editing dives

Dirk Hohndel dirk at hohndel.org
Mon Oct 20 20:01:14 PDT 2014


So this is just a silly story about what I worked on today and how I get
from A to B to C...

When in Duesseldorf, I heard second hand over lunch that our manual dive
entry is frustrating for "power users" who have a larger number of dives
they want to enter - as appealing as the graphical editor may be for the
beginner, once you want to enter dozens of dives a table driven
alternative (just as we have in the planner) would be better. So I figured
I work on that.

So I played with the planner to see how things work there, saved the dive
and switched to add dive mode. Ha. We show deco. Oh, of course, because
this is right after the deep planned dive I just saved. So let's change
the start date. Oops, that doesn't work, it always reverts.

I tracked that one down: silly thing, if and only if we edit a manually
added dive, marking the widget as changed ended up recreating the dive
from the diveplan and the diveplan from the original dive - which reverted
the start time. 

So what was I doing again? Right, I wanted to change the start time of
that second dive so that it wouldn't show deco. Turns out that didn't
work, either. Changing the start time needs to trigger recalculating the
deco (because the pre-existing tissue load could change).

So I fixed that. Tomaz, please take a look at that commit - emitting a
"dateTimeChanged" signal seemed like the logical way to do it, but I'd
like to make sure this doesn't make you throw up...

So this is how I ended up fixing the two bugs I fixed today. Needless to
say, the original problem that made me start looking at that is still
present.

Why am I telling you all this? Because it shows that we need more people
playing with all kinds of different scenarios and filing bugs when things
don't work. I think part of our problem is that for most of us here on
this list (more than 100 people by now) the things we do every day just
work... so we only test the things we do and miss a lot of the bugs that
are still in the code...

/D

PS: Tomaz: yes, you are right. Moving points in the planner has gotten
brutally slow. I didn't bisect this buy my guess is that something in the
CCR patches caused that...


More information about the subsurface mailing list