On Wed, 2004-06-09 at 17:56, Arnold Krille wrote: I think we've (perhaps?) finally figured out that we can't really have a "standard-LAD-GUI-elements-set". It will just turn into another LADSPA-GUI war, nothing will get decided, and nothing will get done. For example, you would like to volunteer to create one for Qt. Great! I hate Qt - even having the abomination installed on my machine is a comprimise as far as I'm concerned (just to give you an idea of how some people feel, I'm not trying to start a flamewar here). Some people would want it in Gtk, the Qt people would refuse.. Some people would want it in C++, the C people would refuse.. Some people would want it in some scriptey thing like Python, the Real Programmers would refuse.. etc. etc. etc. I think at this point all we need is a mechanism for a host to say "plugin, show your GUI now". (Incidentally what DSSI does as far as I know, which isn't very much yet). No embedding, no standard widget set, no crazy event loop stuff. Let's (gasp!) actually do something that will get done. (Of course noone is stopping anyone from making an audio-app-gui-widget library, far from it) GMPI will come along eventually and make it all but a memory anyway, non? -DR-

There is chasm both broad and deep between "plugin, show your GUI now" and an actual implementation of such functionality. Even on windows/macos, where the toolkit issues do not exist in the same way, its still the *host* that creates the top level GUI resources and integrates them into its own scheme for display management. steve knows all too well from the gmpi list just how thorny these issues are. "let's (gasp!) actually do something that will ge done" ... nice try there dave, but it just *isn't* simple. The LADSPA XML DTD that was posted/edited/discussed here 2 years ago would do a lot of what people have described here. But its dead in the water because it can't do it all. There is no magic bullet here that does not involve putting a gun to people's head and saying "do things this way". Oh, and the gun doesn't exist. --p

On Wed, Jun 09, 2004 at 08:41:47PM -0400, Dave Robillard wrote: why dont we specify behaviour and a gfx format for control animations and then implement the widgets for gtk and qt ? you volounteered for qt. i can do the gtk part. anyone for fltk ? i dont see a problem. there is knob code for every toolkit on sourceforge. lets unify the gfx data have a widget for every toolkit. -- torben Hohn http://galan.sourceforge.net -- The graphical Audio language

On Thu, Jun 10, 2004 at 12:50:19PM +0200, torbenh(a)gmx.de wrote: Brilliant, but get every aspect of the behavior down. For example, something that no one has discussed yet: if the knob is pointed at 3:00 and I click-and-hold on it at 9:00, does the knob jump to 9:00? I've seen knobs do that, and I have seen knobs which don't. I personally can't stand knobs that jump when I click them. It is so frustrating.

for DSSI, this seems great. but for LADSPA it seems like a major change in philosophy, one that would turn LADSPA into DSSI. i am not opposed to that, in fact i think its a fairly good idea, but i think we need to be clear about that if that is in fact the general idea. yeah, returning an XID sounds like a nice touch there. btw, i guess we should get busy with LADSPA v2 soon, eh? --p

On Thu, Jun 10, 2004 at 12:25:46 -0400, Dave Robillard wrote: You have to care about MIDI to some extent. Unless you actually want some of the features sticking with LADSPA seems like a better idea. The DSSI API is a bit heftier than LADSPAs. Not technologically, its just a name. I have some things that would be better done in DSSI that LADSPA, but mostly things that tend towards synthyness. - Steve

On Thu, 2004-06-10 at 17:00, Steve Harris wrote: As long as you never, ever have to look at it. I could actually read the XML stuff in the old Ardour presets files. From what I can see the RDF/N3 format is somewhat similar to cuneiform or possibly Sanskrit ;-) Jan

On Fri, 2004-06-11 at 04:02, Steve Harris wrote: Given a choice I'd rather have "easier to parse" as long as all of the tools needed for people to deal with it are written. Jan

On Thu, 2004-06-10 at 10:01, torbenh(a)gmx.de wrote: I kind of agree - I don't really see the point of embedding plugin GUIs in anything. A modular with the controls like SSM maybe, but you wouldn't be able to see anything, modules being that big. But I probably just havn't thought of it. Screenshots? (other than reason :) ) -DR-

On Thu, Jun 10, 2004 at 04:01:25 +0200, torbenh(a)gmx.de wrote: Never really thought about it to be honest. Ah, I didnt realise the GUI had to do anything special, thats significant - what are the requirements or limitations? Sure, I understand the motivation, I think its a good idea, just never occured to me. - Steve

Dave Robillard wrote: But as far as I understand, it's not about GUI-elemt toolkits but about a list of GUI elements and a (or some) ways to place them. Then every host can implement its own set and / or we can provide GUI libraries for the different toolkits. Uwe -- voiceINTERconnect www.voiceinterconnect.de ... smart speech applications from germany

On Wed, Jun 09, 2004 at 11:56:59PM +0200, Arnold Krille wrote: Other way around? Define a graphics system so I can add my own knob-graphics, and you make them work.

On Thu, 2004-06-10 at 10:22, torbenh(a)gmx.de wrote: What about MNG? It's basically just animated png, is it not? I know there's a libmng, but I've never used it so I don't know if it's any good. -DR-

On Wed, Jun 09, 2004 at 04:23:32PM -0700, Tim Hockin wrote: It seems like there are as many preferences as users. How about ~/.knobrc ;) - Steve

On Wed, Jun 09, 2004 at 08:40:09PM -0400, Dave Robillard wrote: I just can't get used to radial turning, I find it obnoxious, and know many who feel the same. Linear was obvious to me. Radial was very non-obvious. no! Having the two axes mean different things is bad. No one mouses in a perfectly straight line. There really needs to be some global option for radial vs. linear knobs. I want all my knobs linear, always. so I see: * left-click = coarse asjust (radial or linear) * ctrl+left-click = fine adjust * double-click = edit I don't like a 'reset' button. Reset to what? 0? the preset? the last save-point? If there has to be a 'reset' it should be in the context menu (though I guess I don't feel too strongly about it :) Too many clicky options is confoosing.

On Wed, 2004-06-09 at 19:23, Tim Hockin wrote: Okay, (combining your post with my last one on the same subject...) how about this: I think radial adjustment is the clearest, most obvious way for a knob to move.. it is a knob after all. A newbie sitting down at the machine isn't going to try to drag it vertically, they're going to try to "turn" it like a real knob (UI people have some term for things that don't act how they look, but I can't remember what it is.. it's bad design though). But I can also definitely see reasons for linear adjustment (ie adjusting with your eyes closed) How about radial being the default (as it's clearer and the most obvious, as well as being good for coarse adjustments), then hold ctrl to fine adjust linearly. Of course, this is leaving out linear coarse adjustment, but can we really come up with a way of having all of them? Yet another hotkey :/ We could have left/right be coarse and up/down be fine.. it seems pretty arbitrary but the more I think of it the more I think it wouldn't be that bad. Confusing newbs isn't an issue if it's hidden under a hotkey... thoughts on this? I really do think radial should be the left-button default though, it's just what a knob visually suggest it should do. If linear should be the default, the widget itself should be a fader IMHO. Think of the poor newbs! :) -DR-

On Wed, 2004-06-09 at 21:03, Tim Hockin wrote: If it's determined that linear is indeed better I think the knob should change into something that makes it clear that up/down linear movement is what's required then. Grabbing a knob and moving the mouse up and down just isn't intuitive, and smashes the hardware metaphor to tiny little bits. -DR-

I for one dislike the knob design, you just cannot make fine adjustments easily. Someone said that you can take the mouse out to a greater radius from the knob, but that is just plain silly. While you move the mouse out from the knob, you're bound to make small adjustments to the value of the knob, unless you're a super-human, ultra-precision mouser! This means that you're original value of the knob has been lost, and you'd have to live with that, or move it to it's original position first, and then make your slight adjustment. Also, it's just not natural for me to move my mouse in a circle, the natural movement of my hand is not a circle, try to draw a perfect circle in the Gimp sometime by mouse! This means, that at different parts of the knob tweaking, you'd be changing value by an unpredictable amount depending on how good your circle is. This means also that left-handed mousers are going to change values at the opposite sensitivity level at different parts of the motion. There is no easy solution as far as I can see, but what about a few physical knobs, cheaply built, that send some kind of data via USB or something that can change values. That's the best way I reckon, except for the cost factor. I'll keep thinking about a purely software solution though. Dan -- Dan Harper http://danharper.org

On Wednesday 09 June 2004 21:54, Dave Robillard wrote: } On Wed, 2004-06-09 at 22:25, Dan Harper wrote: } > Also, it's just not natural for me to move my mouse in a circle, the } > natural movement of my hand is not a circle, try to draw a perfect } > circle in the Gimp sometime by mouse! This means, that at different } > parts of the knob tweaking, you'd be changing value by an unpredictable } > amount depending on how good your circle is. This means also that } > left-handed mousers are going to change values at the opposite } > sensitivity level at different parts of the motion. Gimp does have a circle tool. It's designed to get around the limitations of the mouse. Vector drawing programs have managed to get past them for vears. } > There is no easy solution as far as I can see, but what about a few } > physical knobs, cheaply built, that send some kind of data via USB or } > something that can change values. That's the best way I reckon, except } > for the cost factor. } > } > I'll keep thinking about a purely software solution though. } } Well, the hardware is always an option of course (a fader/knob box is } top on MY list-of-things-I-can't-afford anyway) but the UI still has to } be as good as possible. } } Can't map a MIDI controller to every single knob you need to tweak! You could if you mapped them to spheres on a 3d grid... { you could indicate relative frequencies with color or with some sort of spatial thing... like a 3d model... spheres would take on a hotter color to indicate higher frequencies or you could simply manipulate them in space... given a nice open gl box... you could set it up fairly easily. {Sgi... digital performer} Maybe you could use both color and space. Besides being accurate... it would be fiun {see blender}}} across a bezier line, a bezier plane or variously in 3d space. You might check the STK pages for some nifty alternative controls. They do something similar to the 3d control I've described above. You might also check the eyesweb pages... they take the thing to extremes. I can move an optical mouse fairly accurately... I can move the wheel on it pretty precisely. If I were to click on a control to activate it and then turn it with the wheel... it would work just fine. If you could place control points across a line and manipulate it with a bezier handle you'd be able to get a fairly high degree of control. If you could map that into an additional, dimension you'd be able to make adjustments through time with a very high degree of accuracy. -- If I had saxophones / Big baritone, cleanin' up the muddy breaks If I had Saxophones / I could get some recognition from that Mobile Alabama DJ {J.Buffet}

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Thursday 10 June 2004 03:03, Tim Hockin wrote: Radial movement on control elements often confuses me. If there was a line drawn from the center of the knob to the mouse pointer, maybe sporting arrows in the directions to move the mouse would make two things obvious for me: - - to move the mouse radially - - that there's different scaling available The display of that "helper line" could be configurable and even defaulted not to be used. Absolute values could also be displayed next to the mouse pointer. The scaling issue may be obvious for most people, but it wasn't for me until this thread. I just hated radial knobs. - - Burkhard - -- Libre Audio, Libre Video, Libre Software: www.AGNULA.org Public key available here: <http://blackhole.pca.dfn.de:11371/pks/lookup?op=get&search=0xFD82303B> key FP 0A65 5E83 F44F 47A5 3DFC 19C5 7779 E411 FD82 303B "Kenneth Brown is president of the Alexis de Tocqueville Institution and director of its technology research programs. [...] He is reportedly "not the sharpest knife in the drawer," but nevertheless is able to converse with many intelligent people, and is accepted at fine restaurants and hotels around the world." - <http://www.adti.net/samizdat/brown.reply.june.04.html>

On Fri, 25 Jun 2004 17:28:09 +0200, Thorsten Wilms wrote ... This is an interesting topic, I think I've missed most of it though :( Personally I'm a radial knob kinda guy - I remember using the scaling-by-moving-the-mouse-further-away thing years ago with spiralloops. The other advantage of knobs is that they take up less space than sliders, and to me collectively look more like a musical instrument controls than a bunch of sliders (I don't know where this comes from) I'm not anti sliders as I use a lot of them - although partly because they are better supported in toolkits (I'm currenly using tk (yes tk!)) and I don't have a choice. I like your fan idea Thorsten, but I also think it could work invisibly - ie no need for the transparent overlay. This would take a bit of learning that it was there to begin with - but transparent graphics like that are expensive (yeah, GPU's I know - but its something to keep in mind) and unnecessary after learing how to use the control. It would be good to collect a repository of existing GUI elements in commercial packages - dunno what the legal implications would be, but it would be useful. cheers, dave

On Fri, 25 Jun 2004 18:54:20 +0200, Thorsten Wilms wrote You would be able to get a feel for it by the change in the motion of the slider control - I think it would become quite apparent once you knew it was there. I quite like hiding this sort of functionality in user interfaces for users to discover :) annoying - but agree that it's sometimes required. I think it needs to be tried. It's not just the individual widgets that are important but the whole design. I always post this, but: http://girl.yowstar.com/app.gif and of course: http://www.ableton.com/pages/products/live/live4tour/featurelist/Session-fu… are often cited from users as "ground breaking" interfaces (others hate them :] ) I agree that all these designs simply rehash the old themes of interfaces - I think we should explore this more as free software types, as we can take risks with this sort of thing (but rarely do). cheers, dave

On Fri, Jun 25, 2004 at 06:00:42PM +0200, Fons Adriaensen wrote: You again. If the user has not even an idea, that there is some special functionality, he has no reason to even look for documentation. Besides, we were talking about widgets. When even single widgets would require to RTFM, what would that mean for a full app? Users should be allowed to concentrate on the problem space of the task at hand as much as possible. Advanced options / complex functionality that one needs to read documentation for is ok, but let's keep simple things simple. Happy now? --- Thorsten Wilms

On Fri, Jun 25, 2004 at 12:09:24PM +0100, Dave Griffiths wrote: I'm all for trying out new things (that should be obvious by now). But a widget should not need documentation, users should be able to figure all major features out by themselves. Additional features accessibe by modifier keys are the only thing where I think it's ok if there's no visiual hint. --- Thorsten

On Fri, Jun 25, 2004 at 08:29:44PM +0200, Thorsten Wilms wrote: Yes, me. I was told by some members of this list only a few days ago that sarcasm was OK. If there *is* some documentation, I see no good reason why it shouldn't be read. Unless the user thinks of himself that he's so bright that he will understand everything instantly. Educated people usually are a bit more modest, and don't mind some effort in order to learn something. Yes, thank you. Just as happy as I was before. -- FA

On Sat, Jun 26, 2004 at 09:58:19AM +0200, Thorsten Wilms wrote: Yes, even if a are a good idea, that's a weak point. I see no easy solution for that, except by providing some form of documentation, be it a text in a status line, a 'tip of the day' in popup, or a manual. Curiosity ? It only kills cats, and then only occasionally. Nothing. It's just not always possible. If a manual is well written and explains both the concepts behind an application and it's modus operandi, the answer is probably 'yes'. At least I do. There have been some references, both direct and less direct, to my 'strange elitism'. I want to put this straight, once and for all, even if is quite OT. Let me start with an example that will put us right at the heart of the matters that I'd like to explore. A few weeks ago I bumped into a journalist I used to know when I was working for public broadcasting years ago. Meanwhile he has become one of the editors of the most popular TV news show here in rainy Belgium. We had a drink, and this is part of our conversation: "... whenever we invite someone into the news studio, he has to make his point in 12 seconds. If I know he can't, I will not even consider him." Why not ? "Beacause it's bad TV, and our market share will go down. Our viewers just don't want to concentrate on any issue, they want to be entertained. Anyway that's what our advertisers tell us: we want the largest segment, which are the people who don't want to think or do any other mental effort." This is just one of the many things I learned over the last years, and which all point into the same direction: the main social dividing line in most western societies these days is one that reflects education levels. It determines lifestyle, consumption patterns and political preferences, and it is much more influential than financial status or the old social classes. There is in the western world a growing number of people who just want to be taken care of in all respects, be it material or spiritual. For this group, everything has to be *easy*. They will refuse any mental effort, are unable to concentrate on anything for more than 12 seconds, and will certainly not read a manual. They will also (as one Belgian politician recently stated, and that made him very popular) see 'intellectuals' as the root of everything that goes wrong in the world. They will never accept any personal responsability - if I burn my lips drinking a coffee that was served hot, I just sue the bar owner or rearrange his face, if I get blocked in a traffic jam that's the fault of all these other people who can't drive and should take the bus, if I lose my job that's because of the moslim immigrants. I will not speculate about the causes for this. A more interesting question is the classic 'cui prodest' - who is likely to profit from this ? The commercial world is one - easy money can be made from ignorant consumers that just want their impulsive needs to be satisfied instantly. Politcians are another - in a democracy a lot of power can had by just telling people what they like to hear, no matter how idiot their ideas are, and that can be done in 12 seconds. Both these parties have a short term interest in stimulating a tendency to dumb down society as a whole, and short term interests is all the care about. More and more, as an 'intellectual' I find myself in a position that comes down to this: either you budge and dumb down, or you'll be excluded. This is just one step from what happened during the Nazi regime, the 'Cultural Revolution', or the Pol Pot government, where everyone who dared to think was just eleminated. As someone who has had the privilege of some education, I wil not look down on those that have been less lucky, and I am quite capable of loving and respecting any person for his human qualities - these have nothing to do with formal education. But I *will* look down on those that give up their basic human dignity - the possibility to learn, explore, advance, change your mind, move your limits, to be free - in exchange for easy comfort and instant satisfaction, while three quarters of the world's population is in dire need of better education and *is* willing to invest some effort. If someone wants to call me 'elitist' for that, please go ahead. Going back to our original point: if a user is too lazy to read a manual, I can't be bothered with his problem. And if someone proclaims that aversion to reading documentation is 'normal', I will disagree, and now you all know why. -- FA

On Sun, 2004-06-27 at 18:22, Fons Adriaensen wrote: Once again, I've been writing graphical editing software for use by scientists for 26 years. I have Mensa members that don't want to read a manual. Why? Because they're scientists. They want a tool to do something with and the tool shouldn't be hard to understand. This is the same situation that you're going to get with musicians. They want a tool to make music with and it needs to be fairly easy to work with. Granted, something like Ardour will be difficult to master but you should at least be able to use it at a rudimentary level without having to read the manual (since there isn't one at present this had better be the case ;-) When you get farther into it you may need to delve into whatever documentation is available. Aversion to reading documentation *is* "normal", at least among most scientists and musicians that I know. Hence the common phrase - "When all else fails, read the documentation". Jan

Pete Bessman wrote: [ ... ] Fons is an intelligent human being, and even if he is being somewhat more "elitist" than some may consider necessary, there is nothing gained by berating his viewpoint with sarcasm. That's a straw man too. Fons' point that you appear to disagree with is "a (volume) knob may indeed have some functionality that is not easily discoverable by playing with it". I personally am undecided about the merits of such a design - i don't like knobs much at all for any purpose - but it would aid the discussion if it could stay focused on the contents rather than the moral elevation (or otherwise) of the participants. --p

At Sun, 27 Jun 2004 23:03:15 -0400, Pete Bessman wrote: So it seems it wasn't clear that the above was a humorous jab, not meant to be taken seriously. My bad. My sense of humor must be "special." This *was* a debate about the performance requirements of Thorsten's fan sliders. I personally think that machines which can't deal with the sliders would be better off running trackers; high-end audio already comes with a high-end performance tag (relatively speaking), and if you can afford a 1Ghz/512MB system, you can afford a GPU. Really, you can: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&category=40160&item=5… The argument got "ugly" when Fons took a point of Thorsten's out of context (if you define "ugly" as the point at which I got pissed and started flinging argumentum ad hominem). "Requiring the user to read documentation to learn about functionality he would not even expect is not an option," is clearly wrong on a general level. Therefore, Fons is technically right; the only way such a requirement could exist is if the human desire for knowledge has dropped to a pathologically low level, either from lack of education (*snort*) or other causes. However, the point is clearly right as it applies to the fan sliders. A user with an IQ >= 100 shouldn't need to consult a manual to spin a knob or flip a switch, and if you think otherwise you should bar yourself from participating in UI discussions. Skewering a point on a literal and pedantic level is a sure-fire way to get labeled an elitist. Moral of the story: make a snide and barely on topic remark, get a face full of flame. -- Pete www.gazuga.net

On Mon, Jun 28, 2004 at 08:47:48AM +0200, Jens M Andreasen wrote: I don't live in your house. -- FA

On Fri, Jun 25, 2004 at 03:38:10PM -0400, Lee Revell wrote: See my other reply. Yes, that's true. I remember driving in the US, and how I found it very odd for all the roadsigns to tell me in plain English what I was or wasn't supposed to do. But 'graphic' signs can be completely useless as well. At work I have a telephone with something like 20 extra buttons besides the numeric keypad. They all have mysterious signs on them that defy both my imagination and logic reasoning. I'd much prefer for them to be labeled 'Hold', 'Transfer', etc. than by the totally uncomprehensible things there are now. If you have a mousewheel :-) -- FA

On Sat, Jun 26, 2004 at 01:33:59AM -0400, Lee Revell wrote: I wonder if such an icon can be found for e.g. 'Hold'. There are some other examples, such as transport controls. "<<" and ">>" are clear enough, but what about '||' for pause and a square for stop ? I'm never quite sure about them, even if these signs are everywhere. I'd much prefer just the text. Of course, then the user needs to know the language, but OTOH icons and colors are not culturally neutral either, and most activities and trades have their 'preferred' language. I agree it's a very natural way of controlling a radial control or a spinbox. -- FA

On Fri, Jun 25, 2004 at 06:15:24PM -0400, Pete Bessman wrote: The most avant-garde music is enabled by very dull interfaces, in my experience. They do, but maybe not in the direction you imagined. And cro-magnon times are well in the past. -- FA

On Fri, Jun 25, 2004 at 07:55:50PM -0400, Pete Bessman wrote: What is "normal" ? I have an increasing difficulty in just understanding what you try to say. Could you explain the terms - mouth-breathing - crow-magnon (sic) music - shiny-quarter interfaces - A-OK (is that the opposite of OK ?) Where is the criticism ? Where is the sneer ? I made the observation that educated people usually do not mind having to learn something. So if there is a widespread aversion to having to learn and read a manual, that seems to indicate that education levels have gone down. -- FA

On Sat, 2004-06-26 at 02:36, Fons Adriaensen wrote: I have been writing graphical editing software for scientific data (sonar, magnetics, gravity, lidar) for over 20 years. Everyone that uses my software has at the very least a BS degree. Many have one or more MS or PhD degrees. I put context sensitive help in every piece of software I write. You can click on any graphical object and get *very* detailed instructions on how to use it and what to use it for. These people do not have an education problem. Do any of them read the context sensitive help, much less the documentation (1 button press on the desktop), that I write? Maybe 1 out of 100. I feel that I have failed in GUI design if the user has to read documentation for code that I write. Granted there are advanced things that you can do that you need to read the context sensitive help for but simple use of the tool should be self evident. My two cents ;-) Jan

On Sat, Jun 26, 2004 at 11:50:29AM -0400, Pete Bessman wrote: You will not trick me into the same abuse. Bye. -- FA

On Sat, Jun 26, 2004 at 01:41:17AM +0100, Dave Griffiths wrote: I agree 100%. it's part of life, and of being a human. -- FA

On Wed, 2004-06-09 at 19:39, Dave Robillard wrote: How about shift left click to zoom the fader in and shift middle or right click to zoom it out. Center around the current value. I'm thinking about trying this with JAMin soon (if I ever get time). Jan

On Thursday 10 June 2004 02.53, Tim Hockin wrote: In MusE I use the mouse wheel to edit both sliders and knobs. I find it to be superior to clicking and dragging. Sort of a poor mans control surface. Ctrl+wheel is used for more finegrained operation. /Robert -- http://spamatica.se/music/

On Thu, 2004-06-10 at 05:33, Alfons Adriaensen wrote: Way too finicky and difficult IMHO. -DR-

On Fri, Jun 11, 2004 at 12:58:27PM +0200, Jens M Andreasen wrote: Indeed it's the same as radial when you are close to the control, and turns into linear as you move away from it. There is no need to make 'circular' mouse gestures. -- FA