Posted on 10/01/2002 1:05:54 PM PDT by ShadowAce
Probably the only reason I'm not running Lycoris is my familiarity with the Red Hat Way; Debian was just a bit too alien. I miss Lycoris' easy integration into the Windows net at home though...
On my last business trip I was thoroughly annoyed that on my Win laptop I couldn't distribute the various apps and docs I was using across multiple desktops, the way I habitually do with KDE. Every time I had to look at something else I had to rearrange the lone Win desktop. Grrr...
In the "Five Things We Hate About Linux" section, I would have called it "Six Things We Hate About Linux", and I would have added:
Lacks a common "look and feel".
One of the advantages of Windows and Apple systems is that all of their GUI programs are operating from the same set of widgets (For example, the Win32-API on Windows). The menus, buttons, scroll-bars, etc. all look and operate the same for all programs; and therefore the learning curve for operating the various programs is not as steep.
In Linux, there is an abundance of different graphical widget sets. There is Motif, Lestif, Xt, Tcl/Tk, Athena, and many more that I just can't remember. Switching between programs which use different widget sets make the user experience more difficult and confusing.
On the other hand, I don't know what the solution to this problem is. The linux community is anything but united; and solving this widget problem would require a united decision.
There will never be a "unified widget set" for Linux because the GUI runs in userland; there's no way to enforce unification. This is a consequence of the architecture and there's nothing to be done about it. Maybe some day Gtk and Qt will be able to look the same and even use the same themes, but I doubt the two camps will ever settle for identical feels.
If you want unification for the sake of consitency to make things simple for users then by all means, unify away, they're your boxes. Give your users all Gtk or all Qt apps if you prefer and they need never know there are alternatives.
Bite me. You didn't look very hard. Gnus blows everything away.
There never will be. We like it this way. Your solution would be to pick the one you like the most and throw everything else away. Choice is a Good Thing because one size definitely does not fit all.
I wasn't talking about "enforcement". But, obviously, there would need to be some agreement among developers.
There will never be a "unified widget set" for Linux because the GUI runs in userland...
I'm not sure what your point is here. If your point is to contrast Linux GUI's with Windows GUI's, I would point out that there is nothing inherent in Windows which prevents the development and implementation of alternative widget sets. Borland's OWL is an example.
I content that the domination of a single widget set in Windows is due to one reason only - Windows application developers made the decision to have their programs "look and feel" as similar to other Windows programs as possible... thereby increasing marketability. I should also credit Petzold's excellent Windows programming books for adding momentum to the Win32 widgets.
If you want unification for the sake of consitency to make things simple for users then by all means, unify away, they're your boxes. Give your users all Gtk or all Qt apps if you prefer and they need never know there are alternatives.
Sounds like you don't think it's a problem, since your solution is to dump it back into my lap. :-)
As early as 1994 (and perhaps earlier), the very-professional-looking Motif widget set was available for Linux. However, it was not free. (The CD cost around $150.) The crappy-looking Athena widgets, however, were free. As is so often the case with OS developers, many developed programs back then using the Athena widgets (and other free widgets) simply out of "principle". (I would say "cheapness".) Now, because of this lack of early agreement, the various applications have a variance which goes beyond "interesting" and more toward "annoying". At least now, there are several good-looking (and free) widget sets available.
I have always considered GUI programming to be the most selfless task that a programmer can undertake... with complete focus upon the users' experience. But if Linux is to expand into the "non-geek" domain, we Linux programmers are going to have to do some selfless soul-searching and decision-making regarding our GUI's. We need a more unified "look and feel".
"We"? Who are you? Linus Torvalds? :-)
I wrote...
On the other hand, I don't know what the solution to this problem is. The linux community is anything but united; and solving this widget problem would require a united decision.
Your response...
Your solution would be to pick the one you like the most and throw everything else away.
C'mon now, altair, in retrospect don't you think you've "slightly" distorted my argument?
Choice is a Good Thing because one size definitely does not fit all.
Don't let the java guys hear you say that. :-) Seriously, though... Are you saying that a common "look and feel" is beyond the capability of the open source community? If you are, I think you underestimate its capability.
I would contend that the Win32 API and widget set has done a pretty good job of "one size fits all" in the Windows operating environment. I can't think of any program which you couldn't write using Win32, and as a result it has set the standard for user interfaces. After all, even KDE offers a Win32 "look and feel" for their desktop environment.
I'm an Open Source developer (for 15 years) and I've worked for a Linux distributor.
C'mon now, altair, in retrospect don't you think you've "slightly" distorted my argument?
Not really. I'm typing this in a Konqueror window in a KDE desktop. I have an XEmacs window open to read email. XEmacs has a distinctly different look and feel than KDE/Qt and has not been ported to use Qt (though it can use Gtk). I don't have a problem with that.
You wrote:The linux community is anything but united; and solving this widget problem would require a united decision.
You are correct. We are anything but united. Some people weren't happy about the pace of Emacs 19 development and XEmacs was born. Some people weren't happy about the Qt license and Gnome was born. There's room for both Emacs and XEmacs. There's room for both Gnome and KDE. We can't even decide on a common language. Qt is written in C++, GTK is written in C. I don't see this as a problem. We also agree to disagree on licenses. Some insist on only Open Source (preferably GPL), others don't. I'm not a purist like some. I have fancy licensed Japanese fonts installed, I have a bitkeeper window open browsing kernel source and an Adobe acroread open to a manual.
Are you saying that a common "look and feel" is beyond the capability of the open source community? If you are, I think you underestimate its capability.
Yup. We can't even decide upon a common set of editing keys. I can use both Emacs and vi style keybindings for editing, though I prefer Emacs style. Microsoft Windows guys use a system of keybindings that drives me crazy. So, yes, it just isn't possible.
A war story: once upon a time the default background color for XEmacs was eye-searing white. Everyone hated it, so I decided to change it to the same color gray as Netscape. There was weeping and gnashing of teeth. Although some people liked it, I was flamed to a crisp. It was finally decided to use a lighter shade of gray. I hate the lighter shade of gray, but that's what a .emacs is for. We cannot agree about colors, we cannot agree about keybindings, and we certainly aren't going to be able to agree on The One True Widget Set.
I can't think of any program which you couldn't write using Win32, and as a result it has set the standard for user interfaces.
Possibly true. But the problem is that Win32 stinks as a user interface.
After all, even KDE offers a Win32 "look and feel" for their desktop environment.
I guess that's why the defaults all stink. Fortunately, KDE can be customized to something slightly saner.
Look. I said that I didn't know what the solution to the problem was... and so you just "created a bogus solution" and made it mine. That's distortion. Really.
I'm an Open Source developer (for 15 years) and I've worked for a Linux distributor.
Yeah. I was pretty sure you weren't Linus. :-)
But the problem is that Win32 stinks as a user interface.
First time I've ever heard anyone make that claim, except for the Mac fanatics. I suppose that if the Win32 user interface "stunk", then nobody would be asking the theme question of this thread's article: "Can a frustrated Windows user dump Microsoft?" Heck, everybody's grandma and grandpa is using a computer now because the Win32 interface is very easy to use.
We like it this way [lacking a common "look and feel"].
I don't doubt that you like it this way. But not all in the Linux community like it this way. Furthermore, the other 99% of computer users who are on Windows or Mac have come to expect a certain consistency across their applications. The fact that Linux is different from Windows is not the problem. People switch from Windows to Mac and vice-versa every day with minimal problems. In either environment, after one learns how to work that first program.... learning others is rather easy. Not so with Linux.
Again, I contend that the lack of a common "look and feel" is one of the biggest impediments to mass migration from Microsoft to Linux. Like I said earlier, I'd put it in my "Top 6".
I dislike the Mac user interface too.
I suppose that if the Win32 user interface "stunk", then nobody would be asking the theme question of this thread's article: "Can a frustrated Windows user dump Microsoft?"
Most people don't know any better. Microsoft Windows suffers from what I call "you are trapped in a twisty maze of GUI windows, all alike syndrom". Also, like the Mac interface, it insists on using a broken mouse. Xerox Parc did years of research before deciding upon a three button mouse.
In either environment, after one learns how to work that first program.... learning others is rather easy. Not so with Linux.
You might have a point there with some people, but not all. Things in the Un*x world are certainly much different than they were 20 years ago. The first time I ever used Emacs I couldn't figure out to quit, so I finally had to pull the modem jack and relog in. It's much easier now. I've run in a Japanese locale and KDE for several years now. I am not a good reader of Japanese, but I haven't had too much trouble going about my business. I'll admit that I cannot install a Linux system in Japanese. I tried that once and it was a disaster.
Again, I contend that the lack of a common "look and feel" is one of the biggest impediments to mass migration from Microsoft to Linux.
O.K. I'll grant your point. I can't speak for anyone else, but I do development for me. I want an operating system that no one can ever take away from me and I can fix problems myself if I have to. It does not matter to me whether there is a mass migration to Linux or not. I don't care.
Personally, I'd say it's a very courageous thing to bet against a community with that sort of track record.
For anyone who concludes that since Linux doesn't have the desktop today it never will, I can only shake my head and wonder.
Computing has only been a 'mass market' thing for what, at most ~5 years? It's a very immature market. Anyone who thinks the way it is now bears any resemblance to the way it will look as the market starts to mature sure doesn't have much of a sense of history, I think.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.