A slick KDE Traffic #53 is out. This one features a single long-standing topic, the release cycle of KDE 3.2. Read all about it at the usual location.
Dot Categories:
A slick KDE Traffic #53 is out. This one features a single long-standing topic, the release cycle of KDE 3.2. Read all about it at the usual location.
Comments
> Oh, but then you work with a widget set that is much more third-party than ours, so you probably can't avoid that sort of problem.
Yup.. at various times in history, kde and qt's themeing wasn't even the same.
Hey... I can respect the notion: KDE can't ship a "stable" release and have it fall flat on its face... we all want continuous improvement, no doubt about it.
But I have to agree with Binner/ Kainhofer:
added time != improved reliabilty
Keep the scheme of recent release-schedules(post KDE-2) of fairly regular/frequent releases. I don't really agree with the dev-releases because that's just CVS no matter how you look at it. Just wrap up the changes sometime soon (maybe after Nove Hrady) and announce a feature freeze.
I know everybody wants to get in their new wiz-bang features, but every few months (definitely less than a year) the code needs a good clean up and then released to the users. More time before a release just allows more time for CVS to become totally different from the previous release... it doesn't make it any more stable. (From my viewpoint anyway).
Oh come on anyone who thinks KDE is easier to use than gNOME is kidding themselves. In terms of usabiltiy GNOME wins hands down, and usabiltiy isn't just for dumb users it is for everyone even smart users like havin a usable desktop, this does not mean taking away a lot fo features, it just means reimplemented them in a easier to sue way. IE: DE's clock.
KDE is really far behind in usability and without something as good as GNOME's HIG (written by SUN) anda QA team taht needs to validate KDE applications ebfore shipment things won't improve.
Users want more feautres, but they would much rather have an easy and antural desktop.
Your text's readability sucks as well. ;)
> Users want more feautres, but they would much rather have an easy and antural desktop.
That's not true at all. People (who matter-- non-computer "newbies") usually want features, not usability. If this weren't the case, Apple would have 90% market share, WordPerfect would be the most popular wordprocessor, Lotus 1-2-3 would be the most popular spreadsheet app, etc..
Anyway, most users screamed for features from the time before KDE 1.0 to around KDE 3.0. There were less usability concerns back in KDE 2.x and 1.x because there was less configurability.
In fact, back then, a lot of people said that KDE was less tweakable than GNOME 1.2 and 1.4. KDE 2.0-KDE 2.2 probably were actually.
> KDE is really far behind in usability and without something as good as GNOME's HIG (written by SUN) anda QA team taht needs to validate KDE applications ebfore shipment things won't improve.
Nah.. KDE 2.x proved that things could remain usable than a rabid usability team that continues to strip out or leave out perfectly usable programs, as is the case currently in the GNOME world. This might be fine in GNOME development, but there are simply more volunteers in the KDE side of things than vice versa, and I don't think turning down people's work is quite healthy for KDE right now.
> KDE is really far behind in usability and without something as good as GNOME's HIG (written by SUN) anda QA team taht needs to validate KDE applications ebfore shipment things won't improve.
Talk about a poorly formed cheering section. KDE had design guidlines when GNOME was making a big deal that they were going to have them and asked KDE to sign on. Not only that developing on KDE automatically supports this where as GNOME developers *have* to be asked to do this because of architectural differences. Of course SUN is known for being a powerhouse in desktop market share. Now KDE needs a QA team? How about advertising, public relations and cost accounting? The thousands of users on CVS as well as the millions on official releases using the bug tracking system affords it's own QA. However I'm wondering how a sentence started on usability and gravitated towards QA which is not directly related.
I think the crime of all this propaganda from the GNOME camp on misguided usability ideas is that it first assumes that usability is a "one size fits all" proposition. If that is what open source is now about then we really do need the advertising and accounting departments to sell that BS. The next one is that fewer options make any difference at all to the vast majority who never even look for them if there are good defaults. Finally there is the fact that for experienced users who know what they want, not allowing them to do what they want does not enhance their user experience.
Just because someone subscribes to an invasive perspective that tosses out software contributions in favor of rewrites because of "too many options" does not mean that such a narrow view has any merit for those people who hold the belief that flexibility and ease are not mutually exclusive. In case you're missing the point there is a difference in serving a button down world of lazy sysadmins and real people who actually enjoy a menu instead of the old SNL "cheesburger, cheesburger, cheesburger, cheesburger, cheesburger".
You are absolutely right. Glad to see such perspicacity.
Gnome believes that less options mean better usability; and better defaults are good for more users.
While these points are very valid, I know very very few people that acutally can and want to work with a Gnome desktop. Most people just get bored and reboot to Windows.
I can't tell you why; but Gnome is not as good as you claim
Well, yes in the literal sense, but than again would you want applications where you can do only the simplest of tasks like in WordPad? I doubt you would and I would not want to have such dumbed down programs myself, I believe ther eshould be a good balance and msot features should be kept uness they are compeletely bogus like KDE's background program which only takes a screenshot ofa website you point it to and refreshes the backgroudn every so often without eltting you navigate the website at all.
However, what I do want is to have all the feature rich applications in KDE like Konqueror, the control center, Kdevelop etc. completely undergo a UI check in which the usability of these applications can be improved through contructive criticism and work from the developers. As it is, many of KDE's applications are very powerful yet not as easy to use as they should be. For example as many users have complained, Konqueror's context menus are just TERRIBLY stupid and will often display irrelevant options and miss the most obvious ones like "copy", the "actions" menu should improve usability a little though.
I know power users don't care much about usabiltiy even though they do like better designe dUIS, btu teh thing is that normal suers wh aren't linux fans and just want to try it will quickly be turned off. Especiallly by the isntallation fo programs, I know I was, I do hope autopackage.org will fix this issue plaguing all Linux's.
Anyway, i am very pleased to see that some UI sisues are being fixed, like the KDE clock for example. =) This si a nice start.
It would be great if KDE would design quality UI guidelines like GNOME's hig though.
> It would be great if KDE would design quality UI guidelines like GNOME's hig though.
KDE has already had these guidelines for over four years. Unfortunatly, the problem is that
> However, what I do want is to have all the feature rich applications in KDE like Konqueror, the control center, Kdevelop etc. completely undergo a UI check in which the usability of these applications can be improved through contructive criticism and work from the developers.
Btw, it helps if users contribute by giving mockups of how things should look. Qt Designer really makes this easy.
> For example as many users have complained, Konqueror's context menus are just TERRIBLY stupid and will often display irrelevant options and miss the most obvious ones like "copy", the "actions" menu should improve usability a little though.
And as MANY other users have said, it's been fixed in kde-cvs.. now, Konq's and Nautilus's are comparable in terms of items. Nautilus still doesn't have an actions submenu, while Konq does, but on the other hand, Konq doesn't have a "Open in" submenu, while Nautilus does.
> KDE's background program which only takes a screenshot ofa website you point it to and refreshes the backgroudn every so often without eltting you navigate the website at all.
It's not meant to-- it's mainly for displaying one page things like satellite images and such-- at least that's what I used to use it for. It's not really completely useless. Will most people use it? Probably not though.
I tend to agree with Mosfet on this one. KDE needs to have some kind fo regular realeases like GNOME has snapshots with well documented changes and a few screenshtos and other goodies,
Developers should each have a list of what they have done for KDE so the changelogs could easily be compiled.
What I think would be nice to help make this happen is for anyone to be able to click a poster or reporter's name and see a littleinformation about the person and their positiion not jsut their e-mail. For example
name
personal information (e-mail, interests, lcoation etc.)
bugs reported in last 7 days
-
-
-
-
-
-
-
view all bugs reported by ----
comments by -- from last 7 days
- (links)
-
-
-
-
-
-
-
view all comments by --
patches by --
you get the idea.
I think taht would really be an improvement
>KDE needs to have some kind fo regular realeases like GNOME has snapshots
Wait... not so fast. I personally hate the way GNOME releases code, so I'd rather not take them as a good example. Why do I say that? Easy: Distros would start putting code that is "the one that works".ç
In next versions we could have either KDE 3.1.92, or 3.1.93.4, or whatever, if they see any problem in 3.2 or they just cannot wait until 3.2, and that would be a mess.
As an example, do you actually know *anyone* that knows what version of Gnome they are using?
I only have rpm versions of Gnome, but my system says:
libgnome2_0 -> version 2.2.0.1
libgnome-vfs0 -> version 1.0.5
libgnome32 -> version 1.4.2
libgnomeui2_0 -> version 2.3.0
libgnome2 -> version 2.2.0.1
So... am I using 2.2.? 2.3.0? 2.2.0.1? I don't care, but I would care about my kde version.
bugs.kde.org's query interface and http://lists.kde.org/?l=kde-cvs&w=2&r=1&s=cvsuserid&q=b are your friends.
Good idea regarding your suggestions for bugs.kde.org. I'll look what feasible ways realizing something of that kind are there.
I can't wait tos ee what you come up with =)
Also thanks for the tip .... annonymous ?!
One of the biggest things KDE fans are waiting for, in regards to KDE 3.2, is the new stuff in khtml from Safari. While I understand that a lot of really, really great ideas are going to come out of the conference coming up, I don't think you should let the hormones and insight completely take hold of the release cycle.
A 4.0 release is something that would take a significantly longer period of time. It might be six months. It might be three. It might be a year. We won't know, and only the developers will have any idea. With an incremental release such as 3.2, we have reasonable expectations that a release will come out sooner rather than later.
Konqueror, and by extention KDE have a lot to gain by using the momentum from Safari. Having to tell users that yes, the changes from Safari will come back into Konq, but then having to tell them that it won't be for a long time will have a worse effect than never having the help from Apple at all. It's all about higher expectations.
About promoters: Promoters are not magical people who can open up your brain while you sleep and download a new list of features that are in CVS. Nor should people who are primarily promoters be required to read every line of code in the KDE project; that's what developers are for. Honestly, if developers want the promoters to spread the word about new features, developers are going to have to meet them half way and write in depth and accurate changelogs.
I'm not even an official promoter and I know it would help *me* when I extoll the virutes of the KDE project.
Promoters should be able to organize themselves and give the developers the time to actually improve KDE instead forcing them to spend time doing something nearly everyone else could do as well when taking enough time.
I myself am not a developer, I consider myself a promoter. ;)
There's two substatements implicit in your statement:
1) That writing comprehensible changelogs is beyond the responsibilities of a coder.
Personally, I feel that is about the same thing as saying that coders should spend time coding and not documenting their code. Changelogs explaining what exactly it was that they added is part and parcel documentation. It could very well be that these changlogs are already in there somewhere, at which point it becomes a technical problem.
2) Project Promoters are an organized force, just like developers.
While there is a kde-promo team, they aren't the only ones doing the promoting. The folks sending out promotion info to companies may be an organized force, but the KDE guys I meet with at the pub on Thursdays certainly aren't. Nor should they be, because then it starts to feel like astroturf.
I personally don't have the resources to get involved with kde-promo officially, though I certainly talk about KDE wherever it's appropriate. I've made a fair number of converts. More documentation about what's coming would certainly help. If KDE is actually interested in the number of people it has using it[1], they shouldn't discount this sort of help.
La la la, I'm (most recently) a project manager. I think it shows. :)
[1]: It could very well be that the KDE people don't really care how many people are using their desktop. If that's the overall belief, I can dig it. However, if that really were true, what's the point of having a kde-promo group in the first place? Surely it's not just to bat away the GNOME trolls. :)
1) The the most extensive documentation of changes in KDE possible is the KDE CVS list, Derek Kite (a single person) shows week for week how useful it is for documenting changes in KDE. Now think about two additional components improving on this: firstly tell developers to add more descriptive text to their CVS commits, secondly let some people dig through the KDE CVS list, Derek's weekly summaries and personally check out the latest CVS head. Can only developers do that? Imo not at all. Is it hard to do? Imo not at all. So where's the problem? =P
2) We shouldn't think about any group as organized force in the sense that it's a closed strictly organized force (most of the time it just works mutually), what we should do is encourage everyone to do something about something as soon as she/he feels like it instead giving others the impression that contributing is very hard and others should do it, then having them begging for this and that. I'm glad you already noticed one doesn't need a promo force to convert a "fair number" of people. ;)
At http://datschge.gmxhome.de/support.html I've created a one-stop page showing all ways anyone can support KDE. I'm considering adding a section for changelog creation as soon as a respective website/mailing list exist (I think a wiki site would be great for this purpose, anyone willing to donate one? ;).
1. very nice list on that website.. needs to go somewhere on kde.org imho :)
> I'm considering adding a section for changelog creation as soon as a respective website/mailing list exist (I think a wiki site would be great for this purpose, anyone willing to donate one? ;).
Perhaps set something up on developer.kde.org?
I just want a slick looking itnerface which si almsot there, only these issue sned to be fixed: http://www.kde-look.org/content/show.php?content=3910
And most importantly, I want an intuitive user friendly, and consistent desktop. KDE greately need to imporve usability, many of its base applications such as the clock with 5 tabs http://bugs.kde.org/show_bug.cgi?id=58775 or KDE's bulky find tool also with far too many options shown at once. Even Konqueror is a mess when it comes to logical usability, often listing dozens of options in its context menu which are completley irrelevant to the user's actions and yet are still present and not even grayed out.
For a good example check out the find tool in GNOME 2.2 , as powerful as KDE's yet infinetely easier to use. Try it out yourself.
Well, if you really want dumbed down user interface which still manages to look confusing and definitely not sleek, then maybe you should use GNOME?
I installed MDK9.1 to one of my computers yesterday and while I was compiling KDE from CVS (didn't install MDK KDE packages), I tested Gnome 2.2 that was included with MDK.
First I tried to configure panel, there were only very few options and the configuration dialog looked like it was designed by some six year old in playschool. Then I tried to configure the clock in the panel. Same thing happened. Not configurable at all and configure dialog still looked clumsy. Great indeed.
I certainly don't want that to happen to KDE.
I hope KDE never ever try the same way that gnome go.
I heard from more people asking :"Where can i configure this and this feature" then the "Configure Center" is confusing me.
I don't want a second "Windows Regiestry" where you must search for this and this feature.
Sorry for my bad english *g*
I think we need to be balanced about this. I have no problem with KDE staying fully configurable. I don't really buy the "configurability confuses users" argument. But KDE clearly could be greatly polished without losing a whole lot of functionality.
First of all, toolbars need to be rethought. Many KDE apps come with 3(!) toolbars. Very Microsoft-ish. In comparison, most GNOME apps start with a single toolbar. The default toolbars should only come with the functions that absolutely need quick access. For example, in my Konqueror setup, I have all of 4 buttons in my single toolbar.
Second, right click menus need to be cleaned up. Right-click menus are supposed to be context-specific, not general purpose! You should not have to actually read the menu like a list to find out what action to invoke. There needs to be some intelligence applied to right-click menus. For example, the right-click menu for images in Konqueror (as of KDE 3.1) has 21 items. Almost all of them are useless.
First, there are navigation items. If I wanted to navigate, I would click on the nagivation buttons on the toolbar, or use a shortcut key!
Then there are three items for opening the image in various tabs and new windows. If those are absolutely necessary, they should probably be moved to a submenu.
Then there is a non-functional copy item.
This is one item that really *should* work, but doesn't!
Then there is a "add to bookmarks" item. How often do you add an image to the bookmarks list?
Then there is an "open with" menu. Okay, this one can stay.
Then there is a copy-to menu. This can also stay.
Then there are save link as and save image as items. How often to you save the link to an image that you can't copy the image URL and save it manually? And the save-image item is redundant given the copy-to item. Having both to do very slightly different things is very wasteful.
Then there is a select-all item. How do you select all of an image? Well, this item doesn't. It selects all of the text on the page. Isn't an RC menu supposed to be context specific?
Next copy image location and copy link location. Okay, these are necessary.
Then, stop animations. If this really needs to be accessed quickly, it should be a button on the toolbar instead of cluttering up every right click menu.
Then view image. Don't we already have an open-with?
Then, view source, view document information, security, and encoding. What in god's name do these have to do with the image in question?
Most egregious of all, there is no properties menu. Konqueror as a file manager has a properties menu, and the image is just a remote file, so it should have a properties menu! And if you had a properties menu, you could move less often used items like image location to that.
Okay, let's take a look at what this menu could be:
- Open
- In new window
- In new tab
- In background tab
- With viewer
- Copy
- Save
- Image As
- Link As
- Properties
Now, the menu is reduced from 21 top-level entries to 4. You don't loose much important functionality, and make the menu finally useful for what it was originally meant to do --- provide quick, no-think access to context-specific features. Big win for speed of use. Best of all, if you allow these menus to be configured, the enterprising user can always put in exact what he uses most often.
Now don't get me wrong --- I love KDE. It's my only desktop. I use KDE's features extensively, to the point where I have it 100% tweeked to work exactly the way I need it to. They'll take away my configurability over my cold dead body. But I don't see any reason not to at least make the default setup more clean, for the sake of efficiency if not ease of use.
That is waht I was saying in my posts all along. KDE does not need to abandon a lot of configurability, only a few options.
What it does need to do is have smart context menus, better layout of features and ebtter documented features, Users shouldn't even need to see options that are irrelevant or can not be used in the context menu.
I use save link as because kde has io slaves. I can go to another app and give it that url and work with it and then save it back to the server using something like webdav or sftp. Also when I wan to tell a friend about an image and I email them about it I send a link not the image.
Right on!
Some other things: The throbber on konqueror. I've used it to spawn new windows since kde 1.x and found it very useful. Unlike the throbbers in other browsers, this one doesn't take you to netscape.com or mozilla.org or kde.org. It has a cool function, but this has become a little redundant with tabs coming in the feature set. I rarely, if ever, use it now. It could have some other neat functions like:
1) The Home button, unlike other browsers, takes you to ~. Of course it is needed in the file manager, but there is place for the Home page too. The throbber could be a place. This way, at least it will behave like other browsers. This could be configurable though.
2) The throbber indicates activity. How about a stop function during animation and reload during no animation? That would decrease clutter in toolbar, getting rid of two buttons. Sure there are some problems associated with a stop/reload combined button, or so I've heard, but it sounds good to me.
Look at KDE CVS head, some of your suggestions are already implemented there afaik.
> Look at KDE CVS head, some of your suggestions are already implemented there afaik.
Indeed.. Konq's context menus have already been cleaned up quite a bit-- however, it's toolbars still have not (only apps that should have more than one toolbar in KDE is probably kdevelop and koffice apps.. nothing else should, especially konq)
Don't be ungrateful for the multitude configurability. That is what seperates KDE from the rest. The power it gives to its users. How many Desktop Environments offer as much configurability as KDE does? None. The fact that I can get KDE looking like MAC in a minute, windows in another, gnome with a few clicks, icewm if I wanted, fluxbox if I tried, UNIX if I felt like it, says wonders about KDE. It is built for the user in mind. It respects our choices, our uniqueness and our differences. It empowers us. See if you'd get that if all those configuration options are removed? Point is, you wouldn't.
The Woes
The KDE KERAMIC Default style is clunky, bulky, bulgy, unnatural and just plan horrible. 'Sexy, Simple, Sleek, Slim, Speed', what ever happened to that interface design philosopy? That was all kde was about until keramic reared it's ugly, slow, unresponsive, memory choking, annoying head in. All that glitters is not Gold, you can achieve much powerful, fast and eloquent user interface designs if we learn to keep it simple, like the latest release of Gnome is doing.
Outside the above, KDE is great and it remains the best! And whoever said GNOME is more functional that KDE most be smoking something, stones perhaps. The only GNOME oriented program I'm jealous of is GIMP. Which reminds me, is there a KIMP package around?
Yes, keramik style looks very clumsy and outdated. I personally prefer dotNet nowadays.
Maybe we need a vote on the slickest style out there for kde ?
My first choice is the "Alloy"-style:
http://www.kde-look.org/content/show.php?content=6390
Agreed! Ilike the smooth toolbars and most widgets of Keramik/Keramik II, but buttons are horrific. I wish buttons were smoother/thiner. Something like in the mozilla theme "Pinball": http://www.deskmod.com/?state=view&skin_id=9363
Someone need to post screenshots of the new Keramik in CVS head, it's much better imo.
I support Kde's look, it makes KDE stand out and most people think it's a nice looking style, myself included. It's not for everyone, but it can be easily changed. The most important thing is people will see a new install of KDE and instantly recognise KDE because of Keramik.
I'd be most upset if Keramik II wasn't adopted as the default style for KDE.
"The KDE KERAMIC Default style is clunky, bulky, bulgy, unnatural and just plan horrible."
I agree. I'm using KDE 3.1.2 with KDE 2 style/icons
"The only GNOME oriented program I'm jealous of is GIMP."
I'm using Abiword 1.9.1 instead of KWord 1.2.1. The RTF import and export filters work better IMHO.
I agree with the poster who says it is what makes KDE what it is. Keramic II,is what should be the new default. Keramic has become what XP's deafault Luna or Mac OSX's Aqua is. It sets it apart from the rest. As for Gnome,it too is great. Using Gnomes graphic login manager,you can change the login screen as much as you can in XP,but for free. And it is the closets thing to looking exactly like Mac OSX when you add the right skin/icons. KDE always retains that clunky,boring Windows95/98 login. Even when another is added,you get that KDE or KDM login manager. Sure,you can add a different wallpaper to it,but it's still boring compared to Gnomes graphic login screens. Once KDE gets rid of the boring login screen,and uses Keramic II,I'll love it.
kdm is slated to have gdm theme support in kde 3.3.. see kdenonbeta/kdmthemes in CVS
When will we get autoreload feature in Konqueror ?
Waiting, waiting....
What is autoreload ?
Reagrds,
Dawit A.
Autoreload? www.faceoff.com/nhl any game. As the game is played, the score and stats are updated. In konqueror. I use it all the time. It is responsible for most spelling and grammar and html mistakes in kde-cvs-digest.
Go Anaheim.
Derek
Assuming you mean the same thing as I think you do, the answer is... you don't have to wait, just install the Konq plugin I've attached to this post. :-)
Rich.
*LOL*
I love it...
An updated version of this plugin is now in cvs in kdeaddons/konq-plugins/autorefresh, enjoy.
Rich.
Is it automatic refresh of the page afte ra givn amount of time, is it that fun game or .... AHH!
I agree that a feature freeze would be good.
As mentioned bugs are not fun to work on and doing a feature freeze would force devs to work on them.
I always thought that once the all items in the todo list for 3.2 were green we would get our release
I'm also wondering if a lot of apps shouldn't be sent of to kde apps( eventhough the interface is but ugly in my opinion. they should remove the frames...).
By having so many apps it seems you add so many more bugs and delays...
I keep imagining a software item in the control panel where I could select extra components from the KDE project. For exemple in kde multimedia i could remove noatun and select kmplayer and juk...
But then I think of so many reasons not to do it :-)
Quite frankly - I couldn't agree more with Mosfet. Pulling things from CVS all the time is a pain in the ass if you just want to check things out. Compiling everything every few days is a pain in the ass.
If KDE could prepare/release an "in progress" release once every month - with the only "requirement" that it compiles on x86 - then I'm sure some distros would start shipping these in-progress releases.
Furthermore, more people would be inclined to test it. I know I would. I don't care if a few features are broken.
Stable releases are well and good. Don't change the way you release them -- but "along the way" releases _REALLY_ should be made. The way it is now is quite frankly sucky in that aspect.
Here comes a shameless plug:
Gentoo lets me have a stable kde and cvs kde available and working, selectable on login just like selecting gnome or kde.
If people want an unstable release of kde, then maybe it's up to them to chose a distro that will make it easy for them.
Derek
What's the easiest way to do that? I want to do it on my Gentoo machine too. :)
Get the ebuilds from http://cvs.gentoo.org/~danarmak/kde-cvs.html
Detailed instructions are there.. it's pretty damn cool too-- let's you have both KDE 3.1 and KDE-cvs within portage. It even lets you pick where you want to put the cvs'd code in, and if you have a KDE cvs account, it even let's you use that, along with ordinary anon-cvs.
These packages with ccache making getting and maintaining a recent copy of kde-cvs extremely easy within gentoo. It even let's you have qt-copy, qt 3.2-b and qt 3.1 installed at the same time :-)