World: r3wp
[!RebGUI] A lightweight alternative to VID
older newer | first last |
Ashley 27-May-2005 [1059] | Can anyone else verify this with 1.2.113 on WinXP. I think it may be 113 / pre-XP specific. I also noticed that reducing the width now gives the following error: ** Script Error: Invalid graphics face object ** Where: action ** Near: show pf/pane/:pane-idx data-idx: data-idx + |
Allen 27-May-2005 [1060x3] | I'm on XP. Not seeing shadwolfs issue at all. |
Invalid graphics face, yes seeing that. Another 0x0 face I guess | |
I've asked Carl. if View will be forgiving of this 0x0 case again, or if we need to test/handle for all guis. I'll post here when I know the answer. | |
Anton 27-May-2005 [1063x6] | I think it's really bad to not support 0x0 sized faces.... |
I've been checking out RebGUI. A number of very good design decisions were made. I've ported my pair-edit style to be rebgui widget. http://www.lexicon.net/antonr/rebol/rebgui/demo-pair-edit.r (It needs editing to locate your rebgui directory.) | |
Actually, locating rebgui is a problem. How does a script such as the above get the latest distribution and unzip it ? I would prefer the rebgui files to be uncompressed (or compressed using rebol, at least) and there should be a standard way of determining the latest version. | |
So a script should be able to check for the latest version and download it into the public cache (which may be successful or not, depending if net connection is available), eg: do load-thru http://www.dobeash.com/it/rebgui/get-latest.r and then just do path-thru http://www.dobeash.com/it/rebgui/rebgui.r to initialize (which automatically falls back to any previously cached version). | |
Another issue: Excuse me if this has come up before.. The archetypal View face has some of its facets set to none, which makes using VID styles alongside RebGUI more difficult (because the VID-based styles are not expecting those facets to be changed.) Suggestion: create widgets/rebgui-face and encourage widget creators to make from that. | |
And now a question: widgets: make widgets [ my-widget: make face [ ... ] ] the best (or only) way to add your widget to rebgui, making it available to DISPLAY ? | |
Robert 27-May-2005 [1069] | table: Is the table widget missing an action when clicking a line? Of how do I add this feature? Is there a way to color odd/even rows differently? |
Volker 27-May-2005 [1070] | mixing vid with rebgui has also some api-problems. some i found: /action: has in vid this args: [face value]. value is something usefull, like area/text, slider/tdata, first text-list/picked etc. maybe its not needed, "value" saves not much to "face/text". but supporting that arg would be compatible /init: in vid this is a block and 'done. in rebgui it is a function. vid has the advantage that init can be easily extended in substyles, just add to it. rebgui has locals because it uses a function. /init2: not called by vid. resizing: not in vid, but one could write a /resize and plug it in the vid-face. also there are some global changes which may break other things (removing event-func and changing global face IIRC). If there is interest in making such small changes, i would help with making the rest compatible. adding #xh - parsing to 'layout and such. for "keep small" rebgui could have a "load-time-flag" to decide about using vid-stuff or all own. What i would really like is rebgui for complex style-building (area with slider etc). this is closer to raw faces. and vid for quickly writing "pages", IMHO it has a rest of advantages there. (maybe not, i am a rebgui-beginner). |
shadwolf 27-May-2005 [1071x2] | Robert looks to my listview code ;) |
http://shadwolf.free.fr/rebgui-list49.r | |
Robert 28-May-2005 [1073] | I already did ;-) and got it mostly working. but not that straight forward as I like to see it. |
Ashley 28-May-2005 [1074] | Anton 1) Design decisions: thanks; if you think any of them were *bad* then let me know, that's the reason it's still alpha ;) 2) I like the edit-pair demo, I'll look at adding a cut-down version of it to %tour.r 3) Versioning / caching was something I was going to look at once RebGUI goes beta and / or we have more of a handle on how REBOL/Services works. 4) "local" face definition: good point, added to next build. 5) Extending widgets object: Yes, RebGUI doesn't have a "stylize" system so you have to extend the widgets object when you need to add your own. My hope is that with a functionally rich set of widgets this will be the exception rather than the rule! Robert Action handler not yet added to table widget. I'll add odd/even coloring to the ToDo list. Volker [face value] - less important for RebGUI as face/text or face/data covers almost all cases; but I'll think on it a bit more as the compatibility aspect is worth considering. /init - I'll look at making this a block in the next build. event-func removal - I'm considering a "load-time-flag" as you suggested system/standard/face - as per comments to Anton above "If there is ..." - agree fully with everything stated here ;) RebGUI development is taking a slight pause until View 1.3 is released, as I want to baseline against that and remove all the pre-1.3 exceptions that have crept in. |
shadwolf 28-May-2005 [1075] | Ashley you are right we need to base RebGUI on REBOL/View 1.3 that's the way it must be done ... |
Graham 28-May-2005 [1076] | What's the time line on focus and tabbing between gadgets? |
Anton 30-May-2005 [1077x2] | Ashley, I had also wrote a VID integer-edit style, then I decided to make a single function to generate pair, integer and decimal-edit styles. I could convert this in future. It kind of violates your "each widget self-contained" design feature, but the code shrinkage was too hard to resist, and the potential is there to support more datatypes... |
5) Extending widgets object: I was looking for a doc to show how to create the simplest widget ever. If you are happy keeping it this way (ie. "do it yourself") then I could write the example script and accompanying doc, to add to the Widget Designer's Guide. | |
Ashley 30-May-2005 [1079] | ... simplest widget ever ... How much is there to write about? widgets: make widgets [my-widget: make face []] Or am I missing something? ;) Your VID edit styles sound interesting, but I'm looking for a more universal solution to "field input masks" and "field input validation". Something that would let me define a telephone number mask for example. |
shadwolf 30-May-2005 [1080x9] | some thing like a ###-#### for fone number for example and ??##??? for 2 strings followed by 2 number then by three strings |
like 555-1234 | |
or AD21SQF | |
rafinement could be a coded like mask like 3#-4# (for fone number mask ) and 2?2#3? (for the 2chars then 2 number then 3 chars mask ) | |
as fone number are different betwin areas it would be nice to can specify localisation for masks rules in the localisation file ;) | |
so ? = what ever a to z ot A to Z character and # = What ever 0 to 9 char all other caracteres in it are considered as always needed things. Repetition can be simplify be using numbers in the mask | |
like 2?-3#-5? | |
I think parsing rules that complains to my description is not so hard to do ;) | |
when an entry doesn"t feet to the mask a message must be displayed showwing and automated example of what must be a good entry to warn user :) | |
Ammon 30-May-2005 [1089] | Nice, try just typing rebgui into the location bar on FireFox... |
shadwolf 31-May-2005 [1090] | Ammon :when I type rebgui in the firefox 1.0.4 location bar I get the redirection to the rebol france article (in french) I wrote ;) king of little presentation for french readers |
Ammon 31-May-2005 [1091] | Hm... You must be in france and Google's france branch must be redirecting you differently... |
james_nak 31-May-2005 [1092] | Ashley, very nice work. I have probably asked everyone in this group at one time or another about how VID works (or doesn't seem to). I really appreciate that you are working on something like this. I had all but given up trying to get my interfaces to function the way I wanted them to. At some point one has to just give in and focus on the actual program. Now I'm thinking I can have my cake and eat it too. Thanks big guy. |
Anton 1-Jun-2005 [1093x2] | Ashley, I felt the need for this kind of validation / masking as well. Just never got around to it. It's not simple. |
Simplest widget ever : you got it. :-) That saves a lot of time getting started. | |
james_nak 1-Jun-2005 [1095] | What is the mechanism to know which line in a table has been selected? |
Ashley 1-Jun-2005 [1096] | Not implemented yet. |
james_nak 1-Jun-2005 [1097] | Thanks. |
Ashley 2-Jun-2005 [1098] | Latest build available at: http://www.dobeash.com/files/RebGUI-024.zip Highlights include: - New edit object / feel added in a separate file, %rebgui-edit.r - Documentation on above at: http://www.dobeash.com/it/rebgui/edit.html - Face redefinition localised to widgets object - Radio-group color toned down from green to leaf Known issues: - auto-fill broken for the time being - edit feel needs to support tab / shift+tab into and out of "grouping" widgets - cursor down goes to end of text instead of next line?! (yell if you know the answer as I'm stumped on this one) On the drawing board - area widget needs a scroller added to it by default - field format / validation masks - inline spell-checker |
Allen 2-Jun-2005 [1099] | ashley I'll take a look at auto-fill on the weekend, and try and get it to behave like the VID version does. anything to look out for? |
shadwolf 3-Jun-2005 [1100x3] | request-color ** Script Error: in expected object argument of type: object port ** Where: wake-event ** Near: if in pop-face/feel 'pop-detect [event: pop-face/feel/pop-detect pop-face event] >> |
drop-list ** Script Error: in expected object argument of type: object port ** Where: wake-event ** Near: if in pop-face/feel 'pop-detect [event: pop-face/feel/pop-detect pop-face event] >> | |
with rebol/VM 1.2.116.3.1 and windows XP SP2 ( i don't experiment those bugs by my own there where submitted in the Form of rebolfrance.org website by cr8825.) | |
Ashley 3-Jun-2005 [1103] | Allen: Thanks, should be a bit easier now that we have a [relatively] complete feel / focus system that is similar to VID's. ;) shadwolf: these occur with latest View releases. They'll be corrected once View 1.3 is released. |
ChristianE 3-Jun-2005 [1104x3] | Regarding use of color in RebGUI: I don't think color choices for LEDs in RebGUI are good, which is namely because the default colors are red and green. That not only excludes the color blinds, but even the red-green-blinds (sort of lesser color blinds), which applies to even more people (mainly male ones). Using blue and yellow would make things slightly better, but would still be a hurdle for full color blinds. Probably different shapes would be more usefull here, as they'd give an additional clue. And, finally, the colours are named wrong in the LED group: check-group 60x4 data ["Red" true "Green" false "Blue" none] Believe me, I had to measure the valve values to find that out ;-) |
And personally, I find the visuals of the sliders and draggering irritating, too. The gradient in the sliders container IMHO doesn't integrate in the overall RebGUI appearence - since it's (button left aside) nearly the only pseudo-3d effect - and the visual contrast between dragger an container IMHO isn't very strong. I usually find me clicking in the container when I expected to grab the dragger. The gradient in itself is somewhat unsual, too, since it makes the container appear standing out of screen, if one assumes that light falls from upper left corner (which is how most pseudo-3d screen elements are designed). | |
draggering = draggers | |
Brock 3-Jun-2005 [1107] | I second Christian's thoughts on the dragger and gradient. I also found myself reaching for the gradient section thinking it was the dragger. Because it has more graphic detail, it attracts the eye first... that should be left for the dragger/button itself in my opinion. The background of the slider area should be flat. |
Robert 4-Jun-2005 [1108] | Yep. And the current style doesn't fit to the rest of RebGUI. |
older newer | first last |