World: r3wp

IIRC RebGUI was supposed to be more memory savy, which does not seem 
to be the case. What is also interesting is, that rebol 1.2.47 consumed 
less memory ...

I'm experimenting a little now, but with multiple resizable panels, 
I think it might be easier to generate layouts on the fly rather 
than keep a static layout that one updates various fields.

Petr, you will have a baseline memory consumption but it's not going 
up a lot. So make your app more complex ;-)

think this might be classed as a bug in the layout 


display "" compose [ button "testing" 15.1 button "again"] do-events

the first button does not display if there is a decimal size

Because Rebgui uses a  default pixel size that can be eg. 4, you 
can't easily get pixel precision in layouts.

correction .. the second button does not display

Robert - there still has to be a leak in REBOL! Even if you try primitive 
hello world example, you start with 6MB of RAM, you do nothing, and 
after do-events it grows 3 - 4 MB, slowly, but constantly!

Out of interest, the memory figures on OS/X using rebface are:

	rebface	1.8Mb
	VID		3.2Mb after loading %view.r
	RebGUI	3.5Mb after loading %rebgui.r


but these are for the respective defaults (VID apps often require 
additional patches/fixes and styles, RebGUI apps rarely use *all* 
the default widgets). I should also add that RebGUI (or indeed VID) 
have ever been fully profiled and optimized.

So the correct route is to build custom rebgui.r, including only 
widgets I am using?

Is there a way we can "robustify" rebgui so that we trap all errors??

make object! [
   code: 312
   type: 'script
   id: 'cannot-use
   arg1: 'multiply
   arg2: 'decimal!
   arg3: none
   near: [freedom * face/data]
   where: 'wake-event
]

Yep;
go: does [display ...]
while [[error? err: try [go]] [alert mold err  go]
Along those lines... I haven't got the code in front of me

[[ ]] should be [ ]  ... umm, mental methane

Hmm.. I already trap the errors ... I would much prefer that they 
didn't happen!

ie. they are trapped inside the rebgui code and dealt with there.

I've put a check in the engage event to see if face/data is none?, 
and if so, make it 0.  In case someone like me has set it directly 
to none ...

this is my new appointments display ... is there a way I can drag 
the buttons around? http://screencast.com/t/ou1MwDdS

so I can move an appt from one slot to another

Looks good Graham.

Cyphre has a drag-panel style that allows drag and drop of various 
objects on a window.

Does it work with Rebgui?

I don't know.  I just remember he had the style.  Don't know how 
well it works as I've never used it, let alone with RebGUI

I've got some custom panels that I have a resizing routine written 
for them.  how do I trigger that to run when the main window resizes?

graham i like your app it's very beautyfull  ^^

Thanks shadwolf :)

Graham - nice interface - what widget is used for the scheduler displaying 
time slots?

no widget .. it's all created on the fly

Basically it's tab-panels, panels, and buttons assembled as needed.

And all the ideas were taken from the MS silverlight medical demo

aha ... I like collapsing/expanding capability. ...

It just regenerates the layout

and then I use tab-panel/replace-tab [ new layout ]

the MS demo expands in-situ which would be nice to do ... but I recreate 
the whole thing

Anybody out there using RebGUI ?  I'm just starting to use REBOL 
and have delivered a small app using VID  to a volunteer group who 
are collecting local history articles. And I'd like feedback on any 
gotchas with RebGUI vrs VID. Thanks

Well ! There seems to be !! It took a very long time to display any 
messaged when I visited the group, so I though there wern't any :-) 
Cheers.

http://www.dobeash.com/rebgui.html-> docs etc...

to get rebGUI  you simply write those lines into a ".r" file or you 
go into the rebol/desktop in the REBOL/DEMOS folder

REBOL [Title: "RebGUI Demo"]

do http://www.dobeash.com/RebGUI/get-rebgui.r
do view-root/public/www.dobeash.com/RebGUI/tour.r

I'm trying to do some 'field navigation' within a rebgui form and 
cannot figure out how to get teh filed cursor to 'goto field'

label "Publication : "   

  publication: field  100x8   font  [  size: 18 color: black  shadow: 
  none  ]
					on-unfocus [ uppercase publication/text 
							     ; I'd like to jump to "page" field here
							   ]
		label "Date :"     

  date: field  30x8  font  [  size: 18 color: black  shadow: none  
  ]
		
		label "Page: "  

  page: field  20x8  font  [  size: 18 color: black  shadow: none  
  ]
		return

how about  ... set-focus page true

I've tried that, all i get is a stack error:  as soon as I press 
TAB to enter the on-focus , and the program aborts thus ** Internal 
Error: Stack overflow
** Where: unfocus

** Near: if all [face/type <> 'face get in face/action 'on-unfocus] 
[
    unless face/action/on-unfocus face [return false]
]

maybe it's because you're doing stuff which resets the focus back 
to the publication field?

this works 


do %rebgui.r

display "" [

label "Publication : "   

  publication: field  100x8   font  [  size: 18 color: black  shadow: 
  none  ]
					on-unfocus [ uppercase publication/text 
							     ; I'd like to jump to "page" field here
									
							   ]
		label "Date :"     

  date: field  30x8  font  [  size: 18 color: black  shadow: none  
  ]
		on-focus [ set-focus page false ]
		
		label "Page: "  

  page: field  20x8  font  [  size: 18 color: black  shadow: none  
  ]
		return
] do-events

Mmm ! Your code sort of does the job, but your forcing the field 
Date to be skipped by it own on-focus trap
which would skip Date always,
what I'm trying to do is have Description

control the next field to 'visit' depending on what ever test I code 
in the on-unfocus
 attached to Description field.

for eg: ( pseudo code )

label "Publication : "

  publication: field  100x8   font  [  size: 18 color: black  shadow: 
  none  ]

  on-unfocus [ if desc-flag = 1 [ set-focus new-description ] [ set-focus 
  old-description ] ]

====

Your observation of 'reseting' focus back to Description field is 
an interesting one which I had not considered.

well, until Ashley tells us how to do it, you could hide a dummy 
field behind the date field to take the focus from the tab, and then 
make the decision from that.

Now thats a interesting idea - bit of a kluge tho.


I'm a bit surprised that this 'obvious' programming need is causing 
a problem ! 


I can envisage many situations where a program needs to set the cursor 
focus upon any 

of the available fields displayed depending upon a fields content.

I'm guessing it's a bug ... set-focus is triggering an on-focus event 
and so your code keeps triggering causing a stack overflow.

I think it might be the 'fact' that the on-unfocus never get to see 
its true/false return because my code wants to "go over there" before 
that happens, only Ashley can tell.

Correct ... the set-focus "eats" the return state.


set-focus was intended more for use either within a display's do 
func or from outside the display. Tab order cannot be set programmatically, 
although it should be noted that tab order is the same as the widget 
specification order so you can at least sequence your widgets in 
the desired order. There is no easy way to conditionally skip widgets, 
although build#106 added the set-state func which lets you toggle 
a widget's info state. As for this being an "obvious" programming 
need, it depends on the type of GUIs you wish to write ... you are 
the first person in over two years to have hit this issue (or at 
least report it) ... although I'll freely admit that keyboard navigation 
is not RebGUI's strong point ;)

well, I have had a similar problem.  I have a drop-list with title 
in it followed by a radio-group with gender, followed by occupation. 
 I want the user to skip the radio-group which is set based upon 
the droplist value.  Eg. drop-list is "Mr." and so set the radio-group 
to "M" and then skip to occupation.