Posted
by
michael
from the includes-two-of-every-application dept.
An anonymous reader notes that OFB has a short blurb about a new Linux distribution, Ark Linux, based on Red Hat and chasing the ever-elusive goal of being "easy to use for the masses".
This discussion has been archived.
No new comments can be posted.
Ark Linux did not exist four and a half years ago. In fact, we didn't want to go public yet - we don't like preannouncing vaporware (though the current alpha version is actually usable, if you can overlook the rough edges) -- so there wasn't much need to update the website etc. until someone decided to notify the media of our existance.
The 4 1/2 years reference from the original poster refers to the creation of the Mandrake distribution, which started with similar goals, but developed into a different direction.
Mandrake is a good distribution, and suitable for desktop use if you know what you're doing a bit - but Mandrake is an all-purpose operating system, and therefore too complicated for many newbies (this may be hard for us geeks to understand, but people do get confused at the notion of 50 editors, 4 desktop environments and 20 MP3 players).
Ark Linux will focus on being a home user OS, and just that. IMO, if you want to please too many totally different needs at the same time (and "server", "techie workstation" and "newbie home user" are 3 _very_ different needs), you have to make too many tradeoffs.
Obviously these folks need to read the Open Source HOWTO [slashdot.org].
What Linux needs is more announcements of distributions "focussed on ease of use for desktop users." There's no need to actually ship anything -- just making announcements seems to be sufficient to get a Slashdot story. And you can spend hours posting here, rather than staring blankly at ark.c in Pico!
It's a pretty interesting environment which allows new companies (distributions) to get an instant running start by basing a new distribution on someone elses. It reminds me of the whole "on the shoulders of giants" saying.
With all these easy to use Linux versions around, a problem arises.
Basically, Linux isn't THAT hard to use for real Newbies (people with very little Windows experience, but aren't afraid to learn), but it is hard to use for many because it is *unfamiliar.* Having many different easy-to-use versions may be a problem in exacerbating that problem. Lets face it, Windows is easy to use mostly because since Microsoft has a monopoly, it is what most people are familiar with.
Basically, Linux isn't THAT hard to use for real Newbies (people with very little Windows experience, but aren't afraid to learn), but it is hard to use for many because it is *unfamiliar.*
There are some reasons that Linux is hard to use that have nothing to do with being different than Windows.
For one thing, Linux lacks a consistent definition of how a GUI should work, so you don't get the kind of consistency you get with, e.g., MacOS. For instance, some of my apps use the middle-mouse-button method for copying to the clipboard, while some use the Win/Mac method. Some try to implement both, and fail. Inconsistency is probably unavoidable considering the open-source development model.
Another problem is shared libraries. Shared libraries have some real advantages, but they create lots of problems with installing software. How are you going to explain to a naive user that they can't run a certain app because it requires libfoo 2.0, and they have libfoo 3.0 installed?
When will developers finally realize sharing libraries is really retarded? EVERYTHING should be self standing. Yea, maybe 10 years ago conserving disk space mattered, but c'mon. Now, drives are $1 a gig and doubling like mad. Quit it with the shared DLL's and shared libraries, and shared anything. Fuck installers too. Just put it on a CD like this
/OpenOffice (executable) /data/-everything it needs-
Then there's no making sure someone has the right version of anything installed.
Don't give me some excuse. Most Mac software I've seen is like this. If there's some technical reason you can't do that, FIX IT. This is OpenSource right? You have no-one to blame but yourself, just fix it so you can just put the app in a folder on any machine no matter what.
You compile all apps -static and you get 50 copies of malloc() in your memory (RAM+swap, but eventually you would want them to be in RAM) simultaneously! (actually, ps -ax|wc showed my that I have 89 processes running now, and I suppose most of them would use malloc(), right?)
Next, factor in cache trashing -- that shared malloc() code could've occupied couple lines in a corner of your instruction cache, now you need to flush/reload them every time application context is switched.
Yea, maybe 10 years ago conserving disk space mattered...
Besides size, the other advantage of shared libraries is that if there's a bug in the library, you fix it, and the fix automatically propagates throughout all your apps.
Rather than asking developers of end-user apps to stop using shared libraries, I think it might be more on target to ask developers of libraries to stop changing their APIs. If they never change the APIs, then they never have to break binary compatibility.
The problem is that this is fairly easy to do with C, but requires a heck of a lot of discipline with C++. In C++, almost any change you make to a class that's exposed in your API will break binary compatibility.
I suspect that a lot of application developers could easily build monolithically linked binaries, but they don't because they know it would increase the load on their servers by an order of magnitude, which would cost them money.
It's also worth noting that this whole thing is basically only a problem with C. In Perl, there is really no such problem, because you don't get an incompatibility as a side-effect of changing a class.
The best way to make linux "easy to use for the masses" is NOT going to be by adding yet another distribution of it. Personally it would help if all the "desktop linux" companies pooled their resources and made one, standard linux desktop running on one, standard gui interface. Adding another linux distro just makes things more confusing for the masses.
More choice isn't always good. I would rather see ONE good desktop linux package than ten substandard ones.
There is a user running around here which has a sig which goes kinda like this: Tesla: Had Edison thought out his work and spent more time in preparation, he would not sweat so much.
This is not an interesting comment, it is woefully repeated trite. This sort of comment is dredged up any time there is any thread about Linux and the Desktop.
Why don't we just get all of the soft-drink manufacturers to get together to make ONE good drink to rule them all. While we're at it when are the clothing companies going to get together and mass market grey jump suits so we can really move into the twenty-first century?
Soft drinks (including juices) are available in tetra packs, glass bottles, plastic bottles, squeeze bottles, cans with ring pulls, cans with button tabs,... etc.
Customers have no trouble learning a new interface. They do it all the time.
Customers have no trouble learning a new interface. They do it all the time.
Amen to that brother. That statement reminds me of a "debate" (actually more of an argument) I had with my best mate a few years ago, a die hard Mac head. We were discussing interface consitancy, with Paul attempting to show that Linux would never go anywhere because it had 2 major widget toolkits that had different themes, no universal HIG etc. He ended up trying to prove that people couldn't use other peoples mobile phones.
The hilarity that ensued kept us amused for a looong time. The sight of him walking around the 6th form centre with his phone asking people to open up the address book on it, then looking dismayed when after studying it a moment they did it as fast as he would have done, was excellent. The "what are you smoking, man" looks he got were even better.
People are smart, people are adaptable. Yes, a completely consistant UI is good for user efficiency which is why people are trying hard with unified themes and user interface guidelines etc, but at the end of the day with a few notable exceptions (software installation/resolution switching) it's been possible to figure how to work Linux for a long time now and anyway no other OS has a completely consistant GUI, least of all Windows (not even the mac).
Let me know when setting up a stable, usable* Linux box is as easy as pulling the tab off a can of soda, or twisting off the lid, or squeezing the bottle, etc.
Will do! Now, let me know when you find a can/bottle that can play counter-Strike, fetch my email, format papers for printing, browse slashdot, etc.
That would be counter to the aims of the companies though, especially if they have stock holders... if they were to cooperate, then they'd be opening themselves up to a suit by their stock-holders for not acting in their best interests.
Further, one company would be easier for (MSFT|SUN|APPLE|$Proprietary_Software_Production_C o_Bent_On_World_Domination) to smite (TM).
Basically if you want a unified easy to use distribution, you're better off making your own (non-commercial distribution), and hoping the commercial distributions standardize off you.
You would rather see one good linux package than ten substandard ones. Well, first of all, they can't all be substandard. Whatever is their quality is the de facto standard. But I digress. What I am trying to say here is that it would not be in the interest of the Linux community to combine all of the distributions into one "good" one. Each distribution has its own advantages and disadvantages, many of which are mutually exclusive. Combining Debian, Slackware, Red Hat, Mandrake, etc would not yield one system with the advantages of all and disadvantages of none. It simply wouldn't work.
One of the biggest draws of Linux and the Open Source movement in general is that there is so much choice. Not only the freedom to choose a specific distribution, but to create a new one if you feel there is an unfilled niche. Combining distros cannot work, so these people felt that they could fill a niche by creating their own distribution.
Here's the deal. RedHat Linux isn't anymore difficult to install than WinXP (or really any other previous) incarnation of Windows. There really isn't a MS Windows installation for the masses if you really get down to it. If you want the masses to use Linux then you will have to have more computer manufacturers that will bundle Linux already on the machine.
I am so tired of hearing this and so tired of this comment being moderated up as interesting. It is not interesting, it is not informative, it is not insightful. There is no reason to mod up this tired old argument anymore ok people.
Agreed, and I'll add one more thing. The first distro that adds Windows Domain support - COMPLETE support will be the one I implement here.
My users get mapped drives and shares when they login to the domain. I need my users to have centralized information stored on a Windows-based server. The software we use have few analogs in the open source world, so our backend is likely to remain the same for the forseeable future.
All of this needs to be seamless, and it needs to happen reliably. I haven't seen that yet. The pieces parts are there, certainly, but no one has gotten it right yet.
My guess is, this is EXEACTLY what Red Hat is planning to do with their next Workstation version.
Seems like a nice idea to me:
Take the good base of RedHat and replace their castrated Desktop with a working one. Since Bero is responsible for this it might even turn out pretty good...
Is another "easy to use" Linux distro what we really need? I think the last thing I want to do is be able to take a program off of my Mandrake box, take it over to my Ark box and not be able to run it because they are two different distros.
And also, who has not seen this "make linux easier for the desktop" thing before? There are about a million and one distros who purport to do this. Why does everyone feel the need to reinvent the wheel?
There are about a million and one distros who purport to do this. Why does everyone feel the need to reinvent the wheel?
Though I can't speak for the producers of ArkLinux, I will speak as someone who has been involved with them early on.
I think one of the prime motivations is to act as a 'concept' vehicle to offer innovations to the larger distributions. In Detroit, it's often difficult to get the Big Three automakers to really cut new ground and try something risky. However, they love to put a bunch of wacky ideas into concept cars, and then slowly evolve their best ideas into real products for the street.
The well-known distros like Red Hat, SuSE, Mandrake and Debian are analagous to the Big Three automakers in the Linux distribution space. However, they're not in a high-markup tangible goods market like Chevrolet, GM and Ford. Thus, they can't afford to make their own 'concept' tools and services to help their own evolution.
I see all of the smaller distros as helping the evolution by giving each great idea (and tons of mediocre and bad ideas) a public forum in which they can prove themselves, and be cherry-picked by the powerhouses of Linux adoption: the Big Distros.
And frankly, Detroit isn't being robbed of available talent whenever some kid puts together a supercharged Dodge Charger with neon all over it. Likewise, Red Hat doesn't sweat it if some afficianados take a different path and try a few new things.
You better hurry and volunteer to help with Ark Linux. There's only two positions available. There's a catch, however. You must have experience in python, awk, squeak, moo, bark, and oink.;)
Just so this doesn't mislead anybody... The above post was intended to be a joke, because of the name 'Ark Linux' (All the animalstwo-by-two, etc). It backfired, because somebody modded it 'informative'. =)
It seems to me that this debate will rage on and on about which desktop is the best etc etc. I don't know much about this ark linux offering but I do like the idea of someone else taking a crack at it. its a worthy goal to reach for.
it seems to me that the only way this will be achieved is by a distro making some very draconian decisions about what to include. I for one would be delighted to see a very lightweight distro that did just a few things very well.
what I mean is if there was a distro that included a web browser, mail client, open office, mp3 player & a terminal program, and they all looked good and were consistent in look, feel and function, then this would be a major step forward. you dont need to include much more than that to have a successful easy to use distro. no point in including tons of compilers and dev tools if you want to make something easy for non-techies to use.
hopefully this is a goal that will be reached someday. it is so possible to do that I'm really surprised to see so many failed attempts. All I'm suggesting is a lightweight distro with a select few apps that are heavily polished and work well. If you did that you could gain a following and maybe get enough credibility and respect that other app developers would code to your human interface guidelines.
With the exception of Open Office, you've just described Mac OS X.
And if only these open source developers would stop with the "distribution of the week" approach and direct their efforts towards something like a simple GNUstep [gnustep.org] install, they'd actually have a shot of getting some OS X software brought over to Linux and FreeBSD.
As it turns out I use OS X as my primary desktop OS. I love it.
GNUstep does not get enough cred. If you had a distro running GNUstep w/ a few polished apps you'd open yourself up to developers releasing apps for both your distro and OS X at the same time. On top of that you'd have a Linux desktop capable of running any other Linux app.
Umm, he said next software announcement. That was the last one. The interesting thing will be to see if they release an entire office suite to compete head on with MS Office. Safari and Keynote were the next software announcement until they were announced, at which point they made people anxious for the possibility of a new AppleWorks.
Ark Linux will not be a Windows clone, though we don't hesitate to clone parts of the Windows UI where it makes sense.
"Microsoft does XXX" is neither "XXX is good" nor "XXX is bad".
Ark Linux aims at providing everything the average home user needs - the desktop, office suite, media player, net access tools, a couple of games, etc.
For developers, there's the Ark Development Suite, a collection of development tools and programming languages (about 200 MB) - comparable to Visual Studio (with obvious differences - e.g. we support additional languages like Python, Objective-C, Perl,..., but don't support C# (yet - Mono isn't ready for prime time)).
Of course, Ark Development Suite is 100% free too - the only reason it isn't included in the base OS is that 95% of the targeted users won't need it.
Linux, BSD, and everythign else need one thing, just one thing that embedded systems like Zaurus and others do successfully to make it easy for the masses. Remove everything technical. All the drivers should be an LKM, there should be something like a "Control Panel" to choose which ones load on boot-up. Users should be allowed to pick and choose their WM, but it should be easy. Gnome Control Panel did this I think. There should be _no mention_ of the term: 'recompile the kernel'. There should be binary packages available so people can download, right click, and select "Install..." to install a program. It should boot-up graphically into X-Windows and should only have one xterm-like replacement(or even xterm itself) hidden, deep in the menus.
Does this take out everything we love? Yes. I'll gladly stick to FreeBSD with my rxvt's and my gettys, my kernel compilation and my make worlds, thank you very much. The beauty of Windows is that it takes all of the fun out. But what we consider fun others consider a PITA.
If a distribution did this, took everything out, made it all LKM's, and took all the technicalities out, then we would have a system for everyday users. They don't want the command line. Come on, some of these people can only do 5 wpm! When starting up your web browser because you don't know about tab-completion and you can't type takes 15 seconds before the program is even executed, you have a major problem there. It should be click(or, optionally, click-click) and it should run.
This needs a team of dedicated people to run a free system like this, but more likely, a team of employees from a company will do this(and many are trying to) and to maintain binary packages, an easy system to install/update/delete packages. That's what we need.
What is wrong with Red Hat and Mandrake and the others? Simple. Whenever they do something to make everything more simplistic, the community(read: many of us slashdotters) makes a backlash saying the system lacks this or that functionality because they did this or that. I say let them. We always have Debian, Slackware, and other distributions. Let a few of these become end-user-never-going-to-type-in-a-command-in-their -lifetime-bought-the-damn-thing-at-wal-mart-for-ch rist-sakes distributions and leave it at that. Applaud them for taking out even the ability to re-compile the kernel. Applaud them for it.
Yes, many of these companies are doing idiotic things, but we're part of the community and they look to us for guidance. Let's tell them, not yell at them, to take apache out of the main distribution and for god sakes disable it for regular desktop users.
Desktops should be so easy an 8 year old should be able to set up. Servers should be difficult to set up.
Sorry, I had to rant. and yes, I know it's the same rant everyone else on slashdot says, but this time I get to say it.
I never tried it so I don't know for sure, but isn't Lindows trying to be exactly what you're describing? I mean, maybe not every little detail (or maybe even every little detail, as I said, I don't know for sure), but certainly the spirit, no?
Well, to tell you the truth, I am a Slackware user because I gave up on the distros that PRETEND to be user friendly, so why not just go for the real thing and get my hands dirty, but at least I know what the hell is going on. And I have yet to see a distro that is so user-friendly both in the interface as in configuration and application installation as BeOS -still- is.
I haven't tried it either. I'm glad you see my point though. I think there should be different distros. Some for the more experienced user(Slackware, Debian, Etc) and some for the fresh-from-the-womb (Redhat, Lindows, Mandrake, Ark, Etc..). Choices are a good thing sometimes. Fragmentation is a good thing sometimes. If everyone pooled all their resources imagine how pissed the 'linux-must-stay-pure' people will be when the 'linux-for-the-masses' practically disable getty and remove every terminal(xterm,rxvt,eterm) etc.
Configuration is a big baddie because you only have to do it once. And the more experienced users are ok with a more difficult install.
The biggest problem with lindows is that they want you to pay for the most basic apps. Half the reason for switching to linux (especially for a home user) is that all your basic essentials are free. Nobody expects you to have to pay a basic program everyone needs on a day to day basis. Commercial apps take care of commercial problems.
I don't know about you, but I feel lindows has strayed far from the community and only attempts to provide a semblence of support for it. Right down to providing source code they've fought what community is about tooth and nail. The CEO isn't interested in the community, he's intrested in exploiting it.
I agree with you in principle, but why must such a division be made in Linux? The divide you are trying to create between the user-distro and the hacker-distro is gaping. Couldn't it be possible to make it easy for the new user but leave in real functionality in case they want to learn? After all, if you can only convince someone to move to Linux because everything difficult has been removed, then they won't want to move to real Linux where the useful things are still there. If it becomes prevalent, there would be a massive base of binary-only systems that can't even compile programs that they download. If something like this is implemented and takes off to the extent that would be considered a success, there would be the same amount of division and condescension between old Linux and new Linux as there is now between current Linux and current Windows.
That would not be beneficial to anyone but the Windows-haters. Linux-lovers should not jump at something like this.
Well, my idea, which I didn't state well enough, was to have binary pre-compiled packages easy enough. I want the gap, I want a end-user linux. But I want them to be able to go to the web site, download eterm, right click, hit install, and run eterm. That act alone would open a new world to the user. And same with gcc. There is no way it should be necessary on a desktop. But it definately should be there to download and install, with the punch of a button. Then users could go from there. (Mind you, the gcc package should be called 'develoeprs' or somethings and come with everything for configure, etc..)
--redhat 7.1 had a good gnomerpm that all you needed was a right click to install an app. It was great, worked well, now it takes three different screens and a lot of clicking to do the same thing. So at one time it existed.
Coming from a mostly never ever touched a cli since the 80's on dos which I detested as just stoopid, I found (still find on 7.2) redhat to be quite easy to use as long as I stick to redhat brand rpms and not generic rpms, that one threw me for a loop "why" that was when the "non standard" rpm happened to me. Seems as a bit of courtesy that whomever took the label "rpm" and made an app that was labeled that would have followed the originator's standard in what file went where. If they didn't like iot they should have called it something else. To me that's like buying a non oem aftermarket alternator that is labeled "belchfire" but it won't bolt onto a belchfire. Sorta kinda dumb and rude there.
Free as in speech and in beer doesn't necessarily make the speech outstanding oratory or the brew outstanding beverage. So I agree, quality, consistency, ease of use. A few dozen apps that work well, not a few hundred that are so-so and require weeks of code wrangling. The main bitching I see here is "oh no, can't lose our elite tweaking ability". Who says you have to? Duh, any distro can add in the dang shell and console, but not being absolutely required to use it makes it or breaks it on "the masses" useability scale.
Another really off the wall side issue. Apps included in an officially released boxed set distro, and what you pay for the distro, there should be micropayments to the app maker guys to go with it, if they are releasing as begware. that will make what's included or not be a lot easier to sort out. If the apps aren't begware, and they are toting 100% of the note and don't care, swell, their bandwith and coding time. I don't mind begware and would like to support them (the apps I use and find useful obviously), but sending one dollar via online payment schedule to a zillion app orgs is just silly. And I think the clone copiers and resellers should send a buck per cd to the distros they are cloning. And here's another kicker sure to be controversial. the gpl says it has to be "free", swell, the information is free, the bandwith isn't. People should help pay the bandwith and associated costs for downloading ISO's. That bandwith is not the "source code" that is free, you still get the data, but those servers take wires and electricity and buildings, which costs money. It's just as much a real expense as it is to make cd copies and ship them, albeit less money, it adds up. Call it a "shipping and handling" charge for the "free" stuff you are receiving.
The entire idea of linux for the masses is linux for everyone. The only way to do this is to take out the difficultires of it. I think if a linux box has a web browser, a media player, a GUI, an e-mail client, and some games, it's reached the usefulness of most people. Not the die-hard gamers, not by a long shot, but by many people. What we consider useful is something that is considered pointless by many. That is why I think it should be so easy to install and use that people who want terminals won't mess with it. It should hide everything about it that is linux, and just be a functional, enjoyable, stable, useful computer. With all the free software we have available, with binary packages that are easy to install, if the OS was easy to install and get up and running, I think it could be a contender for world domination. Damned if I wouldn't be handing out the CDs at the mall, or leaving some for grabs at the local coffee shop.
But the gap is what I want. We may call ourselves 'users' but most of us and what we do are development. For those who aren't, many do not need the command line at all. Windows was doing a good thing hiding it, the problem is they wanted to eliminate it when some of us needed it. It should be a download and a few mouse clicks away to install rxvt, but it doesn't need to be base install for those who don't need it.
And you are right. Although I wouldn't say "Windows-Haters." We want the Windows-Lovers and the Windows-Neutral people also. Linux Lovers probably shouldn't use it, but support would be good if they got Linux running on more and more machines. Especially if it means games would come out for the Linux platform. Lord knows I'd be happy with a Linux Morrowind. =)
The whole idea of choices leave linux-lovers choosing debian, or brewing their own, but this would make Linux more mainstream and on the desktop, but the way free software works assures us that it would not be the only choice. So linux lovers should support it, they can always have what they want. But most people just want a web-browser with flash, an e-mail client, a word processor, and solitaire. Not something that takes a book to install, several hours to configure, and a day or two to go to cnn.com from trying to understand everything. It would take many of us less than an hour to get to that point, but someone who thinks a computer should just handle most of that stuff itself(which, look all around us, it _should_), will get upset and switch.
I fear that the only way to make something that easy to download and install (binary-only) is to make sure that the system for it does not change very much. Also, it would be necessary to ensure that the creators of the projects would make a pre-compiled binary for that system so that it could be used. The user-Linux would have to be unchangeably standardized and impossibly popular from the very start. What I mean is that in order for such a thing to gain users, it would need developer support, and in order to get developer support it would need a large userbase. Typical spiral. On the other hand, a system that compiles all of its software needs only the code (not specific developer support). Some possible workarounds for these issues is a binary archive at the distribution's website that takes projects from across the web and compiles them so that they can be downloaded and used on the new system, or a installation system that does not use a binary, but rather compiles and installs the software with one click. I know I would look forward to that kind of functionality. If everything could be specifically compiled just by right-clicking and selecting "Install," it would remain invisible to the user and it wouldn't be necessary to distribute everything as platform specific binaries.
If I'm not mistaken the GPL allows binary distributions, as long as it is made clear the source code is available upon request. That would be easy to do.
Unfortunately, having it compile when you right click and select install would be very visible to the user. Several minutes visible, and many people would get tired of waiting. Just imagine OpenOffice compiling on the system, while you wait, after downloading that ungodly large file already.
I don't see a problem with platform specific binaries. But yes, you need someone on the backend to compile those and make them available, that is why I suggested a team of dedicated people or a team of employees to handle that.
You would need not only a dedicated team, but a dedicated supercomputer to compile enough software such that everything is available. A lot of Windows programs take several minutes to install, so as long as there is something that takes up a lot of the screen and has pretty colors on it, people will understand that it is doing something important (installing). They wouldn't have to be informed that it is compiling, because they probably wouldn't know what that means, just that it is installing. Everybody knows what that means.
OpenOffice doesn't really take that long to download with a broadband connection, and I wouldn't recommend downloading anything larger than a few megs with a dialup connection. Just have your friend with cable burn you a CD-RW with what you want downloaded. It takes less time than downloading it at 2-5KB/s, and it's an excuse to get away from the computer and hang out with your friends for a while. And that is what a lot of people that this is aimed at would be interested in, much moreso than many geeks.
Desktops should be so easy an 8 year old should be able to set up.
This is why I think that some Debian projects like Debian Jr. [debian.org] ("Debian for children from 1 to 99") are particularly interesting. I don't care about kids programs, but I think they will face some UI design issues that will help make Debian easier for all (non-guru) users.
...are the ones that have a stated purpose, and dedicate its entire user interface towards that one purpose, for example ClarkConnect [clarkconnect.org] is used by many of my non-Linux/Unix using friends, and the only questions they have are always in regards to using it in a manner it was never meant to be used.
Enough, that is, of distributions that are "for the masses". It should be clear to everyone by now that this phrase is utterly meaningless, since it encompasses a huge number of possible approaches to the problem of making lusers happy with Unix. I propose that this phrase and all similarly generic phrases be officially declared Fucking Useless, and anyone who uses them be savagely beaten until they come up with a particular differentiating feature for their distribution.
So what is special about the distro of the week? Hardware autodetection? Careful customization of packages to provide a uniform and sensible default UI? Good paper documentation?
Oh, Jesus, if I just stop there, someone will moderate this up. Do you people realize how pathetic you are, that you're reading this? Writing it was bad enough (shame, shame, shame!), but reading it... can't get read again. Come on, eat me! Burn, karma, burn! SLASHDTO DEITORS SUX0000RZ1!!1! Bibbity bibbity bibbity!
It's not only a useless phrase, it's actually actively harmful. It leads people to believe that it's a possible or desirable thing.
This guy said it best:
The idea that an arbitrary naive human should be able to properly use a given
tool without training or understanding is even more wrong for computing than
it is for other tools (e.g. automobiles, airplanes, guns, power saws).
While true, I don't think that applies here. A computer is not a tool we use directly, but rather a tool that we build layers of tools on top of. Building a layer of limited, easily-comprehended functionality for lusers on top of the same basic tool that has the powerful, complex functionality i need does not harm me. (It has consequences that irritate me, because the lusers think they're using a tool remotely comparable to what I use and yammer ceaselessly at me, but it doesn't harm me directly.)
If you understand Linux well enough, it *is* easy to use. I find it very easy to use, even when doing something unusual that hasn't been designed ahead of time for me. I think the goal of theses "easy to use" Linux distros is really to be "easy to learn". And in this world, "easy to learn" means creating an interface similar to what people already know to shorten the learning curve.
most people say they "know PC's" of course meaning windoze. however, most of them have neither a) installed windows nor b) know but a few apps and don't know how to do the most basic things.
ask 100 windows users how to
change network settings
see what programs are running, and system resources used
add/remove programs at start up
99 of them will give you blank stares and tell you they have no idea. the problem is not that windows is "easy" because it isn't. it is what people know, and there's a HUGE difference.
linux doesn't need to be easier, because it is more than enough. no, i'm not talking about installation. we just need to get the camel's nose under the tent. most people who are familiar with computers and are not scared of them will be able to pick up quickly how to do things in linux.
C:\My Documents =/home/user
other than that, mozilla, OO, etc., the apps are more than sufficient.
The web site claims that There are already plenty of good Linux distributions for servers and for advanaced users out there - but so far, nobody has tried making an easy to use version of Linux for the masses, even though the KDE user interface makes Linux very easy to use. It is not our intent to take users away from other Linux distributions. It is our goal to provide a good desktop operating system.
But neither the article or the original project site do not disclose by which means the new distro will make Linux easy to use comparing to other linux distros.
Is it just new set of very well debugged installation programs very well integrated with GUI based wizards? If so, what's the difference with Mandrake, Lindows, SCO and Redhat?
Or it integrates some new end-user applications (like ximian evolution)?
Especially I wonder, what is done or is going to be done to improve the user experience from usability prospective? For example, what specifically is done to explain the user what makes this distro so easy to use?
Take off the tinfoil and a quick read of the site..
- 4 clicks to install = good - giving simplification a shot = good - developed own installer (i guess) = good
- no info about what to do with an iso = bad - explained in terms of red hat = bad - unanswered questions on page = bad - needs:
screenshots
info on hardware requirements
info on supported locales (or is it just English)
info for developers on "why develop for distro x"
needs a "why use our distro" page for users
- might be interesting if you could do work just by popping CD in (without partitioning or doing a big install).. or is that what it does? dunno.
Well..
as already pointed out by bero, we are still in a developement stage, and that's why we didn't any public announcement so far.
Our highest focus atm is the codebase, the distro itself.
That's why the website is still a bit "poor".
I hope we will be able to work on it soon... just remember that Ark Linux is an "open" project, therefore everybody is free to contribute anything.
I really hope to get more people involved as time goes on.
Let's see
I wish these guys the best of luck, anything that brings Linux to more people is a good thing, but...
IMHO all attempts to make computers 'easy to use' 'for the masses' have failed. Just ask any joe public who wants to actually use their computer, rather than spend endless hours tinkering with it or being plauged by niggling bugs. Computers are extremely complex things, and the critical mass of knowledge required to make them run smootly makes it impractical for every user to become a system administrator.
I think what's needed is a shift in perception and the model we use to sell hardware and software. Rather than just selling boxes (containing PCBs or CDROMs), computers should be rented. This would mitigate the endless upgrade cycle and (with the internet) allow trained professionals to administer customers' boxes, all included in the price. Linux is an ideal OS for such a distribution method.
Do I think this will happen? Not yet. For this model to be economically viable it would require specialist tools that would let admins look after huge amounts of boxes. A major shift in public perception would be required, especially after all the 'so easy to use even your granny can do it' ads. Finally, I think that it would take a lot for users to hand over the control of thier computer to anyone.
Redhat based distro's are old and everywhere. Why not base a new distro on something sexy like Gentoo or Freebsd.
I love the gentoo package managment. It is better then pure sex and yes even apt-get.
I found the distro too bleeding edge in some area's *cough devfs * cough and it takes a whole weekend to finally configure everything. The latest version was so buggy that I could not get my/boot to mount properly and my Microsoft USB keyboard would actually freeze the system so bad during reboot that I had to unplug my system just to have my keyboard respond. Weird. All the newer distro do this and I think its a kernel bug.
But its lightning fast and more bsd like in which all the actual config files go in/etc and not just symlinks that point to god knows where *cough redhat *cough which make it difficult to learn and administer.
Freebsd rocks too in which sysinstall configures things pretty good but it takes alot of tweaking to get the following Xfree86 to work with my geforce4, the beta nvidia drivers, java, and sound support with my soundblaster live.
I wonder if FreeBSD has support for kernel modules so a recompile isn't necessary just to use sound. If not then an easy to use FreeBSD distro would be hard to make.
Strange, I have had a great deal of unhindered success with Gentoo. Strange I say about devfs, that has been flawless. Devfs is a damned good and correct idea, only see entries for which there is usable hardware, and don't waste inodes on dev entries. Very cool.
Now/boot, if you will look, by default of noauto in the fstab options. For most efficient use of space,/boot is ext2 (at least for me), and/boot needs to be mounted only at kernel changes, so it makes sense. Sorry if I insult your intelligence, but that is the unique thing about typical/boot in Gentoo.
I'm surprised fewer Slashdotters even looked at Ark Linux' web site before engaging in the usual jabber about easy-to-use Linux distros. The site itself has no info about the distro. But if you go to the FAQ, lo and behold you get a bunch of references to "goatsex", an ASCII art dirty picture, a few hundred lines repeating the word "faggot", and a dirty story. Not much about Linux though.
I suspect the whole thing is a put-on by somebody with a serious lack of taste. Certainly if it were a real distro, and this stuff were hacked onto the site, then its server security wouldn't be adequate.
What you've seen is the result of someone from the IP 68.13.232.26 (ip-68-13-232-26.ok.ok.cox.net), located in Atlanta, using Phoneix 0.5 on Windows 2000, abusing the fact that we've tried using an open support system.
The idea behind the system was simple - anyone can ask questions, and anyone can reply. Pretty much like a Wiki. So in a way, we got hacked - but since there was no protection for this area of the website, I wouldn't call it a security problem in the distribution. And we learned from it - the support system is now censoring bad posts. I find it sad that these things are necessary, when ideologically, we'd much rather fight censorship.
We made one mistake - namely that we trusted people wouldn't abuse it. This guy used malicious HTML tags to redirect the support system to his crap site.
Dear "hacker", you can be proud - you just circumvented nonexistant security blocks! I'll vote for you at the l33t h4x0r of the month contest.
I've found Lycoris to be the absolute easiest Linux distro to use. Not only does it have a huge user support base, but even support for those of us wanting to try it and learn it before switching over from Windows, using VMWare.
While the installation is about as simple as it gets, the low click install is lost as Lycoris fires up a game of Solitaire for you play while waiting for it to complete.
I don't know why the hell I'm even typing this -- it's doomed to get buried in the avalanche of other posts and never get seen...
I have installed and am using Red Hat 8.0. It was
very easy to install, and is very easy to use. It was more than four mouse clicks, but their installer is really smooth. If you just want a quick, easy install you click one of four or five radio buttons for the type of machine you want (server, workstation, etc.); but you can also switch all the packages independently.
The OS seems to come pretty well secured by default; and of course there are the requisite "control panel" windows (which actually work!) and OpenOffice links. The terminal is buried waaaay down in the menu structure -- a bit disconcerting at first. But most of the little apps that we all know and love are available in the pull-up red hat menu already.
They've eliminated the notion of window manager from the basic graphic OS install -- there's no reference to it at all. You can switch window managers, but you have to know unix to do it. That, IMHO, is a Good decision.
It seems to me that Ark Linux is a tempest in a teapot -- Red Hat already did all the work; these guys can't possibly be doing much more than smoothing over the install and putting a different theme on the desktop.
Why do I need 50 editors? Is that emacs thing good? Should I use vi?
User administration? What the ****? This is _my_ computer, I don't want a *****ing login! I don't need that in Windows, therefore it's easier to use!
Why can't I edit the menus?
And that's just 4 of the comments you _will_ get.
Also, if you look at the technology behind it, you'll notice it's suboptimal.
Red Hat 8 is an interesting idea, but I think they got it pretty wrong, and made the OS worse by stripping it of some of the best tools and leaving in vastly inferior ones.
User interface wise, Ark Linux does a lot more than putting another theme on top of Red Hat 8 - we've replaced all their UI stuff.
Try it out for yourself and you'll see the difference.
Not really. The only reason why certain versions of windows crash so much is that MS took a massive amount of kludge consisting of bad, unchecked buggy code and stapled a pretty interface on it to make it look nice. Underneath the hood, Windows is a mess.
I'm not saying Linux is any better.. just pointing out the fact that ease of use does not imply instability. Linux obviously has similar shortcomings to that of Windows, but this can be avoided if they work from the ground up and organize things better.
On a side note: Linux's directory structure seems to be quite sprawling (/lib,/usr/lib,/var/lib... etc). Is there a good reason for this, or is it simply done that way for the sake of tradition?
my rule of thumb about linux/unix dirs./lib - libs for binaries for single user mode (ones in/bin)./usr/lib - libs for binaries for multi-user mode (ones in/usr/bin)./var/lib - data files for various applications.
That's just not true. You're confusing the paradigm with the method of implementing it. What's so wrong with having a coherent system of sharing system resources such as a clipboard, fonts, UI widgets, etc? Nothing. A computer should be easy to use. That may, of course, include "easy to customize."
The fact that Microsoft has been alternately lazy and incoherent in implementing its model of "ease of use" shouldn't sway other developers from trying to accomplish the same goal in a more effective manner.
Hell, I've installed a number of Linux distributions. I've tweaked them. I've fiddled with them. But not one has even approached the ease with whick I can accomplish tasks using Win2k. Perhaps on a technical level, Linux is more stable, more customizable, and more secure than Windows, and certainly the open-source ideal is admirable, but when you consider task-based computing where the main focus is on getting work done (which is all that matters to most end users), the mishmash of current Linux builds is just a pain.
I am NOT trashing Linux. It's an amazing accomplishment, and the improvements in UI and functionality (both at the command line and in the UI) over the last few years are encouraging. But there's work to be done.
Perhaps on a technical level, Linux is more stable, more customizable, and more secure than Windows, and certainly the open-source ideal is admirable, but when you consider task-based computing where the main focus is on getting work done (which is all that matters to most end users), the mishmash of current Linux builds is just a pain.
I agree with you here. Technically, Linux has owned Windows for years but we're only now beginning to make inroads that target end-users.
the mishmash of current Linux builds is just a pain.
This is the point I want to discuss. While today, the different distros are probably confusing to and alienating potential end-users (due to their task-based nature), tomorrow those "confusing distros" could become "viable inter-operating alternatives". That means competition and competition means jobs because multiple companies are able to capture niches of the market. I like Red Hat 8.0 for its easy install and slick GUI. I like SuSe for its easy install and snappier GUI. I like Debian for its packages and I like FreeBSD for its security. See, each one of these distros fills a niche. They scratch an itch for each individual customer.
What we have to work hardest on is overcoming the real barrier-to-entry: mindshare. Microsoft has ruled the roost for so long now that most people don't even know they have options and the non-geek people that have heard of Linux think it's a "hacker's" OS.
You're right. We've still got a lot of work to do, but it's not just writing code...it's changing minds. And you don't have to be a code-hacker to educate people about their options.
It was still as stable as Linux, but it was like running last month's version of Debian/Unstable. The packages where cutting edge (unstable), but still old.
What makes Debian so great (inspite of it being a bitch to configure) is the packaging system. I can always get a package and my server (Debian/Stable) has never had a single package conflict.
If all rpm-based distros would band togethor to share packages and package servers they might be able to shake a stick at Debian. If they can pull it off.
once it's as easy as windows it's going to crash like windows too.
No, because we're trying not to repeat the design flaws of Windows.
Some of the things that make Windows unstable, and what we're doing:
In Windows 95, 98 and ME (the most unstable ones of the bunch), every user can overwrite/delete system files. Installing application X overwrites the DLL application Y installed, causing application Y to be unstable. soname versioning is a vital part of all Unix-like OSes, and we're definitely keeping that, avoiding the windows DLL mess. As for overwriting/deleting system files, it's a security vs. usability tradeoff, and I think we've found a good compromise: The system runs as a normal non-root user with special privileges (via pam) to run package installation tools and some system config tools as root without being prompted for a password.
All drivers etc. run in kernel space, frequently causing a badly written driver to crash the entire system. Ark Linux uses the same drivers as any other Linux out there - so we won't lose the stability. The biggest part of the graphics drivers etc. is in userland, so it can't crash the OS
Windows is not open source, therefore its code does not get any peer review. Ark Linux is, and will always be, Free Software. In fact, it's a not-for-profit community project.
Yes, the announcements on slashdot, osnews, ofb.biz and pclinuxonline came as a total surprise to us.
We aren't 100% ready for the user base we're trying to address yet (there are a couple of installer bugs left, and we're lacking a good internet access config tool -- that's why it's called an alpha), so we tried to remain low profile [and didn't put much effort into the website] -- but now that we've been taken to the public, there's not much of a point in continuing along those lines.
There's also not much of a point in putting up screenshots if you know the look will change before you intend to tell the public.
Who knows, maybe we'll find some new contributors (maybe even for website design and graphics?;) )
This is true; an easy to use desktop is like a scope with the volts per division and timebase fixed so people don't have to worry about learning what they are.
The funny part is that if you want to make Linux easy to use, creating yet another distribution is not the right way to go. This has been attempted a lot of times, and it usually just makes it harder to do anything except a few basic things the distro creators want you to do
Debian is Debian, RedHat is RedHat, Slackware is Slackware, 5 is 5...
However, Debian is NOT Linux, RedHat is NOT Linux, Slackware is NOT Linux, 5 is NOT Linux...
There is not a single distro out there that is Linux. There are however, many *nix distributinos that use the Linux kernel. But they are not by definition Linux, but rather distributions of Linux. Linux is a kernel. Distributions are packaged chunks of software that form a complete operating system including a kernel and tools to make the system useful.
Linux isn't an OS. That's the point. If you say Yoper is just Linux, then it's way too incomplete to be used, because Linux is just a kernel. Everything else in a Linux distribution is a separate piece, even the libc libraries.
To have a successful Unix OS for the masses, you need a developer community that places serious value on providing the end-user with a quality user experience.
Mac OSX has succeeded where linux has failed because it really has an entirely different developer culture than linux (or really, most of the unix developer community, for that matter).
Linux has an uphill battle because for so long its developer community has called end users stupid and has refused to believe that its interfaces have serious usability problems.
The point of my subject is that the mac developer community is really so radically different from the traditionalist unix developers in their belief system that you can't call Mac OSX a unix desktop; it's really "a desktop that just happens to use unix". Which is a good thing.
It's also a bad thing, as soon as even one of those user oriented developers determines that an advanced option isn't worth the confusion to the user or for some very wrong reason decides that user friendliness in some way is equal or even more important than functionality it becomes a bad thing.
Remember unix based systems are more stable, secure, and powerful precisely because the developers do not take this view. I don't mind clicking an advanced button on the interface, or switch to advanced mode. But in windows and even worse macOS (worse because there generally is no advanced button) the advanced button hides basic settings that you should not be using the program if you do not understand rather than advanced options. The advanced options are simply not provided except in extremely expensive "professional" packages and often not even then.
Just because you break something down into stages doesn't mean that any of those stages won't have serious usability design flaws.
For example, the Red Hat installer has a wizard, but they still do this confusing, ambiguous, and non-standard hierarchical radio button thingy in one section. Mandrake installer has a wizard, but they still denote progress through the wizard by whether a button that denotes each stage is green or red (which kind of sucks for people with red/green color blindness).
I'm not saying wizards, when properly implemented, are a bad idea, but they are by no means a panacea for the stuff that really confuses end users. While many linux distros now have graphical stuff, the graphical stuff is still designed by programmers without any design sense at all. And this is really why so many people still find linux so hard to use.
1. stop basing them on red hat. If red hat is so great, use red hat.
Red Hat has a great base system, but is not the best distribution over all. Red Hat Linux is, in my opinion, a lot like a rusty old car with a great engine (think glibc, gcc,...) in it. So what would you do in that situation? Right, you'd salvage the engine and put it into a better car. That's what we're doing.
2. Gime me some actual reason why this is better than red hat, mandrake, suse, conectiva or the red-hat flavour of the week.
It's not better, it's for a different type of users. If you intend to set up a server, don't use Ark Linux.
Ark Linux is an operating system for home users - something the distributions you've listed have failed to achieve (it's VERY hard, if not impossible, to write an OS that's a good server, a good home OS, and a good corporate workstation - there are totally different needs for those 3 areas. Our philosophy is to pick one, and do that one well, instead of doing all 3, and doing them just ok.
Four years and half too late. (Score:4, Funny)
Re:Four years and half too late. (Score:5, Insightful)
Re:Four years and half too late. (Score:2)
Re:Four years and half too late. (Score:4, Insightful)
From the web page: TODO: Insert Ark Linux screenshot here
Click on the 'Why Linux?' link, and look at the end of the third paragraph.
Re:Four years and half too late. (Score:5, Informative)
In fact, we didn't want to go public yet - we don't like preannouncing vaporware (though the current alpha version is actually usable, if you can overlook the rough edges) -- so there wasn't much need to update the website etc. until someone decided to notify the media of our existance.
The 4 1/2 years reference from the original poster refers to the creation of the Mandrake distribution, which started with similar goals, but developed into a different direction.
Mandrake is a good distribution, and suitable for desktop use if you know what you're doing a bit - but Mandrake is an all-purpose operating system, and therefore too complicated for many newbies (this may be hard for us geeks to understand, but people do get confused at the notion of 50 editors, 4 desktop environments and 20 MP3 players).
Ark Linux will focus on being a home user OS, and just that.
IMO, if you want to please too many totally different needs at the same time (and "server", "techie workstation" and "newbie home user" are 3 _very_ different needs), you have to make too many tradeoffs.
Re:Four years and half too late. (Score:2)
What Linux needs is more announcements of distributions "focussed on ease of use for desktop users." There's no need to actually ship anything -- just making announcements seems to be sufficient to get a Slashdot story. And you can spend hours posting here, rather than staring blankly at ark.c in Pico!
This is how Mandrake started. (Score:5, Interesting)
Hmm... (Score:2)
Basically, Linux isn't THAT hard to use for real Newbies (people with very little Windows experience, but aren't afraid to learn), but it is hard to use for many because it is *unfamiliar.* Having many different easy-to-use versions may be a problem in exacerbating that problem. Lets face it, Windows is easy to use mostly because since Microsoft has a monopoly, it is what most people are familiar with.
Linux IS hard to use. (Score:2)
There are some reasons that Linux is hard to use that have nothing to do with being different than Windows.
For one thing, Linux lacks a consistent definition of how a GUI should work, so you don't get the kind of consistency you get with, e.g., MacOS. For instance, some of my apps use the middle-mouse-button method for copying to the clipboard, while some use the Win/Mac method. Some try to implement both, and fail. Inconsistency is probably unavoidable considering the open-source development model.
Another problem is shared libraries. Shared libraries have some real advantages, but they create lots of problems with installing software. How are you going to explain to a naive user that they can't run a certain app because it requires libfoo 2.0, and they have libfoo 3.0 installed?
Re:Linux IS hard to use. (Score:2)
Then there's no making sure someone has the right version of anything installed.
Don't give me some excuse. Most Mac software I've seen is like this. If there's some technical reason you can't do that, FIX IT. This is OpenSource right? You have no-one to blame but yourself, just fix it so you can just put the app in a folder on any machine no matter what.
It's NOT disk space, it's memory footprint, silly! (Score:2)
Next, factor in cache trashing -- that shared malloc() code could've occupied couple lines in a corner of your instruction cache, now you need to flush/reload them every time application context is switched.
Paul B.
Re:Linux IS hard to use. (Score:2, Insightful)
Yea, maybe 10 years ago conserving disk space mattered...
Besides size, the other advantage of shared libraries is that if there's a bug in the library, you fix it, and the fix automatically propagates throughout all your apps.
Rather than asking developers of end-user apps to stop using shared libraries, I think it might be more on target to ask developers of libraries to stop changing their APIs. If they never change the APIs, then they never have to break binary compatibility.
The problem is that this is fairly easy to do with C, but requires a heck of a lot of discipline with C++. In C++, almost any change you make to a class that's exposed in your API will break binary compatibility.
I suspect that a lot of application developers could easily build monolithically linked binaries, but they don't because they know it would increase the load on their servers by an order of magnitude, which would cost them money.
It's also worth noting that this whole thing is basically only a problem with C. In Perl, there is really no such problem, because you don't get an incompatibility as a side-effect of changing a class.
Used to produce BeroLinux (Score:5, Informative)
counterproductive (Score:5, Interesting)
More choice isn't always good. I would rather see ONE good desktop linux package than ten substandard ones.
Edison's Approach (Score:5, Interesting)
99% perspiration, 1% inspiration.
Are there 99 not-for-the-masses widely distro'd linux's yet?
Re:Edison's Approach (Score:3, Funny)
Tesla: Had Edison thought out his work and spent more time in preparation, he would not sweat so much.
Re:counterproductive (Score:5, Insightful)
Why don't we just get all of the soft-drink manufacturers to get together to make ONE good drink to rule them all. While we're at it when are the clothing companies going to get together and mass market grey jump suits so we can really move into the twenty-first century?
Re:counterproductive (Score:4, Insightful)
Customers have no trouble learning a new interface. They do it all the time.
Re:counterproductive (Score:4, Interesting)
Amen to that brother. That statement reminds me of a "debate" (actually more of an argument) I had with my best mate a few years ago, a die hard Mac head. We were discussing interface consitancy, with Paul attempting to show that Linux would never go anywhere because it had 2 major widget toolkits that had different themes, no universal HIG etc. He ended up trying to prove that people couldn't use other peoples mobile phones.
The hilarity that ensued kept us amused for a looong time. The sight of him walking around the 6th form centre with his phone asking people to open up the address book on it, then looking dismayed when after studying it a moment they did it as fast as he would have done, was excellent. The "what are you smoking, man" looks he got were even better.
People are smart, people are adaptable. Yes, a completely consistant UI is good for user efficiency which is why people are trying hard with unified themes and user interface guidelines etc, but at the end of the day with a few notable exceptions (software installation/resolution switching) it's been possible to figure how to work Linux for a long time now and anyway no other OS has a completely consistant GUI, least of all Windows (not even the mac).
Re:counterproductive (Score:4, Funny)
Will do!
Now, let me know when you find a can/bottle that can play counter-Strike, fetch my email, format papers for printing, browse slashdot, etc.
Yeah, sure. (Score:2)
Re:counterproductive (Score:2)
Further, one company would be easier for (MSFT|SUN|APPLE|$Proprietary_Software_Production_
Basically if you want a unified easy to use distribution, you're better off making your own (non-commercial distribution), and hoping the commercial distributions standardize off you.
Re:counterproductive (Score:3, Informative)
One of the biggest draws of Linux and the Open Source movement in general is that there is so much choice. Not only the freedom to choose a specific distribution, but to create a new one if you feel there is an unfilled niche. Combining distros cannot work, so these people felt that they could fill a niche by creating their own distribution.
Re:counterproductive (Score:5, Interesting)
Re:counterproductive (Score:2)
Windows Domains... (Score:2, Interesting)
My users get mapped drives and shares when they login to the domain. I need my users to have centralized information stored on a Windows-based server. The software we use have few analogs in the open source world, so our backend is likely to remain the same for the forseeable future.
All of this needs to be seamless, and it needs to happen reliably. I haven't seen that yet. The pieces parts are there, certainly, but no one has gotten it right yet.
My guess is, this is EXEACTLY what Red Hat is planning to do with their next Workstation version.
Non castrated RedHat... (Score:3, Interesting)
Take the good base of RedHat and replace their castrated Desktop with a working one. Since Bero is responsible for this it might even turn out pretty good...
dev0
Re:Non castrated RedHat... (Score:2)
Re:Non castrated RedHat... (Score:3, Interesting)
More fragmentation (Score:5, Insightful)
And also, who has not seen this "make linux easier for the desktop" thing before? There are about a million and one distros who purport to do this. Why does everyone feel the need to reinvent the wheel?
Re:More fragmentation (Score:5, Insightful)
There are about a million and one distros who purport to do this. Why does everyone feel the need to reinvent the wheel?
Though I can't speak for the producers of ArkLinux, I will speak as someone who has been involved with them early on.
I think one of the prime motivations is to act as a 'concept' vehicle to offer innovations to the larger distributions. In Detroit, it's often difficult to get the Big Three automakers to really cut new ground and try something risky. However, they love to put a bunch of wacky ideas into concept cars, and then slowly evolve their best ideas into real products for the street.
The well-known distros like Red Hat, SuSE, Mandrake and Debian are analagous to the Big Three automakers in the Linux distribution space. However, they're not in a high-markup tangible goods market like Chevrolet, GM and Ford. Thus, they can't afford to make their own 'concept' tools and services to help their own evolution.
I see all of the smaller distros as helping the evolution by giving each great idea (and tons of mediocre and bad ideas) a public forum in which they can prove themselves, and be cherry-picked by the powerhouses of Linux adoption: the Big Distros.
And frankly, Detroit isn't being robbed of available talent whenever some kid puts together a supercharged Dodge Charger with neon all over it. Likewise, Red Hat doesn't sweat it if some afficianados take a different path and try a few new things.
Hurry! (Score:5, Funny)
Re:Hurry! (Score:5, Funny)
Re:Hurry! (Score:2)
Re:Hurry! (Score:2)
How about Bison and YACC?
how to make linux desktop good for masses (Score:5, Interesting)
it seems to me that the only way this will be achieved is by a distro making some very draconian decisions about what to include. I for one would be delighted to see a very lightweight distro that did just a few things very well.
what I mean is if there was a distro that included a web browser, mail client, open office, mp3 player & a terminal program, and they all looked good and were consistent in look, feel and function, then this would be a major step forward. you dont need to include much more than that to have a successful easy to use distro. no point in including tons of compilers and dev tools if you want to make something easy for non-techies to use.
hopefully this is a goal that will be reached someday. it is so possible to do that I'm really surprised to see so many failed attempts. All I'm suggesting is a lightweight distro with a select few apps that are heavily polished and work well. If you did that you could gain a following and maybe get enough credibility and respect that other app developers would code to your human interface guidelines.
just a thought.
Re:how to make linux desktop good for masses (Score:2)
With the exception of Open Office, you've just described Mac OS X.
And if only these open source developers would stop with the "distribution of the week" approach and direct their efforts towards something like a simple GNUstep [gnustep.org] install, they'd actually have a shot of getting some OS X software brought over to Linux and FreeBSD.
I couldn't agree more (Score:2)
GNUstep does not get enough cred. If you had a distro running GNUstep w/ a few polished apps you'd open yourself up to developers releasing apps for both your distro and OS X at the same time. On top of that you'd have a Linux desktop capable of running any other Linux app.
Oh wait, I just described OS X again.
Re:how to make linux desktop good for masses (Score:2)
They just released a new browser for Jaguar, Safari [apple.com], based on Konqueror.
Re:how to make linux desktop good for masses (Score:2)
Re:how to make linux desktop good for masses (Score:3, Informative)
"Microsoft does XXX" is neither "XXX is good" nor "XXX is bad".
Ark Linux aims at providing everything the average home user needs - the desktop, office suite, media player, net access tools, a couple of games, etc.
For developers, there's the Ark Development Suite, a collection of development tools and programming languages (about 200 MB) - comparable to Visual Studio (with obvious differences - e.g. we support additional languages like Python, Objective-C, Perl,
Of course, Ark Development Suite is 100% free too - the only reason it isn't included in the base OS is that 95% of the targeted users won't need it.
Linux, BSD, and everything need one thing.... (Score:4, Interesting)
Does this take out everything we love? Yes. I'll gladly stick to FreeBSD with my rxvt's and my gettys, my kernel compilation and my make worlds, thank you very much. The beauty of Windows is that it takes all of the fun out. But what we consider fun others consider a PITA.
If a distribution did this, took everything out, made it all LKM's, and took all the technicalities out, then we would have a system for everyday users. They don't want the command line. Come on, some of these people can only do 5 wpm! When starting up your web browser because you don't know about tab-completion and you can't type takes 15 seconds before the program is even executed, you have a major problem there. It should be click(or, optionally, click-click) and it should run.
This needs a team of dedicated people to run a free system like this, but more likely, a team of employees from a company will do this(and many are trying to) and to maintain binary packages, an easy system to install/update/delete packages. That's what we need.
What is wrong with Red Hat and Mandrake and the others? Simple. Whenever they do something to make everything more simplistic, the community(read: many of us slashdotters) makes a backlash saying the system lacks this or that functionality because they did this or that. I say let them. We always have Debian, Slackware, and other distributions. Let a few of these become end-user-never-going-to-type-in-a-command-in-thei
distributions and leave it at that. Applaud them for taking out even the ability to re-compile the kernel. Applaud them for it.
Yes, many of these companies are doing idiotic things, but we're part of the community and they look to us for guidance. Let's tell them, not yell at them, to take apache out of the main distribution and for god sakes disable it for regular desktop users.
Desktops should be so easy an 8 year old should be able to set up. Servers should be difficult to set up.
Sorry, I had to rant. and yes, I know it's the same rant everyone else on slashdot says, but this time I get to say it.
Re:Linux, BSD, and everything need one thing.... (Score:2)
Well, to tell you the truth, I am a Slackware user because I gave up on the distros that PRETEND to be user friendly, so why not just go for the real thing and get my hands dirty, but at least I know what the hell is going on. And I have yet to see a distro that is so user-friendly both in the interface as in configuration and application installation as BeOS -still- is.
Re:Linux, BSD, and everything need one thing.... (Score:2)
Configuration is a big baddie because you only have to do it once. And the more experienced users are ok with a more difficult install.
Re:Linux, BSD, and everything need one thing.... (Score:2)
I don't know about you, but I feel lindows has strayed far from the community and only attempts to provide a semblence of support for it. Right down to providing source code they've fought what community is about tooth and nail. The CEO isn't interested in the community, he's intrested in exploiting it.
Re:Linux, BSD, and everything need one thing.... (Score:3, Insightful)
That would not be beneficial to anyone but the Windows-haters. Linux-lovers should not jump at something like this.
Re:Linux, BSD, and everything need one thing.... (Score:2)
gnomeprm/"free" software (Score:2)
Coming from a mostly never ever touched a cli since the 80's on dos which I detested as just stoopid, I found (still find on 7.2) redhat to be quite easy to use as long as I stick to redhat brand rpms and not generic rpms, that one threw me for a loop "why" that was when the "non standard" rpm happened to me. Seems as a bit of courtesy that whomever took the label "rpm" and made an app that was labeled that would have followed the originator's standard in what file went where. If they didn't like iot they should have called it something else. To me that's like buying a non oem aftermarket alternator that is labeled "belchfire" but it won't bolt onto a belchfire. Sorta kinda dumb and rude there.
Free as in speech and in beer doesn't necessarily make the speech outstanding oratory or the brew outstanding beverage. So I agree, quality, consistency, ease of use. A few dozen apps that work well, not a few hundred that are so-so and require weeks of code wrangling. The main bitching I see here is "oh no, can't lose our elite tweaking ability". Who says you have to? Duh, any distro can add in the dang shell and console, but not being absolutely required to use it makes it or breaks it on "the masses" useability scale.
Another really off the wall side issue. Apps included in an officially released boxed set distro, and what you pay for the distro, there should be micropayments to the app maker guys to go with it, if they are releasing as begware. that will make what's included or not be a lot easier to sort out. If the apps aren't begware, and they are toting 100% of the note and don't care, swell, their bandwith and coding time. I don't mind begware and would like to support them (the apps I use and find useful obviously), but sending one dollar via online payment schedule to a zillion app orgs is just silly. And I think the clone copiers and resellers should send a buck per cd to the distros they are cloning. And here's another kicker sure to be controversial. the gpl says it has to be "free", swell, the information is free, the bandwith isn't. People should help pay the bandwith and associated costs for downloading ISO's. That bandwith is not the "source code" that is free, you still get the data, but those servers take wires and electricity and buildings, which costs money. It's just as much a real expense as it is to make cd copies and ship them, albeit less money, it adds up. Call it a "shipping and handling" charge for the "free" stuff you are receiving.
Re:Linux, BSD, and everything need one thing.... (Score:3, Redundant)
The entire idea of linux for the masses is linux for everyone. The only way to do this is to take out the difficultires of it. I think if a linux box has a web browser, a media player, a GUI, an e-mail client, and some games, it's reached the usefulness of most people. Not the die-hard gamers, not by a long shot, but by many people. What we consider useful is something that is considered pointless by many. That is why I think it should be so easy to install and use that people who want terminals won't mess with it. It should hide everything about it that is linux, and just be a functional, enjoyable, stable, useful computer. With all the free software we have available, with binary packages that are easy to install, if the OS was easy to install and get up and running, I think it could be a contender for world domination. Damned if I wouldn't be handing out the CDs at the mall, or leaving some for grabs at the local coffee shop.
But the gap is what I want. We may call ourselves 'users' but most of us and what we do are development. For those who aren't, many do not need the command line at all. Windows was doing a good thing hiding it, the problem is they wanted to eliminate it when some of us needed it. It should be a download and a few mouse clicks away to install rxvt, but it doesn't need to be base install for those who don't need it.
And you are right. Although I wouldn't say "Windows-Haters." We want the Windows-Lovers and the Windows-Neutral people also. Linux Lovers probably shouldn't use it, but support would be good if they got Linux running on more and more machines. Especially if it means games would come out for the Linux platform. Lord knows I'd be happy with a Linux Morrowind. =)
The whole idea of choices leave linux-lovers choosing debian, or brewing their own, but this would make Linux more mainstream and on the desktop, but the way free software works assures us that it would not be the only choice. So linux lovers should support it, they can always have what they want. But most people just want a web-browser with flash, an e-mail client, a word processor, and solitaire. Not something that takes a book to install, several hours to configure, and a day or two to go to cnn.com from trying to understand everything. It would take many of us less than an hour to get to that point, but someone who thinks a computer should just handle most of that stuff itself(which, look all around us, it _should_), will get upset and switch.
Re:Linux, BSD, and everything need one thing.... (Score:2)
Re:Linux, BSD, and everything need one thing.... (Score:2)
Unfortunately, having it compile when you right click and select install would be very visible to the user. Several minutes visible, and many people would get tired of waiting. Just imagine OpenOffice compiling on the system, while you wait, after downloading that ungodly large file already.
I don't see a problem with platform specific binaries. But yes, you need someone on the backend to compile those and make them available, that is why I suggested a team of dedicated people or a team of employees to handle that.
Re:Linux, BSD, and everything need one thing.... (Score:2)
OpenOffice doesn't really take that long to download with a broadband connection, and I wouldn't recommend downloading anything larger than a few megs with a dialup connection. Just have your friend with cable burn you a CD-RW with what you want downloaded. It takes less time than downloading it at 2-5KB/s, and it's an excuse to get away from the computer and hang out with your friends for a while. And that is what a lot of people that this is aimed at would be interested in, much moreso than many geeks.
Re:Linux, BSD, and everything need one thing.... (Score:2)
Desktops should be so easy an 8 year old should be able to set up.
This is why I think that some Debian projects like Debian Jr. [debian.org] ("Debian for children from 1 to 99") are particularly interesting. I don't care about kids programs, but I think they will face some UI design issues that will help make Debian easier for all (non-guru) users.
What? (Score:2, Insightful)
Noah's List (Score:3, Funny)
hey I could not resist.. I doubt if Mad Dog reads this
The Easiest to Use Linux Distros... (Score:2)
Aw, darn, a few months too late. (Score:2, Informative)
Enough! I say (Score:5, Insightful)
Enough, that is, of distributions that are "for the masses". It should be clear to everyone by now that this phrase is utterly meaningless, since it encompasses a huge number of possible approaches to the problem of making lusers happy with Unix. I propose that this phrase and all similarly generic phrases be officially declared Fucking Useless, and anyone who uses them be savagely beaten until they come up with a particular differentiating feature for their distribution.
So what is special about the distro of the week? Hardware autodetection? Careful customization of packages to provide a uniform and sensible default UI? Good paper documentation?
Oh, Jesus, if I just stop there, someone will moderate this up. Do you people realize how pathetic you are, that you're reading this? Writing it was bad enough (shame, shame, shame!), but reading it... can't get read again. Come on, eat me! Burn, karma, burn! SLASHDTO DEITORS SUX0000RZ1!!1! Bibbity bibbity bibbity!
More to the point... (Score:2)
It's not only a useless phrase, it's actually actively harmful. It leads people to believe that it's a possible or desirable thing.
This guy said it best:
Re:More to the point... (Score:2)
easy to learn != easy to use (Score:3, Insightful)
Intuitiveness == Familiarity
Just my 2 cents.
Re:easy to learn != easy to use (Score:4, Insightful)
ask 100 windows users how to
- change network settings
- see what programs are running, and system resources used
- add/remove programs at start up
99 of them will give you blank stares and tell you they have no idea. the problem is not that windows is "easy" because it isn't. it is what people know, and there's a HUGE difference. linux doesn't need to be easier, because it is more than enough. no, i'm not talking about installation. we just need to get the camel's nose under the tent. most people who are familiar with computers and are not scared of them will be able to pick up quickly how to do things in linux.C:\My Documents =
other than that, mozilla, OO, etc., the apps are more than sufficient.
Comment removed (Score:5, Interesting)
Re:Knoppix is the easiest to use. (Score:3, Funny)
what specifically makes it different? (Score:2)
But neither the article or the original project site do not disclose by which means the new distro will make Linux easy to use comparing to other linux distros.
Is it just new set of very well debugged installation programs very well integrated with GUI based wizards? If so, what's the difference with Mandrake, Lindows, SCO and Redhat?
Or it integrates some new end-user applications (like ximian evolution)?
Especially I wonder, what is done or is going to be done to improve the user experience from usability prospective? For example, what specifically is done to explain the user what makes this distro so easy to use?
<mass-mode on> (Score:5, Informative)
- 4 clicks to install = good
- giving simplification a shot = good
- developed own installer (i guess) = good
- no info about what to do with an iso = bad
- explained in terms of red hat = bad
- unanswered questions on page = bad
- needs:
screenshots
info on hardware requirements
info on supported locales (or is it just English)
info for developers on "why develop for distro x"
needs a "why use our distro" page for users
- might be interesting if you could do work just by popping CD in (without partitioning or doing a big install).. or is that what it does? dunno.
Re: (Score:3, Informative)
Our highest focus atm is the codebase, the distro itself. That's why the website is still a bit "poor".
I hope we will be able to work on it soon... just remember that Ark Linux is an "open" project, therefore everybody is free to contribute anything.
I really hope to get more people involved as time goes on. Let's see
Arc Linux? (Score:2)
That's some seriously awsome support. (Score:2, Informative)
Nice idea, but... (Score:4, Interesting)
I wish these guys the best of luck, anything that brings Linux to more people is a good thing, but...
IMHO all attempts to make computers 'easy to use' 'for the masses' have failed. Just ask any joe public who wants to actually use their computer, rather than spend endless hours tinkering with it or being plauged by niggling bugs. Computers are extremely complex things, and the critical mass of knowledge required to make them run smootly makes it impractical for every user to become a system administrator.
I think what's needed is a shift in perception and the model we use to sell hardware and software. Rather than just selling boxes (containing PCBs or CDROMs), computers should be rented. This would mitigate the endless upgrade cycle and (with the internet) allow trained professionals to administer customers' boxes, all included in the price. Linux is an ideal OS for such a distribution method.
Do I think this will happen? Not yet. For this model to be economically viable it would require specialist tools that would let admins look after huge amounts of boxes. A major shift in public perception would be required, especially after all the 'so easy to use even your granny can do it' ads. Finally, I think that it would take a lot for users to hand over the control of thier computer to anyone.
Easy to use gentoo or Freebsd?? (Score:2)
I love the gentoo package managment. It is better then pure sex and yes even apt-get.
I found the distro too bleeding edge in some area's *cough devfs * cough and it takes a whole weekend to finally configure everything. The latest version was so buggy that I could not get my
But its lightning fast and more bsd like in which all the actual config files go in
Freebsd rocks too in which sysinstall configures things pretty good but it takes alot of tweaking to get the following Xfree86 to work with my geforce4, the beta nvidia drivers, java, and sound support with my soundblaster live.
I wonder if FreeBSD has support for kernel modules so a recompile isn't necessary just to use sound. If not then an easy to use FreeBSD distro would be hard to make.
Re:Easy to use gentoo or Freebsd?? (Score:2)
Now
Ark Linux web site is a farce with pr0n on it (Score:2, Informative)
I suspect the whole thing is a put-on by somebody with a serious lack of taste. Certainly if it were a real distro, and this stuff were hacked onto the site, then its server security wouldn't be adequate.
Re:Ark Linux web site is a farce with pr0n on it (Score:5, Informative)
The idea behind the system was simple - anyone can ask questions, and anyone can reply. Pretty much like a Wiki.
So in a way, we got hacked - but since there was no protection for this area of the website, I wouldn't call it a security problem in the distribution. And we learned from it - the support system is now censoring bad posts. I find it sad that these things are necessary, when ideologically, we'd much rather fight censorship.
We made one mistake - namely that we trusted people wouldn't abuse it. This guy used malicious HTML tags to redirect the support system to his crap site.
Dear "hacker", you can be proud - you just circumvented nonexistant security blocks! I'll vote for you at the l33t h4x0r of the month contest.
Ironically (Score:2)
While the installation is about as simple as it gets, the low click install is lost as Lycoris fires up a game of Solitaire for you play while waiting for it to complete.
Red Hat 8.0 *IS* ``easy for the masses'' (Score:5, Interesting)
I have installed and am using Red Hat 8.0. It was very easy to install, and is very easy to use. It was more than four mouse clicks, but their installer is really smooth. If you just want a quick, easy install you click one of four or five radio buttons for the type of machine you want (server, workstation, etc.); but you can also switch all the packages independently.
The OS seems to come pretty well secured by default; and of course there are the requisite "control panel" windows (which actually work!) and OpenOffice links. The terminal is buried waaaay down in the menu structure -- a bit disconcerting at first. But most of the little apps that we all know and love are available in the pull-up red hat menu already.
They've eliminated the notion of window manager from the basic graphic OS install -- there's no reference to it at all. You can switch window managers, but you have to know unix to do it. That, IMHO, is a Good decision.
It seems to me that Ark Linux is a tempest in a teapot -- Red Hat already did all the work; these guys can't possibly be doing much more than smoothing over the install and putting a different theme on the desktop.
Re:Red Hat 8.0 *IS* ``easy for the masses'' (Score:3, Informative)
Show it to any newbie, and you'll see.
And that's just 4 of the comments you _will_ get.
Also, if you look at the technology behind it, you'll notice it's suboptimal.
Red Hat 8 is an interesting idea, but I think they got it pretty wrong, and made the OS worse by stripping it of some of the best tools and leaving in vastly inferior ones.
User interface wise, Ark Linux does a lot more than putting another theme on top of Red Hat 8 - we've replaced all their UI stuff.
Try it out for yourself and you'll see the difference.
Re:good luck (Score:2, Insightful)
I'm not saying Linux is any better.. just pointing out the fact that ease of use does not imply instability. Linux obviously has similar shortcomings to that of Windows, but this can be avoided if they work from the ground up and organize things better.
On a side note: Linux's directory structure seems to be quite sprawling (/lib,
Re:good luck (Score:3, Informative)
Re:good luck (Score:2)
nothing is cut in stone but that's what people *should* do. you can help enforce that by patching or not using s/w that's not like that.
Re:good luck (Score:5, Insightful)
The fact that Microsoft has been alternately lazy and incoherent in implementing its model of "ease of use" shouldn't sway other developers from trying to accomplish the same goal in a more effective manner.
Hell, I've installed a number of Linux distributions. I've tweaked them. I've fiddled with them. But not one has even approached the ease with whick I can accomplish tasks using Win2k. Perhaps on a technical level, Linux is more stable, more customizable, and more secure than Windows, and certainly the open-source ideal is admirable, but when you consider task-based computing where the main focus is on getting work done (which is all that matters to most end users), the mishmash of current Linux builds is just a pain.
I am NOT trashing Linux. It's an amazing accomplishment, and the improvements in UI and functionality (both at the command line and in the UI) over the last few years are encouraging. But there's work to be done.
Re:good luck (Score:5, Insightful)
I agree with you here. Technically, Linux has owned Windows for years but we're only now beginning to make inroads that target end-users.
This is the point I want to discuss. While today, the different distros are probably confusing to and alienating potential end-users (due to their task-based nature), tomorrow those "confusing distros" could become "viable inter-operating alternatives". That means competition and competition means jobs because multiple companies are able to capture niches of the market. I like Red Hat 8.0 for its easy install and slick GUI. I like SuSe for its easy install and snappier GUI. I like Debian for its packages and I like FreeBSD for its security. See, each one of these distros fills a niche. They scratch an itch for each individual customer.
What we have to work hardest on is overcoming the real barrier-to-entry: mindshare. Microsoft has ruled the roost for so long now that most people don't even know they have options and the non-geek people that have heard of Linux think it's a "hacker's" OS.
You're right. We've still got a lot of work to do, but it's not just writing code...it's changing minds. And you don't have to be a code-hacker to educate people about their options.
--K.
Re:good luck (Score:2)
Ok, I'll forgive you for being a law student (j/k! we need more lawyers like Lessig).
We are indeed in agreement, and thanks for helping fight the mindshare battle.
--K.
Mandrake to Debian (Score:2)
It was still as stable as Linux, but it was like running last month's version of Debian/Unstable. The packages where cutting edge (unstable), but still old.
What makes Debian so great (inspite of it being a bitch to configure) is the packaging system. I can always get a package and my server (Debian/Stable) has never had a single package conflict.
If all rpm-based distros would band togethor to share packages and package servers they might be able to shake a stick at Debian. If they can pull it off.
exactly (Score:2)
Re:good luck (Score:3, Informative)
No, because we're trying not to repeat the design flaws of Windows.
Some of the things that make Windows unstable, and what we're doing:
soname versioning is a vital part of all Unix-like OSes, and we're definitely keeping that, avoiding the windows DLL mess. As for overwriting/deleting system files, it's a security vs. usability tradeoff, and I think we've found a good compromise: The system runs as a normal non-root user with special privileges (via pam) to run package installation tools and some system config tools as root without being prompted for a password.
Re:What? (Score:2, Funny)
Re:What? (Score:5, Informative)
Yes, the announcements on slashdot, osnews, ofb.biz and pclinuxonline came as a total surprise to us.
We aren't 100% ready for the user base we're trying to address yet (there are a couple of installer bugs left, and we're lacking a good internet access config tool -- that's why it's called an alpha), so we tried to remain low profile [and didn't put much effort into the website] -- but now that we've been taken to the public, there's not much of a point in continuing along those lines.
There's also not much of a point in putting up screenshots if you know the look will change before you intend to tell the public.
Who knows, maybe we'll find some new contributors (maybe even for website design and graphics?
Re:What? (Score:2)
Don't worry. I don't anticipate too many people rushing out to change distros any time soon.
Re:You can't.... (Score:2, Insightful)
The funny part is that if you want to make Linux easy to use, creating yet another distribution is not the right way to go. This has been attempted a lot of times, and it usually just makes it harder to do anything except a few basic things the distro creators want you to do
Jason
ProfQuotes [profquotes.com]
Slackware isnt 'linux' either. (Score:3, Insightful)
**everything** else is just based on it.. So slackware is no more, or less, 'Linux' then redhat, or caldera or any other of the bazillion distros..
( well not BSD, VsTA, etc.. but lets stay serious here )
Uhm, no distros are "linux" (Score:2)
RedHat is RedHat,
Slackware is Slackware,
5 is 5
However,
Debian is NOT Linux,
RedHat is NOT Linux,
Slackware is NOT Linux,
5 is NOT Linux
There is not a single distro out there that is Linux. There are however, many *nix distributinos that use the Linux kernel. But they are not by definition Linux, but rather distributions of Linux. Linux is a kernel. Distributions are packaged chunks of software that form a complete operating system including a kernel and tools to make the system useful.
Re:Uhm, no distros are "linux" (Score:2)
Mac OS X is not unix on the desktop. (Score:2)
Mac OSX has succeeded where linux has failed because it really has an entirely different developer culture than linux (or really, most of the unix developer community, for that matter).
Linux has an uphill battle because for so long its developer community has called end users stupid and has refused to believe that its interfaces have serious usability problems.
The point of my subject is that the mac developer community is really so radically different from the traditionalist unix developers in their belief system that you can't call Mac OSX a unix desktop; it's really "a desktop that just happens to use unix". Which is a good thing.
Re:Mac OS X is not unix on the desktop. (Score:3, Insightful)
Remember unix based systems are more stable, secure, and powerful precisely because the developers do not take this view. I don't mind clicking an advanced button on the interface, or switch to advanced mode. But in windows and even worse macOS (worse because there generally is no advanced button) the advanced button hides basic settings that you should not be using the program if you do not understand rather than advanced options. The advanced options are simply not provided except in extremely expensive "professional" packages and often not even then.
Re:How is one going to go about doing this? (Score:2)
Idiot.
Re:Support needs support! (Score:2)
Do not meddle in the affairs of Wizards (Score:3, Interesting)
Just because you break something down into stages doesn't mean that any of those stages won't have serious usability design flaws.
For example, the Red Hat installer has a wizard, but they still do this confusing, ambiguous, and non-standard hierarchical radio button thingy in one section. Mandrake installer has a wizard, but they still denote progress through the wizard by whether a button that denotes each stage is green or red (which kind of sucks for people with red/green color blindness).
I'm not saying wizards, when properly implemented, are a bad idea, but they are by no means a panacea for the stuff that really confuses end users. While many linux distros now have graphical stuff, the graphical stuff is still designed by programmers without any design sense at all. And this is really why so many people still find linux so hard to use.
Re:same bero from RH? (Score:2, Informative)
I originally started working on Ark Linux as a proposal for a Red Hat home user edition - but obviously they didn't care.
Re:Another one? (Score:3, Informative)
Red Hat has a great base system, but is not the best distribution over all.
Red Hat Linux is, in my opinion, a lot like a rusty old car with a great engine (think glibc, gcc,
2. Gime me some actual reason why this is better than red hat, mandrake, suse, conectiva or the red-hat flavour of the week.
It's not better, it's for a different type of users.
If you intend to set up a server, don't use Ark Linux.
Ark Linux is an operating system for home users - something the distributions you've listed have failed to achieve (it's VERY hard, if not impossible, to write an OS that's a good server, a good home OS, and a good corporate workstation - there are totally different needs for those 3 areas. Our philosophy is to pick one, and do that one well, instead of doing all 3, and doing them just ok.