Linux has a unique opportunity to take advantage of its open source nature by offerings its users greater integration than even proprietary software can due to the fact that developers have access to both the code from the operating system as well as a large assortment of bundled applications. The following osViews editorial contribution is the fifth piece to a multipart series, (  |  |  |  ) which outlines the framework for a consistent, full-featured Linux desktop platform that would help grow the operating system’s desktop install-base among new users and Windows converts.
GNU/Linux Home Desktop Kit PC Project Part 5: User Defaults
Submitted by Kelly McNeill 2004-01-28 Linux 5 Comments
After all it’a a UI article!
The author makes a great point though…much of what is difficult on the Linux desktop isn’t system related at all, but attention to detail of the little things. He has a point. It’s not like there’s a super corp pushing the standard fare desktop down your throats. Considering Linux is Open Source there’s no reason many of his ideas couldn’t be implemented tommorrow…especially in a liveCD distro like Knoppix or Mepis.
Of course windows has many of the same UI flaws…OEMs don’t really have much option to tweak past the defaults….Linux distros DO! That’s a big advantage…to have all the apps work perfectly. Everyone praises Linux for being just the way they want it, but nobody actually SHIPS a verson that way. …Why not? More than that why do we feel the need to fry those who simplify things and at least make an attempt?
… you know the drill
All this backs up the reason why there are “so many” (desktop) distributions. If you customize a (desktop) distribution or make a new one, why not share it? If you want to make profit, why not roll your own and polish it, giving support to your customers, and make profit? It’s all possible, but you can expect others to do so to. Talking about competition! Also, one can already notice this oppurtunity in practice. Take for an an obvious example SJD.
“Current GNU/Linux distributions fall far short of the integration and polish and there is no reason they should. Open source software allows for much greater flexibility in this regard, but this advantages is rarely utilized by the various distro’s”
That’s exactly what RedHat tried to do, right?
The problem i foresee with these projects is maintenance. This costs time, energy and dedication of multiple individuals who are gonna work together just like it is the case with other distributions, (Free) software projects, *BSD’s, etc. Since GNU/Linux and the DE’s it runs on are constantly in development in features as well as security and reliabity fixes it has to be either maintained or it dies. When forks are issued instead of patches by upstream it all sums up it takes even more time, dedication and energy. Also, isn’t this what commercial (desktop) distributions like RedHat, Mandrake, SuSE as well as hobbyist/individual/community (desktop) distributions like Knoppix -to name a few- have tried for years? That makes it sound easier than it really is. Why would this succeed where others have not? Why not adjust the DE’s, art, programs itself by contributing art & code instead of creating a new distribution? If done, the _current_ distributions benefit directly, taken they want these features.
For those who are interested in this project, hosting ISO’s is NP on say sunsite.dk or sourceforge.net. Recently i helped a Free software for Windows maintainer obtaining an account on sunsite.dk; it was as hard as sending 2 mails.
“Also missing is a resumable download manager (like Download Accelerator Plus) for users with dialup ISP connections. This may already exist in some form but not to my current knowledge.”
Downloader for X? Adding this functionality in ie. Konqueror?
These “download accelerators” which put up multiple connections to one HTTPD are horrible anyway, it is highly server unfriendly as well as unfriendly for the rest on the user’s ISP environment, and not always worth the hazzle anyway. Not even for dialup users, mostly only for cable users.
Is that users will be quick to jump onto someone’s case if they ship with defaults other than those the writer intended. Redhat does this and has recieved much flak for doing so. Ximian also does this. They have the doorman program which can always reset settings to the original Ximian settings.
Usually, it is a matter of bundling a good configuration file with the app. GNOME achieves much of the authors objectives by having simple apps with simple configuration. Windows also does this in much the same way too. Their picture viewer does not do too much and no one complains. Anyone who wants a better picture viewer goes and downlads their own, like ACDSEE for example.
The moral of the story is that distributors should only include simple software in their packages. At least, tehy should not install emacs by default. kedit, gedit and so on are usually more than adequate. Whilst maybe everyone changes something to suit them, most people will work with 90% of the apps their distributors shipped with the release. Having these in good places makes much sense.
“The reason for using these settings is that elderly and vision impaired individuals can’t see the smaller default fonts”
…he writes, in an illegibly tiny font.
Also, I find that Debian apps come with very decent default settings for their apps. The bulk of them work straight out of the box, and are immediately useful (i.e. without reconfiguring first). When configuration needs manual intervention, the installer asks the user to supply the required values.