Pie menus
How about implementing pie menus?
Pie menus are menus which have the options displayed in a circle, instead of a list, as in traditional menus.
Pie menus allow faster selection of items compared to standard menus when using a mouse.
How about implementing pie menus?
Pie menus are menus which have the options displayed in a circle, instead of a list, as in traditional menus.
Pie menus allow faster selection of items compared to standard menus when using a mouse.
Comments
Pie menus
That sounds like a really neat idea! I'd vote to include it in R2 as a native widget--something to distinguish Haiku in yet another useful way, if used in moderation.
I think if the default Haiku menu was (what I would call "properly") categorized, with maybe 4-8 main categories, this would make a neat default thumb-button or middle-button assignment, seeing as you'd want to call it up in the middle of the screen rather than where the Be menu is now in the corner.
Kev
Pie menus
I think that we would need some examples/mock-ups. I know that Comcast has The Fanâ⢠and I think that it is _terrible_! No lables, hard to know what you would want to look at, etc...
In my mind, that would be a great example of what _not_ to do.
Re: Pie menus
I personally tried the RadialContext menu for Firefox and was appalled at how annoyingly difficult it was to use.
While it may seem like a good idea for some context menus (especially the smaller ones) - i think most people would still prefer the default line-item menu (after all, people have been working with lists/lines of text for pretty much forever)
Pie menus
I certainly don't have to credentials to know if this is fesable, but it would be interesting if Haiku was built so that this was an option, or even a way to install a right-click menu component which would be equal to and replace the standard.
I'm tired and now I don't think that made any sense. Bacon is good and I love Jolt Cola.
Pie menus
Kev: yeah, that sounds like a good idea. Although it might be a better use of time to get the concept working rather than doing a mockup.
umccullough: The RadialContext implementation is somewhat of a pain, I agree.
I think it could be done much better.
1. Display text instead of icons, easing learning.
2. It's too easy to accidentally open a submenu. Open submenus with the same action as regular items.
3. Delay showing the menu until the mouse button is released, lowering intrusiveness for expert users.
4. Put frequently accessed submenus at North, South, or East, making access easier.
5. Sometimes it's too context sensitive, hiding very frequently used items.
6. Lack of keyboard acceleration.
All of those issues are with that particular implementation, and not pie menus in general.
You could try the Easygestures extension it is somewhat better than RadialContext. But it has the issues mentioned above, except #2, #4 and #5, and two more:
1. Submenu implementation is pretty bad.
2. Total lack of context sensitiveness.
I'm looking into correcting the faults, hopefully I'll finish it soon. Please give it another try once I finish...
I think pie menus should be the default, as they offer much faster selection than standard menus.
hank: that certainly would be possible. But I think it would be better, when a particular pie menu has already been laid out, to only use that one.
Pie menus
I'll give that a try... are you involved in the development of easygestures?
Pie menus
Not offically, but both of the extensions are open source, and my understanding of Javascript is decent. I'm modifying RadialContent instead of EasyGestures, because the code seems much cleaner.
Pie menus
Okay, I finished some modification.
To install, dowload the below file, tools -> extensions, then drag the file to the extensions window.
http://mercury.walagata.com/w/yahoolian/RadialContext_firefox.xpi
Still more to go though.
Pie menus
Pie menus as the default interface (in a specialized version of Haiku) would make us one step closer to Haiku on the iPod 8)
Pie menus
The more i read the more i prefer the extension-based concept of chrisari... instead of integrating it into the OS; it should be offered as an alternative thing that the user can install to his wish.
This is because pie menus take time to get used to; and also are generally less descriptive to guess. Optimoz pie menus for firefox is cute though
Pie menus
A better example of pie menus (aka marking menus, spider menus, compass menus, etc.) would be Alias Maya or AliasStudio. They're actually very intuitive.
In order for them to really work correctly, it's best to use 8 main entries in the menu (N, NE, E, SE, S, SW, W, NW) and have overflow on a list below the pie menu. It's easy to use for a beginner because you can hold it open and look at the entries. Or, if you've memorized the locations, it's quite easy to simply flick the mouse and select an entry gesturally.
Again, when done correctly, pie or "marking menus" (as Alias calls them) can greatly accelerate workflow, yet keep all of the functionality of the regular context menu. There's a great paper written about them online somewhere... used to be on Alias' site before they were bought by Autodesk. I'll post it if I find it.
Here's an image:
http://www.gnomononline.com/tutorials/selections/selections.jpg
If anybody lives in Southern Calfornia, I'd be happy to give a demo on Maya. :D
Greg
Pie menus
I'd prefer six main entries, or four or three. triangular, square and hexagonal menu still is clear, I don't think one with eight choices is clear anymore.
yahoolian, I must thank you, this form looks exactly something which could be working if you prefer doing graphical icons. I think I'll be using pie menus in one of my projects very soon.
Pie menus
I'm hoping full skinning will be available in R2. If it is, this would easily be added as a skin. Want pie menus? Install a skin that uses them.
Pie menus
Not to sound like I want to start a flamewar, but I really hope that skinning is not available in R2 any more than it is in R5 and Zeta. Skinning is a feature that is really cool-looking and lends itself to some bragging rights, but there is so much potential for something that makes the OS much harder to use. Take a look at Linux. Also, most graphic artists, while the have the capacity to make an OS absolutely gorgeous, are clueless when it comes to usability.
Pie menus also have their uses, but what about for a menu that has more than about 6 items? Once the number of choices gets above 6 items or so, they're actually harder to use. Pie menus are good for specialized purposes, but not across the operating system, IMO.
Pie menus
You pretty much nailed how I feel about it also :)
Pie menus
Hmm, not sure I understand what you mean. As long as there is an easy-to-use skin available (one that implements R5's UI, for instance), what is the problem? The implementation would have to be standardized of course. Applications would be completely naive of what skin was being used, etc.
I don't know enough about Linux or the way it implements skinning to know what problems it brings (though I'm sure there are plenty!). Would you explain or give an example?
All that said, if skinning wasn't in R2... I agree with you on the pie charts. I wouldn't want them across the OS. Apps that want to use them could use them by implementing as a custom widget or something, right?
Pie menus
OK, you asked.... :o
Well for starters, there is the issue of branding. A regular person on the street can't necessarily look at a screenshot of something running in X and immediately recognize it as Linux. OTOH, even Joe Average can instantly recognize a shot of OS X or Windows.
A lack of consistency is another reason. Take a look at these 5 WinAmp skins.
http://darkwyrm.beemulated.net/usability/skinarticle/winamp/winampdefaul...
http://darkwyrm.beemulated.net/usability/skinarticle/winamp/sonyblack.jpg
http://darkwyrm.beemulated.net/usability/skinarticle/winamp/kaolla.jpg
http://darkwyrm.beemulated.net/usability/skinarticle/winamp/sonympfx.jpg
http://darkwyrm.beemulated.net/usability/skinarticle/winamp/raydreamtsun...
They are all laid out in the window in pretty much the same way, but kaolla and raydreamtsunami are very different, especially if you focus on the mixer controls.
Also, skinners tend to fall into the 'graphic artist' kind of person. An artist can make an interface that looks so good that you say to yourself, "Wow! That's *so* cool! I have *got* to try that out!". The only problem is that when it comes to software, the proof is in the use, not the look. Few artists know much about usability or take it into account. Unfortunately, this results in controls that are hard-to-find or hard to click on or other issues. Two specific examples:
1) Look at the kaolla example. Because the skinner didn't take into account that you need contrast to be able to read text and controls, everything in the inner part of each window is either light gray or slightly-darker gray. There isn't enough contrast to be able to find any controls or read any text without working really hard to figure out what everything is first. This has to happen every time you work in some other app and switch to it again.
2) In each picture, look at the window control buttons in the top right corner. With the exception of the default skin, you can't just look at the window and figure out what each one does. The default one gives you a little clue as to what each one does and raydreamtsunami at least differentiates the close button from the minimize and windowshade buttons. There is no standard and far too many skins make all the buttons look exactly the same.
A final reason is something that is inherent in people, something I call habituation. People are good at making a task into something that requires no or little thought after they've done it enough. Tasks that fall into this category are things like walking, changing gears on a manual transmission car, or clicking the Yes button in the 'Are you sure?' window when I delete a file in Windows 98. Skinning disrupts this. Try changing the window decorator you use and use it for a couple hours. If you use R5, use the decorator easter egg to change to the Amiga or MacOS 8 decorator. Under Zeta, pick one quite different from the one you normally use -- Baqua, Origin, or OriginSimple are probably good choices. Although you can get work done after changing the decorator, it does make it harder because you can't just go on "autopilot" when you need to close a window.
Pie menus
Thanks for explaining. 8)
I understand this. But I don't think it's a good reason to not want skinning. Users should have the ability to make their Haiku look like Windows if they want, IMO. What's underneath is the important part.
I absolutely agree here. I always use the classic Winamp skin; I can't stand 3/4ths of what is out there--too confusing to use. The classic skin gets the job done. :)
That said, why should you or I care if Joe Blow uses some weird skin that you or I wouldn't even know how to operate? It's Joe's Winamp--not ours.
Again, I absolutely agree. But again, that said, who cares if there are hard-to-use skins out there for Haiku? Usability is the utmost importance for this project, but only for the basic Haiku. If skinning was added in R2, I'd say ship an R1-like skin as the default and allow users to install 3rd party skins if they want.
It does disrupt it... but not for someone using the default skin, or a well-designed skin. Again, Joe Blow can toss habituation out the window for all I care.
Do you see the point I'm trying to make, DarkWyrm? My friend uses some weird, completely wild skin for his Winamp. Like I said before, I use the classic one... so when I'm on his computer I'm totally lost on his Winamp. Big deal. It's his Winamp, not mine.
Don't get me wrong; I agree with everything you say, but I don't think these are good reasons not to have skinning in R2.
You seem to be a pro when it comes to usability and all that, but I think you're forgetting that customization is a part of usability. If I wasn't able to customize Windows, I probably couldn't use it very well. Skins are just another type of customization. Maybe my friend Joe will find a 3rd party Haiku skin that actually improves Haiku's usability for him.
Anyway, I'm not trying to change anyone's mind, or flame or be mean or anything like that (so don't take what I say that way). :)
Pie menus
Actually, I'll agree with DW here - branding, and ease-of-use is EXTREMELY important in a project like Haiku. Just imagine if Joe Blow with his horribly-disfunctional skin has his wife/girlfriend, or various friends come and use his computer for a few hours.
Even if it looks cool, there is a strong chance that they'll walk away with the impression that Haiku is hard to use.
Another example is various publicly-accessible machines with Haiku installed - I don't know about you, but I CAN'T stand when i use a public computer and someone has crapped all over it with their themes and colors - skinning would be even worse!
This is the type of viral marketing that Haiku will need - and can't afford to lose early in the game. R2 may even still be too early for these types of features - but ultimately we'll see.
I would say Haiku needs nothing more than what Windows XP provides for it's default theming - and that is more than even I would like.
Anyhow, back on the topic of pie menus - I still haven't found these to be terribly useful whenever I use them. I see them as mostly a novelty-concept simply to be different - but I don't find myself attracted to them for usability. There needs to be a good defined line between "looks cool" and "works well".
Pie menus
Theming is terrible, except when used for accessibility. I really think Haiku should follow Apple's lead on this one - allow some basic colour scheme changes and nothing else. The sort of users that want theming can always download a third party application to make it so anyway.
That's what drew me to BeOS in the first place - simplicity. Theming only complicates things. IMHO, one of Haiku's key goals should be to keep things as minimalist as humanly possible. Distribute it with no applications aside from one of each of the basics: text editor, browser, mail client, instant messaging, calculator and a few little knick-knacks. Leave everything else like games, office suites, media players to the users. Personally I want Haiku to be the sensible alternative to the ever-increasing bloat and cruft of Linux and Windows: small, fast, easy to use, consistent and unobtrusive. I think theming contradicts those goals at a fundamental level.
Pie menus
"Distribute it with no applications aside from one of each of the basics: text editor, browser, mail client, instant messaging, calculator and a few little knick-knacks. Leave everything else like games, office suites, media players to the users. "
I was with you till media player. the age of digital media is kicking off big time now. A lot more than when BeOS was trying to be the "Media OS" Failure to include a fairly comprehensive and easy to use media player would be a major ommission. Especially given Haiku's/BeOS's so called media heritage. Put simply, the competition is Windows Media Player and iTunes. But yeah, download/purchase an office suite, game, CAD app if you need it.
However, this is a bit OT. Apologies to the original posters.
Pie menus
Yeah, good point, a basic media player is a necessity.
Pie menus
I think we should have a media_player_server, even. Make playing MP3s as integrated as copying files... with file system attributes, I can just keep my playlists as smart searches, anyways.
Pie menus
I agree with all of this in principle, however I'm afraid we'll lose many more potential users to a lack of skinning than we would from people getting a bad impression from using a poorly-designed skin.
Another problem I forsee if we don't take the first step on implementing skinning is that we'll end up with the problems Linux faces: a million different ways of doing the same thing; we'll start seeing 3rd party app servers (uh, or whichever server is in charge of the UI) pop up that have no standardization. Different distros will be bundling different app servers, etc.
Personally, skinning isn't something I use, or would use. I just hope that whatever is decided will be the most beneficial to the project in the long run. That said, I've got confidence the devs will make the right call.
Pie menus
I don't see how skinning would prevent the creation of different distros, if the default UI is well designed then why make an other just for the sake of it? The solution is to make a god UI from the beginning, not skinning.
If people want skinning then there is other operating systems they can use, imo skinning is against Haiku's ideals of usability, consistency, simplicity and lightweightness so why waste resources on supporting it.
I also belive that there are a lot of people, out there, that are tired of lazy and bad GUI designers that makes so terrible default UI's that you are FORCED to skin it youreself to make it useable (like WMP), so they would happy to see that is at least one OS that priotize usability above skinability.
Pie menus
BeOS R5, the epitome of usability, consistency, simplicity, and lightweightness, supports skinning with 4 builtin widget and decorator themes. Yet, anybody who even glances at a yellow tab along the top of a window will instantly recognize it for a BeOS tab. Having the ability to change the skin certainly doesn't seem to have let the graphic designers of Be Inc. get lazy with the default appearance, either.
Is skinning really all that detrimental to an operating system's usability? Some users prefer a certain appearance to work best, and to not provide the ability to use that appearance would actually detriment Haiku's usability.
I think that providing a system-wide skinning utility or system before anybody else creates one would help promote consistency. In BeOS R5, several applications have been created to skin the widgets and coordinate the backgrounds and colors and icons, such as BeTheme and Theme. Both of these are meant to supplement BeOS's native skinning abilities with different colors and icons. Both Living Dangerously and WindowShade hack BeOS's app_server to change the colors of the window decorator. There's an entirely separate fork of OpenTracker whose sole reason for existing is to use the CrystalSVG icon set from KDE. Deskbar Fun and Deskbar Fun Reloaded and Haiku Deskbar Logo all exist to merely change the icon in the Deskbar for the Be Menu. Can the end users make their demand for customization any clearer? If BeOS R5 had provided all of these features from the start, would so many different programs have been created to perform the functionality that can be in one program? Wouldn't the desktop environment be more usable and consistent if all of this functionality was combined into one simple program?
Instead of having to download several applications, the user should have the ability to change the look of his desktop by merely downloading a theme or theme pack, perhaps through the Haiku online content distribution channel. How much simpler can that be?
And since the app_server has to have a builtin theme by default, in order to know what graphics to draw, a simple change to the app_server to change where to get the source graphics from and at what offsets they should be drawn will not add any amount of bloat.
Personally, I love Zeta's theming system. Unzip one file to a certain directory, and a new theme shows up in the Appearance settings, and the change is shown instantly when the user selects a different theme. Is this functionality and ease of use so contrary to Haiku's vision of functionality and ease of use that the possibility of this feature is being destroyed?
Browsing around the SVN repository gives hints that Haiku will plan to provide themable decorators, so a natural extension of this idea is to provide builtin utilities to modify the appearance of Haiku in other areas. Having a utility to change the colors to a certain scheme certainly wouldn't add any extra bloat at all. Changing all of the default icons to a certain set would only require the added bloat of one tiny utility. A utility to coordinate the backgrounds of all of the workspaces would be incredibly tiny. Some of the utilities mentioned above could be merged into an Appearance settings application, similar to what Zeta has. One extra program, providing all of these benefits, could hardly count as bloat, and could easily be downloaded separately if the developers really feel it is too big to fit in the image. It would be better if there was one program to provide all of the appearance changing functions available in Haiku, compared with BeOS's situation of at least 3 different categories of programs.
In summary: I feel that supporting skinning will not harm Haiku's goal of usability, consistency, simplicity, and lightweightness, and in fact can even help usability, consistency, and simplicity.
Pie menus
One critical thing you're missing here is that currently, the Haiku theme is mostly drawn with LINES, not with graphics... this means it can be easily accelerated by most 2d drawing engines/hardware without needing blit graphics all over the place which would probably be a lot slower.
Zeta's theming system is already IMO, horrible. It's slow by comparison. Sliding tabs are gone with essentially no potential to re-appear (because the graphical theming basically makes them not possible).
Basically, this is my opinion... but if given the choice, all I would care to "skin" is the colors and maybe fonts - but certainly not the graphical look of the windows, widgets, menus, scrollbars, etc.
Pie menus
What will this theme give you? Do you really want that we (the developers) work on a feature which:
* does not improve ease of use
* does not make you more productive
* potentially decreases productivity and usability
* adds more configuration options (increases UI complexity)
* wastes computer resources
* wastes developer resources
* makes some users waste time on searching for a nice skin
* makes bad designers waste time on useless skins
* only gets used by a small minority of computer users (almost all users stick with the default XP theme, for example)
And all this just to get your *personal* desktop experience, so you can stare at a few shiny buttons the whole day? Is that really what you want?
If you seriously think about it: This is one of the time-wasting things in life that have absolutely no use.
If someone designs a better theme then it should *definitely* become part of the official Haiku distribution.
Don't you think that it should be our responsibility to disallow making Haiku less productive? This includes adding configuration options which decrease ease of use and productivity.
Pie menus
Waldemar for President!
Pie menus
One question: How easy/hard would it be to pull the widget- and decor drawing code out into separate libraries that anyone could write drop-in replacements for?
Pie menus
Nos gusta complicarnos la life
Pie menus
Depends on the theme at hand.
Depends on the theme at hand.
Depends on the theme at hand.
Conceded, however is not a good point against the notion itself.
Should be insignificant if approached and implemented right.
I personally wouldn't want the core devs working on this until waaay down the road, like... when we have R1 shiny and we're all using it.
One man's waste is another man's treasure. Please don't think you have authority to speak for every computer user out there.
What you call bad another might not. What you call useless another might find very useful.
For the regular users, I agree, most of them do stick with the default (only because they don't know how to change it). Every moderately-fluent user I know of does not use the default theme. Nevermind the fact the default theme for XP sucks while the default for BeOS/Haiku rocks. ;)
And all of this recreating BeOS so we can just stare at a 5-year-old OS that not many people used in the first place?
Come on dude... Haiku is recreating BeOS because a LOT of people liked it. Implementing skinning sounds about right as a LOT of people like it. Maybe not you or I or Urias, but they are out there. Many more than ever used BeOS. (Sad but true.)
That's your opinion because YOU have no use for skinning. You are not the only potential Haiku user.
You're assuming implementing skinning would make Haiku less productive. But that aside, mark my words, if Haiku becomes popular you will see 3rd party hacked app_servers that break compatibility with the official Haiku app_server (or worse, entire incompatible distros), and we'll be facing much of the problems Linux faces.
You can't disallow anything regarding Haiku; it's open source. People will come along and modify it... whether or not those changes are incorporated into Haiku's official development tree.
Your general users want customization; either they'll get it via a standardized skinning API or they'll get it from an unofficial, incompatible app_server. If the devs don't care about Joe Blow, then forget about skinning--I'm sure all of us who liked BeOS could care less. But if you want to attract a broader userbase, you cannot deny that they will demand customization.
Pie menus
So, you want to say that a system with a complex UI is better than a system with a simple UI?
I have the impression that you want to change Haiku into a new Linux. This will hopefully never happen. Otherwise, creating Haiku would be pointless because we already have Linux. What do you like about BeOS? I really can't see how it can make you happy...the whole point in BeOS is its speed, ease of use, and simplicity.
You should really think about Haiku's target audience. It's not a geek OS. We want to reach end-users who think that computers should just work and never get in your way. Theming is definitely a feature which gets in your way. It doesn't improve ease of use or productivity and it adds *modding* options which are of no use for your daily *work*.
I think that if theming will ever become a high-priority item on our TODO list Haiku has failed its goals or reached perfection. There are *so many* things that we need to research and improve and all of them are many times more important than theme support.
This argumentation leads us to Linux where everyone has maximum choice. If we start with themes we will have the next minority saying they need more configuration options for X. When do you think should we stop adding useless stuff? We should better never start with this.
Strange...I don't know *anyone* (I'm serious!) who has a different XP theme. Some just switched to the Win2K mode because it uses less resources. The only thing people do is change their desktop background.
Your claim that if skinning were easily accessible most people would use it is just speculation. Most people don't use themes because they really could care less about computers. Maybe your "modding" friends won't believe it, but most people use computers for *work*.
Come on dude... ;) I'm not talking about the R5 look. It's terrible. I know that myself. Please think of OS X and tell me why anyone would want to change the theme.
I only have one explanation: those people primarily use their computer for entertainment (games) instead of work, they have too much free time, and they don't know what to do with it.
Haiku should (will) have a very good theme, so how do you think can theming improve the way you work with your computer? You will basically need a completely different UI concept if you want to improve anything.
If theming was so popular then why do I really not know anyone who uses that feature? Even Microsoft wanted to remove theming support in Vista because their surveys showed that almost nobody uses them. Also, look at OS X which doesn't have theming support. I haven't heard of anyone who rejected to use OS X because it doesn't support themes. ;)
Seriously, a themeable app_server won't break binary compatibility. Even if it does, it's not really our problem because it's the mistake of the people who did the modifications.
Linux has much bigger problems because there are no real standards on libraries (and their versions). Haiku does have a standard. Linux doesn't have a standard directory structure. Haiku does. Linux doesn't have a standard UI API or media API or ... Haiku does. Linux has no official distro that most people should use. Haiku does.
Exactly. So what? That's why I think it's better if gamers and time-wasters use the super-customizable skinnable ultra-complicated "Haiku Xtreme" distro. It will hopefully not be officially supported by us, but as long as it is compatible to the official distro the "Haiku Xtreme" users won't care, anyway.
You really have a very "easy" standpoint. You *want* something. But what if I (and other Haiku users) reversed your argument and said
"I will not use Haiku or create a new minimalistic Haiku distro if theming will be added!"
Do you want to stop me, too?
I can already say that if Haiku ever gets as customizable as Linux I will leave this project. I believe in simplicity and from my discussions with other Haiku developers I got the impression that I'm not alone.
You make it sound as if >50% of the world population seriously can't live without skinning. Maybe 40% of the *geeks* really need theming because otherwise they don't know what to do with their computers:
http://channel9.msdn.com/ShowPost.aspx?PostID=129777
Note that this poll doesn't represent the average user because only computer experts visit msdn and normal end-users didn't take part in this vote.
What I want to say is this: not even half of the geeks really want theming. Moreover, Haiku is not a geek OS, so why should we care? And why should we create a distro that doesn't fit our goal: making an easy to use OS that just works without ever getting in your way.
You obviously can't make *everyone* happy. We can't listen to everyone's wish and implement everything someone might need. This will result in a product that makes most people unhappy because it gets too complex. Of course, those who want a feature will shout very loud and those who just want computers to work will not care to take part in discussions, so it will always seem like most people want a certain feature, but we (the Haiku team) must use our brains and make sure that only useful features get added. As you have seen from this discussion, the Haiku developers seem to think that theming is useless and evil. DarkWyrm even took the time to give a list of very good arguments why theming is evil. Unfortunately, you don't want to listen. This will be my last reply to you.
Pie menus
A complex UI is not inherently any better or worse than a simple one. It depends on how it's organized, who will be using it, what it's for, etc.
How did I give you that impression? In any case, I absolutely do NOT want Haiku to be another Linux. I loathe Linux as a desktop OS.
Amen, brother.
Simplicity from a user's and application developer's perspective wouldn't change by implementing app_server-level skinning, would it? We'd have one more Preflet for choosing the skin, and applications would be completely ignorant of which one is installed.
This is exactly my perception of Haiku's target audience, and has been from day one.
From who's perspective? You OS devs? Sure. The end user? Absolutely not. Application devs? Nope.
So you're telling me that the UI has absolutely no impact on productivity and use in daily work? With this logic, you're implying there's no performance gain over BeOS's UI versus Windows 3.1's.
Skinning allows modification--that is, detrimental and/or beneficial changes--to the UI.
I agree 100% with the exception of the first statement.
Actually, my ultimate point (if you had read my posts waaay back in the thread, in addition to my conclusion) is that the way to prevent us being another Linux where there are a million different components and dependencies and incompatible hacks is to nip it in the bud by implementing a standardized skinning API.
If you don't agree with that ultimate point, fine.
Hmm, that's interesting. I can't use the default XP theme; my productivity decreases as I'm used to the classic Windows theme... most others I know are in the same boat.
I don't think I ever claimed most people would use skinning if it were more accessible.
Look or usability? R5's look is like soooo 1990s, but the usability still matches modern OSes; that was my point.
Never used OS X, but I've heard it's got a nice UI. (I'm not dodging your point, I'll get to it later on in the post.)
I disagree. Now, I admit, I am NOT an application or system developer; I do Web development, but I do know that different skins can increase or decrease the usability of the product that is skinned. A good example is Winamp; I like the classic skin because it does what I need and is fairly straightforward to use, while there are many Winamp skins that make it harder to use, less intuitive, etc. Why do you think this cannot be applied to an OS UI?
I also know the value of skinning in Web development; they can completely destroy or create usability of the product at hand.
First, just because everyone you know doesn't use it doesn't mean a large section of users do not.
Second, Windows built-in theming support is NOT skinning. The customization that can be gotten from it is laughable; to do any serious UI customization, one must get a 3rd party program like WindowBlinds, etc. I'm guessing their surveys inquired about Windows theming and not 3rd party skinning.
Though I've never used the system myself, I also have heard it's got a great UI. That said, just because you haven't heard of people rejecting the system solely on its UI, doesn't mean they aren't out there. There could be thousands of Windows users who contemplated OS X but decided against it for lack of theming. (Hey, it could be possible. :D)
Assuming we do start seeing skinnable app_servers that break compatibility, it does become the Haiku community's problem: we turn into the mess that is Linux.
I agree that they are much bigger problems, but they are down the same line as having a non-standard skinnable UI API.
Haiku won't have an official distro anymore than Linux has an official distro; that's the nature of open source: we will see dozens of distros, some "Haiku-certified", and some not. It's speculation on whether or not the majority uses a certified distro. My guess would be everyone will at first, but the more the system gets popular the more distros there will be then who knows.
If it's compatible so Moddy Modskipper can just drop it in his Haiku(r)-certfied distro and it's good to go, then that would be great. I just am not optimistic enough to believe this will be how things will go; I guess I'm pessimistic and believe eventually the incompatibles will be a force to contend with, and that's when we turn into Linux.
I'll put aside the fact I do NOT want skinning for myself (as I have said many many times throughout this thread).
I don't want to stop anyone. I am on your side, Waldemar. I want to see Haiku succeed. We all want what BeOS had and hinted at, but to do that we have to build a userbase. I think, perhaps incorrectly, that skinning is something that will be required to earn a significant userbase.
Hopefully I'm wrong on that premise; hopefully most people don't care about skinning, and we'll have an army of users who don't care about it and have proactive desires not to have it. I would be peachy. But I'm not that optimistic.
And as a user, you are not alone. I don't want Haiku to be as customizable as Linux--I'm only talking about a standard skinnable UI API.
I hope you're right then.
You really think by R2, grandmas are going to be using the system? Be realistic; it is a geek's OS. I certainly hope that changes, but for now and--as far as I can see in the moderately near future--that's what it will stay as.
Absolutely.
I think you've confused me with someone else.
I've read every post on this thread and listened. DarkWyrm's reply, if I remember right, was at my request to explain usability issues.
I don't use skinning, nor do I personally want it. I only keep arguing for it because I think if we don't standardize skinning, we'll turn into a mess like Linux. I hope I'm wrong, I really do, on that premise.
Please don't insult me by implying I think my wishes are more important than everyone else's. I'm sorry you seemingly took offense as that was not my intention.
My conclusion on the topic of skinning (since it seems to have escaped most people, probably in part due to me):
If the lack of a standardized way to skin Haiku were not to lead to the headaches associated with the lack of standardization Linux has, then I do not want Haiku to be skinnable. If, however, the lack of a standardized way to skin Haiku were to lead to the types of aforementioned headaches, I want Haiku to provide standardization on this. That's all I'm--and have been--getting at. I don't want us to be another Linux; I can see how implementing skinning would seem to put us in that direction, but I can also see how not standardizing it could put us in that direction.
Pie menus
I think there's some bleeding of the terms here. "UI" is being used to mean "skin", where really "aesthetics" should probably be used. I think we all agree that the windows funiture of classic BeOS is a little old now. I like it, but we've moved on from 8-bit now. Maybe it's me, but "UI" means more than the look of an interface, it's how it functions, and how many steps a user has to take to achieve a result. The "look" of a UI can play a part in the semiotics of the interface, but it's only one part of the equation. However, I admit, this is quite subjective.
Do I think we should have a theming API? Yes, but not just to pander to geek users. No, I think the more important use of a skinning/theming API is for things like corporate identity (mass deployments of Haiku, all using the brand colours, logos, etc), and to also design themes with children (as an early learning aid) and the visually impaired in mind. NOT, to have "Geiger Alien Theme", although it's kind of inevitable that such a thing will happen, either 3rd party of from Haiku.org itself. If some of the more sane uses of "skinning" as I've mentioned above, could be achieved without also allowing some of the more asinine abuses, I'd be all up for those options being explored instead.
disclaimer - I've never felt a need to change the "theme" of an OS. This is on all Windows, Classic and OSX Mac, BeOS and Linux. I've always used the default. I do however, tweak settings to best suit my tastes, as do most folk I guess.
Pie menus
Hehe, you could just have replied with this last paragraph and I'd have understood what your whole point is. Sorry. :)
As expensivelesbian pointed out we were talking about different things. You can't have OS-level theming that is as flexible as that of WinAMP. It will mostly be limited to "Star Trek theme" and "OS X" theme (like WindowBlinds). At least, it can't fundamentally change the UI.
Usability improvements should really be done in the official UI and this is a lot more than simply changing the look to match OS X or Vista. Usability and productivity improvements mean things like replacing installation dialogs ala:
-----
License text......
-----
Do you agree with the license?
(o) I Do Not Agree ( ) I Agree
[Cancel] [Next]
into something more effective like
-----
License text...
-----
Do you agree with the license?
[Cancel] [I Agree]
We hopefully will some day have settings for visually impaired and color blind people (higher contrast, less colors, bigger fonts, etc.), but this doesn't require full theming support.
Really, the minority of people who must have theming/skinning support can use Linux or XP. We can't care for everyone. And if those guys create their own Haiku distro we can pretend it's yet another Linux. ;)
If we ever find that most of our users switch to a Haiku Xtheme distro and nobody really likes the official distro then we could add official theming support, but I think we shouldn't do this in advance just because something bad *might* happen. Let's first concentrate on doing things right and later make it worse if this is absolutely not successful. ;)
Pie menus
I quite like the R5 look... and I run Shapeshifter on my OS X computer to make it look mostly white (and remove all traces of brushed metal). I suppose that means I'm the odd man out. TBH I like R5's look better than OS X's. The only thing I miss on R5 interface-wise from OS X is click-through.
I think we should be able to change colours in Haiku. I'd like the yellows darker and the greys more blue, f'instance... but I really don't think we should be able to change the fundamental look of title bars, though, like you can in Dano (if my memory serves me). I see no point in being able to move the close button etc, either.
-Paws
Pie menus
I figure most users don't care about skinning and the ones that do can install a third party theming app. I'd be really disappointed if Haiku included theming (outside of accessibility stuff like high contrast) out of the box.
As for the current look, it's mostly fine by me. The grays could be a little lighter, the keyboard icons in the menus updated (will they be a part of the new icon theme? I hope so) and buttons made a little softer (that semi-3D look is pretty 1999) but apart from that it's great. Really icons, fonts and backgrounds are what makes the difference, and Haiku has already pulled through with fonts (better than Linux and Windows, that's for sure), looks like it's going to pull through on icons and I have faith that there will be some excellent backgrounds sorted out and included in the first release.
Pie menus
It's definitely going to pull through with the icons. There were so many good entries, I was really impressed at the quality of some of those sets.
Haha, absolutely. :lol: I'm probably just being pessimistic, the more I think about it the more I doubt it'll be a big issue anytime soon if at all.
Pie menus
one argument against skinning is that it can take out functionality such as the sliding tabs. it is important for haiku to be in control of the entire UI for maximum speed and less bloat.
What i do like in firefox are not the themes but the extensions. Even then you have to be careful because it's getting to be a big mess. When that happens, people switch to something like Flock (see flock.com) which integrates lots of features and the look in a unified simple way