Lately, we’ve all read a lot of articles about desktop Linux – so many that it’s getting hard to tell them apart. One says “Why Linux Sucks,” the next “My Success With Linux.” Even Michael Robertson of Lindows.com joined the fun with his “Why Desktop Linux Sucks, Today.” But very few people have proposed anything radical, and I believe that’s what’s needed to take GNU/Linux to the next level.
The polish that Red Hat and Mandrake have added to their consumer-grade Linux offerings have meant a lot to users like me, those who use it as their main desktop OS. Finally, these distributions are starting to look like a professional quality operating system. One of the main problems, the way I see it, is what I call “marketspace crowding.” There are so many distributions, so many flavors that are, when it really boils down to it, just semi-clones of one another with one good tool or one little tweak added or modified. The idea of roll-your-own distribution has left the market saturated with many similar distributions, and even those “in the know” don’t know how to properly choose a distribution. The solution should be to make your distribution stand out, and in order to do that, you can take advantage of the opportunity to do some things right.
So, if I had the capital and connections to make a new distribution, I’d take full advantage of the opportunity. The market doesn’t need another carbon copy of the current Linux market leaders. It needs something new, something fresh, something that really rocks the Linux community’s world. So I’m in charge of my own personal Linux distribution, DistroX, and here’s what I’m telling my staff and engineers.
Let’s start by delving right into it: let’s address some Unix-wide issues, that will make our development unique. The file system is a nightmare for a normal user. This has been covered in exhaustive detail by hundreds of articles, but if I’m going to run my own distribution, I’m going to make a sensible directory structure: /users, /apps, /system, /hardware, /downloads, /logs, /servers, /shared, and more. Then, using symlinks, we’re going to recreate the current basic layout of the standard Linux/BSD filesystem to assist developers in porting applications. For example, our system would probably include the following the symlinks:
/home -> /users
/var -> /logs
/var/www -> /system/websites/
/etc/httpd – > /servers/httpd/
/dev/hdc -> /hardware/cdrom/
/mnt -> /hardware/drives/
This will encourage new users to explore their computer, not fear it, as many do. Apple has this right – Mac OS X’s file system is human readable, and not intimidating at all. [Note: The above symlinks are just used to illustrate the mimicked filesystem, not final decisions. Please don’t comment on the actual links referenced above]
The next idea is fairly controversial, and it comes in two parts: first, our goal is not necessarily to follow LSB guidelines, nor to be compatible with other distributions. Our goal is ease. Dropping legacy support and not being dependant on RPMs will open up the doorway for my engineers to revolutionize Linux software installation. Again, this topic has been discussed in detail in the past, so I won’t get into the specifics, but a drag-and-drop software install/uninstall is clearly the preferred method by desktop users. I barely know hown to uninstall software on Linux past “rpm -e program” and on Windows, my control panel used to be filled with a number of programs whose uninstall failed, and in the process, wiped out that single .ini file that directed the uninstall (to get rid of them, registry tweaking was necessary). The second part of this step is to set up an actively developed software repository on the internet. Users aren’t stupid, they don’t need a hand holding front end, but they do need easy access to programs, and an online facility dedicated solely to my distribution is definitely a plus for me as a potential customer. Incidentally, I don’t know why this type of thing isn’t more popular with today’s large distributions – it would be nice if there was a single site where you could go for install packages specific to your distribution that were guaranteed to work.
In following that, a good amount of resources will be allocated to developing an online community a la Gentoo. Forums would be well maintained and cared for religiously. Help would be granted unconditionally. Contests would be held regularly and other promotional stunts to get and keep people interested. An active community means more developers, more porting, and more word-of-mouth. Conferences, training, certificatons would all be investigated. Anything we can do to make people try our distribution would be considered.
Choice. It’s what Linux and BSD offer that, to some, is the biggest draw. My developers are going to meet and agree on ONE desktop environment. Yes, we’ll include the libs for the other major one we leave out. But my distribution will not, as many large companies have, concentrate all of our resources on making one environment look nice and then include a second that pales in comparison. I’m going to remove a lot of choice from the user, because, to many, it’s more a gamble than a choice. If I customize Gnome to be unique for my distro, I don’t want people to unknowingly choose KDE. Or vice versa. If you’re an expert, and require the another DE, you can download the source and try to compile it yourself, visit our forums, or even assist us in porting it so we may offer a downloadable package customized for our distro. I’ll include one IM program (gaim), one FTP program (gFTP or Kbear), one e-mail client (Evolution), one office suite (OpenOffice.org, but also kedit and kwrite, which serve different functions). I may choose to include two or more browsers – as they are frequently used and that choice is not likely to cause confusion. I won’t list all my application choices, but the point is, I won’t include applicationss for the sake of including applications only. Only those that offer a different and distinct experience or that offer something unique will be considered for inclusion.
Installers have been a hot topic of discussion lately too – mainly because they’re all getting pretty good and now. It’s a race to who can make the one that the crowds feel make the most sense. Anaconda has come a long way and leads the pack, SuSE YAST has received praise, and Mandrake’s redesigned install is clean, fast, and easy. New breed distros like ArcLinux, LindowsOS, and Lycoris have really gotten the rapid install process down pat. I believe the perfect is somewhere in the middle – graphcal, heavy on eye candy, with few visible options but lots of “Advanced” buttons. Full control is necessary for experts, but simple steps for newbies. Some essential tools: NTFS resizing, coexisting with other Linux installs, and a configurable boot manager. I’d really like to see Linux boot on ATA RAID. When this project is more stable, we’ll incorporate that ASAP. Lastly, in the installation category, each step will have easy-to-read documentation on screen, such as ‘Which file system should I choose?’ along with recommendations – “DistroX recommends UFS file system for all desktop computers. This file system provides crash recovery/protection and has been thoroughly tested. If you intend to run your computer as a server, click here for more suggestions.” People should never get scared during an install, but I don’t want to remove all options from the install process. None of the BSD’s use graphical installs yet (though the OpenBSD project does have GOBIE under development). Many complain that installing a BSD with a DE is still one of the toughest tasks out there. Ours will be fully graphical and thoroughly tested. Public betas will solicit feedback until we’re convinced it’s as easy as a Windows install.
On the subject of installs, I’m not going to worry about “bloat.” I’m already stripping out a number of apps, so what I’m not going to worry about are libraries and system files. Even the minimal install will include every common system tool my develops can think of. We don’t want anyone, anytime, to have dependancy problems. We’ll include everything we can think of, and if we come upon and app that needs something not in the default install, we’ll include it in the install files on our application website. Dependacy problems are an issue of the ’90s – there’s no excuse for them today. On my home machine, I’ve come across this scenario: I download program A. In order to compile it, I need lib X, Y, and Z. I download them all. Library Y fails to compile because it requires B and C. I download B and have to upgrade C. But B depends on E. And E seems to be installed. 45 minutes later, I quit. On my distro this will never happen, because our install files will include everything. Hard disks are large and cheap these days. We will simply maintain: a worthwhile desktop system needs disk space. Broadband is increasingly common. Packages should include all files necessary for use.
Another thing I’m about which I’m concerned is after market add ons. Red Hat ships without mp3 support and may not be the last distro to do as such. I’ve maintained that the workaround here is to add a menu entry to the launch bar: “Install MP3 support.” The script that runs fetches the mp3 plugins, installs them, and then, after confirming successful install, removes the entry from the menu. Same with DVD plug-ins, nVidia drivers (if an nvidia card is detected), and other after-market add ons. This will allow further compliance with licenses. Perahps a “Run Once” menu should be present, or even a script that installs all the necessary after market add ons and then removes itself.
My distro would be available for download on throttled public FTP servers and also from some sort of contributor/buyer FTP servers. My contributors and paying customers can also download my apps at high speed. I don’t mind people seeding BitTorrent or anything, I just don’t want to pay for the bandwidth. Last I checked, availability was rarely the reason people did or didn’t try a distribution. On that note, my source code CDs are not available for public download either. That’s bandwidth I don’t care to pay for. Source, or what we release of our source, will be available to those that request it for postage only – I’ll GIVE it to anyone who wants it and is willing to pay for the postage – the CD’s are on me. The cost of CDs is negligible, and the cost of postage will weed out the “getting it ’cause it’s free” crowd. I don’t want to milk the developers, I just don’t want the whole world downloading 2 or 3 extra CDs by mistake. As a company, we want to be smart, we want to share, and we want developers to feel loved, but most importantly, we want to make money to stay afloat. So we need to be smart.
Defaults are very important – one thing I’ve noticed is that there are few system-wide defaults with most Linuxes. For example, click on a link in Evolution, I get Mozilla. Click on a link in gaim, I get Konqueror. Want to change that behavior? I change it in gaim. But Evolution’s default won’t change. And Mozilla always defaults to MozillaMail. There is a way around this. The way I understand it, pretty much everything on a GNU/Linux system runs commands terminal-style. Perhaps the way to get system-wide default is to have a given directory, say, /system/commands, that appears to be the equivalent of /usr/local/bin or /usr/bin – that contains the executables from the command line. Except, in our distro, the real files are kept in /system/bin. /system/commands is full of aliases. Then, when you change your default browser through our control panel – all the known browser commands – konqueror, mozilla, opera, galeon … they all change to aliases of your selected browser. Now, obviously, menu commands have to map to the hypothetical /system/bin, since we don’t want to effectively stop those programs from ever launching – just from launching via standard systems commands. This is all very complex, and would surely be an engineering feat, but until Linux/BSD as a whole is cleaned up and standardized, it’s a possible solution.
If this were reality, and I truly had a venture capitalist’s backing on this project, I should mention that I’d probably have to reconsider Linux as a whole. We’re a company first and foremost, and so our goal is to stay in business and make money. We’d have to achieve this in two ways:
a) first off, we should probably base our distribution on the FreeBSD system. FreeBSD has many drivers, boots much quicker than Linux, is very clean code, and most importantly, has a much more liberal license. The GPL provides little room for making money on software, only support (see many comments by RMS, Bruce Perrins, and Eric Raymond, as well as the history of Cygnus for more on this). In order to develop an OS and provide the right environment for it to grow, we need to be a viable commericial company. In order to do this, we must protect some of our code. The second part of this scenario, strangely, is the exact opposite.
b) we’re going to share some of our code. As much as we will borrow from other coders, we will share some of our improvements with the community. As we all know, companies that are reluctant to share source code are frequently shunned by the knowing community. We will be good open source citizens, but we will protect some of our investment. The decision to use FreeBSD may not be a popular one at first, but it would ultimately be what sets us apart from the pack. Our distribution will be based on years of proven rock solid stability fused with a modern approach to user interaction.
Now, let me be the first to say that much of what I propose here may not be technically, or more likely, financially, feasible. However, I believe that the open source community gets things done by discussing what we’d like to see and eventually, the most popular opinions make their way back to the developers and the commericial companies. Clearly, Linux and FreeBSD on servers is working and popular. Linux on the desktop has come a long way, but even companies with a novelty – Xandros and their integration into Microsoft domains, LindowsOS and their Click-N-Run, etc, still aren’t really making any noticeable dent in the desktop community. Linux can’t be scary, and all we’ve seen so far is a pretty face to the undecipherable mess underneath. If you don’t agree with me, show the filesystem layout to your mother and ask her where an application might be located. Even I couldn’t tell you for sure – /usr/bin? /usr/local/bin? /bin? ~/.appname? /opt? /etc? It’s a total guess. DistroX is going to make sense because it’s going to be built that way from the get go.
I’m of the belief that someone is going to have to do something drastic to see real change. There should be a real, distinct difference between a “packaging company” and an “operating system company”. The first step, in my mind, is to decide you’re not just going to release “YALD,” or “Yet Another Linux Distribtion.” What do you think?
About the Author:
Adam Scheinberg is a regular contributor to OSNews.com. He recently moved to Orlando, FL, where he is looking for a network administration position. He uses Mandrake Linux 9.1 as his main desktop OS.
The author wishes that Linux would boot off of ATA RAID. He went on to choose FreeBSD for it’s liberal license. I think he’d be surprised to find that FreeBSD already boots off of ATA RAID.
That’s exactly the point. The future will not be hirarchical. How exactly this will be done remains to be seen but one thing is fact: Only if we abstract the underlying filesystem as much as possible, we even have the _chance_ to do the convertion.
Making the filesystem easier to read (for english speaking people) will do nothing but make a flaw a bit more convenient (and introduce foreseen and unforeseen complications).
A small first step in GNOME land could be seen with the Epiphany browser’s bookmark handling. This is just so much more powerful than a hirachical tree of bookmarks, even though the current implementation is still very rough. If it wouldn’t crash constantly on my system, I would use it just for this feature. It will be interesting to see how this all works out but I’m quite confident that this will be one of the major advancements in the next years and this is the only thing where I’m afraid that Microsoft could get a huge lead if we don’t hurry now. Will Linuxland be ready for such a massive change? Or can it be done without breaking too much existing stuff? I realise that I yet have to learn a lot about this topic.
Tux: You could really help the Linux community with this.
Help the Linux community by replacing Linux with FreeBSD, great idea! 😉
BTW: It never fails to amaze me how readily people bristle over their “territory”, and fling insults at posters who obviously have good or, at worst, neutral intentions. “Can’t you get it through your thick head”, “You obviously didn’t read the article”–such a drag to step in these dogshit comments, even if there is a gleaming dime of relevant info in them.
Here’s my actual post. I’d like to switch from Windows to Linux. I know nothing about Linux, and know far more about Windows than I ever wanted to (because it is accident-prone). I hate MS, and their obsence “success”, and for that reason alone I want to switch. But, even if I hate to admit it, because the polarity of “hatred” feels good sometimes, ambivalence being a difficult and unsexy state to maintain, Microsoft was a necessary evil for the world of computers, and much progress, in chips, peripherals, software, etc, owes itself to MS. But now it’s time to move on, away from this monster against which all OS standards are measured, and who’d own even the Internet (italics mine) if it could.
I like Adam’s article. Thank you, Adam. It presumes that true progress on a distro will be made on behalf of people like me, who want a great OS but who don’t want to get cozy in the codey innards of our systems. The article starts with the premise that a great distro has to give people what they’d expect from Windows. Of course. Duh. That’s where we live, the critical-mass. The article addresses some of things such users shouldn’t tolerate. I, for example, don’t like the absurd naming of Linux apps, which are themselves an analog of Linux’s lack of code-transparency. And I don’t ever want to see feel hear or smell the word “dependencies”. I have other work to do, so I want less “choice” upfront–just give me the best, up front. If I don’t like it, make it easy for me to get something else, and get rid of what I didn’t like.
I also like the idea an early poster made that such a product might more resemble Mac than Linux or Windows. Fine with me, Mac is better than both of them. But Apple is also a monopoly, and Jobs also is way too rich. It’s time for him and Gates to |just leave|, italics mine.
What MS did for the world is to show us what is needed in a personal computer, both by giving it to us, and by failing to. It didn’t get so stinking big because of its products, but because it had products, and standards, albeit proprietary standards, to offer just when the foolish world was clamoring to start the “computer age”; another case of being in the right place at the right time.
Adam, I don’t think making money is the only model that would produce a great OS (distroX). Once something–personal computers–becomes a commodity, and almost an essential (alas) for modern, that is, consumerist, life, then that commodity has to taken out of the hands of those who would sink their needlelike teeth into you and never let go until you can’t pay them anymore to suck your blood. This also applies to “broadband”, cellphones, and even cable TV–all way overpriced, even at “entry-level”. (I say this as someone who can afford these things) This fever has to break. Since so many earn their livings from computers–doing something other than building OSs–the OS has become an essential commodity, and I for one am tired of owing any “living” I make from my computer to Bill Gates, or any facsimile thereof. But, then again, we’re all working for the landlord.
This may seem off topic, but too many “consumers”, already haggard and hyper from dishing out too much cash for music, games, software, etc, are all-too willing to steal a song for, well, a song. They steal from Michael Jackson, they steal from some poor punk hoping to “earn a living” from his music. Look at the results. Napster, Kazaa, et al, caused such an uproar among the media powers, including MS, that these companies seek to, and will succeed at, destroying the ability of electronic devices to make copies of digital media files. Every electronic machine you buy will have “their” spy-chip in it. In the name of profit. Guess the “honor system” would never have worked with downloading music. Not with this populace.
But it can work in the open source community. I believe the “honor system”–donation, if you want–is fundamentally connected to open source. Build me something good, and I will pay you. You won’t even have to ask.
Thank you, folks, for reading. I’m sorry if I rambled, but there is unavoidable politics involved with Linux, or open source–we have yet to see it in full force–and I wanted to touch on certain aspects of it.
After trying sevral other distros, I really like knoppix. It has it’s drawbacks, but for the most part it just works. If you’re looking for a newbe Linux, then start there and address the issues present.
Knoppix is really just a custom Debian on a bootable CD. You can load it up, tweak it how you want, and burn another. It has about half of your list done for you. True, it uses the standard layout + some stuff to be bootable, but it’s fairly standard–most linux help docs will get you going. It has a standard desktop. You can change it, or not, but it starts with the same thing unless you change it–good first step. Hardware detection is great. Everyother distro I’ve had to go without something basic, mouse, video, sound…The basics work and work well. Yes, it’s a read-only CD, but that means users can’t mess it up and can always go back–best part is without loosing data!
The only thing missing with Knoppix is an easy way to install new programs. He’s necessarily catering to 905 of the people, but everyone seems to have their own 1-2 key programs that breaks their use of it. What Knoppix [and any other Distro] needs is a surefire way to install programs. Compiling a la gentoo is great for the Pros, but the rest of us just want to run them. Frankly program-in-a-folder would seem to be the best route for all involved. Develop an online app that detects your system [I know already done!] but installs the dependancies to the program folder. For something like Knoppix, the current version of the disc is the standard layout and everyone can just adapt to that. each prog would run in it’s own “sandbox” uneffected by the others. Again, I’m not a zelot or anything, Knoppix is just the best place to begin such a project because it’s basicly locked down and install-less.
I totally agree. Especially with that arcane file-system crap that I have to put up with Linux nowadays.
What? You think I’m some sort of illiterate? Got my start way back in the Apple/Commodore days and been everywhere from Mainframe, *NIX, Mac, OS/2, Windows and more.
POSIX? Screw POSIX compatibility. It doesn’t make sense to the user. ONE DISTRO and ALL APPS specifically tailored to that ONE DISTRO.
Let me give an example of how stupid things currently are. RedHat 7.2 installation. How do I allow another computer to print through my computer? Well, just “man lpr” or “man lpd”, right?
WRONG! Because RedHat (not picking on RedHat, every distro does it their way) puts the conf file SOMEWHERE ELSE. And it is not even named like it says in the man page. Cause all they did was pick up the man pages from somewhere and slap it on. Check on-line? You get 5 different answers.
THIS SUCKS. Linux expert or no. Where did POSIX get you?
The article was good, but the writer did not appear to have a deep knowledge of kernel implementation. That’s not a big problem because the article gets you thinking, but if we are going to make a better platform, for it to really happen, than we need developers to want a new platform. We need developers who are knowledgeable enough about all kinds of modern and legacy software to be able to focus on what worked, and what didn’t work. We need a generic model and a open source model that are co-operative.
It’s good to see someone actually acknowledge that not
everything about linux today is perfect. (Many people
I talk to claims so…)
The file system is a living hell for a new user. I have
seen a lot of people not finding their way around it
(including myself). It’s not always obvious that dev
stands for device. (DEVelopment strikes me as a first guess)
And in my personal experience, I have to navigate through
a lot of it to make things work, even though, it seems to
me now, that it shouldn’t be neccessary for me to do so.
And a good, informative graphical installer that does not
scare first time users is a good bet as well.
The distro described here would be great to attract new
users.
If this is already taken care of, my apologies, but the
last time I tried Linux (red hat 8.0, I think) there seemed
to be a lot of different places to set up more or less the
same things. This is more confusing rather than an option.
S.
I must disagree with those saying POSIX compatibility is useless. They say “Where did POSIX compatibility get you?” Well, it gets you a wealth of programs that will compile with virtually no changes on Linux, BSD, Solaris, and most other Unices. This is important. It means when one Unix gets an app, it can be shared with the rest.
The article starts with the premise that a great distro has to give people what they’d expect from Windows. Of course. Duh. That’s where we live, the critical-mass.
I agree with you this is Adam’s assumption and the usual one on OSNews as well. I’m not sure its true however. I don’t know that systems built for the computer ignorant are the future. In terms of computer ignorance its been the assumption for decades that computers would have to become as easy to use as telephones or televisions. In reality however during the last 20 years telephones and televisions have become considerably harder to use. Meanwhile the population has become vastly more computer literate than it was.
Take typing: 20 years ago most office workers did not know how to type. Today some basic typing skills are assumed from educated person. Another example is the distinction between writing and typesetting. 20 years ago this was poorly understood and wordprocessors operated in terms of pages. Today they operate almost completely in terms of streams of text, people now write a text stream which will only bind to a position on the page as part of the typesetting (usually seconds prior to printing) processes. Both are major conceptual shift in the broad populations.
Another conceptual shift has been that people now understand the distinction between application and applications specific binary data (i.e. a .doc file vs. word.exe). 20-30 years ago this was a very complicated notion to explain to people and was considered very techy.
It may be that computer knowledge becomes a human skill like reading or driving which take a great deal of time and effort to acquite and even more time to achieve profeciency.
I don’t know that the “The critical-mass” is really as bas as people say. What people (especially on OSNews) seem to mean are
a) American / 1st world
b) computer ignorant and proud of it
c) not terribly well educated in general (i.e. HS-college)
Look at the 3rd world and you have vastly more information workers who are:
a) have more time for training since the hire / layoff rules of America don’t apply
b) have staff with more education and intellegence
c) Are mote intellegent and more educated
d) Are embarrased by lack of knowledge and usually seek to address these failings
I could see Linux doing quite well as the desktop / corporate OS in the 3rd world. If Linux ends up winnign the desktop wars in Peru, India, China, Russia… that’s huge progress for this decade. The US corporate culture is Microsoft’s home turf and most important battefield. Linux’s key advantages (freedom, configurability / scriptability, cost) don’t nearly as much impact here and Microsoft’s (genericity, standardization, ease of training) are more imporant. I would expect the US corporate desktop market to be the very last dominos to fall not one of the first.
… if you’re talking about binary packages.
As much as I hate to give them any credit, typical Windows applications come with everything they need. There are some exceptions to this — many a game come with DirectX on the CD but are not installed as part of the normal install process.
How does this work? A program is installed in a self-contained directory, and a shortcut to the program is placed somewhere accessible (like the Start Menu). If a program needs Qt (for example), it comes with it and places it in the directory where the program is installed. The OS will look in the binary directory to find needed runtime libraries.
Sure, there are some global system files (windows/system/*.dll), but for the most part these are relegated to drivers, not application level code.
Can this work under Linux? Sure — I write an app that requires Python and Qt, and the installer includes the runtime libraries. The problem is that Linux will not look for .so’s in the binary directory, so the program users run is actually a script which adds to LD_LIBRARY_PATH. Now was that so hard? Just make the OS’s loader do the same, and you’ve fixed that problem so all programs can work this way.
Worried about disk space because you’ve got 20 copies of some library floating around? That’s what the disk space comment he suggested was for — 20 copies times roughly 10MB a piece is 200 MB. Not too bad to alleviate all versioning issues, huh?
Now if that library is something like Qt where it is part of a window manager (e.g. KDE) and very commonly used, maybe that one library deserves promotion to be “part of the OS”, in which case the version had damn well better be stable over time (read: binary compatibility). Otherwise, one copy per package, please….
Have only one entry in the menu! Jesus, how hard is this to do? If you want more apps, the “power” users can just run a “UnHide the other apps” menu entry….Keep everything installed, but keep it simple for his audience I guess.
This article could be correct if only you turned it around and applied it to why Mac OSX is not a serious Unix as it stands and how what it really needs to take some pages from what the Linux community does right.
Modular, flexible, powerful systems that run on the fastest hardware available.
Your symlink idea gives you away, you have obviously never managed more than your desktop.
One of the single biggest problems with MacOSX is its over use of /Library and inconsistent FS with every published unix standard. The LSB is there for a reason.
Need I even mention that Mac OS X uses spaces in its file and directory names all over the place? Try scripting that correctly. The people that designed that system have obviously never written a shell script and no, apple script does not count.
MacOSX is a unix file system created by someone who never leaves the finder. That may work for a single at home desktop but it scales terribly for those of us that have to managed hundreds of systems.
Speaking of scaling, bsd over gpl? The biggest problem with BSD license is that it does not scale. Any time a project gets big enough to be useful some company proprietizes it and the inevitable fork occurs. Now you have two projects that become increasingly more incompatible until you might as well choose a completely different application. The GPL is scalable fork protection. Don’t comment on licenses until you get yourself some education on the long term benefits of the GPL on Open Source project health.
As far as package management, score another thing that makes Mac OS X scale terribly, not to mention causing no end of headaches and running around for support personnel.
RPM is not perfect, but tools like yum and apt-get rpm make it very effective for managing large numbers of systems and keepin them up to date. No going to the finder and clicking update either, no user intervention required as the systems update themselves, no reboot required.
Perhaps you should take some time to understand the power that a proper package management system gives you. Your comparison to Windows installers is a good one, but only because it maps to Mac OSX and its 1980’s era installers. Want to know how to make a unmanageable mess of a computer? Install software by dragging it and causing hidden scripts and programs to make alterations to your system. Please, that might seem like high tech to a desktop user, but to someone who manages clusters of linux boxes it is almost enough to make you cry.
Libraries? Do you even know what you are talking about? Have you ever tried to compile and install OS packages on Mac OS X? No elf file, no LD_LIBRARY_PATH, no ld.so.conf…they use some crackpot scheme for shared libaries that could only been created by a marqui de-sade type. They know it to, and they are changing it.
Look I do not mean to bash Mac OS X. It runs my G4 powerbook wonderfully and its integration of wireless is very well done. However it has a long way to go before I will consider it a serious Unix worthy of something more than a quicktime server or an ssh console.
The nearest thing to this already is Linuxstep.
http://www.linuxstep.org/
I would comment on the article, but it’s pointless as all the issues and their solutions have been thrashed out a long time ago.
The problem is that everyone has their own idea of the best possible distro, but very few can write code. Therefore we get a lot of wish lists, but very few real projects.
There are already countless ‘ideal’ package managing solutions, filesystem layouts, dependency resolving mechanisms etc, and to what benefit?
The ‘innovations’ are the *cause* of incompatibility not the solution!
OSX gets on perfectly well with /usr /var /bin /dev /etc, the difference is that apple understands the important part is making sure the user never has to deal with them directly.
You can rename entries in the root fs how you like, but if the <=average user has to navigate and comprehend them *at all*, you have already lost them.
I don´t want Eugenia to shut up, ever, even if I don´t agree with her sometimes. In fact the only people I think should shut up are those that try to shut up others, like you, Joe.
BTW, thanks Eugenia, that was an interesting thread.
“part of the OS”, in which case the version had damn well better be stable over time (read: binary compatibility).
At the highest levels (i.e. the kernel guys) binary compatability is not seen as an important feature. The kernel isn’t binary compatable over time, there is no way the desktop will be. Linux developers are in broad agreement that they are not willing to take the hits required to maintain long term binary compatability.
So your setup would require the distribution maker to release a particular build and then recompile an app a 1/2 dozen times for each environment and then distribute those binaries. Not undoable but expensive. They are never going to get the upsteam support for anything more than this.
Don’t want to discuss your proposed changes to the filesystem? Fine…I can syslink mine any way I like too.
Hmmmm, well I’m all for breaking LSB if it means putting a stake through the vile heart of RPMs once and for all. There’s a cold, cold place in Hell for RedHat on that one thing alone. But the fact that the author can’t figure how to uninstall software past “rpm -e <program>” is a big slug from the Foam Cluebat. However, after 20+ years of computer experience — on both sides of the support fence — I can say without fear of contradiction: Users are morons. Period. If they weren’t they wouldn’t need sys & net admins to bail their asses out every 5 mins.
Choice, especially on such a hot-button issue like “KDE vs. GNOME”, is critical. Deciding on one or the other and then throwing all your resources behind it is just assinine. Build distro-specific themes, submit tons of bug-reports, do whatever it takes, but do *NOT* reinvent the wheel by picking a favorite now and then regreting it later when the winds of change…well, change. Both projects are currently trying to work on a set of guidelines that just might actually help them settle on a common back-end — *FINALLY!* Oh, and BTW, I personally enjoy tweaking FVWM2 til it bleeds…all the extra GUI garbage just slows things down and makes the “eye candy & twitch reflex” game of user interface just plain useless.
Speaking of GUIs…
Which is easier: having a user chase pretty little buttons around the screen all day long or just having them answer a simple series of questions — while providing them with relavent information in order to make informed decisions? As a user of the oldest living Linux distro (guess…) and OpenBSD, I have to say that a simple text installer makes things much, much easier. The machine tells me what it needs to know next…instead of guessing which shiny red button to press to continue. Further correction: OpenBSD is *NOT* playing around with a graphical installer (even the GOBIE web page states it’s currently two [!] French CS students at the moment)…expect Theo to pay you a visit real soon now and God help you.
The issue of bloat has been beat into the ground, so I’m gonna just say “to each their own.”
<gut-wrenching vitriol deleted>: “The way I understand it, pretty much everything on a GNU/Linux system runs commands terminal-style.” Gee, no foolin’? When did that happen? *sigh* Moving on…
And well…the rest just tires me out…<licensing war rehash>…I prefer Freedom (beer and speech whenever possible), but I’m still doing a re-evaluation of GPL vs. BSD, so that’ll have to wait for another day.
Oh, and in conclusion, your perfect OS has already been built — nearly 20 years ago: Plan9.
Google is your doppleganger,
Kingstrum
Kingstrum, I agree totally.
is that it’s all done by programmers! We haven’t the foggiest idea what a regular joe user needs…we make things that work logically for us. I know how to enter my credit card on a web purchase form…but my sales doubled when I put up a picture of a cc and identified the various information they need to type with a legend. Linux is too smart for the desktop. dumb it down and it will sell like chitlins.
Linux desktop is almost already what we need but not in the main distros. Since main Linux distros such as overrated Red Hat and nice but after you manage to install SuSE 8.2, or Mandrake that is getting there but still needs several improvements, are not the answer to an average desktop user who patiently waited for years for those bigger names in Linux business to hear her/his needs.
Now those companies say they are going to pay more attention to desktop, still when you try to install their newest version something is missing, say, “Red Hat ships without mp3 support”, SuSE 8.2 is difficult to install on some hardware and Mandrake just makes cosmetic changes without addressing a larger question of running MS apps, “NTFS resizing”, or having “integration into Microsoft domains”.
Yet, we all have our gods in Linux desktop. One can argue that his/hers is the best, the most potent and…forgiving of all Linux gods. But we’re looking for some universality here, right? We need a desktop Linux to do a lot of things and yet be simple and cheap. Hey, that’s an ideal desktop Linux god. So, while big names in Linux have long history of not looking after desktop users needs, some of us turned into less known distributions and discovered that there are other gods besides big gods as well.
And those smaller distors are something of a discovery within a discovery. First you discover Linux, an idea, a god, but then you discover that that god has its flows and maybe even stinks, so you moved to other big and known Linux gods and the same thing happens. If you didn’t give up smaller gods are there to help, even so they anticipated you were searching for some other solution, they set their priorities in desktop development to begin with. Let me tell you about my gods in Linux desktop.
I need to resize NTFS partition, I need not just see the “c:” drive but be able to open files that are there and copy their content say from MS Word into some Linux app while still in Linux, I need also to run MS apps in my Linux environment such as Photoshop or Visio, I need to have good integration into MS Win domains and their networks without doing a lot of Samba configuration, I also need to install Windows apps with a simulated Windows reboot in just few seconds instead of waiting two-three minutes – hey, it’s a litany of needs.
But folks that’s already been done in Xandros Desktop Deluxe 1.0 that includes CrossOver Office, which by itself costs over $50 as a separate product while Xandros includes it in their distro. Xandros has all those godly features mentioned in the above paragraph, so obviously for me that distribution is my religion and I’m using it and I’ll recommend it for deployment to my customers who are thinking and deliberating which Linux god is the best for their desktop needs.
I find it strange that so many people whine about their current distros, invent new file system hierarchy, and nobody cares to help existing projects like LinuxSTEP, which are right on your target …
Here’s an article ( http://www.kuro5hin.org/story/2003/4/7/1599/91200 ) I think you should check out, at the very least for the portion titled “Change the filesystem layout.” I didn’t really bother reading the rest of the article because that first page turned me off, and a quick glancing/perusal of the other two yielded a sort of “What do I care?” response from me what with all these articles about Linux desktop this and Linux desktop that. I’m A-OK with people writing a ton of articles on this subject, but not within this one and a half or so week period. It’s like a flamewar with no official sides, and it just makes both sides look idiotic.
I’ll keep my dual-boot and enjoy the best of both worlds, thank you.
All these articles on the problems with Linux… sounds like a bad breakup and a desire to paint the other person with dark and ugly colors.
The interesting stuff (the betrayal of Linux by Linus) only gets one article, essential just lip service.
Even the perennial whipping-boy of the server world, Sun, got one article… an interview with some total idiot who obviously has never stepped outside of the fine timeshare cubicle Sun provides him.
OSNEWS has turned into some strange degenerate version of Slashdot… just the ugly OS wars and no cool science articles.
I hope we can get this out of the OSNews system soon and move onto better and more interesting things.
Having only read part of the article, I may be mistaken, but, with the fact that you are proposing things done with Open Source software, you could always roll up your own distro. You would probably be breaking a lot of compatibility, and POSIX compliance, but I don’t know of anything that is stopping you, or are your suggesting someone else build the distro of your dreams? The beauty of Open Source and Linux is, if you want to do something with the stuff, do it!! If you want someone else to do it for you, well, either wait until someone else makes the distro you want, or, pay someone to do it for you. I happen to like the way Linux is, and if a person can’t handle it, well, there are alternatives for them, such as OS/2, or even CP/M………
So your setup would require the distribution maker to release a particular build and then recompile an app a 1/2 dozen times for each environment and then distribute those binaries.
I think you must have missed part of my point, as that is exactly what I was trying to propose a solution to. Let me attempt to clarify: I’ll continue to use Qt as an example.
The “normal” approach for developers is to use whatever the hell version they want of Qt, and just redistribute the library as part of their package. No app will expect any one version (or any version at all) to be installed system-wide, and thus there is no reason for the developer to create multiple binary distributions.
If it turns out that Qt support is going to be widely required of most apps and binary compatibility is supported by the library developer (and Trolltech has actually been somewhat good about that), only then can it be promoted to OS-level library status and made available system-wide.
Compare this to the Win32 system libraries, and even DirectX. DirectX is in general *completely* backwards compatible, so upgrading to a new version still allows old software to function.
So to summarize again: anything guaranteed to maintain binary compatibility and that will be widely used by apps can be made system-wide and in general safely assumed to already be installed as “part of” the OS. Any other library that an application needs should be bundeled with the binary distribution of the app.
The problem you describe is what exists out there now, simply because the libraries are not distributed with the apps. The app I work on has *one* linux binary distribution, in a .tar.gz (no install script or package), and to my knowledge it works out of the box on every linux distro it has been tested on in the past couple years (incl. RH, Mandrake, SUSE).
There is NOTHING gained by promoting application libraries to “system-wide” status.
One of the BIG PROBLEMS with Windows is the COM deadlocks caused by having many Office libraries promoted to “system-wide”… which were never really designed to be system-wide. DLL-based COM bugs are insidious and very difficult to track down. Add that to the thousands of “wrong version” DLL bugs out there… and time has shown the “system directory of DLL’s” is not a winning formula for a quality operating system.
If ALL apps followed the approach you suggested, with each app including the parts it needs, Linux would be super reliable. There is no upside to following the broken model that Windows imposed.
The Windows central system directory of DLL’s is there just to make it more difficult for third party software developers. It is not there for any useful technical purpose.
Please let us not copy Windows just because some idiots in the Linux community blatantly replicate Windows and call it “innovation”.
Probably not the best place to post this, but I keep wondering how hard it would be to build a MacOSX emulation layer on Linux and/or BSD.
It might involve ditching X and recreating the vector-based Aqua stuff.
Show your mother the file system of any Operating System and have her guess where the program is. That is just a stupid comment. Show your mother the inside of the Space Shuttle and ask her where the main electrical breaker is. My mom knows where her files are, because I showed her. If you remember, when you first started using a computer, no matter what OS, you didn’t know shit about it. You learned, why can’t your mom? If you think POSIX guidelines are wrong, write your own guidelines, and build your company around them. There is nothing stopping you from rolling your own distro of FreeBSD, if that’s what you want to do. If you want someone else to pay you to do it, start asking around, otherwise, in your spare time (and I know there’s probably not much of it) start up the distro of your dreams, and someone may join you in developing it. I won’t, because I happen to like Linux the way it is, with way too many choices of some things, and very little of others, but the source code to all that I get on the discs.
To create a better OS X than OS X, simply license Display PDF from Adobe and go from there. There is very little ingenious technology in OS X. Most of it is just remanufactured NeXTware bolted to some new parts from Adobe and other companies that Apple has bought.
If the Linux effort had a central brain, making something superior to OS X would happen very fast.
Instead the father of Linux is off on a mission to add DRM and other police state features to Linux. Obviously Linus has decided Fuhrer is a better title than father.
The last time I checked, all these 3 distros – RedHat, Mandrake & Suse are already heading toward the same goals in many instances that the authour mentions, as are other distros.
There is a long way to go, but by developing your DistroX, you are indeed doing exactly what you proport to disagree with – YALD
I remember trying out RedHat 6.0 some time ago and I can say the difference between RedHat 6.0 and RedHat 9.0 in terms of user friendliness is remarkable.
I guess what I’m trying to say is that the current leading distros are improving on each release, bringing out another Linux Distro that doesn’t borrow from current ones is impossible or at best a waste of time, so why not just contribute your ideas for the ultimate distribution to a leading distribution ?
Why not present the user a nice friendly view through the Finder (a la Windows Explorer, HP-UX’s SAM or the CDE explorer) and leave the filesystem as is?
I mean, when the user opens the DistroX Explorer, show him ícons saying ‘hardware’, ‘my documents’, ‘my songs’, ‘my pictures’, ‘my programs’ and whatnot. The / directory can be accessed by clicking on ‘my system’ -> ‘geek stuff’ -> ‘you dont want to click this’ -> ‘root filesystem’.
Think about it: how many times do you need to access C: from the Windows Explorer? You click on ‘My network places’, or ‘My documents’ instead, and go from there. If you want to use a program, you click on the ‘Start’ button. Personally, I only click on C: to reach C:TEMP (which I don’t remember if I created myself or came with the OS).
With a cool GUI installer (like GOBIE) a lot of drivers, maybe you could try a way to make easy the use of Portage, using the big gentoo repository…..
bninja_penguin said “Show your mother the file system of any Operating System and have her guess where the program is. That is just a stupid comment. Show your mother the inside of the Space Shuttle and ask her where the main electrical breaker is.”
um, “Program Files”? (as much i think it should be “Applications”)
i no nothing of how *nix works, but would it not be possible to break compatability with POSIX, yet use symlinks for backwards compatability, like the oppisite of what the article suggests, and what OS X does.
Hi Adam,
some things are already there .. some for ages.
The symlinks are part of Nextstep (now Mac OS X) for more then 10 years .. and I think it’s a good idea to use it in Linux too.
The package installer: If you are familiar with Debian and apt-get you know it’s easy to install new Debian packages. You can search for them, you grab only one and it is installing all needed dependencies and you can use FTP servers..
The only thing missing: The Drag- and Drop-Interface. But that isn’t impossible. Let’s do it.
_One_ working desktop environment is a problem for the whole Unix/Linux/X Windows community for ages. OpenWindows, OSF/Motif, CDE, KDE, Gnome etc. It what be nice to have only one.. Your “workaround” to choose one is not the worst idea.
The system wide defaults are existing in Debian. Look to /etc/alternatives – there are links to the real application installed on your machine. But who knows about it? (If you like create a symlink /System/Apps to it;-)
I really enjoy working with FreeBSD – I am a systems admin and like the reliable and clean system. But there is a lack of device drivers especially for “home computer equipment”.
But if you do it it will certainly help the FreeBSD project because the team is “server focused” (for good reasons)
But anyway,
thanks for your interesting article
Peter
He is prefectly right in making links such as /hardware. The underlying structure will remain the same. So if new hardware is installed, its file will enter /dev which is linked to /hardware. Pretty simple!!!!
its file will enter /dev which is linked to /hardware. Pretty simple!!!!
Not that simple, if that’s so then, say you type “cd /hardware” and then you find out you’ve been taken to /dev. What the?? How did I get here? Would a normal user ever think that’s a shorcut to /dev? Again, you have been taken to an unknown world. Or, say you double click the file/shorcut that looks like a folder in Explorer (**cough cough** Konquerer) and again the user gets transfered to /dev. If somethings like this is to be done, you better make the real thing, not decoys…Even if it breakes other applications, create a brand new distro that will work this way and then all we need to do for the rest of the apps to work is simply modify them a little bit (the parts of the code that tell the app where to find the required libs) and recompile it and there you go, you will make a version of your program for this brand new distro.
Where do i order?
If hardware manufactores would just deliver drivers for linux and not just windows. The software companies are allready(more or less) supporting linux, but importent hardware manufactores as creative labs should make drivers for linux, so the costumer could benefit the advanced features of their sound cards(like EAX). I like what nvidia and ati are doing..giving us drivers and actualy updating them
Why ain’t manufactores like logitech, creative, microsoft(NOT gonna happen). Another annoying thing about drivers is that after produkt launch there will go like 3 months or so before drivers might begin showing. come on, a little more support and alot of people will be very happy.
Firstly, I don’t think that Apple or SUN are any better than MS…
That said…
I think that Linux should concentrate on:
a) being an OS for low end hardware
b) being an OS for “hardcore” g33ks
c) being a top class server OS
Leave the consumer end to the people with the time to do it. Leave it to Apple. Concentrate on making Linux better for the people who will value it. DO make it easier to install so that a hardcore g33k can spend his time doing cool things rather than trying to get his Winmodem to work…
Alex (The Original) said “Not that simple, if that’s so then, say you type “cd /hardware” and then you find out you’ve been taken to /dev. What the?? How did I get here?”
well, can anyone please asnwer my query? would it not be possible to break compatability with POSIX, yet use symlinks for backwards compatability, so this:
/home -> /users
/var -> /logs
/var/www -> /system/websites/
/etc/httpd – > /servers/httpd/
/dev/hdc -> /hardware/cdrom/
/mnt -> /hardware/drives/
would become this:
/users -> /home
/logs -> /var
/system/websites/ -> /var/www
/servers/httpd/ – > /etc/httpd
/hardware/cdrom/ -> /dev/hdc
/hardware/drives/ -> /mnt
a solution to combatting symlinks showing up in file explorers when they shouldn’t would be easy if there was feature (and i don’t know if there is) where the user (or install) could set files to different levels of hiddenness, and the user could select what level they wanted displayed. this would help if a user wanted to try and fix something with their system by being able to unhide normally untouched system files without unhiding unused/redundant/whatever files at the same time. just an idea..
kind of like ‘layers of importance’
No developer would support your *incompatible* silly distro. I think this says it all — how many things you can afford to `change radically’, support them w/o any help of upstream manintainers, and not go bankrupt right away? Either you want *NIX or you want something different, but having Unix and pretending it’s something different is hard (look at OS X). IOW, try it and I’ll enjoy watching you failing 😉
To even propose development of a OS that would compromise 30 years of evolution and hard work with standards should be regarded as a crime, a proper punishment for this crime would be life in prison. In order for us to have systems that can help us with our daily stuff we need standards, if we don’t have standards we will invent the wheel three times a day and we will loose our productivity. Not another system that don’t allow us to interact in a common way please, we already have the guys at Redmond trying to force us in to a corner.
As soon as you use a license that can be commercialised and controlled you return to the state of play that Microsoft has. The most popular system will if it becomes commercially (monopoly oppotunity) to do so will close up and keep its captive market.
The GPL and the freedom it protects makes Gnu/Linux the most suitable option for a general purpose world OS because
1/ It is/can be very low cost, hence everyone everwhere can legally have it without the rampant illegal copying that Microsoft turns a blind eye to whilst it gets benefit.
2/ It cannot be co-opted
3/ Everyone can see it fully to write truly competitive apps
and other things…
Let’s make such a general purpose desktop OS that fills the vast majorities needs. It’s like building roads and letting everyone use them. It’s infracstructure and should be a public work.
This applies to many core applications like e-mail and common data formats.
This whole approach does not remove the specialist softare from being a profit center (think Matlab, etc…) But even those are being showen to be in danger by the business models of companies like MySQL.
So you can make money in the application space. Just not the greed dripping amounts formerly made by monopolistic exploitation.
Meanwhile if using the system of the commons is to common for you then f**k off and buy the latest whizzy OS on super hardware. That’s a little like the realm of cars where only a few people drive Ferraris and only a very small number of the custom built numbers actually use the abilities (think formula one racing) the rest do it because they wouldn’t be seen in a Ford, Skoda etc…
This is a word I’d like to send to all the `Oh! Great ideas! I can’t write a shell pipe but want to use Linux’ types.
Please stop trying to make Linux `something else’. Linux is a free UNIX-like OS (this is The Definition). It should stay so, because it’s great in this respect. If you don’t like *NIX, use something else, or invent another OS. Linux then perhaps won’t conquer the world, which is OK, and won’t be flooded by BFUs, which is really great.
Anyway, whinning is pointless. You have the power of GPL. Want to change it? Change it!
Its interesting how things go round and round…. some OSes resolved these things decades ago time moves on and then they recreating the same mess…
Asking the old questions:-…
keep everything the same
keep everything as it is but only show “certain parts to the enduser”
keep everything as simple as possible
I think the best thing to do is take a stance and stick with it for a few decades no matter what it is… it has to be consistant.
The question of the desktop user and how to gain create a bigger push over to linux… ” I hear some windows users say I have windows working as I need why should I change? ”
The answer has to be in whats developed under Linux if all the apps that run are just replacement for Windows apps or look the same then for that user theres not much reason for change…
There needs to be a push into new ways of using a desktop, interacting with a desktop… The reason for keeping 2D enviroments over exploring 3D “virtual reality” enviroments is that doing most things would not be an improvement from the current ways of doing them etc… To write a letter the only way people are thinking is through 2D eyes and placing this into a 3D enviroment…. *may not help you do the letter faster or better but may have other unseen advantages*
There needs to be either killer apps “the old way of thinking what pushes OSes and enviroments” or maybe its time to explore fundermentals a lot more…. and changes in how we use linux in a everyday useage and create something totally new and fresh… that people will see the advantages and what to use and adopt..
simple examples from time
– punch cards to keyboards
– 70’s/80’s move from command line to GUI…
– windows 3.1 to windows 95a *mostly same product just a different
way of looking at it and thinking about how to interact with it”
Microsoft have always let others create then borrowed… *I wont go down this road no more as it serves no improvement to Linux*
If no one or nothing is ever creating whats going to be pushing the boundries of Computing…
In twenty or fourty years time we will still have a 2D monitor, keyboard n mouse… with “Application menu” in the bottom left hand corror…
…………………………..maybe u get my point maybe u dont…………………………..
Hello all,
Where to start? Ok ill just jump right in then. Filesystems: Leave the current FS alone please, although it may seem very cryptic it does work very well for the pro’s. What many of you are upset about is being forced to attempt to understand all of its different aspects. As Eugenia and several others have stated with a good UI that becomes a non issue. There honestly is no need to break compliance in this area. Another point is that there really isnt a need to get radical at all, this would imply that *M-A-N-Y* hundres or thousands of Dev’s before today have not already Done So
What I mean is simple, the useability you are asking for is Mostly All Here already. Both simple users and Power Pro’s ™ can both be served well. Take for instance that although im relatively new to linux I strictly avoid booting straight to GUI. Why? because Run Level Three has CLI tools that allow me to UNFUBAR my freshly broken system. Then again if I wanted to boot straight to GUI thats very easy to do too. And a Good UI makes doing so Trivial.
Another Point there must be a good filesystem that CAN be used if need requires on both CLI and in GUI mode, linux has both right now, many doezens of times over. What is needed is a serious UI GURU to step up to the plate and Re-Package the plethora of great UI fragments we have alread into a Co-Hesive Whole. That ladies and gents is the prob as stated several times. Its not that linux is hopelessly broken it is that linux is hopelessly fragmented and cannot find a “central UI personality” if you will…
A quick and dirty example of a DE: fluxbox with RoxFiler and a few addons in it. Believe it or not its DAMN flexable, fast, easy to use in many areas…What is !!NOT!! is even re motely easy to Get Too That Level. Its a bonifide MOFO to assemble it to BE that way. A good UI team of Dev’s / UI designers could take this simple example and make it one seriously kick ass usable GUI. Please dont get me wrong by thinking im proposing that this formula be The Holy Grail ™. Im not. It’s just an quick and dirty example. As Eugenia said, It is about a Cleaned up and easy to figure out UI design. Thing im trying to get across is its nearly all here NOW but its simply fragmented to hades. Whatever else you do DONT break Linux as its current woes are by far more a symptom of a dysfunctional G-UI / Common Tool Set than a broken <arch>. If you take a step back and a couple to the side I belive you will see a new Distro with a serious UI clean up (Not a super heavy GUI) that adds in a great User Interface is what is needed. Abstract the confusing-to-newbs FS with the very clean UI and you *WILL* find very quicly that nothing drastic really needs to be done. I will grant that both Driver add in / removal needs more work, and that package handling both adding / removing needs some massaging. But look around and you will very quickly see that Package management is broken more by fragmentation than by lack of viable current solutions.
So to re-iterate in conclusion: 1) address simple problems under the hood (common app fonts, apps handling, driver/device addin removal ect..), 2) Decide on a CLI toolset that maximized the CLI power-AND-Allows a very clean UI to be implemented so that a very usable GUI results. Its the Creative Discipline in GUI design that is badly broken / missing. NOT the Talent in Developers!!! If Dev’s and UI pro’s were to shake hands and roll up their sleves in 12 months or less a new and truly INOVATIVE Linux distro would see the light of day… think seriously on this folks. It is not nearly as far away as many may think it is. It only requires thinking outside of “the box” in a creatively dispassionate manner for a short time.
Cheers…:)
Those suggestions are about one percent of all stuff that needs to be either implemented or modified in Linux distros. Everybody just sit on their asses contemplating Linux philosophy and write another clone to kill time…
Fact is that Linux distros are Unix (as in mainframe) with a GUI slapped onto it. No more no less. There are a few usable applications, yes, but everything else is lacking.
Calling any distro an OS or desktop is a joke. RedHat, Suse plus some others made a real effort recently (8.x 9.x), but that was what any reasonable human being would have expected a year or two after their appearence on the market. Too little too late. If they don’t keep up that recent pace forever, then they could just aswell throw in the towel now.
Microsoft looks around and snatch things already existing, Linux developers try to implement a few percents of what M$ mimics. Get a grip and do something new and original for a change.
This is no Troll, I just have a different perspective; I still use AmigaOS (along with XP and RHL), which is still maintained by it’s users. Our shareware/gpl developers manage to produce quality software. F.ex. a scalable GUI engine was implemented by a user already in 1992 (MUI). It’s of “corporate” quality (no, I don’t mean bug infested :-), yet free shareware. We have a single free fileserver with mirrors all over the world (no login required, anonymous ftp accepted). It still puzzles me why Linux developers can’t manage to do stuff like that. All projects are tiny details and nobody is prepared to do anything major (read: usefull) and everything is hosted on personal webpages. And to be honest, the distro installers would have been written by a user years ago if we were talking AmigaOS. We have a wide variety of _working_ software, not a wide variety of cloning.
Linux developers are among the most conservative I know of.
AmigaOS still have many features not stolen by M$ yet, so why not implement those in Linux? Start with smart icons f.ex (with command line arguments and app settings, all in plain text, accessible by right-click/icon settings). File comment are not unique, but Linux is unique in not supporting them (very handy when they read: http://www.fileshare.org/application_name/unreadable_short_filename…), so you can either identify what it is, or resume the download without specifying anything but the filename. And, yes, this is standard behavior in AmigaOS.
Linux is an excellent example of having freedom and not using it. You have the freedom to kick Posix and you don’t. You have the freedom to make GUI’s work faster and you don’t. You can make it userfriendly in notime and you don’t.
Start grabbing your freedom. You can’t keep on limiting the desktop by imposing server specs as a bad excuse of not adressing real problems.
Get some imagination for gods sake.
Don’t get me wrong; I like using Linux, I’m not so fond of XP, but I love AmigaOS; it’s got features where it counts. Linux and XP are just so low-tech when it comes to the desktop. I travel back through time I use them and in some ways even their applications.
Some random thoughts:
The filesystemlayout: Quite frankly this does normally only concern people who work on the command line and then only whose $PATH is not set up good enough for the command found automatically. On a graphical desktop the user should not have to care wether the program is in /bin or /usr/bin. That’s what the icon in the start menu is for. IMHO the filemanager should do the abstraction for the user (I don’t recall if the KDE filemanager this already). In the directory tree he should get /yourhome (maps to /home/$USER) /system (maps to / if you really have to look there) /cdrom1 etc (the various removable media his computer supports) and maybe some /storagex entries (for free disk space outside his home he can access to dump his MP3s). You don’t have to rewrite the filesystem layout, you just have to hide it successfully. (On a sidenote: I would vote for some small change in the UNIX Filesystem convention: could someone please rename “passwd” to “user” and “shadow” to “passwd”? This way it would more resemble the tools that work on these files (groupadd/group, useradd/user, passwd/passwd))
Making it a desktop distro: A desktopdistro IMHO is not about choice. It’s about “It works (for me)” ™. Therefore simplicity in the install is key. “few visible options but lots of “Advanced” buttons” is definitively the wrong approach I would say. And “use this filesystem if you want to use it as a server” sounds to me like he wants to build what we call an egg-laying-wool-milk-boar. They don’t want to build a server, otherwise they would have gotten a different distribution. Just give them Reiser and 90% won’t ever know the difference (or XFS or JFS, whatever).
About software install: To put it one way: It’s not about RPM but how you use it as a distribution builder. What’s probably most frustrating with software installation in a an environment like Debian (I know it’s DEB not RPM but lets leave that out for the moment) is the lack of a hirarchy. Assume you are a desktop user and what to install a nice text-editor (from all the myriad text editors in the repository) but you don’t know the name of one specific. How do you find one? Is there a category /editors/graphical/gnome ? With the gnome maybe highlighted because the system knows you have gnome installed as you current dektop? No, but there should. I haven’t used SuSe in recent years but from what I recall, it did have some thematic structure that i felt more comfortable with from a desktop perspective.
Dependency hell is not a matter of the package mechanism but more of distribution maker that did not consider these problems carefull enough. A common library should never be further away than the installation CD. A program that needs it’s own, not provided by the distribution maker, librarys can install them in a special subdirectory in the program directory and should be compiled that it finds them there without having to bend the $PATH.
I had some more thoughts about how a desktop distri should be but that would better fit a whole article that just a post. The question was always “Is Linux ready for the desktop?” the answer is Yes/No I would say. The technology is there. Somebody just has to build something out of it that works in a desktop way and not in a “It might also be a server if you try hard enough to bend it this way”. A desktop is not about choice but to give the user something that just works for him.
BTW: I think this article was more about how to run a company that builds a desktop OS based on Linux/BSD that how to build a desktop distribution.
Nice work Adam. I read this piece as an attempt to get people to start thinking in a different direction (as opposed to YALD, as you mentioned). The specifics of your DistroX aren’t as important as the thoughts and concepts behind them. It made me think “outside the box” and that is how true advancements and innovation happen.
Thanks,
Joe
I can’t beleive that everyone is so negative about the progress that Linux has made on the desktop – it’s improving at a huge rate of knotts.
I recently installed Linux on a file server at work and all my workmates were very impressed with the GUI (RH8). Especially impressed with multiple users remotely using a desktop on the machine.
IMO the linux desktop is good now and improving rapidly… so I’m satisfied, at least. And contrary to the article, I don’t think that closing source and charging for software is the way to get the best out of OS developers! I’m sure there are a lot of other developers there who think “right, you want me to contribute code to your system, but you’ve got the core bits under lock and key… Riiiiight, and can see what a GREAT deal this is for me!”
Code freedom is the best thing about linux and the reason it is successful *despite* not having the most polished desktop interface ever.
Peter.
TUNES is radical.
Why do people continue to ignore me when I point out that appfolders are not workable? My mind is open, but nobody has been able to deal with these arguments I put forward in my FAQ on the subject (see autopackage.org). To be fair to him, Thomas Leonard is playing with new ideas to try and eliminate the concept of software installation, but they don’t resemble appfolders all that much.
I think it’s pretty funny that people pointing out problems with the approaches set out in this article are decried as “visionless”. The fact is that people who know what they’re doing *ARE* trying to solve these problems, there was a post earlier moaning about clipboards and lack of desktop standards. Hello? Freedesktop.org? Clipboard problems means you need to upgrade by the way (or stop using mozilla, which doesn’t follow the standards because the module owner has got used to the broken way).
Really, I think if you want to write an article like this, you should be somebody who has sat down and followed the debates on the various Linux forums over this and that, tried to understand all the issues (not just end user simplicity, but also manageability, security, backwards compat, i18n etc), taken part and then you can make suggestions about how to do things. As it is, these articles repeatedly make suggestions that were written off years ago for valid reasons, simply through lack of research.
I am glad someone brought this up.
Seems like Apple is the only one thats able to do it right the first time. I have been waiting for Linux to “rule the world” since 1998 when i was a 100% BeOS user.
What has really changed?
Fundamentally at its core, nothing has changed. You still have dependency problems, while flaming MS for its DLL hell.
You flame the teletubbies GUI of Windows XP, but havent for the most part even got the fonts straighten out OOB.
Still, the visionless stay behind defending Linux while waiting for the newest kernel because its free while much fewer developers are working on other operating systems and have gotten miniscule communities that are fundamentally much farther ahead of you.
Linux brags about having millions of developers, but i gotta say that they are doing a piss poor job if there are really that many.
Just put 1000 developers on f.ex openbeos, Amiga, SkyOS, AtheOS, Syllable or all the other much smaller operating systems being mentioned at Osnews for 18 months, and i personally believe either would kick the crap out of Linux as a desktop operating system.
1. Dependency problems? Have them automatized away forever. Use apt-get with the stable Debian repository.
2. Wrong application bindings? Solved with /etc/alternatives in Debian.
3. Complicated install? Pick the simplest one on the market and let users choose their software themselves. Just provide one big button labeled “Get Software” that uses apt.
4. Need acceptance? You can’t buy developers with a community if you require them to use snail mail and refuse them some parts of the system they want to tweak. Above all, do not steal from your users the software they bought from you.
5. Need to make money? Sell something you do best. It cannot be a distribution; that market is saturated.
Don’t think in terms of a distribution. What you are calling for is tool integration. Make a product of that, not of the software. You can’t compete with free. Be the first real contender in the total tool integration market.
Bye a mac and start using Mac OS X instead and you have ease of use.
I do not understand why people even want’s Linux. I do understand that people wants to get away from Windows but Linux is to me the wrong direction. The idé of open source is great but it doesn’t work.
Ok a mac is not cheap and it’s not the fastest computer in the world. But Apple got something that the open source community totally lacks, as I see it. The open source community don’t have a CEO like Steve Jobs who has the power to pull the whole community towards his vision on how the future computing should lock like. The Open Source community simply have to many wills and most of them are just copies from other Operating Systems. So why not just get an mac and forget about dependency problems etc. etc.
And If you don’t like Steve Jobs, hes time of retirement will come some day to. But admit that he’s way better to the industry than Bill Gates.
By the way I have no problems at all with Steve Jobs, I like what he does.
One of the bad things that linux for the desktop shares with Windows is the anomalous way the user is expected to understand the filesystem.
All of these systems have a directory that corresponds to the Desktop (/home/<foo>/Desktop in KDE, /Documents and Settings/<foo>/Desktop in Windows). I have watched several new users get totally confused by this and the way it violates the directory navigation metaphor. I’d love to see that changed.
Why do we still have this huge box that sits on our desktop with an architecture that is 20 to 30 years old?
Linux will win as the lines between embedded systems and desktop pc’s start to blur.Settop boxes have more brains than pc’s. With more and more broadband services becoming available and wi-fi becoming more feasible, things such as stand-alone game stations, pc’s, settop boxes, tv’s, audio equipment will begin to be combined into “smart” systems.
The manufacturers of thess new systems (Sony, Philips, etc.) will be using a Linux OS and not Windows.
“I can’t beleive that everyone is so negative about the progress that Linux has made on the desktop – it’s improving at a huge rate of knotts.”
i agree…the only 3 issues which still need heavy improvement imo are installation of programs, printing (drivers and integration) and (some windows-wannabe-distros are reported to have that) functional easy-to-use connection to a windows-network.
except those contructionsites, it simply needs more polishing, integration, even more graphical tools, maturity…
imo 1-2, max. 3 years from now.
I liked the ideas of /conf and /user/asau. In addition /sys/info for list of devices and /sys/conf (I don’t like it to be in /proc). I also think we should kill all that /usr/bin /usr/local/bin /usr/local/local/bin etc. Should be one /bin and one /apps or /app or /opt (though “/opt” should not be). In addition, it would be good to become able to create temporary symlinks in root directory like “/my” (-> /user/asau), “my0” ( -> /asau/TeX/src ) etc. One abbreviation of “~” is not enough (but Midnight Commander must die, for not to come into dreams at midnight, brrr). OTOH, I don’t think there is necessity to separate /logs from /var. Are you going to look at logs so often? On your desktop computer?
—-
About installation and package managing… KILL ALL “DEB”-s and “RPM”-s! Why you need yet another packaging and compressing format?! All installation should be done the right way: mkdir /app/APPNAME; cd /app/APPNAME; tar xvfz /tmp/APPNAME-X.XX.X.tar.gz THAT’s all. “TAR is all you need!”
If your program wants a library, it should have a place (in /conf/APPNAME/libs, /var/APPNAME/conf, /app/APPNAME/libs or somewhere else) to write associations like “libc –> /lib/glibc-X.XX.X.so” or save its libs in /app/APPNAME/lib/ . If you run out of space you can replace real /app/APPNAME/lib/libc with symlink to /lib/glibc-X.XX.X.so Nothing is more simple. If you don’t want to package libc, place a hook: “Notification: Replace /app/APPNAME/lib/libc with symlink to glibc-X.XX.X or compatible. Application stopped.” And don’t forget to make link from /app or /desk (like a shortcut):
cat >/app/start/APPNAME <<END
#!/app/APPNAME/EXECFILE
END
chmod a+x /app/start/APPNAME (Or even simpler: make a (sym)link!)
Just KISS. Nothing can be so simple. You get rid of all those dependencies in one moment.
BTW, Plan9 has nice feature to mount some different directories to one point (something may overlap, but this can be resolved).
Sometimes, I just don’t get this “community”. It seems that at times the “Wall” has been hit as to how much needs to be done for the benefit of everyone. Let’s assume that somebody’s got a point that software installation on Linux is a problem. Then, just for the sake of research lets actually delve into and read or scour the net for issues regarding dependancy issues. Does anyone really understand just how many HOURS of human resources are wasted on this crap. Does anyone truly care? And that’s the point. Linux people (myself included) have a tendency toward having a holier than thou attitude about computers. Trust me, I DON’T want to believe in some weird stereotype of the glasses wearing geek that weighs either 86 lbs or 323 lbs hanging out all day talking to themselves about code, and sitting in a room piled high with books and arcane junk to the ceiling, never getting out in public and has an insecurity problem. Is that where all the intellectual snobbery comes from insecurities? Point is : Make it work. Make it work right. And make it work for all of humanity.
If Linux users come off as snobs it’s only because so many of them hear the same drivel over, and over, and over. You would get aggrivated too if you had to deal with it. I’m not like that, but that’s most likely only because I’m a casual Linux user that doesn’t have the know-how to attract the newcomer’s attention. I don’t really get a whole lot of people asking me things like “Well why isn’t it just done *this* way? It would be a lot easier!” It’s the whole thing about Linux traditionally being for people who enjoy the system and want to have a sense of accomplishment for having done what they’ve done. Traditional users tend to be do-it-yourself types, and these new desktop users that have been flooding the arena lately demand, demand, and demand. They don’t offer anything, they don’t care about how it’s been done or why it’s been done (Standards be damned, get rid of POSIX! That’s funny, if I recall correctly, Microsoft is required by law to include POSIX compatability if they want to do business with the US government. I know because I’ve disabled POSIX on a few 2k boxes to get as much performance as possible out of a gaming box). They’re in for a rude awakening when they find that open source developers don’t respond well to ignoramuses spouting bull. It works in the Windows world because a lot of developers sell their software and *have* to listen to customer feedback.
Don’t ever forget this: In the open source world, expect a handful of developers to tell you to go to hell when all you can do is complain without ever lending a hand. This isn’t Windows-land.
The first question is how radical do you really want to be? As some keep pointing out “Linux” is just the kernel. Everything else could be changed or replaced, and it would still be a “Linux distro”.
I would dump X and use an entirely different GUI system, preferably one closely integrated with GUI system tools. Nowadays, there are even several different choices of file systems: Ext2, Ext3, XFS, ReiserFS, UFS, etc. Pick one that would work best with your GUI design.
Of course, the more “radical” the distro, the less compatible it would be with other Linux distros. However, POSIX is not LINUX or UNIX. I would stick with a high degree of POSIX-compatibility unless you had very good reasons for not doing so.
File system hierarchy? Again, it’s a matter of how compatible you want to be with other Linux distros, or how well you can compensate for the differences. But I wouldn’t put too much faith in symlinks to solve the problem.
Of course, to make this a reality, you need some good developers who can actually create the code, because the more radical the distro, the more new code that will need be to generated. Some existing software could probably be used as a shortcut, but you don’t want to restrict or lock in your vision, do you?
The first thing to do would to take a look at [i]already existing[/] projects and make sure somebody isn’t already doing exactly what you want. LinuxStep has already been mentioned, and I believe there are a few different GUI projects like Fresco around. And Blue-Eyed OS is another project that is using the Linux kernel, although their intention is not really a Linux distro.
But the first step is really to have a good idea about what you’re trying to achieve.
i posted a comment to this article yesterday and now i cant
find it. anybody see it ?
thanks
I’m fairly anti-Mac, but I bought a PowerBook G4 2 months ago because I wanted real UNIX on my laptop (I was getting tied of the limits of CygWin) and the ability to burn DVDs was appealing. I found that it has most everything that you listed in your article.
By the way, the library dependency problems of Linux just don’t exist in FreeBSD. If you want a package that requires other packages, just install the package you want, it will find everything else. Backwards compatibility isn’t so much of a problem because older libraries can co-exist with newer ones. Heck, some vendors ship binaries compiled on FreeBSD 3.0 because they “just work”… even on FreeBSD 5.0.
Sometimes, I just don’t get this “community”. It seems that at times the “Wall” has been hit as to how much needs to be done for the benefit of everyone. Let’s assume that somebody’s got a point that software installation on Linux is a problem. Then, just for the sake of research lets actually delve into and read or scour the net for issues regarding dependancy issues. Does anyone really understand just how many HOURS of human resources are wasted on this crap. Does anyone truly care? And that’s the point. Linux people (myself included) have a tendency toward having a holier than thou attitude about computers. Trust me, I DON’T want to believe in some weird stereotype of the glasses wearing geek that weighs either 86 lbs or 323 lbs hanging out all day talking to themselves about code, and sitting in a room piled high with books and arcane junk to the ceiling, never getting out in public and has an insecurity problem. Is that where all the intellectual snobbery comes from insecurities? Point is : Make it work. Make it work right. And make it work for all of humanity.v
If you want to clean things up, I suggest reading the Unix-Haters Handbook first.
1) Then I want to tell that the ESCAPE key definitely needs to be fixed. As most function keys start with ESC, a program which receives just ESC wait a few seconds for other data, and if that does not come it finally interpretes the ESC. This is awful.
2) The X-Windows system is also not very good. There should be one toolkit. For example, create a GTK+2 or QT3 X-server extension and rewrite all toolkit libraries including Athena Widgets to just call functions from this toolkit.
3) Development for the distro: most newer programs have this great AUTOCONF tool. It is no problem to install any program in any weird directory, just add some options to ./configure. So a new directory layout is possible with just recompiling.
4) People say that /dev is just as good as /Hardware. I think not. /Hardware is much clearer for users. And with autocompletion it does not cost more keystrokes – H-tab is nothing more difficult than d-tab.
5) cp, mv, ls should become copy, move and list. Or should mkdir and rmdir rather become md and rd?
6) A next release of KDE and Gnome might hide ~/Desktop and ~/Trash. Maybe we should make /system/* and /user//*, and thus have /system/Trash and /joe/Trash?
7) Michael: I do not think any particular filesystem will fit a GUI better. In Linux, file access is transparent, from the applications viewpoint it does not matter anything which filesystem the system uses. Besides that, I really like ReiserFS. Mostly since my ext3 partition corrupted when it went full.
8) ASau: I agree on your opinion only TAR is needed. But it needs a graphical interface. Tar-archives should behave like normal folders in a graphical file manager. Like in RiscOS.
9) About Steve Jobs, our PPC 6400/180 is the worst computer I have ever seen. OS 7 runs in a m68k-emulator and the newer OS 8 is too heavy for the system. But OS X seems to be great.
10) “Just put 1000 developers on f.ex openbeos, Amiga, SkyOS, AtheOS, Syllable or all the other much smaller operating systems being mentioned at Osnews for 18 months, and i personally believe either would kick the crap out of Linux as a desktop operating system.” The I guess IBM or such would already have done that. Or well? I am really looking forward to a bootable Syllable cd-rom.
Bon. Ik ga TV kijken. Tschüss.
> Filesystem layout. I would recommend the author to read
> about what POSIX is. POSIX defines a standard layout for
> the file system.
yeah. POSIX says that a system must have the following directories:
/
/dev
/tmp
and the following files:
/dev/null
/dev/tty
Period. Check your facts before posting, please
Daan said:
2) The X-Windows system is also not very good. There should be one toolkit. For example, create a GTK+2 or QT3 X-server extension and rewrite all toolkit libraries including Athena Widgets to just call functions from this toolkit.
That’s right – make one new GUI toolkit with a nice API like in BeOS and make wrappers for all other toolkits.
BTW, if you really want to be radical and do away with legacy you should base your system on some new improved(tm) programming language. C++ is nothing but a hack to make C object oriented. (that’s very off topic, I know)
I would like to point people to the following URLs:
http://www.linuxstep.org/
http://developer.linuxstep.org/
LinuxSTEP is an experiment to not think of Linux as a rag tag collection of applications atop a Linux based kernel, but rather to create a single cohesive environment on top of the Linux kernel using GNUstep as the primary application toolkit, and unification point.
The other major point of this design is to remove the legacy UNIX filesystem cruft, and replace it with more meaningful pathnames meant for user consumption, and to provide user friendly means of managing that information not only at the application level, but at the filesystem level.
I recommend reading the LinuxSTEP Filesystem Heirarchy at developer.linuxstep.org for more information on our revolutionary filesystem design. Of course, you can also stop by #linuxstep on irc.freenode.net and talk to the developers.
-Thom Cherryhomes
http://wwws.sun.com/software/sunone/innercircle/newsletter/coursey_…
Just as I don’t agree with most of the points raised in the osNews article, I think that David Coursey has got it wrong in the article I reference above.
Unfortunately, they’re complimentary. As more and more people have the idea of what would be the Next Great Distro, more of them move full circle towards repeating the history of Unix – binary incompatibility, differing filesystem layouts and increased fragmentation.
If you, YOU the reader, want to see Linux in a dominant position any time soon – or /ever/ – only use distros that don’t try and fragment the Linux user base. Support the LSB. /Understand/ why these are good ideas and why they’re necessary for Linux to win – and even to survive.
From the sounds of it you just described Lindows. Of course you run into problems when you pick and choose what your going to install. If you pick kde and concentrate on customizing that to run on your distribution perfectly, but anything else just kinda works then you are going to alienate all the Gnome lovers out there, and vice versa. Then of course on the install you talk about adding all kinds of ‘eye candy’. Persoanlly, I hate eye candy. I would rather just pick what I want, express install or custom, and get it done so I can start using my new system. I havn’t taken a major poll on this but most people I talk with, in and out of the IS field, agree. So I personally would remove the eye candy and just get the install done as quickly as possible.
Don’t really understand why you would restructure your filesystem. A typical new users, windows or otherwise, more than likely will not go past the gui until they are comfortable with the system. If you really want to make it worth while, I’d leave the installation as close to unix as you can and include a book on the linux system. You’d do well to cut a deal to include Linux for dummies along with it. Concentrate on making the graphical nice and clean, add lots of configuration tools and icons, redesign their ‘start menu’ so that they can find everything. They should be able to do anything on the system from their desktop. Some times are harder to do like creating symlinks to samba shares or things to that nature, but eventually I’m sure that could be worked out as well. With this type of configuration they are up and running on their new linux system, can do everything they need from the gui, AND they have a book that’s simple to read if they want to learn more about the linux filesystem and how the system they are running works.
I agree that when your done with the install you should have defaults set. But you should pick these from your installation. Have a default set that if you pick express install it will set these as your defaults, like Mozilla and Evolution. But if you pick custom then you should at one point get one page that lists all the major clients, Web, FTP, Email, etc with perhaps drop down boxes next to each. Each box should contain all the currently installed packages for that field (ex: Web would list Mozilla, Galeon, etc). So when your done, you have all the main clients set for your distribution, but you allowed the user to pick them.
For the most part, your right that drastic changes would have to be done for linux to make a hit on the deskop for normal windows like users. I personally think distributions like Lindows and Xandros have a good idea and will probably work well to expand linux onto desktop systems. I personally wouldn’t use one of their distributions, but that’s cause I actually LIKE having options on my linux distributions. Just remember though that when you take away things like being able to install anything and have many different options like window managers, you are no longer running linux, your running the next version of Windows.
Aric
And while we’re at it, lets drop the x386 baggage too.
I disagree with changing the filesystem lay-out. it is an amateur techie perspective only. On a good stable distro, there is no reason to browse the filesystem. The filesystem is there to support the system, but all a user needs is to find their document, spreadsheets, images, etc. On a good distro, there is no need to hunt after individual configuration files or mess around in the /dev directory. So, the focus of the Desktop Environments should be to hide as much of the underlying mechanics as possible. And I don’t mean obscure, I mean not needing to see…
Let’s focus efforts on building a good DE rather than dropping POSIX compatibility and use all the effort to update apps for the new filesystem layout.
No mention in your article of bsd/gentoo emerge, linux apt-get or urpmi. How can you, as a Mandrake 9.x user, not
mention URPMI. It resolves the dependency problems of which you complain. No mention of Webmin which does much of the /servers, /hardware aliasing you so desire. Hard to take you serious when you don’t even do the minimum in research.
Why not use BeOS ??
In a few months we’ll have OpenBeOS too!
—8<——8<——8<——8<—
/home -> /users
/var -> /logs
/var/www -> /system/websites/
/etc/httpd – > /servers/httpd/
/dev/hdc -> /hardware/cdrom/
/mnt -> /hardware/drives/
—8<——8<——8<——8<—
Jeg forstår ikke en døyt av det som står over her..
heller..:
/home -> /brukere
/var -> /logger
/var/www -> /system/vevsider/
/etc/httpd – > /tjenere/httpd/
/dev/hdc -> /maskinvare/kpspiller/
/mnt -> /maskinvare/sekundærplatelager
Er du ikke enig??
So you want to make it more easy to understand the paths. For who??.. Not all speak English. There are more people that speek Chinese. So why not Chinese. You understood what I wrote above. It’s a Germanic language. One of the base language of English…
—
Mike Klev
In your article – I read, make the installer as easy as Windows.
– Nobody I know that has on off-the-shelf copy of windows and that has 1 periperal/component not included on the stock CD can install it. And they have comented that it is longer and more complicated to install than whatever Linux distro I have proposed: mostly Corel 1.2 and lately, Xandros. Corel was actually king of the newbie_install :
5 screens & approx 7 clicks and 1 reboot later is was all done.
Let’s reverse the scenario and have Windows do as well….
Cheerios,
SP.
Simple and easy to use, that’s what I want. I don’t understand
tech talk. Just give me simplicity. That’s what makes windows
and Mac so attractive. They make it simple, and the programs all work. Yes, I have to pay for them but I don’t want to run something that only works partially. I just loaded Xandros, recognized all my hardware, nice OS but when I wanted to copy a audio cd with cd roast (their default)it told me that audio copying was presently not supported; back to windows. There is no choice yet for people like me.
People I switched to unix cause i don’t need this shitty things. I WANT ALWAYS HAVE FREE ACCESS TO COMMAND LINE. I cleary understand where to locate system components, I don’t need to rename (or link) /dev /usr /home – they are transparent to me; nor my mother, neither my sister need or want to know something behind fs layout.
I like Unix, you can do what you want with GUI, but don’t force to use it, don’t touch command line, vi, text-based installer. As for GUI i like only blackbox.
It is nothing forbidden to me now, your system or distro tries to limit my freedom I simly can’t permit that!
sounds like a great idea, ive benn using linux for more than a year now, and for many things its still to hard to use.
A free osx on the x86, why not, please free us simpletons from the shackles of microsoft, give us something free, that works, thats designed for idiots like me.
and then youll get idiots like my mother, board directors and the people from IT who have windows in their blood using open source.
(open source is the point here, not just linux)
make it for everyone and they will come.
ps. my mother hates linux because she cant use it.
FOR EVERY ONEsaid in a small spooky voice
As a 20+year user of Dos/Windows I can say that when it comes to downloading rpms/deps and installing them I go nuts! We here do not understand “compiling” or dependancies. I know that these are common in the Linux community and our hats off to those of you who can do it. However our,or most “computer users” need a straight forward way to just use our machines to do work without having to “mount” something or “compile” something. About a year ago when we first contacted the Linux community in forums like this, we had, shall we say, mixed acceptance. Some meeting were very hostile, however before long we talked to other Windows migrators who helped us see the light of Linux! To those we say thank you. I know this will irritate some here, but please bear with me. In Windows to copy a floppy 1.44 disk you simply right click the floppy and select copy floppy. Lindows/Xandros/RH/Mandrake etc. all have a floppy formatter. You click on it and it allows one to format a floppy in Ext2 or Dos. Thats simple. Why cannot Linux etal have a simple program to copy a floppy as good as the formatter and as easy to use? There are still no good PaperPort type programs for scanning and no good CD label making programs. Ever try to use CirclePrint? Linux Distros are so close to allowing us to get away from “Darth Vader” and their activating ways. We as x-Windows users do not want Windows to run on our Desktop Linux Distro, we want Linux-period! Help us by providing something we know how to use. Are there enough Linux Distros for those able and happy to command the OS? If so distribute one which has been proposed and everyone can win.
Thanks for listening
P. Nelson “build it and they will come”