Mailinglist Archive: opensuse (818 mails)

< Previous Next >
Re: [opensuse] Re: Howto repair - corrupt jpeg file - anything new out there?
On 9/21/2011 3:35 PM, Philipp Thomas wrote:
On Wed, 21 Sep 2011 08:58:58 +0200, jdd<jdd@xxxxxxxxx> wrote:

don't forget there are at least 3 fat version (12, 16& 32)

Yes, but that doesn't contradict what I wrote. As far as you format
the drive according to what variant the camera expects it shouldn't
matter who did the formatting.

Philipp

Since when does "should" mean anything in computers? Or anything else technical?

You are really missing the point of the original recommendation I think.

When someone actually asks for advise for something like this at all, they have by doing so, advertised that they by definition would prefer, out of all the theoretical possible ways to do something, the simplest method that is the most likely to always work with the least requirement for them to be responsible for doing something the right way every time. The fewest and simplest rules to follow. The fewest opportunities for a possible mistake or unforeseen issue.

That is absolutely not "Eh just use whatever, it should usually work fine as long as everyone involved in the various different development teams for camera and the desktop software all did what they were supposed to." Yes that's true, but it doesn't come close to comparing to "Let the camera do it. That way it is either guaranteed to work perfectly, or you can safely know either the camera or the card is defective and return it for a new one or another model."

Unlikelihood is often a broken and wrong way to think in computer/tech/science areas. What matters is possible vs not-possible. Likelihood only matters in comparison to some other likelihood to judge which of two or more options presents the least or most likelihood. Then it's still a clear either/or choice, maximal vs not-maximal, if you can't have possible vs not-possible.

If you format using a pc, it's possible to have a problem. It's also possible for the problem to be your fault.

If you format using the device that will use the card most, then it is least "likely" for it to go wrong, but more importantly, it's not-possible for it to be your fault. So if it craps out, at least you have already done the biggest debug step already, it can't possibly be something wrong with your pc or software or your skill or awake-that-day-ness in using them.

There is no such thing as perfection or absolute certainty of course, but there is very clearly a difference in robustness of approaches. There is a maximize-chances vs don't-bother-to-maximize-chances. However pretty-good the less-robust way is, it's still the less-robust way. No getting around that math.

True, for this card formatting question it usually doesn't matter and true sometimes you want to move the card around between several devices and so there is no clear way to pick which one should receive the ideal treatment. Deciding that is a case by case basis. But the question asked was a simple one with a simple set of parameters to consider and a clear more-certain and a clear less-certain.

But you went a little further than that even. You didn't merely recommend "don't worry about it". That's actually fine, because really, it usually is, I don't bother to worry about it myself just for the record and my own recommendation normally would be don't worry about it. But the OP asked specifically if there was a, however small, chance of a difference between the camera vs the pc, and the answer to _that_ question is absolutely yes there is. But you went a tad further and simply stated an untrue fact that there is actually no difference and that there is no such thing even possible as a problem that would be affected one way or the other by the formatting device/method.

Well you're flat wrong about that. Yes I know my example wasn't the result of a difference in the fat formatting itself. Irrelevant on more than one level. It was still a perfect example and the honestly most recent hard personal example I happened to remember having myself first hand. There was nothing wrong with the pc or the sd card. The fat formatting was perfectly within specs and normal, and yet the camera had a problem with it, but if the camera were to format the same card, it would create a 2G filesystem, leaving 2G of unused space, (and unusable by anything else either unless the partition was shrunk and another partition created to use it). The camera would work perfectly with the formatting IT created, and it would NOT work perfectly with the formatting a pc created, even by a quite competent user who knows enough to specify fat16 or fat32, avoid unusual (== unexpected) block sizes, etc.

It not only serves as a _general_ example that things that are supposed to work don't always, as if that should ever need explaining or proving, It actually is an example of _exactly_ what the OP was wondering about. His question was higher level and encompassing more possible types of or reasons for problems than the single detail of possible FAT formats. He just asked if it made any difference to let the camera format the card. For the camera I was tallking about (which happens to be Samsung NV3 from a few years ago), it absolutely did exactly that. Whether it was because of a weird partial size-limit that only affected some routines and not others in the camera, and not specifically a low level fat structure issue, is irrelevant. All that matters is that, without having to know anything, the rule "let the camera format the cameras card" would ensure that camera "just worked", and not following that rule could result in a camera/card combo that doesn't.

Stop trying to defend the indefensible.

--
bkw
--
To unsubscribe, e-mail: opensuse+unsubscribe@xxxxxxxxxxxx
For additional commands, e-mail: opensuse+help@xxxxxxxxxxxx

< Previous Next >
Follow Ups