World: r3wp

Are they all Jaime's

You could always talk to Dave Steinberg at geekisp.com ... [billing-:-geekisp-:-com]

And see if you can take over the account if Jaime has stopped paying 
for them.

Beer, Orca, RebGUI, Synapse were all setup by Jamie AFAIK (or did 
you setup Synapse).

No, Jaime did.

I did have my own account there as well, but closed it down due to 
too much trac spam.

I wonder if there's a better [free] alternative these days?

Yes .... there are.

Can't tind the one I came across a while ago :(

The trouble with free hosting is .. that it often becomes not free, 
or you lose control.

BTW, geekisp.com was charging me $5/month.

USD?

yes

Oh, table error .... how do you know it's data drive?

It's very odd .. the same data works sometimes, and other times it 
craps out.

I've seen something similar before ... solution was to ensure all 
values were strings ... and to copy/deep the data block prior to 
passing it to the table widget. I think it has something to do with 
either a non-string value (e.g. a block) confusing things, or a shared 
series [value] being modified by soemthing other than the table widget.

http://www.assembla.com?

Not free.

I'm going to try copy/deep .. but the errors are occuring in a table 
that I already pre-process.

http://trac.geekisp.com/rebgui
is up again

Hmm, as are the others.

Ashley .. are Rebgui display objects so diferent from VID ones?  
I'm just wondering why henrik's to-postscript function doesn't work 
... see postscript group.

Shouldn't be ... both display and layout just generate a bunch of 
view faces at the end of the day. If it's really an issue you can 
always call system/words/layout from within RebGUI for those layouts 
that must be VID generated.

I thought as much.

system/words/layout would mean I would have to use the VID dialect 
though

What I need is an object browser for Rebgui

A tablet user asks ... how to enable multiselect without using the 
control key ... he's using tablet in slate mode so no keyboard, and 
the virtual keyboard takes up too much real estate.

Map the CNTRL key to one of the freely assignable TabletPC buttons? 
How does he do it currently (e.g. in File Explorer)?

He doesn't normally do this.

I've got three vertically left aligned group-box's and I'm sizing 
them the same by using absolute sizes to also right align them.  
But this means I can't use different gui sizes because then the contents 
of the boxes can exceed the group-box.  But if I use automatic sizing 
using -1x-1, then I can't get both left and right sides of the group-boxes 
to align ... or can I??

Unless I enclose all of the group-boxes inside another grouping widget 
and use the #L directive on all the group boxes?

.

. .

. .

How to easily set auto resizing, for simple scenario as - window, 
two group boxes, in each of them area. I try with various LVHW combinations, 
but can't it get working, group-boxes are overlapping with simple 
#HW:


display "just a try" [group-box "1" 100x50 #HW data [] group-box 
"2" 100x50 #HW data []] do-events

it works, if I change panel 2 to #HX, but then it does not resize 
....

Looks like a resizing bug.

Ashley - how do I overcome this problem:  I need to reduce the record 
block prior to db-inserting it because it contains sub-blocks. I 
would like to use 'next inside for autoincs, however.....

oops - group is one off

I've got about 300 different screens in my application ( apparently 
someone counted ) ...which is a bit of a  cognitive load on the users 
:(

Graham, how long does your app take to load and how are the screens 
initially set up?

I am trying to use the function keys to replace a mouse click on 
the "Save" button.  But the function keys are  global .... so if 
one defines it to have a specific action for one screen, there's 
bound to be a problem where the wrong definition arises.

How long ... not too long. Never timed it.

Screens are all created dynamically.

So, what I need to do is make the F keys stick to a particular screen, 
rather than being global.

the parameters that are available are face and event

So, how can I use the face to identify the particular screen ... 
without too much work!

http://demo-client.s3.amazonaws.com/synapse.exe

I ditched the Windows installer and my app now does the installation 
of all the accessory programs viz. ghostscript, gsview, imagemagick, 
jarnal etc

which is a bit of a  cognitive load on the users :(
 

Yes, Qtask has the same problem...  Having looked at many applications 
that have LOTS of tabs, medical, legal, etc. I don't know yet what 
way around it there is.


One idea I have is that you come in to a mostly blank set of screens, 
with a few tabs, and a button that brings up a dialog.
It is a giant menu of things you can ask to be on the screen.

You pick from this, and at LEAST then everything that is there you 
put there.
Also, you can turn everything on quickly as a feature.

And you will always know if you don't see it, it is probably on that 
menu of things.