Description og the .DAT file?

[johnson]

Hi

I am making a parser that will read the SMS.DAT file, fetch each individual field in it and store the contents in a database so that SMS can be easilly searched.

The format is:

ARCHIV\
1
1
[..PDU..]
2006-09-15 18:34:46

My questions are:
1) What do the numbers 1 and 1 mean and which numbers can they have?
2) Is there any indication if the SMS is a sent one or a received one?
3) Is there any indication if the timestamp the the Sent SMS is a valid one, i.e. generated by MPE itself?


Regards!

[FJ]

I'm wondering why you need a searchfunction. MPE has this already implemented ( CTRL+F ).

ARCHIV\ - message is in Archive, no subfolder
ARCHIV\Testfolder - message is in Archive, in the subfolder "Testfolder"
The second value describes the Phonestorage:
0,1 = Inbox
2 = Drafts
3 = Sent Items

The third value is the MessageID in phone and in the messagearchive it is not important

A timestamp on a sent message is always generated from MPE - there is no possibility to read out the date of sent messages.

[johnson]

Hi

Thanks for answering fast

I need the search function to be able to do more complex searches.

What I mean by a valid timestamp of Sent Items is whether or not the message was actually sent through MPE in which case the timestamp is deamed valid, otherwise it is just the time of the synchronization in which case the timestamp is not valid.

Could the SMS.dat format be enhanced to include information about whether or not MPE sent a particular SMS?

[FJ]

I think this is not needed.

[johnson]

Alright. Sorry to hear that. Itis probably the first time you have ever had such a request

I care much about data integrity and when MPE has to make its own, and wrong, timestamps, the data in the file is tainted and should be marked by a mere 0 og 1 in the file.

Thanks for answering, though

[johnson]

Well, in fact, I thought a little more about it.

I think that the reason why you discard the idea of data integrety so fast is because I have mentioned that I am doing stuff with the data that you feel is unnecessary or something only one in a million would do

But, every single user of MPE who uses the archive, has the exact same problem. The date of Sent SMS is sometimes 100% correct (MPE has sent the SMS), or 100% wrong and the real sending time is anywhere from 10 seconds erlier to several days or weeks earlier.

WHat is the point of an archive and timestamps if these are not to be trusted? Or even worse, sometimes they are, sometimes they are not.

I am not debating on why you don't pull out the timestamp from the phone, I know this is not possible.

All I'm saying is that when presenting data to a user, this data should be valid or marked as tainted.
The user can now either erly on the data or make qualified guesses.

I suspect other users are also reading in here, what do you guys think of this suggestion of a validity mark?

[FJ]

Its a simple thing: Sent items have always a generated timestamp - so you never can trust them.

[johnson]

If I send it through MPE the generated timestamp *is* valid.
Otherwise it is not.

All I'm saying is that a simple 0/1 marker in the file would be an extreme improvement.

Would it take much time to implement?

[FJ]

OK, now i understand - i will see what i can do for you.

[johnson]

Thanks dude

Always difficult explaining techical issues in a language that is not native to oneself.

[johnson]

...If you do get around to fix it, it would be very nice if the timestamp in MPE was presented in italics if the timestamp was generated at the time it was movedto the Archive.

This would make it *much* easier spottingthe SMS's that needed to be edited and have a real timestamp issued.

Thanks again