OSNews

And they had dconf instead of DCOP.

Problem there is that dconf is a really different project

Number of occurences were they spelled Qt right: 0
;-)


where
;-)

There are public forums for each project. For example, for kpdf I found the following: http://openusability.org/forum/forum.php?thread_id=325&forum_id=304
entitled "First usability inspection".

Replying to self: And for other programs the section "Usability Reports" is actually filled, for example for kivio: http://openusability.org/reports/?group_id=44

Why all that fear?

Do you know that these are just ideas yet?
Do you know that you can get involved?
Do you know that you can post your ideas to prevent that what you fear?

Why keep all those ideas locked while you have the freedom to express those ideas?

If you are so very certain that this will fail, then why not write a great text explaining those reasons and, why not, tell how to solve them with your ideas?

You are free to do so.


If you can't come up with specific points to why the current ideas would fail, then I don't get where all this fear comes from.

I know they're ideas, and I'm expressing my opinions about them. I'm not particularly good at writing things, but, ironically, I'm working towards my ideas as we speak. Specifically, I'm working to integrate the linux kernel into a more coherent directory structure, along with writing a completely new environment--new DE, new language, new applications. I don't want to bore you with the details, but hopefully it will be pretty usable when I get some of it done in a year or two.

-bytecoder

"Specifically, I'm working to integrate the linux kernel into a more coherent directory structure, along with writing a completely new environment--new DE, new language, new applications. I don't want to bore you with the details, but hopefully it will be pretty usable when I get some of it done in a year or two."

hmmm... I wish you good luck and long life. Many have had such ambitions, and few have succeeded. Does your project have a name?

Well thanks. It is a very large project, but I don't see myself quiting--I'm very excited about it. I don't have a project page up or anything, but I'm currently going by the name HeliOS (pronounced he-lee-ose); I figure I might as well get something up and running before bringing it out into the open.

-bytecoder

"Specifically, I'm working to integrate the linux kernel into a more coherent directory structure, along with writing a completely new environment--new DE, new language, new applications.

Is this new directory structure Gobolinux-like?

I suppose--it's roughly based around the same principles. This is basically what I've got so far:
/ -
  Applications
  System -
    Kernel
    Libraries
    Services
    Temporary
    Utilities
  Users -
    bytecoder -
      Documents
      Images
      Music
      Settings

There will be no package manager; instead, application bundles would be used, a la NeXT. The user simply drags a self-sustaining application into the /Applications folder to install it, and deletes it to uninstall. I should point out that, to the user (bytecoder), the filesystem is presented like this:

Applications
Settings
Help

Documents
Images
Music

Both Settings and Help are smart folders, which collect the settings program and help files from each application bundle and lists them to the user.

This just scratches the surface of what I'm planning on implementing, but I don't have the time to describe the rest.

(I didn't proofread this much because I'm tired, so ignore any gramatical/spelling errors)

-bytecoder

Ugh, the spaces ( ) showed up fine in the preview...

well, you seem to have a distinct vision for your project and it seems like a good one, if you would expound a little more on the details of the project perhaps you could get some other developers involved, perhaps me.

I'm planning on waiting before getting other people involved, because doing so will require a lot more formality, which would slow down progress immensely early on. I figure it would be better to get a solid base done, then release it to the public and grow it.

-bytecoder

sounds good

Be sure to post on here when you are ready to have the community help, I will be the first in line to help out.

If you have the time, please do "bore" us with the details

Well, part of the idea for Plasma is that you'll be able to bring your desktop "to the front" and then send it away again - rather like the Mac Dashboard but seemingly more generic.

AFAICT the core plan is to unify "dashboard-like" functionality with the desktop and the taskbar to get the same functionality but with greater flexibility. In that sense, this part of Plasma is evolutionary rather than revolutionary (not that it's a bad thing).

Well my resolution is "just" 1280x1024 and I can use my desktop just fine, I'd argue that's not so rare anymore. One reason I like GNOME is, that I can keep most windows relatively small, which allows me to work more object oriented (as opposed to application oriented). On Windows, most applications are much more designed to live in a vacuum with huge MDI container-windows and similar monstrums.

I'm personally quite happy with using the desktop as a virtual workspace and just dumping (temporary) files on it, but I'll remain curious what ideas KDE comes up with. The idea to combine the desktop and panel into one entitity is definitely a step forward. The current separation feels artifical and is not extremely useful.

Quote:
" Using some KDE apps for the first time is like trying to interpret ancient Egyptian hieroglyphics: "uh, it's a man pointing in both directions, and next to that there's a magnifying glass with footprints around it, and then a really big down arrow... what does it mean?"


Ohh come on.
Bad programs is not something specific for KDE.

Even extremely good programs with a high level of usability can be difficult to use for the first time.

Ohh come on.
Bad programs is not something specific for KDE.

No, but it's much more common.


Even extremely good programs with a high level of usability can be difficult to use for the first time.

By definition, programs with a good user interface usually have a medium to low learning curve. It's very rare when a well designed program has a high learning curve.

-bytecoder

>By definition, programs with a good user interface usually have a medium to low learning curve. It's very rare when a well designed program has a high learning curve. <

any powerful program aside from a cd burner, web browser, or a word processor generally has a somewhat steep learning curve for an average user. However, these are what most average users use on a daily baisis

If you took an average user and ploped them down at a comp running adobe premiere and photoshop and told them to edit a video or photo, I find it hard to believe that they would just click their way around and magically edit the video/photo to a decent standard.

the same goes for most any software that has any ability to do anything out of the norm.

the applications that the average user uses on a regular baisis are exceedingly easy to use on kde or gnome.

Apps like Softimage|XSI or Alias Maya have legendary well thought out usability and interfaces. But that does _not_ mean you'll be able to get anything done without reading the manuals or some tutorials - because some people don't think good usability is, like, get simple crap done quickly and without learning, while advanced stuff is hidden (like Gnome's gConf) or missing. For those people, and for me, good usability means: once you get the concept, know the structure and shortcuts, you'll get complex stuff done _real_ fast... :-)

amen brother

Apps like Softimage|XSI or Alias Maya have legendary well thought out usability and interfaces. But that does _not_ mean you'll be able to get anything done without reading the manuals or some tutorials - because some people don't think good usability is, like, get simple crap done quickly and without learning, while advanced stuff is hidden (like Gnome's gConf) or missing. For those people, and for me, good usability means: once you get the concept, know the structure and shortcuts, you'll get complex stuff done _real_ fast.

Once you get the concept, know the structure and shortcuts, gconf-editor becomes very easy to use.

You have to draw a line between specialized applications for the professional sector and tools which exist to get a particular job done. Nobody wants to lose valuable time while working on Project X, because one first has to figure out how to get a simple ISO image burned. Even expert users regularly have to accomplish tasks they never had to do before, and then intuitiveness and ease of use are key.

Also, powerful and efficient interfaces have nothing to do with the amount of customizability or the amount of buttons on the toolbar. A good example would be Blender. It doesn't have a whole lot of customizability and basically forces you to learn the obscure way it's meant to be used. But that's not a bad thing, because it's actually a very well thought out and powerful interface.

Generally, I'd say good usability means thinking things through instead of just piling features on top of each other. Everything else depends on what kind of users you are targeting and what you want to enable them to do.

"For those people, and for me, good usability means: once you get the concept, know the structure and shortcuts, you'll get complex stuff done _real_ fast... :-)"

Sounds like the command line.

> > "For those people, and for me, good usability
> > means: once you get the concept, know the structure
> > and shortcuts, you'll get complex stuff done _real_
> > fast... :-)"

> Sounds like the command line.

Why not? A *good* command line is one example. But it's the same game as with GUIs. There are shells or configurations thereof that are more usable than others (working cursor keys, correct terminal settings).

It doesn't mean that CLI > GUI or the other way around. They're different and have different uses and audiences.

Crap, it seems like I didn't construct my post in order to stimulate *worthwhile* discussion...

So more directly, what do you think about Aaron Seigo being the next Miguel de Icaza? Does this mean that a laundry list of Linux vendors missed the boat by putting their weight behind GNOME? Does Plasma seem broadly applicable to both consumer and corporate desktops?

I'll go first:

1) My money's on Miguel in a FOSS celebrity deathmatch. He's small but scrappy.

2) No, the only way you miss the boat in free software is by not getting on the freakin' boat.

3) No, unfortunately eye-candy has a bad reputation for being unproductive.

> This all sounds great, but how much extra ram
> should I buy and what kind of new cpu should I look for?

The Plasma devs keep performance in mind:
See
http://aseigo.blogspot.com/2005/07/moving-web-servers-what-plasma-n...

Not if you know what you're doing. To use Linux you must know very well what you're doing anyway, so that point is moot.

If I understand this correctly, Plasma is just like Dashboard... Useless eye candy...

No, it's a lot more than that. Plasma encompasses the desktop, kicker, notifications and applets like Super Karamba, and a contextual search and linkage system (Tenor). It is an awful lot more than just a Dashboard.

By the way, what's happened to Gnome's Dashboard these days?

Or work more on a true UI replacement to the old files/folders/desktop metaphor???

It's being done, but it's not practical do just move to some other way of working because people work with file and folders today.

But I will get no use out of them just like dashboard or konfabulator.

It's not Dashboard and Konfabulator. It's about much more and about how all these desktop components hang together. That's what real usability is about, not just human interface guidelines.

By the way, what's happened to Gnome's Dashboard these days?

You mean, this one?
http://24.15.37.191/~bytecoder/screenshots/8-11-05.png

Scribus is a Qt app.

Maybe you mean Sodipodi.

Oh boy, your link is 5 years old! And KDE still lives and prospers despites all trolling during last five years.

Oh boy, your link is 5 years old! And KDE still lives and prospers despites all trolling during last five years.

I think that was the point. I can't believe that's five years ago. It could have been written yesterday.

Yes, I think what I said in that interview stands up pretty well - there wasn't really any effect on KDE at all. I'm not sure how much effect there was on Gnome in the long term either.

It's either complete and bugfree or abandoned, since according to Gnome CVS nothing has happend the in the last 15 months to 2 years. Looks like a hype and drop project. A minor part of it lives on in Beagle.

Actually, I was investigating it a while back, thinking much the same thing. The focus of the project has certainly switched to Beagle, though I'm not sure if that's an abandonment of Dashboard, or just a matter of focusing on getting the search component working first.

The other thing KDE needs to think about is to (by default) provide ONE solution for each task an average user might like to accomplish. e.g. one of each of the following

But it already does ...

- Music player (at least as good as itunes - i.e. amaroK. Can also RIP CD's. No need for a separate ripping program)

Useless, no need for that, as kdemultimedia (I think) already provides you with a KIOSlave to do that : when you click on your popping CD icon, you see several directories (WAV, Ogg, MP3, ...) which contains all your tracks encoded and even one file containing all the CD tracks. Just drag and drop, it's accessible from Amarok too !!
No need for a ripping program ! The preferences for the encoding are in KDE preference dialog.

- Media Player (ties into GStreamer backend for video playback)

My wife uses KMPlayer, but there is Kaffeine which is very good too. And both can use gstreamer or Xine, that you can select at runtime.

- Photo Management app a la iPhoto (including basic editing, like Red Eye reduction, etc).

Digikam is already there and so powerful.

- Groupware (kmail, korganiser, kopete, address book)

Kontact !!! Do you actually use kDE ?

- Office (KOffice)
- Text Editor (KWrite)
- Small selection of _quality_ games.

Already there obviously. The last Ksirk is out BTW, and slibo still works in latest KDE (perhaps one day I will dust it off as it seems no longer cared upon by its author).

- CD/DVD burning tool (k3b for more advanced projects, but basic support for burning files/iso's should be in Konqueror - burning audio/mp3 disks should be in amaroK).

All of this have been there for a long time. You should read the docs that come with KDE some time. The icon is right there on the panel.