[PATCH] Ticket #423 - Import GPX file from SD card

[Venkatesh Shukla IIT BHU]

On Thu, Mar 13, 2014 at 3:51 AM, Aurélien PRALONG <
aurelien.pralong at gmail.com> wrote:

> Venkatesh,
>
> > Firstly, the enhancement is added. That was really necessary.
> Great.
>
> > It might be a good idea to combine both and have something like "sym +
> name" shown as the dive name.
> Indeed. Even better : check if none is empty, then "sym - name". Anyway,
> if it is too much, users may edit the title later.
>

This has been done. The format now is name (sym). Regarding the GPX files.
What should I do for tracks and track segments? In the present scenario,
they are ignored. The subsurface application in its auto add mode, records
the location after a predefined displacement in space (10 m) or time(30
min). Should that somehow be included here?


> > The marker has been replaced in the map picker.
> Great (bis). Now it works as expected. Still, I would remove the black
> band between the title and the map, and replace the title by "Click to
> select". This way you gain some precious pixels on the screen.
>

Done.


> > The location I would go for a dive could be a place with a different
> local time
> Indeed. To resume previous mails, divers prefer to keep local time. So no
> timezone is required. From a code perspective, it means that we save the
> date as if we were on GMT timezone, but with local hour. Eg. if it is 15h45
> at Bucharest (GMT +2), we would store the time as 15h45 GMT. We have some
> time utils in DateUtils for these kinds of manipulation.
>

Added a datetime picker. The resultant dialog looks ugly and crammed in my
android gingerbread but is functional. As  you said, I have ignored the
timezone entirely.


> > So, in order to simplify things and rise above the cloud of my doubts, I
> chose time difference between now and the dive
> > to be the input given by the user.
> I don't know if it will be the case, but if you add all your dives at the
> end of your holidays, removing minutes for several days will be
> complicated. So a time & date picker will be more appropriate. Another
> thing I think disturbing is that you show the title & date popup after
> having closed the map. I would do it only on validate, as it doesn't break
> the context.
>

Regarding the context breaking, I shifted the dialog to the mappicker
activity.


> > - SimpleDateFormat are not thread-safe. Do not make them final.
> Instanciate them each time it is necessary
> You forget some in GpxDiveListAdapter
>
>
FIxed.


> Finally, some words on your commit :
> - You forgot to remove the old org.subsurface.GpxParser
> - It may be personal, but I think that "Improved code as suggested by
> Aurélien<https://github.com/venkateshshukla/subsurface-companion/commit/b107c4248d2be75b7c2645262682e87dab20aaba>"
> looks weird when you do "git log --oneline". Referencing the discussion in
> the subject is good and enough. Just describe the content. Who suggested it
> is not the first thing you want to know.
>
>
Yes. It was indeed naive of me. I'll keep this in mind in future. I have
amended the git commit with more appropriate title.


> > Which of these is more viable for GSOC, given the time period of two
> months? Which would benefit the community
> > more and has greater necessity?
> > Also, I am not a diver and would not have a divecomputer with me. So, is
> it possible to develop libdivecomputer
> > for android without having a divecomputer?
> I'm not a mentor for GSoC (although I'll try to help for Android), so I'll
> let anyone else answer you.
>
> Aurélien
>
>
One last thing. I have sent pull request on github having all these
changes. Could you please respond to it?

Venkatesh Shukla
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.hohndel.org/pipermail/subsurface/attachments/20140315/680b74f5/attachment.html>