[GRASS-dev] Re: [GRASS GIS] #842: wx location wizard: spincontrol busted for UTM zone

[GRASS GIS]

#842: wx location wizard: spincontrol busted for UTM zone
---------------------+------------------------------------------------------
  Reporter:  hamish  |       Owner:  grass-dev at lists.osgeo.org
      Type:  defect  |      Status:  new                      
  Priority:  major   |   Milestone:  6.4.0                    
 Component:  wxGUI   |     Version:  svn-develbranch6         
Resolution:          |    Keywords:  location wizard, utm     
  Platform:  Linux   |         Cpu:  x86-64                   
---------------------+------------------------------------------------------
Comment (by cmbarton):

 Replying to [comment:10 hamish]:
 > Replying to [comment:9 cmbarton]:
 > > Please try the following. Comment out lines 731-742 in the
 > > OnParamEntry method
 > ...
 > > This disables the validation check for the text box of the spin
 > > control. If this works, maybe I can validate this another way.
 >
 >
 > Yes- that lets me spin. If I manually type in 67 or -1 for the zone I
 get a pop up and it corrects me to the nearest valid value.
 >
 > but if I let it spin full-scale a few times in one direction when I
 press continue it complains about an invalid zone even though the number I
 see on the screen is in-range.
 >

 That is REALLY weird! The SpinCtrl is limited in the code but the
 accompanying text box is not. My system behaves as described in the
 wxPython docs. I can spin all I want, but it only goes from 1->60->1->...
 No way to spin it above 60 or below 1. On the other hand, I can type
 whatever I want. If type 80, for example, the zone simply doesn't get put
 in. I will add a validator to the page changing handler, where it should
 not interfere with the spinning. But the behavior on your system is odd.
 Seems like a bug in the GTK version.

 The error trapping I did for bounds is minimal, but it prevents the wizard
 from throwing an error to terminal if the rows or columns are 0. Your
 other suggestions are good. But like you said, should go into 6.5+.

 Michael

 >
 > Replying to [comment:8 cmbarton]:
 > > The other stuff is a bug fix I noticed when I was testing this.
 > > I was oversure of fixing it I guess or I wouldn't have added this
 > > fix to the patch. Have to live with it now.
 >
 > probably just worry about this for the development branches for now &
 when you get a chance, but re. bounds settings a couple of wishes:
 >
 >  - popup error if s>n, or w>e (except LL), or abs(ns)>90 (LL), right now
 you can click the button but nothing happens if its invalid. feedback as
 to what the error is would be good.
 >
 >  - after setting the bounds the new region settings are copied to the
 terminal. left over debug code?
 >
 >  - LL bounds should accept [NnSsEeWw:] as well as [+-.0-9] chars, and
 res should accept ':'; currently it just accepts floats. maybe quietly
 accept 123d45'43.21"S format as well if we are feeling generous?
 >
 >  - it would be really really gee wiz cool if for LL a red rectangle
 would draw covering the region you've selected on the center map. even
 cooler but not that useful in practice would to be able to drag a box on
 it to get the bounds numbers close, rounded to the current resolution.
 >
 >
 > thanks,
 > Hamish

-- 
Ticket URL: <https://trac.osgeo.org/grass/ticket/842#comment:11>
GRASS GIS <http://grass.osgeo.org>