I'm using Red Hat 8.0.

Can I run an application written for GNOME in KDE? For KDE in GNOME? For either in X? For X in either?

It would help if I knew how to edit the contents of the GNOME menu. Instead of having all of the obscure menu entires (including the "Extras" entry, with many sub-entries), I'd rather have a more traditional menu layout.

I'm a native Windows user, and I'm used to the "Start menu" being stored in a location like C:\Windows\Start or C:\Windows\Profiles\cparker15\Start. Going with what I know, I am unable to find my GNOME menu's contents so that I may rearrange them in an order more logical to me.

Also... I think it would be a lot easier if I could right-click on my GNOME desktop and click a "Properties" entry to edit all settings dealing with the desktop and monitor, such as resolution, wallpaper, display driver, screen saver, etc.

Also, it's aggravating to me that when I click on my GNOME desktop, there's only a "Clean Up By Name" entry. There's no "Line up icons" entry, or an "Arrange by..." list. I want all of my mounted drives displayed in one area, my system utilities (Start Here, cparker's Home Directory, Trash, etc.) in another area, and my shortcuts to all of my programs in another area, and I want them all aligned and neat. I can't do that with GNOME. It is very saddening.

If there isn't already some way I could make these tasks easier to get to such as methods like these, how would I go about creating them for my own personal use, and eventually for everyone's use? How do I program to implement the GUI features of the many different GNU desktop interfaces, such as KDE, X, and GNOME? I'm experienced with Java (Notepad), C++ (MSVC++), Perl (Notepad) under a Windows platform.

I basically want to make my own distribution of GNU that is streamlined and user-friendly like Windows or the Mac OS. I'm having difficulty having to learn all of these text-prompt commands just to get around the operating system. The computer should be working for me. I shouldn't have to work for it. At the moment, it's defeating the purpose of having it. It's taking me a LOT longer to get a task accomplished under my GNU distro than under Windows.

You can run KDE apps in Gnome or the other way around. I know with linux you can configure anything and everything just the way you like it, even if it is just like windows.

Nice. I've found that GNOME is much faster than KDE on my system. KDE takes forever to load, and applications under KDE take even longer.

I know this, also. I'm wanting to create an application/modification to the GUI that would put every configurable aspect of the operating system in an easy, user-friendly "control panel" or some sort of a Wizard. I like Windows because I can either use an easy config tool, or if I want, I can open up the config file itself.

I have an idea that would make the usage of GNU/Linux and Windows in the same environment much easier to accomplish than how it is currently.

What about a start-up group? I've been having outbursts of ideas that would help native Windows users migrate to a GNU/Linux system. What about a GNU/Linux distribution that would seamlessly integrate Windows tasks with GNU/Linux tasks? Sort of like already-existant GNU/Linux systems running wine... but let's face it. Wine isn't stable. And it does have its drawbacks. One major one being it's based on proprietary software components. You need to own a Windows license to use the DLL files...

Not emulation... integration.

To me it's amazing over and over again
how people tend to think of things they
use to do the way they do them as intuitive,
completely forgetting the steep learning curve
they had at the time they started using them,
confusing habit with nature.

Can you recall your transition from WinDOHs
version 3.1 to 95? I had more trouble trying
to find my way around with the "stupid start
button" rather than my Program Manager
and its groups. And I had big troubles with
my switch to OS/2's workplace shell. It took
me a while to find my way around in KDE.
I never managed to get myself to use Gnome ;)
Now I'm using fluxbox, which again is different
from all other graphical shells I've ever used
(hmm, maybe not quite - I remember having used
Usher in Windows 3.1 which was quite similar).

Anyway: at the end of the day, I still think that
OS/2's SOM & workplace shell was the closest
to the way my mind operates, and "to me" the
most intuitive to use...

I can't see a point in trying to make Linux more
like something I can't stand, just in the hope that
I can draw some Lemmings to the light...

Once more my 2 cents...

Cheers,
Tink

That is a great idea, I wouldn't mind seeing a control panel to configure everything. SuSE, is better then red hat, it is a great easy to use distro. Hey why don't you guys sign up for my mail service at http://linuxfreedom.zzn.com/ I have a linuxforum to the link is in my sig.

[rant]

No, I can't. I started using Windows when Windows 95 came out. I think I touched Windows 3.1 a couple of times prior to that, but I, like a majority of the general public, hadn't even touched a computer until 95. There are quite a few people I know who are just now starting to use computers with XP.

Isn't there some sort of a standard for information interchange? I'm not talking about ASCII. I'm talking about filetypes and filesystems standards, along with how programs communicate with the operating system/eachother.

I noticed how Windows filename extensions are pretty good with being uniform. I can always tell what a file is just by looking at the text after the dot. However, with a GNU or any other *nix system, I (personally) can't. A file without an extension could be a number of different kinds of files, namely text files and executables. There are also filetypes I've never seen in a Windows system that I've been coming across numerous times in a GNU, *nix system.

I want to be able to make it so all operating systems seamlessly communicate with each other. I think a GNU system should be able to talk to a Windows system, Solaris system, Mac OS 9 system, Mac OS X system, and BSD system simultaneously without all this confusing and system-hogging stuff going on in the middle. You know? Native support for everything. Wine is an add-on to a non-Windows system. What about native support for Windows apps, Mac apps, OS/2 apps, other *nix apps? Build it right into the operating system, and make an easy-to-follow set of steps that will allow the user to modify the GUI to their liking.

What about how Windows stores system information? The System registry? Is there anything remotely like this in any other operating system, namely a GNU system? What if I want to run a Windows-based, closed-source game? I can't make my own port.. and most Windows games look into the registry for storing and reading information. Again, I could run this game with Wine, but wine is awfully cumbersome when it comes to file management, in my opinion.

"Quit whining or do something about it," someone might say. Well, that's just what I plan to do. I'm tired of having to learn a new set of rules for every different operating system I use. It was a minor discrepancy once upon a time... but with each "genre" of operating systems progressing on their own timeline, the more and more different (read: incompatible) the operating systems will become.

I think it's great that Apple has seen this and has implemented XFree86 into their new Mac OS. That is what I have in mind for ALL operating systems. A Mac can run classic Mac-only programs. Since Mac OS X is *nix based, does that mean Mac OS X can run a program written for my Red Hat system? I applaud Apple for their efforts. Now let's everyone else follow suit to help create a inter-compatible society, while maintaining our different tastes by using different systems.

[/rant]

Hmmm ... doesn't change my point, though. May
I ask how old you are, just so I know who/what
I am dealing with here?

As for me: I started programming in about 82, have
seen systems & architectures come and go, and
have done IT for a living for the longest part of my life.

I think you refer to POSIX. And since you mention
it let me tell you that Linux is closer to that than MS
will ever get. And then there was/is a standard called
CUA (common user access) that MS chose to ignore
as well...

Unix/Linux doesn't distinguish between binary/text.
A file-system flag makes the distinction.

Well, can't blame Linux/Unix for your lack of experience.
"I want a car that works like my pushbike!" ;)

You *are* funny mate... again, MS decided _not_ to
follow industry standards (corba!) but implemented
their own version of operating system independent
(which in MS terms mean that you had a choice
between a DOS based and a NT based windows)
and you rave along here asking others to try to re-
invent the MS wheel?

Another fine example of MS stupidty ... store all your
applications information in a single file and find them
all go down the drain together if the box crashes...
I've seen to many re-installations of MS machines
happen...
I hope no one will EVER attempt to replace the congenial
concept of heaps of individual ASCII configuration files
with a monolithic chunk of hard to maintain data.

I can maintain my entire system with ONE boot-floppy,
not relying on any tools but a small text editor. I can
maintain my machine over a 9.6Kbp/s connection. I don't
think MS will ever be that friendly, and I am too old
for an admin to like jogging ;)

Cheers,
Tink

I'll just say you started programming around the time I was born.

It's not just my lack of experience that brings me these ideas. It's the general public's lack of experience. I don't want to just help myself. I'd like to help others, as well.


Sure. If "POSIX" is a standard for information exchange (six = standard for information exchange?), then that's what I refer to.

Maybe it should. To me, it's more logical, growing up in a Windows world (like so many others).

No, I can't, but I CAN help design a car that works comparitively similar to a pushbike.

Do you think I'm "picking Microsoft's side" or something? I'm not raving, asking others to try to re-invent the MS wheel. I'm voicing my concern about just how incompatible different computer systems are becoming. I think they should grow together, not apart. Hell, I'm not even asking anyone else to do it. I'm simply laying my ideas out on the table for others to see. I'm going to make an effort to make these ideas a reality. If anyone else feels my ideas are worth their time, I'd sure welcome them to help me redefine modern computer standards.

I think the "system registry" is the stupidest thing ever. I can't wait until I write something that will replace it. I'm determined. It'll happen. Someday.

Instead, I plan to replace the monolithic chunk with individual files...

I think it's admirable that you know your systems so well. However, the majority of people don't know a fraction of what you do. I want to give those people a chance to be able to do the same kinds of things.

If you throw a non-swimmer into a really deep body of water, they'll drown. If you gradually immerse them in water, they may or may not learn to swim. The method with better likelihood of success? You tell me.

That's like defining the number pi to be 3
because it's easier to remember ... it will
be easy to use in geometry, but it all will
be wrong :}

The point is that there are modern standards that
the system/systems vendor you find good is ignoring.
And as heroic as your approach is, you're going
the wrong way ... the standards are good, the fact that
you don't know them doesn't obliterate them. It doesn't
mean you have to start a crusade against the standards
(alongside MS, even though you don't seem to realize
that) but rather to evaluate your own background. I guess
it comes down to the definition of the word "normal".
I have an rather idealistic approach towards that. To me
normal is not what statistics define, but about ethics/
aesthetics/truth and "Gods plan". In a society where
90% of the population have cancer I'd still keep saying
that having cancer isn't normal for a human being.


You got your example wrong, though...

You're not trying to teach them to swim but using
tons of sand to make the deep water shallow so
they can sit there having wet asses rather than
learning to swim...

We here at LQ (or least a lot of us :}) are trying to
help people learn to swim... some of us will hold
the fledglings waist, others will have him on their
backs, and if the fledgling kicks, screams and starts
insulting the helper or complaining about being in
the water at all some will think about letting them
drown ;) ... none is putting the water aside, though.

Cheers,
Tink

I love these conversation... Hope nobody minds if I jump in...

I believe there's validity in both arguments. However, I would tend toward Tinkster's side simply because I too, am experienced with programming, hardware, and the like.

First, having different operating systems work together seamlessly is a noble goal, but it will never happen. The main reason for that is the motivation for profit. Microsoft, Solaris, or any other proprietary operating system developer will never support a universal-universal standard of operation. Why? What then would they use to distinguish their product? What would make me, as a consumer, decide their operating system is superior to another's? It would commoditize the operating system market, and they would be out of business in a heartbeat. You may say, "well, screw 'em! Let's reverse engineer this puppy". The company won't stand for that. While they may not have the legal ability to stop coders from reverse engineering their product, they can definitely realease a new version that behaves differently. So the developers are CONSTANTLY re-developing code that worked for a previous version. I would guarantee you Microsoft keeps an eye on some of the software projects s they're being developed. In cases where they feel revenue is threatened, they will make the change. The only exception is if the technology is so well rooted that changing it would disrupt an already mammoth-sized implementation of the technology. For instance, SMB support you find in Samba. I'm sure Microsoft would love to change that protocol, but they can't because companies all over the world would be up in arms that upgrading all their file servers to support the new protocol would bring their IT infrastructure to its knees.

As for making Linux more windows-esque, yes it would probably get more people to try Linux. There is a real-world experiment going on doing just that: Lindows. However, you're stuck with a problem. Why go to Linux if it's just like Windows? What's the motivation for Joe Blow to format his hard drive, move his data, and start booting Linux if all he gets is the same interface? There has to be some feature to attract the new user. Some say the fact you can get Linux free is the motivator. For me, and I would imagine others, it's the security of knowing exactly what you're computer is doing. The average user may not know how to read code, but chances are, someone who does has looked over just about every program available. There won't be any spyware programs running without my knowledge. There won't be any backdoors in server software I run. I will know EXACTLY how my system operates if I have the motivation to research it. On top of that, I can either modify its operation myself, or pay someone I trust to do it for me. To obtain this goal, you have to make some trades. For Linux to be different from Windows (to give you those features), by definition, there MUST be some changes. I'm not sure I made that point very clear, but I'll leave it be for the moment.

Standards are good, but they also have the habit of becoming stifiling if taken too far. There is a fine line to tread in creating standards that ensure interoperability, but without limiting the innovation of the implementers. If you are too strict, then there's only one way of doing things. If it's too loose, it's ignored or opens up huge security loopholes. Not only is an individually focused standard subject to this idea, but a collection of standards must tread the same line. If you begin forcing operating systems to adhere to this filesystem standard, this library use standard, this naming convention standard, this networking standard, and a horde of other standard under the banner of a "universal interoperability" standard, you WILL succeed. The problem is, there will be only one operating system.

You are allowed a choice only when there is diversity. Diversity also leads to innovation because you look at something differently.

Personally, I found it frustrating using Linux to start, but I believe there's a saying that goes "nothing worth doing is ever easy". In this case, I believe it's true. You may try to learn Linux and end up hating. I guarantee you though, that you'll look at computers a bit differently. For the most part, Microsoft has given you a tool and told you how to use it. Using the tool that way may very well get the job done for you. When you learn to use the tool another way, you begin to realize the shortcomings of both techniques. There are good things and bad things about Windows. There are good things and bad things about Linux. Putting forth the effort to learn to use Linux is the only way you can truly decide which operating system's strengths and weaknesses are best suited for your use.

If you look at KDE 3.1 you'll notice that a lot of what you are asking for is already there.

There is a way you could add whatever functionality you desire to any of the desktop environments. Head over to http://www.kde.org They are ALWAYS looking for more developers. Most of KDE is done using Trolltech's QT which is pretty simple to learn if you've programmed any Java or C++. This is another reason many people prefer the Linux environment. If one wanted to add or remove a feature, the source code is freely available to examine and modify. Try doing that in MS world.

I know, just from my perspective, that I really like the text-based commands. They may seem clumsy at first, but once you learn them, they are extremely powerful. And as for something as to a registry.. I agree with what was posted above. I hope Linux never goes in the direction. Once you spend more time learning the environment you'll see that config files are much simpler to deal with.

More than welcome (as long as you side with me ;})

Amen :)

Or, because I like the use of allegories so much:
why try to make a beautiful Indian curry look
and taste like porridge, or have everyone eat
ham & eggs instead of trying their sushi? ;) It's
certainly different, but not necessarily bad...


Cheers,
Tink

*rotfl* That's classic.

I agree with Dark_Helmet - loving the thread. And also agree that even trying Linux - or even DOS, if you're a pure Windows user - using any other kind of operating system will make your computer look different. It's good to have and try different OSes.

As far as writing something to replace the registry, btw, there's nothing you can really do - it's either a monolithic binary database or a zillion text files. The one thing I could see doing is making a registry-like interface to the text files, I guess. I personally had a hard time unraveling how the machine started up and what files led to what files - and still know next to nothing. Something as simple as a tree-view interface to load text files into an editor (or a gui system of buttons to make changes to write to the file through the interface if you must) and online help that took a holistic approach to all these files rather than looking up this file in this man and that file in that man... well, something like that would probably be a good thing.

And I have to confess, I don't see the harm in file-type extensions, myself. Linux does use them, so it's not an absolute - and if you're going to use them, you might as well use them thoroughly and systematically. MS doesn't even do this - arbitrarily naming some dll's .exe and some executables .dll and so on, but you generally do have a good idea about what something is on a finer grain than just executable-or-not.

I'm sure I'm completely ignorant here but how do you tell a shell script - which can be edited with a text editor - from a binary, just from a file list, when both are 'executable'? Whereas .bat and .com/.exe is quite obvious - given that you know what .bat and .exe mean. Which is what Tinkster was getting at regarding things being 'obvious' in retrospect based on the system you learned more than being absolutely 'obvious'.

You don't, unless the distro/unix-system you use
follows a few "basic concepts of tidiness" :}

If unsure, do a file <xxx> ;}

As a rule of thumb you'll find elf-binaries
in
/bin
/usr/bin
/usr/local/bin
/usr/X11/bin
and most scripts will reside somewhere under
/etc
(which I personally find a bit odd since
this mixes the location of configuration files
and scripts that evaluate them)

However, cparker15 having talked about
extensions under Linux I was more thinking
of handles to filetypes like *.cc. *.py, *.sxw,
rather than *.doc, *.xls and the like :}



I don't know digiot, that gives the less than
half-experienced user (and by MS standards this
is what they want people to be) a tool at hand to
ruin more than (s)he bargained for... (s)he can change
things (s)he'd probably never even have considered
just because there's a nice point-and-click interface.

As for linux, there's the Linuxconf-project that is trying
to do something similar ;)

Cheers,
Tink