Tuesday, August 7. 2007Amarok Plasmification
hello planet
we've decided that for amarok 2.0 we want to emphasize what used to be called the "context browser" in amarok 1.x speak. we decided to break out the context browser into its own central widget... and have re-used libplasma to implement it. this allows us to use all the goodness from plasma with minimal effort (there is of course some tweaking required). as we do not depend on kdebase, we are currently svn:externing libplasma in our amarok tree and linking to it internally. I understand that post-4.0 (4.1 maybe?) libplasma will be broken out into its own separate library, and we will be able to depend on that. using plasma we get to use gorgeous svg-themed applets, the clean and flexible applet/data engine architecture, plus many smaller but no less important features. so if you're curious about what amarok 2.0 will look like, look no further:
Trackbacks
Trackback specific URI for this entry
No Trackbacks
Comments
Display comments as
(Linear | Threaded)
For those of us who frequently use very large playlists, will it be possible to make the three panes look identical or almost identical to Amarok 1.x's two?
Nah, large playlists are stupid and ugly. And Amarok 1.x is not smart, although not ugly.
Let's use xmms then.
Please, let me be stupid and use playlists are large as I want them to be. Do you think that Amarok has mysql support to manage lists of 10 elements? Just curious.
Actually it has MySQL support in order to manage your collection, not long playlists.
I would also like that. Thankfully it seems like that might be possible from looking at the dots in the separators between the panels.
ok so what I see is interesting , but still along way to go. I can think of tons of opportunities that are open because of the plasmification of the middle pane. I hope that you can sort of plasmify the cover bling idea because it looks awkward having three vertical panels and a small horizontal stuck int he middle. I can't wait until you finish the work on the context browser the possibilities are endless.
my hope is to integrate coverbling into the contextview area as well, although i haven't looked into it too much.
basically, i don't know if it's possible because CoverBling currently requires a QGL-enabled graphicsview, and i don't know how that work work out.
Doesn't plasma depend on X? So porting amarok to Windows will be troublesome...
As for large playlists, I would like that too. As for the Amarok GUI, I feel that most of the elements have nothing to do with the oxygen style, so amarok at some point will be thrown out for a more robust KDE music player, I believe, and hope, if amarok really goes the path of evil I'm sorry if I'm trolling, not everyone has to agree with everyone.
plasma does not depend on X. plasma depends on Qt, uses Qt's QGraphicsScene/QGraphicsView framework, and is platform-independent as Qt is.
our context theme will need to be polished in order to fit better with the default oxygen style; our goal is to make amarok integrate as well as it can visually in a default KDE environment: that means Oxygen. our new icons are also done in the oxygen style.
If you refer to the Ars Technica article: "[the] new Plasma workspace environment will remain X" means that desktop will remain X only, because other environments already have their own desktop workspaces. (if I got that correctly)
I am curious how Amarok's current efficiency, and usability will be maintained. Can we switch back to the 1.x look and feel? What makes Amarok 1.x so great is that you can quickly move around in your music. With collections upwards of 25000 songs the current Amarok makes navigation easy.
The whole point is to see the ID3v2 tags, the context browser is nice but should never be the central focus. The focus needs to be on searching and your current playlist. Also the moodbar is very important for finding duplicate tracks and seeing how your music sounds before playing it. This makes making DJ playlists a breeze. I hope efficiency will take precedence to eye candy. Amarok is great not because it's pretty, but because it works well.
i don't mean to sound patronizing, but i think you talking in absolutes about your use case of amarok is not absolutely true. yes, amarok is very good id3 track manager. no, that is not the foremost purpose of amarok. it is a music player, and that is what we strive for.
Hi,
While I appreciate the point you make don't forget that usability is a major part of software's popularity. A few years back when I chose to use Amarok I chose so because it did a very good job of handling my playlist of (only) 1,000 tracks. You can have the best player under the sun but if choosing the next song requires one to trawl through the file browser it will get boring very quickly. So I would say a smart, quick and easy way to manage (and perhaps synchronize) the playlist are very important features of a modern audio player.
Agree. There are lots of decent music players, what I like amarok for is the easy management of my collection. It makes long playlists easy - although mine is not that long, only 900 tracks. I generally turn off the panes on the side. Just a huge playlist, please, even if that's not the way you envisioned amarok to be used.
I liked the look of old amarok, I hope you could hide this big thing in center that is waisting space on screen and fill it with something useful, like bigger playlist. I have lot of different colums shown in playlist so it has to be large, so you can navigate through easi.ly
Oh!! It'es very nice! I like very much the plasma widget inside.
But, the top menu with sound level and control button , is not structured. It look like a widget without layout with a lot of space between widgets. Maybe it's not the final top menu?
I approve of "plasmafication" as long as it doesn't prevent the porting of Amarok 4 to Windows.
Also, with the enlargement of the context browser, have you considered perhaps showing more content at once? For instance, instead of having "Music", "Lyrics", and "Artist" on separate tabs, perhaps they could be dragable panels that could be whatever size you want? Certainly, it should be made easy to customize what information the context browser displays.
Snapping Plasma into Amarok? Ah, suddenly Plasma makes sense to me!
Keep it up, KDE4 dev!
The main thing I worry about is which info will be displayed in context browser. I mean if it is always info about current playing track/playlist - it is bad. Context browser should be utilized depending what are you browsing now - platlist, collection, music store, etc. And maybe there should be several context browser layouts for this things?
I have big plans for using the central context view in connection with the services stuff I am working on.
For instance, when browsing Magnatune albums, I plan on using almost the entire context view to show artist( album info, related artist, and other fun stuff (such as the mood cloud view I had built on top of the old context view) instead of in that crammed little window at the bottom of the browser that the services currently depend on
I think all this things with plasma can as make Amarok even better, as make it unusable and eventually kill. You should be careful and think a lot about usability firstly as a music player.
Also I think the issue with silly playlist should be solved too. You can tell forever about "rediscovering music", but nothing can replace playlist completely. Maybe there should be two playlists - current one and plasma one, with current one visible only when there are no playlist in selected context view.
But will you be able to drag the plasma applets to the desktop?
But I agree with istdasklar that the top widget looks weird. Another problem I'm starting to have with all these SVG-themed windows contents is that suddenly the rest of the GUI looks drab in comparison. It doesn't always mix well IMHO.
though i think the screenshot looks nice, i do not believe it to be the optimal solution - at least not for me. If you added a (plasma) "amarok 1.x-style" playlist applet to the middle pane, or at least the option to place one there, i would be happy. I believe the current playlist is to limited for people with large playlists/people who want to see all the information about their songs.
Just my €0.02.
Please return back old playlist!!!
Please listen to the people. They really like old playlist. I am afraid that you are trying to hard to invent something new that you forget the the best thing can already be invented. Looking at this new ugly (from my point of view) playlist I can't imagine how I can sort tracks by play count or by my user core. Please don't turn great music player into applet view with ugly and not usable playlis with possibility to play music.
Can you do one thing to you old users?
Could you please just port old version of Amarok to QT4 and leave it as it is? I am pretty sure that users will be much more appreciated for this than for this new limited and absolutely not usable play list. What I can see on screenshot doesn't look like a music player. Unfortunately ...
I am using Amarok for a very long time,
and I love it for its playlist and it its database handling. What I do not like so much, is stuff in the UI that tends to take more and more space from release to release, although I rarely need it. This really not just my opinion. A lot of friends of mine, actually love amarok (despite it's tendency to krash) but had agreed that the new and huge context browser, just seems to eat up screen real estate (a resource that seems to be disregarded too much nowadays.) To not be all destructive and criticical, I'll suggest to separate the plasma context thingy from amarok to a full plasmoid. (I haven't actually rolled around in the code, so I don't know if that's a feasable idea. But from a UI perspective I would like that.) Or, since you probably have the problem that the feedback you get through blog comments (like this) and other usual resources is biased to the most outspoken user group, why not open up a poll (when starting amarok) or use LikeBack (from the guy who has written basket). ( Also the padding of these comment boxes is set to zero, which is horrible )
Nice from a technical point of view. But we should take care to keep a consistant and clean default look for all element that use plasma. A powerful vectorial theming system comes with the risk of kitsch and ill-designed gui.
Have a nice day,
BTW: Thanks for all these blog posts, it is really cool to learn more about the development process and what great progress you guys make!
I love the idea of Amarok with Plasma.
However, I really thing that's too much. Couldn't you make the Context View detachable? so you cab hide the rest and put nice browsing functionality in the Context View (or just making it a Plasmoid), somehow like Foobar2000. Context View was nice in 1.x series, but I didn't use it too much (only for lyrics), I used to have the collection tab active. I find totally stupid to display sooooo much information at the same time, people won't read all that information all time. This is a little extreme, but you could actually erase most of the current layout and just leave the Plasma stuff with all needed functionality, and it'll be still better than this current state. Anyway, Implementing plasma like that is enough for me to send you bigs cheers from Mexico xD
I am another concerned user - I really wonder if any of the KDE usability team looked at the current layout.
It is really strange to see the playlist so FAR away from the collection browser - that makes dragging items into the playlist quite unusable. The idea behind beeing "context-centric" is clear but I can not see how this can be practical in any usabiltiy concepts (Fitts' law anyone?!).... Another thing to care about: On the one hand amarok should be an out-of-the-box experience (like more or less static layouts and no overwhelming customization) on the other hand there is now an plasmoid area which allows free placement of elements (which imply wasted space) which leads to a bit of a contradiction..... So: - Place the playlist next to the context browser. Left or right - most likely left makes more sense as the distance from playlist to context browser and collection browser would be minimal. - Try to find a sort of auto-arrangement of the plasmoids to minimize wasted space (like what happens when you resize the main window - the widgets should match it's workspace size as long as possible). - Try to push the people of KDE usability and openusabilty to improve Amarok 2 - keep in mind that Amarok 2 will hit the OSX world (and the Windows world) so this should also be considered from a usabiltiy point of view Anyway enough rantig and trolling! I love Amarok and you developers do a great job - but as long as there are chances to influence the development it might be good to do so
Of course i wanted to suggest:
- Place the playlist next to the COLLECTION browser.
we are very aware of the problem that the playlist is far from the collection browser. our current plan is to cover the contextview with a sort of "popup dropper" when you start dragging a track from the left browsers---that way, you can just drop it onto the correct action. that would avoid the long drag.
and even though the middle pane is "plasmaified", it's plasma + amarok---that means that the user will not have exact control of the layout, there is a layout manager in the contextview, that ensures that applets use the maximum screen space. i assure you, no real estate will be wasted!
Leo, thanks for the reply.
It's fine you play around with new ideas but please keep some basic principles in mind. The basic drag & drop task is to place an item (or multiple items) to an EXACT location in the play-list. This action requires aiming - hence you plan the mouse movement/route beforehand and start dragging and drop it at your desired place. Something popping up would make this impossible. Even if it is the whole play-list cloned next to the collection browser when starting to drag, it would be confusing and counter intuitive (as you can not really aim before you drag). So I urge you to think twice and do what Fitt's Law concludes. And what is the big problem placing the context view on the right-hand side? This way, the context could still be "expanding" and dominant but you would avoid some obvious usability problem. Nice to hear you plan to implement a layout manager (I have not seen it yet in the recent SVN version or did I miss something?!) - also the screen-shot implies otherwise - besides an simple QLayout* will certainly not do the job but anyway you will hopefully manage this
it's there actually, since yesterday
basically what happens is it makes as many columns as possible based on a minimum column width, and flows your applets into the miniumum height based on the number of columns.
While most programs seem to get better and better, I really have my doubt about the development of Amarok, which looks like it's getting worse every time I see a new screenshot.
I wasn't really fond of the context browser in the middle, but now you've put an entire desktop in it. I don't care if you want to put some extra information for me, I like to see what other songs and albums there are, like the 1.x series, it takes 25% of the screen, just about right. But if you drop all these wiki pages, lyrics etc. in it, please make it optional, 80% of my music doesn't have any info about them on wiki's, and lyrics are even more rare. If you reserve space for that that space won't be used and I get an empty window. It starts more to look like playing music is an "afterthought" in the 2.x series... Anyway, I hope -- like other people here -- there's a way to get my 1.x look back, or I'll stay away from the new version. But from the looks of it, there seem enough people not liking it anyway, so the chance of getting some sort of a fork with the old look is there. And there are always other programs to go to if no fork will happen.
I agree. Pretty much every screenshot I've seen recently has made me want to cry.
Might as well just kiss my good old music player good bye and instead just accept the fact that everyone's trying to make stuff flashier than Apple for no reason. What a nightmare.
I can't believe the reaction to this. I would like to add to the comments in support of the new layout. I think all users should ask themselves, "Why do I use Amarok?" and look at 2.0 with their reasons in mind. I like it for a number of reasons.
1. The powerful collection browser remains relatively unchanged. For me Amarok has always been about having control over a large collection rather than a large playlist. Amarok 1.x as adapted really well to the ever changing definition of what a collection is and nothing of that is lost here. 2. Amarok 1.x has had at its core for some time now a really good use of dynamic playlists as an improvement on large playlists. The area it hasn't handled well is compacting a playlist full of albums and expanding streaming items within the playlist. I think the new playlist will deal with the drawbacks while keeping the benefits of the current playlist. There are superior alternatives to Amarok 1.x for those who use large playlists. If you are using large playlists within Amarok you are already missing the point in my opinion. You should already be looking for another player. Nothing in 2.0 will change that. 3. The only time I spend more than a few seconds looking at Amarok is when I use the current context browser to view lyrics, artist/album info or collection information. Otherwise I interface solely on an audio level or a quick drag and drop, pause or play. That is not only testament to the current usability of the collection browser but also testament to the need to improve the interaction with the context browser. I think the new plasma layout will provide not only an improved context browser but support for the playlist and collection browser plus who knows what other benefits. I do however share some of the usability concerns expressed by kroemer in post #17 but I think it is only fair to wait and see the testing release before getting to worried about that. Good points though. I would also like to suggest that the initial 2.0 release should undergo a further beta stage on Windows before official release for Windows. I think it is important that the app be stable and slick at the point when Windows users give it five minutes to prove itself. Good luck with the work. It is much appreciated.
1. Maybe for you the main point is to manage collection for your music. But for other people the main point can be to manage playlist. For example in iTunes music collection can be treated as large and unfiltered playlist. And this doesn't mean that people who like this miss the point. They can think differently, behave differently or just need something different.
2. Maybe there is a superior alternative to Amarok for your way of usage. It is called xmms. It has similar playlist and can be used in the way you use Amarok. 3. Other people can spend some time creating and modifying their playlist. You can want to change the playlist because your mood has changed, you want to play newly added tracks, or your best rated tracks but with some exclusion. I want to add that I agree with comment #17 in the point of view that playlist should be near collection browser. I hope if Amarok developers will not change their minds about vision of Amarok 2.x there will be KDE developers who can for Amarok 1.4.6 and port it to QT 4.
1. The powerful collection browser remains relatively unchanged.
This is because it hasn't been "innovated" yet. Just wait until it will be a time to "innovate" collection browser.
To Leo Franchi.
>to those who don't know who i am, my name is leo franchi, and i'm working on amarok for my SoC project this summer. I guess now the readers of this blog know who are you. But I wish you never started work on Amarok. Nothing personal but you made it look horrible. Sorry, but I can't stand it. It is just awful.
Wooah dude that was just mean...
Leo I am sure that most people really apreciate what you are doing and we are really excited about the possibilities of this and looking foward to what Amarok 2.0 will bring. I understand that people are afraid of new things, but I encourage you and the whole amarok tema to keep on the good work. I understand that this is a work in progress and it is excited that you are experimenting and looking to INNOVATE.
I really didn't want to offend anyone. Sorry if I did.
But sometimes I have a feeling that some project just try to reinvent the wheel when they do what they call "innovation". Sometimes it happens that the best solution is already invented. And you need to be careful in you innovations. It was just emotional comment. Sorry again .
Wow, so many comments to make a hard-working developer feel unappreciated. To all those developers, especially Leo: keep up the good work! For my part, I'm currently undecided on the exact new (default) layout, but the concept looks interesting, and I'm still hugely excited by Amarok2. Amarok 1.x was so good, how could you complain without waiting for an even remotely finished product? Feedback, yes; 'Nothing personal but you made it look horrible', absolutely not.
As for the new layout: - I'm assuming, with all the plasma goodness, that this is all theme-able. The default will presumably be a large pane as given, but with a pane size that can change, anyone can create a theme that can look like the old if they really want? - As well as the spiffy new expanded description layout of the playlist, there may be a 'simple' or 'condensed' view, ie one line of info? So, when we're in the mood to rediscover our music, we can get the expanded view; when all we want to do is listen to the same playlist for five days straight, we can get a condensed view. One of the great strengths of KDE and all its products is configurability - the default may look ok for some, but more ambitious users can always change it to their liking. As to Leo, thanks for the update on your work. Big changes are always hard to take: so people are just a little skeptical at first (as tikal26 just said, keep trying to experiment and innovate!). Give us a great end version, and they'll change their minds just as quick. The hard work, I guess, is the plasmification and construction - themeable items, while important, can be fiddled with at any time after that.
I also don't think Leo is to blame for the current layout and look (so keep the heads up).
This is a team project so if there would have been big objections then there would be a different layout. Anyway, innovation from a technical point of view is great. Even if I have personal doubts about using Plasma, it might be nice idea (on the other hand an bad idea do basically fork it). But there are some fundamental facts about usability and visual design (even if taste differs) which really need to be addressed by the team. If there is no-one in the team that is experienced in this field then (as suggested) try to push the usability team and some artists. It is clear that there is a priority as things like layout and "themes" are quite easily adaptable in contrast to release critical bugs Nevertheless it might be wise to attract someone who polish things up a bit - it's all about promotion in this case - even if this is not final an aesthetic look will be a major aspect for public relations, as Amarok is so famous that even non-Linux-specific sites are posting screen-shots of alpha/beta releases and hence people might get the wrong "first-impression". Especially end-users may not realize that those screen-shots are only showing some state of back-end development as they are lead by their visual impression. And I doubt anyone reasonable finds the current look attractive (as it is clearly a hack)
Like some people, I'm not liking how Amarok is turning out.
To me this focus on the context view is pure non-sense. It's not because it's cool that it makes sense. The main focus should be the playlist. What's inportant in an audio player and media library manager ? 1. Play audio. And play it damn fine taking into account replaygain and true gapless out of the box among other features. 2. Database management. Kick ass optimized indexing, fast and efficient both speed wise and memory wise. Make this data accessible in configurable ways to the user. Make every Tag accessible, displayable in columns or collection tree etc. Glossy UI while looking cool is a waste of time to me. You're not supposed to stare at your player all day long, do you ? I predict that if playlists are not the main focus of Amarok 2, a lot of users are not going to like it and use something else. Right now I'm using foobar2000 under wine. It's still the best for my needs. It can be as bare or as fancy as you want. It handles multiple playlist and large playlist just fine (which is not surprising since it was designed for that). Running under wine it takes less memory than amarok. I could go on forever on foobar but will stop there. I urge the amarok devs to reconsider some of their design decision...
I think amarok looks really great and I'm really looking forward to use this!!!
Keep up this great work and thank you very much!
Is this a mock up or the real thing? Waaaay too much wasted space on the top, among other things...
The plasma bit looks quite good so far, though it's obvious that some things need to be improved, but aesthetically, the possibilities for it look good.
The top bar, however, with the control buttons and whatnot, looks quite horrible. I don't really have any suggestions for improving it, unfortunately, I think it just looks ass-ugly. I hope it gets a major makeover before release.
People.
Of course this is just a mock-up. The sidebar includes two Playlist entries. Leo is showcasing the new features, thus they are large. Consider that the panes are movable and widgets are also movable. Consider that you could potentially move everything back to where you want it.
People, people, seems like you haven't learned one of the most important things when talking about Amarok.
Some time ago I had a playlist of 3000 songs and I was happy with it and I never ever wanted to change to anything else. In the end something just made me try a dynamic playlist, and I loved it and changed to dynamic playlist. You have to learn that whatever things look like, in the end Amarok developers are right. If I didn't know that, I'd most probably think the new layout is a mistake. But as I know Amarok devs are right, I am pretty sure using it will make my mind dance in happiness. Can't wait till the release to see what new coolness the world's greatest music player has to offer me! ^_^ Great job, Leo! \o/
I can't agree with you.
Nobody can't be always right by the definition. I am a developer myself and believe me that the right time to change something is initial stage of work. Nobody wants to change something radically if he just finished work and implemented different approach. I am absolutely positive that when 2.0 will be final nobody will do major changes to it for years.
Nobody can't be always right by the definition, I'm agree. But the point is that amarok's developers have done important changes before that people could think are bad, but they were right.
Why don't wait to try the new version before you say it's wrong? It's not the first time the amarok team can surprise us. |
Amarok LinksCalendarQuicksearchCategoriesSyndicate This BlogBlog Administration |

