invalid long filename entry for /.Trashes
Remove? yes
** Phase 4 - Checking for Lost Files
4 files, 122728 free (61364 clusters)

***** FILE SYSTEM WAS MODIFIED *****

Repair attempted on 1 volume
1 non HFS volume repaired

--
The iBook saga continues... last updated Thu 7/10/04 (day 73)
Could it all finally be over?
http://ibooksaga.blogspot.com/

Archived from groups: uk.comp.sys.mac,comp.sys.palmtops.pilot (More info?)

In article <BD98ACF5.52C47%chrisridd@mac.com>,
Chris Ridd <chrisridd@mac.com> wrote:

> On 17/10/04 10:18 pm, in article 1gltaqy.1841hj31d8gns6N%me3@privacy.net,
> "zoara" <me3@privacy.net> wrote:
>
> > 2. Next time this happens, what could I do to repair it?
> I'd be tempted to just mount the card on the Mac to recover any files you
> wanted, and to use the Palm to reformat it.

Ditto to Chris' adice! Anytime you have a storage media that displays
problems and then are subsequently able to mount it --- do yourself a
big favor and copy your files off it while you've been given that 2nd
chance! [ I learned that the hard way with a flaky hard drive. ]

Then by all means reformat the storage media. In this case, I'd do that
on the Palm, since that's the native 'home' of your SD card.

As to how it got corrupted: it could be a program on the Palm, it could
be related to how the card spontaneously unseated itself from the T|E -
maybe the card got pushed in and out intermittently: each SD card insert
event normally powers up your Palm.

Finally, there are some free SD card performance programs for the Palm
-- running one of those for some continous cycles would be a good way to
confirm that your reformated SD card is reliably being written to and
read from.

Archived from groups: uk.comp.sys.mac,comp.sys.palmtops.pilot (More info?)

On 2004-10-17 23:18:19 +0200, me3@privacy.net (zoara) said:

> So.... the info on there wasn't that important so I'm not too fussed
> about recovering it (it's just a mirror backup of the Palm and a bunch
> of eBooks),

It's odd, isn't it, that we use as backup the flakiest of storage
media! SD cards seem to fail regularly, and the Palm itself rarely
(knock on wood!!!)

J;

--
Encrypted e-mail address. Click to mail me:
http://cerbermail.com/?nKYh3qN4YG

Archived from groups: uk.comp.sys.mac,comp.sys.palmtops.pilot (More info?)

zoara <me3@privacy.net> wrote:

> I noticed the SD card had popped out of my Palm Tungsten E while in my
> pocket. I popped it back in and got the dialog asking whether I want to
> format the card. Odd, it shouldn't do that.

Exactly the same thing happened to a friend of mine, exept he runs XP at
home. I wonder if the card was being written to at the time of the
eject?


Gwynne
--
My real email is net, not line.

Archived from groups: uk.comp.sys.mac,comp.sys.palmtops.pilot (More info?)

On 18/10/04 6:34 pm, in article 2tid53F2070hcU1@uni-berlin.de, "Jeremy
Henderson" <me@privacy.net> wrote:

> suggested that they have a limited life, so in terms of USENET logic
> that must mean that they are "flaky" :-)

LOL.

Cheers,

Chris

Archived from groups: uk.comp.sys.mac,comp.sys.palmtops.pilot (More info?)

Chris Ridd <chrisridd@mac.com> wrote:

> On 17/10/04 10:18 pm, in article 1gltaqy.1841hj31d8gns6N%me3@privacy.net,
> "zoara" <me3@privacy.net> wrote:
>
> > 2. Next time this happens, what could I do to repair it? Is it
> > possible/likely that Disk Utility screwed things up further because it's
> > handling a non-HFS+ volume? I missed out a few obvious steps (partly
> > because I didn't expect anything to work and wasn't that bothered about
> > the data) but are there different tools I should have used?
>
> As you could see from the snipped output, Disk Utility copes just fine with
> non-HFS+ volumes. Under the hood it calls a different program to check and
> make filesystems for each different "format" it supports.

Well, in the log it *appears* to work. But does it really?

> But there's a chance that Palm's idea of a FAT filesystem is slightly
> different from Apple/FreeBSD's, and that repairing the card on the Mac
> screwed it on the Palm.

That's what I meant. It's Bill&co who decide what's 'real' FAT and what
isn't, right? Is all the other stuff based on officially released specs
or is it reverse-engineered?

> I'd be tempted to just mount the card on the Mac to recover any files you
> wanted, and to use the Palm to reformat it.

Yes. That would have been the sensible thing to do, wouldn't it? After
this and the other responses suggesting that, I slapped my forehead in a
comedy "silly me" style.

-z-


--
The iBook saga continues... last updated Thu 7/10/04 (day 73)
Could it all finally be over?
http://ibooksaga.blogspot.com/

Archived from groups: uk.comp.sys.mac,comp.sys.palmtops.pilot (More info?)

bt schmidt <bt_schmidt@hotmail.com> wrote:

> As to how it got corrupted: it could be a program on the Palm, it could
> be related to how the card spontaneously unseated itself from the T|E -
> maybe the card got pushed in and out intermittently: each SD card insert
> event normally powers up your Palm.

I'd noticed that. Damned annoying. Or rather, it's damned annoying how
easily the SD card unseats on a T|E.

> Finally, there are some free SD card performance programs for the Palm
> -- running one of those for some continous cycles would be a good way to
> confirm that your reformated SD card is reliably being written to and
> read from.

Thankyou. I will try one of them out to see.

-z-

--
The iBook saga continues... last updated Thu 7/10/04 (day 73)
Could it all finally be over?
http://ibooksaga.blogspot.com/

Archived from groups: uk.comp.sys.mac,comp.sys.palmtops.pilot (More info?)

On 18/10/04 10:01 pm, in article 1glv7j2.y3q3hb1r44ztiN%me3@privacy.net,
"zoara" <me3@privacy.net> wrote:

> Chris Ridd <chrisridd@mac.com> wrote:
>
>> As you could see from the snipped output, Disk Utility copes just fine with
>> non-HFS+ volumes. Under the hood it calls a different program to check and
>> make filesystems for each different "format" it supports.
>
> Well, in the log it *appears* to work. But does it really?

The source code's around somewhere, and the logging had terms in it that I
remember from my days playing with FAT filesystems (on the Atari ST!), so
I'd say yes.

>> But there's a chance that Palm's idea of a FAT filesystem is slightly
>> different from Apple/FreeBSD's, and that repairing the card on the Mac
>> screwed it on the Palm.
>
> That's what I meant. It's Bill&co who decide what's 'real' FAT and what
> isn't, right? Is all the other stuff based on officially released specs
> or is it reverse-engineered?

I'm not sure. MS recently got their attempt to patent FAT thrown out, so
that might suggest the specs are in the public domain.

That doesn't stop anyone from implementing it wrong, or omitting something
because "nothing else will ever access the card", or ...

>> I'd be tempted to just mount the card on the Mac to recover any files you
>> wanted, and to use the Palm to reformat it.
>
> Yes. That would have been the sensible thing to do, wouldn't it? After
> this and the other responses suggesting that, I slapped my forehead in a
> comedy "silly me" style.

Accompanied by a "Doh!" :-)

Cheers,

Chris

Archived from groups: uk.comp.sys.mac,comp.sys.palmtops.pilot (More info?)

On Sun, 17 Oct 2004 22:18:19 +0100, zoara wrote
(in article <1gltaqy.1841hj31d8gns6N%me3@privacy.net> ):

> 1. What was the likely cause of this?

I have friends in VERY high places ;-)