EARLIER ITEM

FREE SOFTWARE MENU

LATER ITEM

THE PROBLEMS FOR A DESKTOP GNU/LINUX

Brian Grainger

brian@grainger1.freeserve.co.uk

During the Easter break I spent some time reading, from cover to cover, my magazine guide to GNU/Linux, published by Live publishing. It covered installation of the cover CDs (Mandrake 9) and Demo Linux along with tutorials which included the following:

• Each of the Applications of Open Office
• Using a scanner
• Burning CD ROMs
• Compilation of application source codes

I think you can see it covers all the areas a new GNU/Linux user is likely to want covered. As such it was a very good read and I would recommend it if it was still available in the shops! There is now a slightly bigger magazine in the shops which adds some more articles to those I read, only had one CD attached instead of 3 and charged almost twice the price. Not such a good bargain.

Having completed the read, it is also obvious that the GNU/Linux evangelists, of this magazine at least, are somewhat contradictory in their opinions and that GNU/Linux has a fundamental problem if it is to succeed on the desktop.

Two points come out from reading this magazine. First and foremost is that everyone takes every opportunity to criticise Microsoft. This is a common trait with anyone pushing GNU/Linux. I agree it is a good reason for looking at GNU/Linux in the first place but let us give credit to Microsoft where it is due. This magazine has extensive articles on using the OpenOffice.org (OOO) suit of applications and every one praises its functionality and look. They should realise that all this development, especially of the look, has come from Microsoft. The reason OOO is so good is that it is virtually like Microsoft Office, so you can start using it without much effort. I had considerable trouble getting started with KOffice because it is not like Microsoft Office. Now, I am prepared to spend a bit of time relearning how to do things, but most computer users will not. If they cannot use it quickly they will not use it! Microsoft is often criticised for plagiarism and that they have never really developed anything themselves. I think this is unfair. Most software developers make use of other peoples ideas. It is a fundamental point of the Free Software Movement. What Microsoft do is build on those ideas in two ways. They introduce features people want, so the product get better functionally. Give them credit for this, even though they can be criticised for never knowing when to stop!

The second thing Microsoft does with their products relates to the other contradiction in the magazine. The writers introduce a topic, compilation of software for installation say, by saying people have lots of problems. Then they say the article will demystify the process. They are wrong. After a few paragraphs of technobabble the question on my lips is why do I have to do this, let alone understand it! It puts ME off, let alone anyone I might try to convince to use GNU/Linux. Microsoft have solved this problem with all their products. Ease of use is uppermost in their minds when they develop products. Again they have gone too far, for this nerd at least, but give them credit for doing what they have done. Without them the average man in the street would not be using computers. They have widened accessibility. Unfortunately, the GNU/Linux community do not appear to want to recognise this problem. To be told to read the Man pages or that all you want to know is on a web site is a poor means of helping people who are complete novices to GNU/Linux. I guess ICPUG came about to solve this problem for Commodore computer users. The equivalent for GNU/Linux does not exist. There is a lot of help out there but it is NOT addressing the newcomer. It only addresses members of the GNU/Linux clique.

Why are things so complicated?

Let us be honest. There is one aspect of GNU/Linux that will inevitably be complicated to understand. It is a secure multi-user operating system! I have found, throughout my life, that increased security means increased difficulty. That is why it costs so much more to develop something for the MOD than it does in industry. Windows tries to hide its security model, (yes it has got one), from the home user. Consequently, home users can use it fairly easily. GNU/Linux cannot do this. You HAVE to log on. You will NOT necessarily be the administrator /super user by default. You must understand the concept of file permissions, having read only access, etc. It's tough but its good for you. However, it is a drawback to getting new users to GNU/Linux. Nothing can be done about it except to educate but, as there are those who still open e-mail attachments from people they have never heard of, it is going to be tough!

The one complication I cannot tolerate is that touched on above, lack of ease of use. To succeed on the desktop in the wide community, not just in industry, GNU/Linux must become easier to use and it must start with the installation process. I would like to think GNU/Linux machines will appear with GNU/Linux preinstalled. This will remove the problem initially, as it does for Windows. However, upgrades still have to be installed and applications must be installed as well. Somehow, there has to be a one click installation process.

• If it needs uncompressing it must be done without the user knowing about it.
• If it need compiling it must be done without the user knowing about it.
• If the files need putting in a special place it must be done without the user knowing about it.

It should be easier than Windows, not harder, since it hasn't got that mysterious Registry to worry about.

I have a feeling part of the problem is the multitude of distributions available. Perhaps one has to compile things before installing to make it link to the appropriate distribution. I am guessing. I do not really know. If that is the case then there is only one solution. Like the video recording industry before it, which was seriously hampered until VHS won out, GNU/Linux on the desktop needs one standard!

Unix itself suffered from this. Although at times one version or another threatened to dominate, it did not happen. Mysteriously, GNU/Linux has risen to take the mantle and with good luck should become the One Unix the industry has wanted.

I propose that for GNU/Linux to succeed on the desktop we need a standard desktop.

First of all decide which GNU/Linux distro is to be the One:
Debian, Mandrake, SUSE, Red Hat., ...

Decide which widget library, (from which windowing systems are made), is to be the One:
TK, Kylix, Gtk, Qt

Decide which windowing system is to be the One:
Gnome, KDE, ...

These options are not independent but you get my drift. Once we have decided on the One, the whole community can beaver away making one product better rather than making a multitude of products good, not one of which will be used in great numbers.

When all this has been decided everyone will have a standardised set of libraries and applications can be compiled to the standard rather than a multitude of different options. The easy installation routine should follow.

So, how will we decide the One Desktop GNU/Linux to Rule Them All.

Sadly, the GNU/Linux community is so full of nerds that will not reach out and understand the real world that the answer is not likely to come from there. We need you for the great software though. (Hopefully, that will kill the hate mail!)

Maybe something will turn up that will become a de facto standard. I guess this is the most likely answer although I don't like it much as it may involve a commercial enterprise. Replacing Microsoft with another company is not the answer. It just perpetuates the problem.

Perhaps another Linus Torvalds will come from nowhere and make the decision. I do hope so but such individuals are rare.

Perhaps Richard Stallman and Linus Torvalds should get together and make the decision and organise the resulting community. Both highly respected people with no commercial axe to grind. Well, I can dream can't I?

When all these problems are resolved there will still remain one, maybe the greatest, inertia. People will be reluctant to move if they cannot use their favourite program. Just like OpenOffice.org is a powerful persuader to move people from Microsoft Office we need to be ready to have a list of alternatives for all the other key applications run under Windows that people will be able to use quickly, before they have time to complain.

In this article I seem to be describing some high hurdles for GNU/Linux to overcome before it becomes mainstream. Yet, I have introduced this area of the web site and I am optimistic for the future. Why? Well, I think things are moving in the GNU/Linux direction now and Microsoft have been making too many mistakes of late. Also, I have seen Knoppix! That will be the subject of another major feature in this area.

TOP