[maemo-users] gpe contacts import

From: Mark Haury wolfmane at gmail.com
Date: Wed Apr 23 19:37:42 EEST 2008
Michael Wiktowy wrote:
> Expecting someone to put the bit of effort into detailing their
> problems is the smallest price to pay to have them fixed.
> Bug tracking software allows the developers to be a lot more efficient
> at staying on top of diagnosing issues than juggling a bunch of
> unstructured, vague, ranty emails. The fact that you have to do some
> email validation/registration process (similar to signing up for a
> mailing list) and you can't just reply back to the bug tracker via
> email is an unfortunate consequence of our spam-infested Internet.
It may seem reasonable if you only consider a single bug in a single 
application, but that's not the real world scenario. What is actually 
happening is that the developers have the easy side of the bugzilla 
process, and they're only dealing with the one bugzilla, while the 
average user is dealing with bugs from a bunch of different apps at 
once. Don't try and tell me that's not valid.
> I found it deeply ironic that the OP complained about developers not
> valuing his time while completely disregarding the value of the
> developer's time by transferring the entire burden of
> mind-reading/diagnosing/fixing onto the developer while the OP got to
> vent his spleen in a cathartic insulting email. That is not an
> entirely equitable trade.
Yeah, well, in my experience it's more often the *developer* who starts 
off with the rude "RTFM" and sarcastic comments, not the person 
reporting the problem. Usually the person who has run across the problem 
is simply asking a question. The rant comes *after* they get mistreated...
> As some guidance from someone who doesn't think registering to a
> bugzilla as an insurmountable burden (and to try to steer this thread
> on a productive course), some of the things that, in my opinion, would
> be important to see in this "bug report" (and why they are important)
> would be :
> - What software was used to produce the multi-contact vcards?
In my case it was Palm Desktop on Windows XP (which I had to boot 
specifically for that process - I almost never boot Windows anymore), 
although when that failed I tried some other converters I found online, 
still to no avail. While it's *possible* that the Palm Desktop creates 
malformed vCard files, it's a pretty mature app at this point and 
therefore unlikely. Even if it is the culprit, it's also a very strong 
candidate for making allowances in the importing app's procedures. That 
may not be technically "right", but in the real world it's wise.

> (possible buggy production of vcard file in which case you can only
> expect GPE to not die on garbage input)
> - What OS was the vcard file produced on? (possible issues with
> parsing different end-of-line characters among other things)
Everything I was ever taught about programming was to take such cases 
into consideration and automatically deal with them "gracefully". 
Especially if you're dealing with importing data, it's a no-brainer to 
automatically handle such elementary issues as the differences between 
the text file formats of the common OS's (and their built-in text 
editors). Choking on something that obvious is unforgivable.
> - A test case, aka simplified sample, of the vcard file that is
> causing the issue. (would be able to quickly determine if it was bad
> input, bad parsing, special character encoding, etc.)
This is a lot more work than you think. Especially with regard to 
contact information, you're often dealing with *very* sensitive 
information, and therefore are not at liberty to supply the original 
file. That means that coming up with an example test case involves 
creating a complete new source. That's not nearly as easy or reasonable 
as you make it sound.
> - When the issue started or has it been there from day-one? (to
> determine if it was a regression in recent changes to quickly
> pin-point the source of the bad code)
A perfect example of how the heck am I supposed to know that? I just 
installed gpe, and just tried the import. There's absolutely no way for 
me to know how long the bug has existed, and no reason that I should anyway.

> - What hardware/OS is GPE being used on? (to see if the developer has
> the right platform to test the fix or has to rely on the user to test
> it)
N800, OS2008 #2 (the latest, although the situation was the same with 
OS2008 #1).

> Hopefully the OP can fill in these gaps because no one else can.
> /Mike
I answered these questions in good faith, and I know that there's 
someone on the list who can make use of the information. I'm not going 
to bugzilla with it, though.


More information about the maemo-users mailing list