wishlist

[Dirk Hohndel]

On Tue, 2013-05-21 at 15:58 +0200, Jef Driesen wrote:
> On 2013-05-21 15:22, Dirk Hohndel wrote:
> > On Tue, 2013-05-21 at 11:33 +0100, Pedro Neves wrote:
> > - Tags: possibility to add user created tags (I would add
> > "scientific", "fun")
> > 
> > Definitely on the list of planned features.
> 
> User defined tags are a great way to enter custom information without 
> needing explicit support and/or a custom user interface for each type of 
> information. Do you want to mark dives as a night dive? Just create a 
> nightdive tag. People who don't care about nightdives are not annoyed by 
> a useless (from their point of view) nightdive checkbox. Of course this 
> is not limited to night dives :-)

I slightly disagree. I think having a pre-defined set of tags is a good
start (many users don't know what they could be doing with tags). But we
definitely need to add user defined ones.

That's on the to do list for the Qt port. Just not something we have
gotten around to.

> I would even vote for supporting a hierarchical tag system, so you 
> easily categorize your tags. For example separate group for marine life 
> sightings (with subgroups for species as you wish), another group for 
> dive type/purpose (wreck, training, deep, photo, etc), and so on.
> 
> Long time ago I made a prototype UI for this. The screenshot should be 
> self explanatory:
> 
> http://www.subaquaclub.be/users/jefdriesen/tmp/screenshots/divemanager-dive-tags.png

That's a really 90s looking UI with a high appeal to some type of
users... not sure how much the larger diver community would like this.
We are hoping to go a little more "modern" with the next version.

But I like the idea of hierarchical tags.

> > - Comparison between different algorithms (possibility for the user to
> > compare the same dive plan using VPM and Büelmann algorithms)
> > 
> > I find that of extremely limited value. All these algorithms are wild
> > ass guesses with some heuristics on top and some magic constants to 
> > make
> > them match known good profiles.
> > 
> > But hey, YMMV
> 
> Might be useful to compare the same algorithm, but with different 
> parameters. For example comparing the effects of using different GF 
> factors, using another gas mix, etc.

We actually tried that (well, I did during implementing - can't remeber
if I ever committed it) and it was visually very confusing. But sure, we
can play with that again, maybe the Qt folks come up with better
visualization ideas.

/D