AltME groups: search

Do you mean you want to show windows inside the main window ? Or 
are you happy with panels ? eg:

Why want you do that?
If you want to refine things after layout, you can do instead:
 lay:layout[..]
 do extras
 view lay
I

O'k I somehow managed to view chages in Plugin object frame by using 
new script instead of old one (saveiing old one as old01.r and putting 
it in html page).
But Volker's example does not work for me, fex:

mainWin: layout [
	t1: text "AAA"
	button 200x20 "Atvçrt kârtis" [
		do something
	]
	do [
		t1/text: "Text changed"
		show t1
	]
]

view mainWin

Look here:
	print mold svv/vid-styles/text/init

where it says:
	size: (size-text self) ....

We can do the same:
view layout [
	t1: text "AAA"
	btn "Modify" [
		do bind [
			text: "Text changed"
			size: 1000x200
			size: (size-text self) + (edge-size? self) 

     + (to-pair all [para (to-pair para/margin) + to-pair para/origin])
			show self
		] t1
	]
]

Well, Anton first my mistake was connected with that plugin used 
previous version  of script.
To better see my problem,  than try this:

mainWin: layout [
	t1: text "AAA       bbbbb"
	button 200x20 "Do something" [
		do something
	]
	do [
		wait 10
		t1/text: "Text changed"
		show t1
	]
]
view mainWin


Then you wil of course get text changed but before 10 seconds you 
will only see just "Rebol/Plugin 2.6.2.3.1", but not the layout - 
layout appears after 10 seconds and then of course t1 is cahged.

But I need at first to see layout and then after 10 seconds t1 is 
changed.

hmm, that is guru stuff ... although I do seem to understnad the 
bind, I think I also understand, while less experienced user is in 
hell ... then stylize/master is insufficient imo

well, it is just - stylize/master is here for you to define new style 
... looking into other ones. It should not presume you know about 
where the stuff is bound ... the question is, if it could be cured 
:-) after all - touching stylize/master you are already touching 
the guru stuff, so you better know what you are going to do ... and 
how :-)

Pekr, what does Bobik want to do ? It looks like he wants to capture 
the return key (which is #"^M" and not newline by the way).

I will post it to him ... what does your function do in particular? 
:-)

to your double bind example - what does it do internally? you first 
bind the block to system/View - what happens in that context, and 
what happens in next context - ctx-text?

now I seem to understand, just did some small example myself:

block: [print a]
do block ; 'a is not known
my-context: context [a: 1]
do bind block my-context ; now 1 is printed

Looking at the code I assume what you want to do is initialise the 
text facet and make sure it's shown the first time the display is 
shown.

I have a question WRT changing the text of a shown label. I use a 
way where the layout of my GUI is done at startup-time once. Now 
I need to change the text of some labels at run-time and I don't 
want to do the following:
	- name each label that needs to be updated.

 - write a function that changes the TEXT word and shows the widget


I would like to use a word in the initial context that keeps a reference 
to the text (or function returning the text) which gets reevaluated 
if a SHOW is used.


Do I have to write an own style that supports this or is there a 
simpler way?

What I want is:


1. The exchange rate to be saved in a file, so that it will always 
be loaded upon start the script.

2. The exchange rate to remain in the field through each loop of 
view.

3. And, of course, I want to be able to change the exchange rate 
if needed.


Seems like it should be a common easy thing to do, but it eludes 
me.

Louis, I haven't been following this problem but if you want to just 
load a rate from a text file you could do something like this which 
handles loading before showing the face and allows saving an updated 
rate 

REBOL []

these-styles: stylize [
	lab: label  60x20 right bold middle font-size 11
    fld: field  60x20 font-size 11 middle edge [size: 1x1]
]

    rate-file: %/c/rate-file.txt
    if not exists? rate-file [write rate-file 1.000]    
    rate: load rate-file

view layout [
    across
    styles these-styles
    lab "Rate"    rate-field: fld bold (form rate) return
    btn "Save" #"^s" [save rate-file form rate-field/text]
]

I think I can do this, but it would probably be better for you experts 
to come up with something really simple and elegant  that can be 
submitted to RT for Rebol 3.

Louis, just catching up with your post back on Monday, I wanted to 
say that it's a bit odd to use the FEEL to do init code. It's not 
really intended for that. A way to do custom init code is like this:
	layout [
		field with [
			append init [
				; custom init code here
				text: exchange-rat
			]
		]
	]

and in the above case, just setting the text facet, it simplifies 
to:
	layout [
		field (exchange-rat)
	]

Thanks. I like things to be simple. By the way, I just  looked at 
your styles gallery. Very nice, and gives me examples to do things 
I need to do.

It should work first time. When all is downloaded and run, all the 
needed files should exist in your cache. If you are happy with the 
cached version, and want to run directly from the cache, replace 
the second line with this one:
	do do-thru/args site/do.r [cache-only "gui/style-gallery.r"]

Use this code to check for newer versions and update your cache:


 head clear find site: select load-thru/update http://www.rebol.net/reb/index.r
 [folder "Anton"] %index.r
	do do-thru/update/args site/do.r [update "gui/style-gallery.r"]


Use the code below when you are happy with the cached files you have 
and want to do those only:


 head clear find site: select load-thru http://www.rebol.net/reb/index.r
 [folder "Anton"] %index.r
	do do-thru/args site/do.r [cache-only "gui/style-gallery.r"]

I had another one in between which I didn't release. I think I tried 
to do too much with that one and it got hard to debug.

you don't need to make a style for the with block.  I just did this 
cause I thought you meant that you want the functions to be applied 
to a broad set of gadgets.  this works directly:

view layout [
	button "do" [face/print-lbl]  with [
		print-lbl: func [][print self/text]
	]
]

See also: face/changes dialect [activate minimize maximize restore]

After setting face/changes to one of those words, you must show the 
window face to do it.

stupid question... can't we set the progress bar value directly within 
the dialect?

view layout [progress 0.3]

I find it a bit ugly to have to do this:

view layout [prg: progress do [prg/data: 0.3]]

ok, I read back a bit and I understand more the origin of the discussion. 
 I had assumed brock had an image with an alpha and was not able 
to save it out.  his real question basically... how do I CREATE the 
alpha channel from View gfx.

and you have answered that.

sorry for stickin my head in my ass, I should have read above.

damn anti-aliasing!!  now I have fringing or a halo effect around 
my image due to anti-aliasing.  I'll see what I can do about this. 
 Thanks again.

a good trick is to set the gamma to one end, do the drawing and then 
compensate gamma back with its complement, to get original colors 
back.

I was able to get Anton's example to work in the scenario I had indicated 
above.  I'm open to looking into your suggestion, do you have an 
example that I can use.  I checked some of the rebol/view docs and 
the image datatype doc and don't see anything that guides me in the 
direction you are suggesting.

so my question is, why do you need to make one color transparent, 
instead of using the alpha channel?

yes, just like following 3 years of mechanics will allow me to change 
my pistons in my car myself... the question. is... do I want to? 
 when the whole platform is the one taking away those events!  for 
compatibilitie's sake?

You could express everything that you want but do not have as something 
"taken away" from you. That's the "glass half-empty" way of looking 
at things. Are you a programmer or not ? If you want something, just 
make it !  If it's hard, work hard !  The whole art of programming 
is to make something previously difficult, easy.

I only wish, and I'm surprised you are not appreciative of, the fact 
of allowing REBOL view give me more of what its already doing.  it 
wont add anything substantial to interpreter in file size, wont remove 
any features, won't even affect current VID objects, as they already 
handle the most basic event types.  But at least, I will be able 
to cater more precisely to client's needs without it removing the 
core reason I would propose REBOL for a client in the first place... 
cause I can just do it.


an example, in many apps, having the middle mouse button would have 
been very usefull (especially in my field, where all computers have 
3 buttons), and sorry, but having to try and hack such a thing in? 
 I don't even know where to start, how do I then reverse hack such 
a thing so that it gets into view cleanly?  how will a word be assigned 
in the event/type correctly?  and then linux... my god, do I have 
to understand the whole X,  just to figure out how to add a middle 
 mouse buttons?

Getting REBOL/View to the same level on all graphical platforms, 
that REBOL support, should be a bigger priority! I'm not eager to 
do a lot of View development with the current situation.

I would suggest setting View+ channel. We should voice what do we 
regard being important. E.g. - does rebol app still crashes, when 
changing resolution? That is unacceptable. There should be much more 
precise layer, getting info from OS - e.g. i do normally work in 
2 monitor set-up. I would expect system/Screen-face to reflect that, 
containing measure for both displays etc.

do you expect code block to be printed?

yes I do. I'm not sure it makes sense to visually press the button 
down and then it does nothing.

what does 'cue function do? Does it do action block submitted to 
style? (button here)

vector! is a DT that was being implemented to represent real arrays 
(i.e. so that 10 integers do not take 160 bytes, but just 40). it 
would mainly be useful for rebcode, but there are many other cases 
when it is useful.

How do I resize a text-list horizontally?
I tried this code
view layout/size [
    tl: text-list data ["A" "B" "C"]
    across

    btn "+" [tl/size: tl/size + 10x0 tl/sld/offset: tl/sld/offset + 10x0 
    show tl]

    btn "-" [tl/size: tl/size - 10x0 tl/sld/offset: tl/sld/offset - 10x0 
    show tl]
] 300x300
which works if the text-list is made smaller.
But if it's made bigger I get artifacts.
What else needs to be set in the text list to make it work?

view/new layout [
	button "start work" [work-to-do?: true]
]

do-work: func [][
	; <- do a chunk of the work here
	work-to-do?: is-there-more-work?
]

forever [
	if work-to-do? [do-work]
	wait [0.02]
]

I need some tips on "How to reduce memory" when using 'make face... 
my current application is eating 13 MB of memory with 90 faces...I 
have 80 faces that do actualy the same.. And is it standard that 
view eats this much memory? I compaired it with a using layout but 
the memory is not much different..

17 MB is very low for a rebol script, I think. :-) I've seen them 
eat up towards 1 GB of memory. View does eat a lot. Why I don't know, 
but it must have something to do with buffered uncompressed bitmaps.

i think there was a hack to do that, iirc you can set the caret to 
a string different than face/text or something like that.

I think F10 isprobably because that maps to WM_MENU, so WIndows is 
eating it. For cap/num-lock, you should be able to do that with the 
GetKeyState and GetAsyncKeyState APIs. It is a bit of a shortcoming 
that we only get keypress events, not keyup/keydown as well (if we 
want them anyway).  I can understand the logic, that you won't need 
them in most simple apps, but a lot of people are doing stuff with 
REBOL where they would definitely help.

I gotta tell you. Implementing an IME and showing Chinese characters 
on REBOL/View are perplexing and painful, especially I am doing this 
all by myself. Even worse, considering I have so many works to do 
in my office, I don't really have much time to do this for REBOL/View. 
But I want it so bad, what options do I have? Waiting for REBOL 3.0 
or 3.1? Well, I look forward to it, but I don't count on it. Only 
God knows when it's going to be released. 


As a "Messaging Language" for communication between people and people, 
computers and people, computers and computers, REBOL should have 
supported I18N many years ago.

as an alternative for now, have you considered using an HTML interface 
instead of View? (ie. your app listens to port 80 and you point the 
browser to http://localhost/) It will be a pain too, but it's probably 
easier than trying to do IME from scratch. (OTOH it may depend on 
your needs.)

Gabriele, thank you. Actually I tried to use Flash 9 / ActionScript 
3.0 as the interface and REBOL as the engine. It worked. I _like_ 
ActionScript 3.0, but I _love_ REBOL even more. So, if there is still 
a change to do it in REBOL, I probabily will not give up. The REBOL 
language is too good to give up. Gabriele, please push I18N support. 
I'm willing to help.

And I have to say it again, the basic is to be able display the characters! 
I don't need to count lengths of unicode strings and other related 
actions, if so, I can do it by using my own functions.

and bitmapped font can improve performance, as you don't ned to do 
all the resizing and antialiasing and other actions which must be 
done to display vector font

just one last thought, it would be good to make it independent on 
X-win under linux - As you may want to be able do graphic operations 
on server which is not running x-win (most servers)

cyphre, do you expect that the new and faster SHOW would be postponed 
to R3? is it tested?

Gabriele, 

Actually, Oldes is right. Showing two-byte characters is good enough. 
IME is not necessary for REBOL/View, because every Chinese/Japanese/Korea 
OS has proper IMEs installed. IME sends the codes encoded in the 
OS codepage to the focused window. For Example, If the codepage used 
by Windows XP is Big5 and I type in the Character which means one 
( #{A440} in Big5, #{4E00} in Unicode, see http://www.unicode.org/cgi-bin/GetUnihanData.pl?codepoint=4E00
), my REBOL/View program will get two key events sequentially, which 
are #{A4} and #{40}. REBOL/View shows it as two characters instead 
of one. I hope that REBOL/View can let the OS do the text-drawing, 
like the REBOL/core console does. REBOL/core console doesn't have 
the Chinese-Character-Showing issue, because it basically send the 
#{A4} and #{40} to the console, and let the OS do the text-drawing. 
the OS knows that #{A4} and #{40} should be combine to one Big5 Character, 
so it just show it as one character. Of course, if I type in two 
ASCII characters, the OS is smart enough not to combine them into 
one "non-existing" Big5 Character. CJK encodings are supersets of 
ASCII, just like UTF-8 is a superset of ASCII.


It's nothing to do with Unicode, so it is not too difficult to fix, 
I guess. Please fix this in 2.7.5 or 2.7.6 please ... 

It's on my wish list for Santa Claus this year.

Here's a document (using LAYOUT spec) which tries to explain a proportional 
resizing algorithm.
do http://anton.wildit.net.au/rebol/doc/resizing-algorithm-doc.r

I think it's better to be modular. But anyway, maybe you could make 
a small demo which shows how each of Glayout's parameters work:

min-size, def-size, init-def-size, manual-min-size, static-size, 
elasticity, stretch... and any other I do not see.

ok. Do you preallocate space for all the entries beforehand? So it 
doesn't have to expand along the way.

Scaling factor and direction? I am afraid I dont understand what 
you mean. Anyway, I don't really have time to do the font-compression, 
and it's not practical for me for now. I am trying to change my bitmap-based 
text-rendering REBOL/View component to a vector-based one. I hope 
the performance will not cause too much pain. Considering the complexity 
of Chinese font, the rendering performance is what I am worring about.

in any case, a way to automate this is pretty easy IMHO. just do 
drawing and measure the amount of pixels which overlap wrt previously 
defined recurring shapes.  generate a view app which shows the highest 
matches in order and then a simple yes/no to confirm.

Do we need a "User Interface Style Guide"? It may be part of the 
wikibood. I came to think of it when revising some of my scripts. 
I often program the key "Q" to be used to quit my scripts. But maybe 
it should be <ctrl>-q to quit, like it is with AltME!? Then I thought 
about, if anyone of you has made a style guide regarding user interfaces 
in REBOL? REBOL/View is made for different platforms, each with their 
own style guides. Should we stick to those guides, so our scripts 
have to use different keyboard shortcuts on different platforms? 
No, probably not! A REBOL script should work the same across all 
platforms, I think. Then we need a "User Interface Style Guide", 
so us developers can stick to the standard when developing REBOL 
applications, shouldn't we?

you're best bet is to use a monospace font and do a very fast char 
count instead.

Is it possible to catch program close event - when windows shut down 
and do something (save data)? The same appies when somebody closes 
all program windows or terminate process from task manager.

The line in the TAB-CHAR handler in EDIT-TEXT which is causing all 
the trouble is this one:
	if not event/shift [action face face/data]

This line is gonna get the knife, if I have anything to do with it.

this is a difficult issue. personally, i prefer tab and shift-tab 
to not do any validation, while enter does. it's not always a good 
idea to not let users go away from a field just because it is invalid.

however, since this depends on the application etc., i think the 
best solution is to call the action on enter, and just call face/refocus 
on tab / shift-tab. then you can make refocus do the same as the 
action, or not.

to be safe, you can view/new, change feel, then do-events.

Gabriele, I see your point: The user might want to enter some data 
in a field, but part way through think of something else and leave 
the field to attend to it, then return to the original field to complete 
the data entry. Finally the enter key will do the face action which 
can validate the field.

There should be a ton of placeholders inside the feel code, to let 
you easily set those placeholders from within layout. I know this 
is possible to do.


view layout [field 100 single-key-action [print face/key]] or something

What I just don't know - what if you would like to serve e.g. holding 
down shift key and holding down left mouse button? e.g. for dragging 
something? I mean - you need both functionalities of on-something 
- how do you mix code then in those code-blocks?

1. You have direct event handling with on-XYZ [do-program-code]

2. Then you have dynamic flags for, say, qualifier keys inside that 
program code, from the event system

3. Then you have face flags. Some things like size limiting a field 
should be simple. It can _probably_ be mapped to 1.

I'm not sure either. Replacing the current FEEL system would be a 
tremendous amount of work. I was originally thinking more in line 
of extending it to those placeholders. I'm no expert on dialects 
and combining it with FEEL, but if it's possible to do, then OK. 
:-)

yes, I think gradients are quite hard and non-intuitive to do in 
DRAW

somehow, it seems like its missing a color blend mode selection, 
multiply, add, screen, etc. right now its a ramp which fades to 0 
on each side, such that when two successive colors do not have any 
bit in common it creates an alpha channel premultiplication grey 
zone...

yes, the new gradients will be even more configurable. Regarding 
the access to the AGG  internal representation. I have talked about 
 this with Carl and we agreed in R3 it would be allowed to do functional 
programming of DRAW blocks. But all this needs more thinking...

Maxim, regarding this: "somehow, it seems like its missing a color 
blend mode selection, multiply, add, screen, etc. right now its a 
ramp which fades to 0 on each side, such that when two successive 
colors do not have any bit in common it creates an alpha channel 
premultiplication grey zone..."

blending is not a part of gradient definition, look at Flash or SVG. 
Gradients work the same as in Rebol inthis case.

Just wondering if this sounds like something that a modified desktop 
would be able to do.  I know this is exactly what IOS does, but this 
is overkill for this simple document generating tool.

I am one of the people who can claim a certain level of advancement 
with VID and view, and yet why do I keep fiddling around with such 
things...  there are so many little details to remember that it comes 
to a point where I keep forgetting what I haven't played with within 
a year... but it always comes back to bite me the year after ;-)

I completely agree with Maxim, although I don't forget the things, 
but it still takes a long time to figure out how to do trivial stuff 
with a field.

ImageMagick works well. I've only needed and done simple stuff with 
it--it's so feature rich and complex, I don't have time to dig in 
and do fancy stuff.

well, it won't work here now since I read it requires ghostscript 
to do what I need. I'm not going to balloon a 130 k script to 30 
Megs, just for that :-)

I get this:

connecting to: www.lexicon.net

Include: Couldn't load-thru http://www.lexicon.net/antonr/rebol/gui/menu.r
** Script Error: menu-style has no value
** Where: do-facets
** Near: menu-style

I use do-service to issue a RS request. But it's not blocking my 
GUI. I want to keep the user away from clicking around.  How can 
I block the processing of GUI events?

Interesting, the doc string for do-service in my 19-Nov-2005 copy 
says {Send request to a REBOL service and wait for response. Return 
result.}

so all you should need to do to block the gui is remove the event 
port.

The READ + WAIT should be roughly equivalent to DO-SERVICE in your 
case.

I stand corrected. :-) I should know better than to say "you can't 
do that", because almost anything can be done in REBOL, it's just 
a question of effort and obscurity.

There is a script called update_index.r that probably is supposed 
to do this - but it bombs when not finding "desktop/icons".

Maybe this will help. 
http://anton.wildit.net.au/rebol/gui/demo-drop-down.r

But it sounds like the small size of the parent face (the tab panel) 
will clip the drop-down.

Actually yes, I think what you want to do is modify this function, 
so that the width of the drop-down list (which is a popup window) 
is not determined by the width of the drop-down face which popped 
it.
	print mold get in svv/vid-styles/drop-down 'show-dropdown

Ashley, it depends what you want to do.

If you aren't going to transform the text in crazy ways (or maybe 
even if you are), then I would create a "sizing-face" with the same 
text and font, then use just size-text on it. It's going to be faster 
than scanning every pixel.

How do I do following?

copy-screen: center-face layout [screen code here]


forever [ if new disk is found .... start screen, call copy routine]


And now I have a problem - if I do view copy-screen, call of copy 
routine is not performed. So I tried to put the copy routine into 
do [call copy routine] into the layout code, but those variables 
are not known yet ...


I would like to have some "delayed" run of 'do section. (view/new 
and do-events is not possible, as my copy routine is recursive and 
do-events would be callled multiple-times. Is there any trick? Timer 
probably? Well, but trouble is, that script sits in the background, 
and copy screen should apper only if new disk with appropriate content 
is found, then it should disapper and go to wait mode into background 
....

petr, while you are busy copying, you are not handling events, so 
no do-events there. you can call wait 0 every iteration if you want 
to respond to events while copying.

hmm, don't see much change in a case like this, you still do the 
layout, view/no-wait it (new ref name instead of /new), and do the 
copy. maybe, you'll want to copy in a bg task, but that's another 
story.

Gabriele, I do following:

copy-dir: func [from to][recursive copying function here]
copy-screen: layout [layout of screen code here]

forever [ if new drive is detected and contains particular directory, 
view copy-screen, which should start copy-dir function]


.... somehow can't get it working properly. View/new can't help me 
here imo ....

Gabriele - how could I do it with view/new? I can't start copying 
after it, because I want to update screen. I can show some elements, 
but e.g. buttons will not be active, unless do-events is being run. 
That is why I complicated my script by implementing "delayed" copying 
= view copy-screen, there is an 'init style with rate 1 and engage 
defined. I check if it is a first run, and if so, I set rate to none, 
call show, and that is the place from  where I call my copy-dir function 
.....

you have to do view/new, then inside copy-dir you need to call wait 
0 every now and then (eg. every file you copy) if you need your buttons 
to work.

can I somehow have layout defined in the manner, that I can:


1) reinitialise it (I can imagine something like store-screen: [copy-dir-screen: 
layout []] do store-screen

2) how can I "unset" it, "uload" it, whatever, to be able to reset 
all its words?

If you are *really* constrained, create your UI with draw commands 
and manage events and everything yourself. Of course, you may end 
up with so much more code that it's a moot point, so it depends on 
what constraints you have. 


It reminds me of when VB1 came out, and people tried to use thousands 
of text boxes to create spreadsheets, which was slow, resource hungry, 
and limited in size, because you could only create so many windows. 
The common question was "How does Excel do it then?", and the answer 
was that Excel (v3 at the time I thnk) created only 3 (or so) actual 
GDI windows, everything else was rendered, and interaction handled, 
dynamically

Thanks for your input guys. I am not constrained that much. Our kiosk 
PCs do have 256MB of memory, being older PCs, some Duron 1.2 or so, 
WinXP. It runs presentation in full mode, local Apache based. Plenty 
of photos, simply kiosk presenation, not ordinary web pages. Dunno 
how IE is memory hungry. My script, which will run in background, 
easily eats some 39MB memory. I did some unsets, recycles, so while 
in wait mode, script goes down to some 24MB. It is that I was just 
curious how to aproach that ...

Hi, I'm trying to visualize data I get from a network port within 
a view window. IIRC using wait within a view handler is greatly discouraged 
(and my first tests seem to show it: the window is updated for the 
first info, but the updates stop after the second, or so. )
So, how do I connect to a tcp port from within /View?

data-i-want-to-display: [..............] ;probably a block
window: layout/tight [
	data-box: box 400x200 rate 0:00:0.001 feel [
		engage: func [f a e][

   if a = 'time [do-something-nifty data-i-want-to-display show f]
		]
	]
]
view/new center-face window
setup-port-listener
forever [
	check-my-port-and-update data-i-want-to-display

 ;add some window shutdown code to close ports etc after window closes..
]

Do I need to some how save "gui" face inside the loop?

How do I get y value to reduce before append