Development version - Questions and Answers

[Die Batzen]

I installed the DEV version today and think it's very well done:

• the combined view of waypoints and variables is very clear
• no need to switch between waypoint and variable tabs anymore
• detected variables are listed under the formula
• additional variables are grouped under Unused Variables

Some observations during testing:

1. clicking on a variable always allows any input (text) even if the variable doesn't have a function included; this can easily lead to false input; using edit works as in previous versions
2. the delete button for a variable listed under a waypoint doesn't delete the variable - it just clears its content; on the other hand, Unused Variables get deleted
3. if a formula contains an error then it's getting red (as usual), but no hints what could be wrong are shown anymore (e.g. if the degree sign is missing)
4. consider this formula: N50° 10.XXX E10° 22.XXX and XXX = A + B
   • if you copy/paste that formula then variable X gets detected (clear since var XXX isn't present yet)
   • add a new var XXX
   • edit var XXX: add formula and assign A+B --> now A and B are listed as Unused Variables
     since A and B are not detected by parsing the waypoint formula, that makes sense; but they're still used implicitly to calculate the waypoint which qualifies them to be listed as Used Variables: what do you think?
5. × as multiplication sign: wasn't able to figure out how it should work (for AxB or axb A,B,x or a,b,x are detected)
6. variable functions: pow0, pow1: could you explain how that works?
7. function to fetch cache data from remote server (locus-solver.de): same as previous point - could you explain how that works?

That's all for now, next is to do some field testing.

[gawalione]

1) I copied the behavoir from the the variable view to the new input boxes. It will be in dev2.
2) The variable gets deleted. But you cannot see it, because the parser will recreate it instantly, because it needs the variable.
3) You should see this in the detail view. In the list view is no space for that
4) Yes, they are not unused. But its hard to detect, if a variable is used in a formula, because in the listview i do not know anything about the variable functions. But good point, i have to think about this.
5) × ist not x! I already modified the parser to detect x as multiplication sign, but i'm not sure how to use this. A formula like N51°(AxB) is a valid formula for both parsers.
One would detect A, x, B as variable, the other one A, B and a multiplication. Both could be intended by the cache owner. I think about this, perhaps i have to implement a switch somewhere in the gui.
6) Its simply the same as ², ³, etc but ⁰ and ¹.
7) Currently the solver can detect its save file format, if embedded into the cache description, and load it instead of starting with a clear cache. See https://coord.info/GC279Q7
But this requires help and interest by the owner. The new function can load such a save file from my server, if there is one.

[Die Batzen]

DEV-2:

• 1) I copied the behavior from the the variable view to the new input boxes. It will be in dev2.

  Seems unchanged compared to dev1: the keyboard has all characters available even if just numbers are allowed (for default variables).
  
  I would be nice if the keyboard directly opens when you tap on the variable input field - currently there are 2 taps necessary: 1 for activating the input field and 1 for getting the keyboard.

• 3) You should see this in the detail view. In the list view is no space for that

  It works in detail view for out of range results.
  But if the degree sign is missing no hint is shown in the detail view, e.g.
  • create a new standard waypoint and omit the degree signs
  • back to list view --> both coordinates are red --> OK
  • enter edit mode of this waypoint --> no hint to the cause of the red color is visible --> a hint to the missing degree sign should be visible in found errors section
• in GC5DAV5 coordinates without variables are detected a little "too much":

  

  If you tap on it and assign it to a waypoint then just the north coordinates are parsed.

[gawalione]

Seems unchanged compared to dev1

Yes, your are right. I did it on my laptop and forgot to commit and push it. Its now in DEV3.

I would be nice if the keyboard directly opens

Will be in DEV3 too.

But if the degree sign is missing no hint is shown in the detail view

Will be fixed in DEV3.

in GC5DAV5 coordinates without variables are detected a little "too much":

Yes, but that was already the case in the other versions. But these only highlighted the shortest match, the new version highlights the longest match.
In this cache the problem is in "E11° 02.483. Auf dem von ..." that the parser parses this like this: E11° (02.483).(Auf dem von ...) For it, all the letters after the dots are valid variables, it does not know better. I dont think, that this is fixable without destroying other matches.

[Die Batzen]

Seems unchanged compared to dev1

Yes, your are right. I did it on my laptop and forgot to commit and push it. Its now in DEV3.

Works perfectly in dev3.

I would be nice if the keyboard directly opens

Will be in DEV3 too.

Very nice, thank you.

But if the degree sign is missing no hint is shown in the detail view

Will be fixed in DEV3.

Can confirm that it's fixed.

in GC5DAV5 coordinates without variables are detected a little "too much":

Yes, but that was already the case in the other versions. But these only highlighted the shortest match, the new version highlights the longest match.
In this cache the problem is in "E11° 02.483. Auf dem von ..." that the parser parses this like this: E11° (02.483).(Auf dem von ...) For it, all the letters after the dots are valid variables, it does not know better. I dont think, that this is fixable without destroying other matches.

Thanks for the explanation.
And in the meantime I realized that it's already working well as it is, since in those cases you get a list of possible results to choose from. Just select the one with the coordinates and you're done.


I used the new version now for several cache searches in the field and can report that it's working stable and reliable.
A big thank you for your excellent work!

[Die Batzen]

Bug in dev3:

• in new listview, create var named A
• create var named a
• try to enter a value on either of them (direct input) --> app stops

[gawalione]

Should be fixed in dev4.

[Die Batzen]

Bug in dev3:

• in new listview, create var named A
• create var named a
• try to enter a value on either of them (direct input) --> app stops

Can confirm that it's fixed in dev4.