Remember how Mark Shuttleworth justified moving the window titlebar widgets to the left by claiming the space freed up on the right side could now be used for something else? On his blog, Shuttleworth unveiled what, exactly, Ubuntu’s plans are: window indicators, or ‘windicators’. In a nutshell, it comes down to having a tray area in every window.
“Windicators are indicators displayed in the window title bar that behave just like the indicators in the panel: they have an icon which shows state, and clicking on the icon brings up a menu,” Shuttleworth explains, “Applications can create, update and remove window indicators using an API more or less like the AppIndicator framework first put to use in 10.04 LTS.”
Basically, it would look something like this:
Shuttleworth gives several example windicators, such as a volume control windicator. Thanks to PulseAudio, Linux now has an audio stack with per-application volume controls. Using windicators, you could expose this functionality better: the volume control inside a window’s titlebar controls the application-specific volume, whereas the menubar one controls the overall volume. Other examples include an indicator that tells you whether or not a file has been saved, an online/offline indicator, progress bars, or whatever else you can come up with.
While the idea is nice, I do have a number of concerns, the biggest of which is how this affects the large groups of users who change themes – myself included. The fancy new system tray in Ubuntu 10.04 only works with the default theme – if you switch to a different, supported icon theme (like GNOME Colors), the icons in the system tray misalign. Messed-up icons would look even worse in titlebars.
Then there’s the issue of modifying the titlebar layout. I find the clustering of titlebar widgets quite silly, and wonder whoever came up with that quaint notion; in Windows and Mac OS X, I accidentally click the close widget all the time. In both GNOME and KDE, you can move the widgets at around as you please, which I do happily.
As you can see, I place the close widget on the right, the maximise widget on the left (for symmetry, I don’t actually maximise anything), while minimise is handled by a double-click on the titlebar (for completeness’ sake: right-click titlebar sends-to-back). The close widget is on the right for a very simple reason: menus and the like are left-aligned, so placing the close widget on the right is just asking for trouble.
I’m left wondering how Shuttleworth’s windicators fit into this scenario.
All in all, I guess my concerns can be summed up in two simple points: first, can it be turned off, and two, will this implementation be more than skin deep?
Kwin devs says it best:
http://blog.martin-graesslin.com/blog/2010/05/why-you-should-not-us…
OT: And why is Canonical paying the MPEG-LA for H.264? Those millions could be spend on cool upstream development
I saw that blog post before I saw this.
Even if Shuttleworth wants “windicators”, it can be done using the existing system and adding an extension to the window manager using D-Bus or a similar technology. I can’t think of a good reason for client-side decorations that doesn’t result in its own set of problems.
It seems that these days, innovation means doing things differently for the sake of being different, rather than making solid UI decisions. Here we have yet another example of “innovation” that’s just going to be a waste of time and resources.
Especially Nouveau (FOSS NVidia drivers).
Get this: If Red Hat alone can’t make Nouveau’s 3D drivers work stable enough with Mutter in time for fall, millions of NVidia owners will be greeted with the classic Gnome 2.x desktop instead of the new Gnome 3.0 Shell.
How embarrassing would that be for Canonical if they had to tell all their users that sadly instead of spending money for an additional Nouveau developer, they spend it on a Mac-using new design team (Google the interview for proof) and stupid “windicators”.
Btw: Why can’t the windicators be on the left side? Did the window buttons really have to be moved to the left to make room for windicators on the right side?
Edited 2010-05-03 23:43 UTC
Did we learn nothing at all from the blunders with KDE4’s birthing process? The *last* thing I want to see distros do is rush out with Gnome 3.0 as the default just because it’s shiny and new.
As for the subject at hand, my mind immediately goes to netbooks as the likely inspiration for this.
We’ll see, but I’m quite sure that most distribution *will* rush out with Gnome 3.0 just like they did with KDE4.0 even many ‘users friendly’ distribution.
Most Linux distribution cannot resist the temptation of the ‘shiny new thing’..
Or maybe a Intel GMA500 Poulsbo driver that works?
Ref: https://edge.launchpad.net/gma500
./nalle.
System tray recently went through transition from client rendering and XEmbed to host rendering – client just provides icons and menu descriptions using dbus to the host (plasma in case of KDE). Why Shuttleworth wants to go in opposite direction with window decorations? His ‘windicators’ could be done with server-side decorations and some EHWM additions.
Talk about underwhelming. That’s what his big idea was? Maybe he should go back to space shuttles and let more competent UI designers take over. His examples don’t even make sense. A volume control? Well I don’t know what apps he’s using, but don’t most apps that deal with audio already have a volume control? Rhythmbox, Totem, MPlayer… yup, they’ve all got volume sliders already. Online/offline indicators? Last I checked, there was one in most messenger’s contact list window, not to mention when you right click on the tray icon it already has. Document’s saved indicator? Hmm, well most wordprocessors and things like that have that. That’s what that little * in front or after the filename means, it means modified but not saved. Filling up a title bar with more widgets is just going to get more confusing.
Oh, that reminds me, his other big brainstorm is to eliminate the tray entirely (notification area in GNOME terms) and completely replace it with that indicator-applet thing. Oh, boy…
A volume control? Well I don’t know what apps he’s using, but don’t most apps that deal with audio already have a volume control? Rhythmbox, Totem, MPlayer… yup, they’ve all got volume sliders already. Online/offline indicators? Last I checked, there was one in most messenger’s contact list window, not to mention when you right click on the tray icon it already has.
Well, to be honest, a simple standardized way of handling those would be better than every single application having to develop their own way. A volume control that works _exactly the same_ in every single application that uses audio and is always in the exact same place actually sounds like a terrific idea to me, and heck, it’s a lot easier to remember then even for complete newbies!
Already having something that works is in absolutely no way a reason for not trying to make it work even better.
Already having something that works is in absolutely no way a reason for not trying to make it work even better.
Except I absolutely don’t see the “better”.
Ooh, let’s break 25 years of accumulated muscle memory by placing window buttons left. Then propose to duplicate the system tray a gazillion times in all windows, instead of having just one in a fixed place, neat and organized.
Telling Ubuntu users “STFU, Canonical decides where your window buttons are placed!” was bad enough. I could even have accepted that if the thing on the right side was jaw dropping revolutionary, but… the proposal is one big crapfest of useless clutter, multiplied over all open windows.
Let’s just throw in some superfluous “new” items like:
Online / offline — apps that need this already have developed other accepted methods.
^aEURoeunsaved^aEUR — Every applications I know pops up a warning dialog if something isn’t saved and you click close. If you don’t know if something is saved and you want to keep it, just hit ctrl + s. (Or do the equivalent clicky/drooly thingie.)
Progress indicators — Progress pop-up dialogs have served this purpose well for over 20 years.
^aEURoebasket^aEUR — Online shopping takes place in a web-page and they already have these indicators. No need to plop it in the windowsdecoration as well.
Sharing — Collaborative suites have far better methods of signifying this stuff.
Volume — Every app that needs this, already has this.
It feels like Mr. Shuttleworth is really desperate to “un-Linux” his Ubuntu, so that he can pretend to have a different OS than his co-opetition. I say let him have his play. While he is playing, he is not listening to his users or his upstream. Running roughshod over the people he needs to give Ubuntu critical mass will come back to bite him.
One Apple is more than enough.
Ah, so you want something standardised, that works the same across all applications … which means you don’t want it done in each application … which means you don’t want client-side decorations. You want a standardised protocol, where the client just says “I need X with value Y” and let the host handle all the gory details.
Kinda like what happened to the systray/notifications system, which went from client-side (everyone do their own implementation) to host-side (everyone call a standard d-bus setup and let the host handle the rest).
Modifying each and every client app to work with “windicators” is the wrong way to go about doing this.
Actually, I see an excellent reason for the volume control on each program’s window. You know those crappy web pages on big companies’ web sites, including movie and (music) band sites? You know–the ones that force you to use Flash, and even worse, have obnoxious sound/music that you have to search for the damn “mute” or “STFU” button somewhere in some crappy GUI? Yeah, those.
Set your browser’s application-level volume to minimum, and problem solved. No need to reach for the speakers any time going to those sites. That way, the volume of your music player doesn’t get screwed up, because you’re only changing the web browser’s volume, and the volume is in a logical location–in the browser’s title bar. I’ve been waiting so long for something like this, because unless I’m looking at YouTube or some porn site, I don’t want *any* sound. Period.
Before YouTube and Flash video, audio was almost NEVER welcome in the browser, especially those worthless bouncy, flashy, noisy Flash ads. Youtube changed that, but the titlebar volume control would still be useful.
And this is different how from clicking the volume control in the systray, and setting the volume for the web browser?
Especially considering that the GNOME systray is in the top-right corner, right where this windicator volume control would be.
Last I used Ubuntu, the per-app volume controls were a PITA to find and get to–I just gave up, used the volume knob on my speaker and moved on. The volume control on the panel with easy access is for the master volume, if I remember right–basically equivalent to the speaker knob. Bringing up the “mixer” (easily done by right-clicking the little speaker panel icon) is also not what I wanted, because the PCM/master/left/right/etc. controls were already set. I’m not using Ubuntu (or even Gnome) right now, but that’s how I remember it.
Did you even care to read Mr. Shuttleworth blog post? All the things you’re saying don’t make any sense.
Move menus to the title bar: Okay. Non-standard, but not terrible. When you get to the point where you don’t have enough horizontal space for the title/handle and all the icons, what happens?
Use icons instead of text labels: Great as long as you don’t have any verbs in your program for which there’s no universally agreed-upon image.
Replace status bars with pop-ups that obscure what you’re working on: You have got to be kidding me.
Fewer pop-ups is what we need, not more. One thing I hate about Windows, or at least the Windows program culture, is pop-up balloons and message boxes. All these do is distract you and rarely provide useful information or effective program interaction. And since Windows is really terrible about focus stealing, these usually end up resulting in accidental actions in the pop up windows, such as hitting OK to dismiss an important error or warning message, or entering text in a new IM window (and thus not for the intended recipient).
There are just as many on linux (at least in gnome). The only mainstream OS I know of that doesn’t have them out of the box is OSX, and even then, the first thing many osx users do is install growl, which is a framework to provide those notification popups.
i sure don’t install growl and i hate notifications.
i’m ok with KDE ones tho. they’re hidden. I click them when i want to see. Most of the time i know what crap it’s going to tell anyway.
Like, i just started the computer in the morning:
it’s going to say “you have XXX updates available”
or i just updated w7:
“you must restart your computer”
and reboot it:
“you have had xx programs installed since your restart do you wish to see the list?” (no shit?)
Another one..
“you have connected to the wireless network XXX” (yeh.. i just clicked on it you know?)
Oh this one is fun too:
“you have hidden xxx but it’s still running, just here!”
can’t actually remember of any popup that made sense.
it’s actually pretty useless in my eyes. Reminds me of office’s clipper dude pestering you all the damn time you know “cause it looks cool and people will like it”
except it’s annoying and doesn’t look cool at all after a while.
Those are different. Those are passive notifications. Although they are still annoying, they aren’t as bad as message boxes and dialogs that steal focus.
oh, yee,
on Sabayon linux, KDE is awefull so many stupid baloons….. it just pisses me off
Thank god I use FreeBSD on my desktop PC with custom DE
I would absolutely love it if all the drop-down/pop-up overlays in web browsers would slide over the contents of the page, instead of moving everything up/down the screen. This is the worst feature in Firefox.
This is a lesson that Apple quietly learned. Once upon a time they were very pro-icon, but have gradually dialed it back. There’s a good reason for this: icons are meaningless unless you have the meaning memorized. Few people can memorize very many icons and certainly not for any complicated idea.
“Use icons instead of text labels: Great as long as you don’t have any verbs in your program for which there’s no universally agreed-upon image.”
Yeah, this is a horrible idea. I’ve actually gone exactly the other way, ever since someone put a neat blog post on planet GNOME a few years back which illustrated the possibility – disable icons entirely and _only_ use text labels. It makes rendering interfaces faster, and it’s a hell of a lot easier than trying to remember that $RANDOM_DEVELOPER’s vaguely dodo-looking hieroglyph actually means ‘premuntificate’ or whatever.
Quote
Replace status bars with pop-ups that obscure what you’re working on: You have got to be kidding me.
End Quote
Exactly. My take on this is that it is a precursor to Canonical introducing Adverts everywhere. Why else would you use pop-up’s for everything that is currently handled in such a discrete way.
Because I read this, and I, who don’t use Linux regularly except for server type stuff, think this looks real intriguing. It will at least begin to put a strong personality stamp on Ubuntu and help the GUI stand out a bit from the crowd.
Right now, Gnome still feels like Windows Generic with poorly organized menus, large amateurish fonts, and broken/incomprehensible hardware issues.
The new theme does help though.
The fancy new system tray in Ubuntu 10.04 only works with the default theme – if you switch to a different, supported icon theme (like GNOME Colors), the icons in the system tray misalign.
It’s worse than that. If you set your panel to be located on the left or right of your screen, instead of the top of bottom, the new notification area doesn’t adjust to display the icons vertically instead of horizontally! They’re still “displayed” in a nice horizontal row…right off the edge of the panel. So unless your panel is ludicrously wide (defeating the purpose of having it vertical to begin with), you can only see one icon at any given time, with no way to see any of the others.
Genius.
GNOME’s panels never were designed to be vertical even though it’s been technically supported, this just makes it worse. Meanwhile, Kicker becomes vertical nicely… and people wonder why KDE is liked when it’s so ‘messy’.
Menus, as we know them are just horrible concepts and I would be in favor of having them replaced by anything…
I have a lot of toolbars where any number of buttons are just below the close widget. Never have I or anyone who can use a computer closed a window due their closeness.
Never have I or anyone who can use a computer closed a window due their closeness.
My, my, seems if people disagree with the proposal, they are just labeled computer illiterate and can be ignored…
Menus are not evil and they are more efficient than cramming lots of icons in plain sight.
This is not about accidentally clicking on close. This is about the proposal being a frickin’ stupid way of designing stuff.
This is all very nice as far as it goes but the danger with any ‘new’ design is that it blindly throws away value that exists in current designs simply because the designers don’t know all of the real uses of current designs, or because they discount the value of those uses.
For example, the three listed uses of the status bar are indeed three common uses. A fourth use that is not mentioned: Non-transient status information, such as current page and line number in a text editor/word processor. You cannot put that kind of data usefully into an icon and hiding it behind a submenu makes it nearly useless. This example is real and represents the tip of a dangerous iceberg which one risks hitting any time a radical design decision is made.
You cannot simply say “This is a great design because the ten problems I set out to solve, I solved.” You can create a decent system that way but you are not likely to create a system that is good enough that it should replace the status-quo (this is also the mistake, incidentally, made by people who try to revise the FHS by fiat). In order to be as good as the existing system you must address *all* uses of the existing system, especially the uses you do not replicate well or are inferior in your system. Talking up why it’s good is fine but talking about how it deals with the areas where it is weak is far more valuable.
Until someone sits down and works through how it fails I do not see that you can suggest that it’s worth writing even one line of code. If the design doesn’t stand up to criticism even on paper then there is probably a serious flaw with it. Again, that may be okay for a brand new system but you’re agitating for replacing something that already exists. The burden of proof is on *you*.
I don’t think Ubuntu has the power to coerce developers to make use of an Ubuntu-specific API for an Ubuntu-specific UI element of questionable value to anyone and no value to non-Ubuntu users. Either push this idea at the toolkit level and adopt it when it’s ready or be prepared to maintain a hefty patchset.
Unfortunately, Ubuntu has the power to screw Ubuntu up beyond recognition in just short 6 months from now. Seems like all these ideas converge to making things adhere to a select few people’s idea of “pretty”, usability and common sense be damned.
The notification bubbles started off with good intentions but they messed everything up by not allowing any customization at all. The activity indicator is also currently messed up and unusable. And instead of stopping and working out these quirks, they rush on to completely modified window buttons and “windicators” (there’s an idea born dead if I ever heard one).
There’s a lot of useful stuff coming out of Ubuntu but also a lot of bad ideas. And I don’t know if whoever’s driving these bad ideas will have the guts to stop and say “ok, sorry, we messed up”, then backtrack and fix things. Few big collaborative projects manage to do that.
wirespot speculated…
The problem with your speculation is that Ubuntu ceased to be a collaborative project some time ago and has become Mark’s little fiefdom. Not a bad position to be in if you can afford it and Shuttleworth certainly has the gold enough to make the rules. Only after a while people stop caring about you having the nicer toys if you’re going to be a prick about letting them play too…
–bornagainpenguin
And Ubuntu users have the power to become non-Ubuntu users as replacement distributions spring up like weeds. This is the power of the free desktop: No platform lock in, no brand lock in. Make a mis-step and you could see your user base disappear inside of a year.
\nod
I’ve already moved from Ubuntu 9.04 to Fedora 13 and Sidux, mainly over concerns about the suspiciously-like-a-crappy-iphone-app-store application manager they where trying to push over good old Synaptic. At this point, I’m glad I got out when I did.
I probably shouldn’t say this, but… this is where we Freetards really start to feel vindicated. If one distributor becomes unreliable or abusive, then there are dozens of other distributors we can switch to, with little effort or loss. We’re not tied to the deck of a sinking ship.
As if I didn’t already have enough reasons to keep looking at a replacement distro Shuttleworth comes along with this…
So long Ubuntu and thanks for all the fish!
–bornagainpenguin
The Windicators are pure awesome.
“Why do you need a volume control in the title bar, Rhythmbox and Mplayer already have a volume control in their window”
Ubuntu supports per-application volume controls through Pulseaudio already. Now you can turn your instant messenger’s volume up and down. With a Windicator, you don’t need to take a trip to the volume control application. Win!
“We need fewer popups, not more”
Windicators are widgets in the title bar, not popup notifications.
“Why do you want client-side window decorations”
Program’s logo in the title bar instead of just text… it’s nice eye-candy and it’s easier to distinguish between programs that way too.
“What use is it”
I’m using Opera at the moment, on a 22 inch widescreen monitor. There’s a bar at the bottom of the window that shows Opera Link, Opera Unite, Opera Turbo and the page magnification. This is using up vertical space. Those icons are not suitable as Indicators (there are too many of them and they don’t make sense in a system-wide context) but they are perfect as Windicators. Save some pixels!
“This will all become useless when Gnome 3 comes out”
No, Gnome 3 uses Mutter as the window manager. Mutter is a fork of Metacity, so this could quite easily be ported.
“They should spend money on Nouveau/codec licensing/<INSERT_PET_PROJECT_HERE>”
It’s Canonical’s money and their power to make the decision. The Windicators will be a very visible part of Ubuntu and I’m sure they will be loved.
If the application already includes a volume control, why would you want another one stuffed into the top-right corner … right next to the global volume control? If you go to the top-right for volume … why not just use the global volume control?
Program logos are already in the title bar, in the top-left corner. Have been for years. Client-side decorations will lead to all kinds of horribleness in UI-consistency. After all, what’s the point of a window manager, if not to manage the windows? Now you want all applications to be window managers as well?
Or, stuff them into the toolbar. No client-side decorations support required.
If the application already includes a volume control, why would you want another one stuffed into the top-right corner
The application’s own volume control could then be removed. POOF, problem gone.
If you go to the top-right for volume … why not just use the global volume control?
Seriously, this is a stupid question: you’ve never wanted to adjust the volume your music plays at while still having all your other audio sources play at the same volume as before?
Ah, but doesn’t the global volume mixer include support for each application, thus negating the need for the windicator?
It does, in a similar way as Vista. You pop up the sound preferences tab and adjust it there, it’s not in the icon. Personally, I don’t see anything wrong with that, but shuttleman seems to be going out on a limb to design something for the sake of designing it rather than its actual usefulness.
Well, this particular use-case is one I hit frequently: the ability to adjust volume per-application with a little widget sitting right on the window title-bar is certainly alluring. It’d be a lot easier than going to the system’s volume control every time you want to fiddle with relative volumes of your different applications. It also has the advantage of easy feature discovery — as has been pointed out here before, a neat feature is useless if the user doesn’t know it’s there, and putting a per-app volume control widget right in the title bar lets the inexperienced user know fairly quickly that their system has the capability of per-application volume control.
Also, if the system makes per-app volume controls easy to find, it might also have the effect of sparing application developers the task of worrying about volume controls themselves.
Amusing note: Windows has this feature in theory, but, as often as not, whenever I try to use it, it fails in one of several ways. Either the volume settings are ignored, they cause the sound-generating application to crash, or, on rare occasion, they cause a crash in the Windows sound service, or even panic the kernel.
Ah, but only if it’s not done on the client-side of things. If it’s done in the client, then it’s still up to the individual app devs to add this feature.
If it’s done at a system level, and exposed via something like a d-bus interface, then it relieves the app dev from writing their own volume control knobs and widgets and what not.
IOW, this client-side “windicator” is the wrong way to go about it.
Why? All you’re doing is moving the volume control to a new non-standard place that will fill up with yet more crap.
No. It’s a nice thing to have when you’re at a fixed workstation where you are comfortable where all your volume levels should be at, but if you’re on a laptop where you interchange between headphones or even worse, speakers in an office, it’s asking for trouble. It’s a nice-to-have thing, but a niche thing at that.
The single biggest gotcha with sound not working is either that the volume is too low or that sound is muted. Now extrapolate that problem to multiple applications. Think about it.
Edited 2010-05-04 17:56 UTC
Why? All you’re doing is moving the volume control to a new non-standard place that will fill up with yet more crap.
It is already in a non-standard place. And every single application has their own volume slider thingie.
The single biggest gotcha with sound not working is either that the volume is too low or that sound is muted. Now extrapolate that problem to multiple applications. Think about it.
If that’s a problem then it already exists.
Because you want to change the application volume and not the system volume?
We must been using different windows managers, I don’t have any.
People care about applications and how to interact with them, not about having a windows manager.
Which is accessible via the system volume control.
Wow, that’s sad. I honestly cannot remember a time when I didn’t see the program logo/icon in the top-left of the window. Maybe it’s time to look into different WMs? KDE has had this for aeons.
Yes, and it’s the WM that makes that easy, intuitive, and consistent. Going from KDE to Windows shows just how bad things can be when the WM is short on features and everything is handled by the app (client).
Edited 2010-05-07 04:33 UTC
Not with a single click, and not with the awareness that the current active window is the application you want to change the sound for.
Those window icons -where available- don’t provide the horizontal space for drawing from a a title bar.
[/q]
The WM does not enforce consistence across applications, a clear example is the differences between applications using different toolkits. The applications have the control for most of their interaction with the users and are responsible for consistence and proper interaction. If the developers which work on the applications don’t care about providing a consistent behavior there is little or nothing to fix on the WM side, providing an API which is integrated with the WM and tracking that that API will be used in most applications is a good step to improve with consistence.
GUI consistency, no. Window control consistency, yes.
The borders work the same, regardless of the toolkit. The window management buttons work the same, regardless of toolkit. Snap-to features work the same regardless of the app, toolkit, etc. Moving application windows works the same. Etc.
All the interaction with the application windows is the same, regardless of the application … because that’s all handled by the window manager.
Moving these kinds of functions into the client application will break all kinds of things.
Just look at Google Chrome for a great example of why client-side window decorations and window management sucks. Ever tried to click on a tab, only to have the whole window de-maximised and moved? Ever had a screen full of tabs and tried to figure out how to move the window? Even tried to click on a tab only to have that tab detached from the window and moved? Ever tried to make Chrome look like any other application on your desktop?
Applications control how users interact with the widgets, and gidgets, and doodads within the app window.
But the window manager controls how users interact with the app windows themselves.
Exactly. Provide an API for interacting with the WM, don’t move WM functions (window decorations, for example) into the client.
I have absolutely no idea what those two lines are trying to say.
Edited 2010-05-07 15:40 UTC
…these were the Jaunty days
For some reasons, the Ubuntu devs managed to write their first non-buggy release in a while (contrast with theoretically LTS 8.04 which broke synaptics touchpad support and some sound card support at release time).
That, plus Jockey, the nifty new notifications, and the richness of the repository, made it an interesting distro.
Now, Pardus has got mature enough to be a nice distro for everyday use, and even managed to make a good KDE 4 experience. In meantime, 9.10 broke PulseAudio again and made the distro uglier, and now 10.04 “broke” windows management by forgetting 95% of their users. Reason ? The designers use macs… And what’s those plans for 10.10 ? Breaking it even more ! Making windows titlebars full of crap, so that new users are even more puzzled and wonder where the close button is because some guy found it smart to use a X windows tray icon !
Thanks for helping me exploring the richness of the Linux world, Ubuntu guys ! Maybe, when you’ve grown up a bit, we’ll meet again…
Edited 2010-05-04 05:00 UTC
…sounds like something Pavel Checkov would say.
How about waiting for the implementation before passing judgment?
The only thing knee-jerk reactions do is give you bad knees.
Because it’s damn near impossible to kill an API once it gets implemented.
If something’s going to suck the time to kill it is before it hits the wild.
AaronD protested…
Because the last few releases have been so chock-full of improvements that show the merits of Shuttleworth’s implementations and designs?
Ubuntu’s nonstandard way of doing things has helped so much more than it has needlessly broken things that somehow other distros don’t have any issues with, including those based on Debian and Ubuntu itself?
Because Canonical can’t seem to stop “fixing” things that aren’t broke?
I don’t know, man…I can’t think of a single reason why reactions have been so unfavorable…
Here’s a clue: It’s not a knee-jerk reaction when your reaction is based on past performance.
Hint: The past performance–it’s not good.
–bornagainpenguin
I really do think having a standardized way of handling per-application volume, showing online/offline status etc is a really good idea: f.ex. controlling volume would then work exactly the same in every single application, in exactly the same place, and thus would be very easy to remember, not to mention that you’d only have to fix bugs in one single place and get the bug fixed in all the audio-using applications at once. So yes, props for a good idea.
Now, about the implementation? It seems to me that his idea for implementing this thing is a combination of theme engine and client-side modifications. Ugh, horrible idea; changing theme would break it immediately. If he on the other hand wanted to modify Metacity then this thing would become WM-dependant and would not work in any other DE!
I hope they re-think this thing and implement it in a way that can then easily be implemented in other DEs as well.
Ah, but this would not be “handled all in one place”. You have to modify every single client application to support this. This is all handled “inside the app”. Which means, if you want to change 1 small detail about volume control you have to edit every single application that uses that volume control.
Compare this to host/server-side handling of window decorations: you make 1 change in 1 place, and every single window picks up the change.
What you want … we already have. It’s called a “window manager” for a reason.
AKA, don’t do it on the client side of things. Come up with a d-bus protocol for it, update the WMs to support it, and migrate apps to use the d-bus protocol with a fall-back to the current setup.
Ah, but this would not be “handled all in one place”. You have to modify every single client application to support this. This is all handled “inside the app”. Which means, if you want to change 1 small detail about volume control you have to edit every single application that uses that volume control.
You didn’t really bother reading my comment, did you? I DID say I don’t like their implementation, period. I said I like the general idea, ie. provide a standardized way for controlling such things.
Apple provided it a while ago: the top screen menu. It unifies system menus and notifications with the current application’s.
Even without implementing that, there could be workarounds. The current NetworkManager and volume applet could offer controls per current application too.
What he needs to do is embrace Qt and build a system around cross-platform development.
Sadly, I have to agree – with the subject title, but not necessarily the solution.
What he seems to have done is got himself fixated on Mac OS X as the holy grail for a desktop environment, hence the moving of window controls from the right to the left. It can only be a philosophical reason to do that. More space on the right? Give me a break. There was space on the left to put these daft controls in anyway Mark, you dimwit.
However, Shuttleworth can’t just outright clone Mac OS X so he has to find himself himself some daft and pointless differences, like these ‘windicators’, so the cloning won’t be obvious.
Frankly, it wouldn’t matter if he started using Qt or anything else as a toolkit or starting doing more cross-platform development. This mindset he has got himself into is now firmly deep rooted, and I can only see him digging an ever deeper hole for himself and Ubuntu. Quite how all this deckchair rearranging will allow Canonical to make money out of the endeavour is another matter entirely.
Edited 2010-05-04 17:35 UTC
I’m saying he should quit screwing with the interface and focus on improving the software library.
Not that I actually expect him to do anything of the sort. This is the guy that thought that covering Gnome in brown was the way to compete with Windows and OSX. Who the hell knows what he will do next.
It would have been nice to wait on “rearranging” the icons, until there was a reason to do so, rather than “phasing in the change”. Right now it is an irritant, especially when one (me) happens to use more than one distro.
And rather than rearrange the icon location, why not put the windicators next to the existing icons?
Shuttleworth/Ubuntu needs to stop with the small relativity pointless UI changes(icons on the left, soon to be removal/modification of the systray, and now windicators) and start to focus more on the corporate world. Its time to get Evolution working (or another client) perfectly with Exchange servers. The framework is there with Openchange and the Evolution-mapi plugin but it is still really buggy. (Although Fedora 13 has made several vast improvements of this with actually moving to Evolution 2.30 instead of staying back on the old version like Ubuntu)
I think this is the hold up for alot of techies and some regular users wanting to move away from Windows/Outlook to the linux world in a corporate environment.
QFT. Spinning wheels redoing the UI like this doesn’t help with bug #1, but perfecting exchange compatibility does. Even better would be to clone postpath: Write an open source mail server that Outlook can talk to and think is exchange but that doesn’t royally suck. Switching the back end off of Microsoft products is always helpful.
Doubly agreed. One of my greatest hassles using Linux at work is the lack of integration with Exchange (and Sharepoint and so-forth). Fixing that would be wonderful, and might score some community points for Ubuntu.
…but moving to Evolution as a mail client will cause more pain than it alleviates.
Here’s another idea.
Why don’t we take the menu bar and roll it into a single drop down button on the toolbar. It would be like the orb in Windows Ribbon interface.
Next, make the status bar transparent, and keep all of the “windicators” there. They won’t be in my direct line of sight, or cluttering up the top of the application, but they will there if I need them.
Finally, why is Ubuntu pulling services out of the DE? The DE provides the services, and applications hook into the services. I thought that was the agreed upon standard.
The more I think about this, the more I think this was inspired by web apps. The have a tendency to cram things together like that.
“Thanks to PulseAudio, Linux now has an audio stack with per-application volume controls.” That was available in OSS4 way before i’ve seen something similar in vista, so at least 3-4 years ago.
http://yfrog.com/j1snapshot1oxp
Sssssh, Linux devs don’t want to hear about how far behind they are. After all, all new Linux tech (ALSA/Pulse) is the best, and has features that never been used by anyone else, anywhere in the world.
ALSA is the best. PulseAudio makes it even better. GNOME rocks. Ubuntu will rule the world!!
[yes, that was all sarcasm.]
I am no fan of PA, but the original statement didn’t say “Linux is now superior to other systems because PA implements a brand new feature no one else has” what he said was that, using PA, Linux now has per-app volume controls. While you could technically have had this for a while now if you compiled OSS4 in to your kernel I don’t know of any distributions that did and it has never been in mainline. Thus, it *is* fair to say that this is new *for Linux* and *for most.*
All I can say is … Whoosh!
I don’t think so. I believe I understood your post correctly. You seemed to be saying “Linux developers are only just catching up and don’t like to admit it; ALSA and PulseAudio aren’t unique or special.” Care to enumerate exactly what point I missed?
Without regard to your original post, the grandparent of my post was an appropriate reply to the parent of your original post.
What I don’t get in all this is why does Canonical not communicat their ideas before they impose them on their users and devs?
Are they really that afraid that another distro might release “their” idea befor them? I mean discussing something can really help in finding caveats, in finding better solutions to the problem.
No. Instead they keep quiet and then *boom* the newest thing on the block that just is incompatible to anything.
For a beginning they could have left the close- etc. buttons on the right side and discuss their idea.
All that makes Canonical look like a company that does not really want to play a part in this whole ecosystem but rather make its own thing.
I mean upstart, their notification stuff and now this. Go collaborate ffs!
The only thing remotely useful to me is per-application volume control. But even without those “windicators,” it’s just a couple of clicks away to, say, mute the browser with PulseAudio. Once you’re accustomed, it isn’t really such a PITA.
The problem I have with the idea, though, is the visual clutter that will inevitably ensue. Hey, the taskbar exists for a reason, and now you want to put indicators on the title bars — and every single one of them? Unless the icons are redesigned to be as minimalist as possible and the number to be aggressively restricted, the desktop will turn to be a mess of icons as each new window opens. Even if you do redesign the icons — what about customization? Say goodbye to colorful themes!
Personally I encourage Ubuntu users to switch to Linux Mint, which is based on Ubuntu, just better and prettier with a nicer out-of-the-box experience.
A “couple” is more then one. Changing the volume of a single application is very cumbersome with the current pulse controls. Not to mention that the sound control may get hidden by the notification area if it is not used often (at least in KDE; I don’t know about Gnome).
As for what is useful to implement as a windicator, that’s really a tough question IME toolbars (uim, scim, etc) come into mind if per-application mode is set. However, I am sure application writers will find other, application-specific tidbits to put there.
It’s not a bad idea to provide some kind of mechanism for this sort of thing, where by mechanism I mean API. Dictating the UI is where I draw the line. Things like this should be exposed by the WM in a manner the WM considers best; could be an additional exterior toolbar, could be inside the operations menu (typically accessed by right click), could be exposed by mouseover on some area of the screen or window, etc., etc..
… are starting to become my nightmare. First that move of titlebar buttons from right to left (yes, I did swap them after install). And now this? Windincators?
I am not sure this is good or bad, I am still watching over this hot debate… If this is to fork more resource in my system, it will certainly suck, big time…
However I fear that I may have to switch to some other distro, like Fedora, when 10.10 arrives. This is a delicate step Canonical is taking, and could very well screw up Ubuntu’s worthy mass of followers.
I have that “This is not a democracy” thing still choking my throat.
Wait… someone said in the past that this whole Canonical thing would turn out into pure nighmare one day.
My justification for recommending Ubuntu has always hinged on Shuttleworth having his head screwed on right. What happened? Now I look like an idiot.
If you were to buy a second car, would you settle for one where the pedals were in different order from your other car? Me neither. I wouldn’t want that on my multiple PCs and operating systems either.
Us visually challenged have come to rely on button/icon positioning rather than trying to figure out it’s colors. Mark, you’ve just sent us back to 1984 in our learning curve.
I don’t know really what the motivation is here. Just because you CAN do something, seldom dictates that you HAVE to do it. Stick with the standard layout, please. Add the icon barrage in a floating toolbar if you have to have them. Just make sure that we can also hide it.
I spend my time convincing people to move from Windows to Ubuntu. Why create more obstacles and sources of confusion for the masses that are already scared of Linux because of clearly perceived differences in its interaction with the user?
Calling for return to sanity. You all know how easy it is to jump from one distro to another nowadays. Canonical should know it too. There is a point where pain of managing the nonstandard UI controls exceeds any perceived pain from switching to a different distro. That is the point where you will start losing fans, Mark.
This is really painful. Interesting post here by the CTO of Ubuntu: http://www.omgubuntu.co.uk/2010/04/canonical-cto-matt-zimmerman-tal…
“^aEUR|the decision to move the window buttons was made by Mark Shuttleworth. He explained that he wanted this in order to be able to use the space on the right side of the window for other things in the future. There have been many well-considered objections to this, both from users and from Ubuntu contributors, but he has decided to press ahead with the change anyway.”
We have many dictators in the part of the real world I live in; we don’t need one in the world of Ubuntu.
I have another reason to HATE this idea…. Sometimes I overshoot the “File” menu and click the window menu instead. The last thing I want to accidentally click is the window buttons. Having them on the right makes sense because the only reason I am in that corner is to manipulate the window. Why would anyone want those controls just pixels away from the commonly used menus of the application.
http://www.codinghorror.com/blog/2010/03/the-opposite-of-fitts-law….
“… I can think of a half-dozen applications I regularly use where the ejector seat button is inexplicably placed right next to the cabin lights button.”