Read here for some ideas before you vote:
http://www.lsdev.org/e107_plugins/forum/forum_viewtopic.php?985.0
I mean "simple" as in its just as hard to use as OTS2 compatability is/was :)
Personally, I like the idea of widgets that a themer could include in his/her own theme. But as far as giving control over which widgets are loaded to the user, I'm going to have to say that's making it too difficult for the themer to create a theme that would look unified with the widgets.
Still, the I have nothing against widgets. But I do think it might make give themes less impact if implemented too liberally.
Yes, each theme should have its own widget list (maintained by the end-user)
And widget's position should be store in themeevars.rc to avoid conflicts with the theme's visuals.
Smurth: What? What if a widget is created after a theme? what if there end up being 100 widgets? Your idea is just not scalable IMO.
Maoserr: The widgets *could* skin themselves with GS, and so too *could* the theme, boom unified look ;)
GS? I never heard that before, what's that mean?
"Global Schemes"
Basically its a way to have any theme use any color scheme, and vice versa.
So you can have some color scheme that you really like (goes with the rest of your system or something) , try theme after theme and they will always us the same colors.
Also say you have this really neat theme (you made it yourself and its got all the cool stuff you like etc) , try scheme after scheme and although the colors are changing, the theme itself remains.
http://www.lsdev.org/e107_plugins/forum/forum_viewtopic.php?421.0
and
http://dev.litestep.com/projects/OTS3%20Projects/wiki/GlobalSchemes
talk about it.
the attachments I have released are pretty old though.
fractal.design: there's no problem if a widget is create after a given theme.
Just add it to the theme's widget list and !recycle :)
All I can say is that some widgets will suit a given theme; some other not.
It's up to the user to decide; that's all.
"It's up to the user to decide; that's all." No your saying the themer should decide, by adding a list of widgets that work.
read the thread:
http://www.lsdev.org/e107_plugins/forum/forum_viewtopic.php?985.0
get THC4k's OTS3 prop:
http://www.ls-themes.org/index.php?cat=article&show=ots3
see if his implementation works.
I'm still not entirely comfortable with THC4k's implementation - do we really want to make python AND lua part of the core?
python is just there temporarily, and I am unsure about lua being mandatory also...
Hmm, when I suggested a poll on this matter I got yelled at.
Darn right. Swedish themers should *never* be allowed to conduct polls.
Ya... what would those be like? "Who has the best blonde hair?"
C'mon xcal, haven't you seen West's avatar? He doesn't *have* any hair. And even if he did, after dealing with this widget thing I think he will have pulled all of it out.
Who yelled at ya West? Just tell me and I'll rough 'em up a bit ;)
Well, to begin with there's those two pesky Americans making fun of me. I don't dare talking back to them 'cause they might come and bomb me.
:D
No way, these days we only bomb third world countries that can't fight back.
And Sweden could fight back? I normally hate this overused acronym, but...
LOL
I doubt that the Swedish army is large enough to defend central Stockholm even :)
(if it weren't for the fact that the Swedish intelligence is probably equally undermanned, I'd be worried about being picked up and taken to an undisclosed location for revealing state secrets)
I don't want any widgets. My browser has widgets but I never even use them. I can't find any use for any sort of widget.
IMO, Litestep is just a uber widget handler. Modules are just widgets.
Especially when you look at themes like popsicle ;)
Bite your tongue! Those are "floaters", not "widgets". Floaters are different from widgets because... ummm, well... *West* will tell you why. Yeah, that's it.
Widgets can be anything, I guess, from non-visual stuff like new !bangs to intrusive boxes like meters, web-parser, etc
Hey! My meters are *non-intrusive* boxes. They're really quite well behaved. My web-parser *is* a bit uppity though, now that you mention it.
I guess you could even have specific keyboard model widgets, that supply all the special keys functionality (and load the volume modules etc to do the actions properly).
That could be very handy for teh noobies.
I dunno, but... before we decide on wether we should have widgets or not, maybe we should agree on what widgets actually are? :D
My definition (well, vague notion at least) is that a widget is a more or less standalone mini app that provides one functionality or other. Like, say, a clock, a VWM or a system meter. By standalone I mean that a widget is something that is NOT integrated into the interface, e.g. not part of a bar with other stuff on it. When someone says "widget", I think free-floating stuff that clutter up your desktop. I really don't see how a bang could be a widget. In that case, is a hotkey a widget too? Or a popup menu? *gasp* Maybe LS itself is.... a widget! ;)
Seriously though, this is getting confusing.
Yeah like I said, LS is just a widget loader really. Modules are basically non-GUI-widgets.
I think relating to Litestep, a widget is a theme that only does 1 thing (like act as a QuickLaunch, or a VWM, or hotkeys for you keyboard)
So really, by asking if widgets are possible, in essence we are asking if loading 2 themes at once is possible.
Thats the easiest way I can think of explaining it anyway.
It sounds to me like a widget, at least as smurth envisions it, is "a standalone piece of LS configuration code (and optionally graphics) that performs a useful function." So... a script that gets weather, a volume control "thingy", a package of definitions to make a particular model of multimedia keyboard play nice with LS - all of those *could* be widgets. They could also be part of a theme. I guess the question is, do we want to post/distribute both themes and widgets (AKA parts of themes), or just themes in their entirety, as has been the tradition?
@fractal
Nothing wrong with your explanation, it makes perfect sense. That was kind of the way I imagined it. Anyway, I'm still not too keen on the whole idea but I'm not going to whine any more about it. :)
The keyboard widgets though, that sounds like a neat idea. Would be great if we could get people to submit working key mappings for their media keyboards and ship a number of them with LS (mappings, not media keyboards). Then the users can just select their model from a popup and whammo -- working media keys. Come to think of it, I have a number of old media keyboards of various brands lying around. Plugging them in and creating mappings for them wouldn't be that much work.
@Boreas
Yes, I think we need to be pretty clear on what this is about. Personally I don't think (or like to think) that a script, with no GUI, that does one thing or other is a widget. Because if that is a widget too, there will be no need or use for traditional LS themes anymore. One could just load up a bunch of widgets and have all the functionality you need. Maybe you could even have container widgets that load other widgets, and thus eliminate the need for any themer to create even a bar or a box. What we will end up with is GeoShell, not LiteStep.
(OK, I said I wasn't gonna whine anymore, sorry ;))
I think it's generally a good idea to try and separate function and GUI as much as possible, inasmuch as it promotes code and/or configs that are cleaner and easier to maintain. fractal's global schemes are a step in that direction, as is the xpaintclass/xmodule development, and perhaps "widgets" are as well. OTOH, in some of the sleekest, most gee-whiz LS themes the two are very tightly intertwined. Can we have the best of both worlds?
@West
Personally I don't think (or like to think) that a script, with no GUI, that does one thing or other is a widget.
But for example...do themes really need to include a graphical representation of a VWM? I think not. So, if I were to use a VWM at all, I would be using hotkeys only. Therefore some widgets may have no visual interface.
Hmm, I guess you have a point there. And in that case I guess a script that, for example, grabs weather data and exports it to evars could be called a widget too. But still... I don't feel comfortable with all this prefab stuff. Who are going to create all those widgets anyway? Can anyone create and release them, or are we going to have some Widget Team in charge of creating good widgets and ensuring stability/compatibility etc?
I think the idea is that we will end up abandoning theme-making and just make widgets. :P
In that case I'm gonna make one huge mother of a widget that actually consists of other widgets. I think I'll call it... a "theme".
A widget that consists of other widgets? That sounds like a... fractal design.
Hey, you're right. In fact, I think I'll fire up xprop and zoom in on a few of fractal.design's themes... let's see if they actually have an infinite resolution.
Or perhaps you'll just see layer after layer of spaghetti.
I think we might be on to a whole new field of science here. Pasta-string theory, or something.
Whoa. That would certainly make us use our noodles.
“Freethinkers are those who are willing to use their minds without prejudice and without fearing to understand things that clash with their own customs, privileges, or beliefs. This state of mind is not common, but it is essential for right thinking;”
Its tough to grasp, I know, but I dont think widgets will end LS as we know it.
They were probably saying that back when OTS was released "Standardising theming!? It will crush creativity" etc etc.
Widgets: Work overtop of any "theme"
That also defines what a theme will be: Basically, the shell.
Its the bottom, the base, the structure that we need to build things upon.
And its not really different to how we have it now:
Today themes are released, with many or just a few features. And if somebody else wants a feature of that theme in their own theme, they look at the code, maybe even copy it directly.
This will still happen IMO, its just that, some features might be made into widgets, saving users copying time etc.
Is that so unforseeable?
I mean, if we want LS to become popular again, this is a surefire way of getting more new users.
Or do we not want more users?
I think I do.
I mean, I am not going to help everyone of them kill windows update, but it would be good to see a thriving community again, like more module coders, themers etc
oh and lol at the funny guys ;D