Slashdot Mirror
ESR's Desktop Linux 2008 Deadline
jesboat noted Eric S. Raymond and Rob Landley's essay about what the Linux community must do to achieve dominance entitled "World Domination 201". It says
"Idealism about open formats will not solve our multimedia problem in time; in fact, getting stuck on either belief in the technical superiority of open source or free-software purism guarantees we will lose. The remaining problems aren't technical ones, and none of the interesting patents will expire before the end of 2008. We've got to ship something that works now. If we let this be a blocking issue preventing overall Linux adoption during the transition window, we won't have the userbase to demand changes in the laws to untangle the screwed up patent system, or even prevent it from getting worse. It's a chicken and egg problem, demanding a workaround until a permanent solution can be achieved. We can't set the standards until after we take over the world."
"Idealism about open formats will not solve our multimedia problem in time"
We can have an Open Source Desktop if we just don't make it Open Source! Brilliant!
Why is ESR so hellbent on taking over the world? We have a system that works, and that can play multimedia just fine, albeit illegally, why does it matter how many people use it? I, for one, don't see ESR wanting to take over the world as enough of a reason to cave in and use proprietary technologies...
You just gave the reason why we need more people adopting Linux: what you say is that Linux can play multimedia files just fine, only illegally (I'm assuming you're referring to the proprietary mplayer codecs here). Yet you see no reason to "cave in and use proprietary technologies"? Strange line of thought...
If, on the other hand, a significant number of people used OSS, they would have a lote more weight to lobby software manufacturers for more open-source codecs, native ports of their software to Linux, etc... making using Linux perfectly legal when those codecs are available on your favorite platform.
"A door is what a dog is perpetually on the wrong side of" - Ogden Nash
They like to use history is this essay, but backward compatability is by far the biggest factor in the history of desktop operating system software. This essay hardly mentions it, and not in the context of history. The biggest reason Windows 3.1 won was because of its backward compatability with DOS -- and Microsoft never forgot the lesson. Dos -> Win3.1 -> Win 95 -> Win 98 -> NT 3.1 (sort of) -> Win2000 -> XP -> Vista. Microsoft gives you a relatively smooth glide up the chain so that you don't have to throw away all your existing software -- and hardware. Of course, it's not perfect, but it's sure better than throwing away everything to move to Linux or a Mac.
Sometimes it's best to just let stupid people be stupid.
Linus: Egads, Eric! BRILLIANT! HAHAHAHAHA, Troz!
ESR: Come, Linus, we must prepare for tomorrow night.
Linus: What are we going to do tomorrow night?
ESR: The same thing we do every night, Linus...TRY TO TAKE OVER THE WORLD!
NARF.
The real solution to make Linux more mainstream is to make users more computer litterate.
Good choice of word to misspell. Besides that, "people" in general want to use whatever everyone else is using, they want to use whatever brand name apps they've heard about and most of all, they want it simple. Every company in every line of business wish their customers were better informed and better trained, it's not going to happen. You can teach a monkey new tricks (like that the Intarnets is now the fiery fox, not the blue e) but most people don't want to become "computer literate". Not even the modern kids who MSN all day want to be "computer literate" in the way you think of it.
Want to make inroads:
1. Corporate workstations. That means in particular
a. Exchange replacement
b. Policy management like Active Directory
c. Heavy compatibility work with MS Office
2. Educational facilities
a. Get Linux labs, dual-booting machines
b. Deploy Firefox, OpenOffice etc. as alternatives on all desktops
c. Make sure all internal systems are platform-independent
3. "Family management"
a. More shades between "root" and "user". Waaay too often I get asked for the root password for things I'd like to delegate, but not give away total control. Linux is great when you're either one person or administering a bunch of people that only get approved applications, inbetween is not that great.
b. Security updates that really are without question, so you could set them up to install automatically. I really like apt-get and all, but it annoys me that I don't know if I'll get asked about some config file where the defaults have changed or whtaever.
c. Somewhere to put "common documents" that is somewhat standard and sane. Everybody has their home dir like "My documents", it's not difficult to fix but it's always a custom dir with custom links, don't people have like general data that's shared with all users?
Gamers and people that rely on support lines or local tech shops just aren't cases you'll win. There's so many quirks in changing to Linux, it just gets too expensive to pay for it (and these aren't the people to search online forums). You need someone with Linux sklil in the company, institution or family. To think that any significant share will put in a k/ubuntu CD and install it by themselves, is dreaming.
Live today, because you never know what tomorrow brings
Well... first off, it's got nothing to do with Linux. What we're talking about is a user interface that runs on top of X-windows. As such, it will run comfortably on any flavor of BSD or commercial Unix, and even stranger operating systems.
Second off, we're talking about a vast set of tools. Gnome is nice, KDE is nice, but they're pieces of a larger puzzle that includes X-windowing systems, and all of their assorted tools, drivers, and niceties, window managers, and applications that may or may not be designed to work within the look-and-feel guidelines of anything recognizeable at all. The problem space is way to big for any one person or organization to just decide, "Hey everyone, we're all gonna be doing THIS!"
Open source software grows and evolves as programmers scratch an itch. You can't crack the whip, as the project will just fork as programmers follow whatever their interest is... commercial, educational, political or just for the hell of coding something neat. It would be nice if everyone could assume a role that's perfectly suited for some master-plan to reach some goal... but they won't. Human nature is in the way.
Open Source Software is not a place where a single goal achieved by everyone working in unison is possible. Yes, Linux itself is cool... but how many variants, patches and forks of it are out there? Quite a few... people take what they need, and follow their own interest. This is what open source software is about. Even then, there's more than Linux: there are the three (Four... five?) BSD-based operating systems, and things like SkyOS and Haiku, besides.
In this maelstrom of variation and choice, you want a single standard UI? Not going to happen. What's more, it will likely work against Linux on the desktop rather than for it. Gnome came about because they didn't like KDE, and wanted something with different political and technical goals. KDE came about because the company had a different commercial and technical goal than Motif. Can you imagine how much it would suck if everyone working on KDE and Gnome were forced to work on making a better Motif? We're better off with many projects working for their own ends. Open Source means that the projects cna pick and chose what they like from each other, everyone wins.
Then there's the issue that Gnustep isn't a part of the discussion, despite being an Open Source re-implementation of the UI Apple uses for Mac OS X... so if the best solution isn't going to "win" anyway, it's pointless whining that the third or fourth best solution isn't getting all the attention. (And, as you've figured out, the order from "best" to "worst" won't be the same for everyone... or even a majority.)
In the end, it's up to the commercial distro-makers to decide what works for them, and to pay programmers and project leads and software architects to make it happen. The interface for the OLTP project shows how to get it done, and done on a shoestring budget in a tight time constraint.
### Ease of installation.
/bin, /usr, /var and all those other root-write only directories that have absolutely nothing of valuable data, since it comes straight from the distribution CD and is trivial to recover, if /home/juser/ on the other side says bye-bye and you don't have a recent backup, then you can really be in trouble. On large multi-user installations things look different, but on your average desktop that whole root/user separation doesn't provide much benefit at all. That of course doesn't mean we should get rid of it, but you don't really need much more then a password-less sudo.
I am sorry, but that is just bullshit. Linux has been extremely easy to install for years, it also happens to be a heck of a lot easier to install then Windows and lets not forget we have LiveCDs, so giving Linux a quick try is among the most trivial things you can do. Beside from that, installation is totally overrated, you do it like once in a lifetime and then never ever again, if you have trouble with it, find a friend that helps with it. Installation is a pretty much solved problem, with repartitioning being the only thing that requires some thinking.
The real problem isn't installation, but maintaining an Linux, simply things as installing a piece of software you have seen on a webpage can be extremely hard and time consuming, even for somebody with 10+ years of Linux experience, for your grandma such things are simply totally out of reach. Sure we have apt-get and friends, but those help absolutely nothing if a piece of software isn't in your distribution, which kind of is always the case with new software. Unless that changes and software installation becomes a no-brainer, Linux won't stand much a chance in the mass market.
And speaking about security, that one is totally overrated as well. On a desktop computer there is only one account that matters and that is the one of the user using it, lets call it juser. If root or jusers account is compromised doesn't make a difference, since in *both* cases the intruder has full access to everything that matters anyway. If there is something I really don't care about on my Linux then its
### The real solution to make Linux more mainstream is to make users more computer litterate.
Good luck trying that, it won't work, ever. The simple reason for that is that computers simply don't make sense. You can teach a person math, because math makes sense and is logic, but handling a computers relies in very large part simply on learning the quicks of its broken software, on Linux just the same as everywhere else. So knowledge from 5 years ago can be totally useless today, lots of computer knowledge is already worthless after a year. Computers simply don't make sense and it requires just way to much time for the average person to learn all the quirks and workarounds. The solution to all this is to simply *fix* all those quirks and bugs so that they never ever touch the users desktop. There simply isn't a logical reason why installing a tar.gz requires me to manually track down dependencies, why there is no undelete and why changing the mouse speed requires editing Xorg.conf while changing mouse acceleration does not, its just bugs and history that made the software the way it is today, there is no logical design principle behind all this. Simply fix it and don't try to teach the user why your software is broken and how to work around it, just a waste of time.
The essay is amusing. Not for its content, but for its format. It starts out with a revision history of all things. Only a dweeb would put that at the beginning of an essay intended for public consumption.
Then there's the focus on "64 bit". Microsoft and Apple both had 64-bit operating systems, then backed off. (It was surprising that when Apple went from PowerPC to x86, they went to 32-bit x86, even though 64 bit parts were already out. Which meant Apple users would face an unnecessary 32 to 64 bit transition on x86, and Apple would have to deal with annoying dual-mode issues.)
What does this essay say Linux needs? "Drivers for all existing hardware". "People who buy a new desktop want to plug in their old PCI cards..." Earth to Linux fanatics: 80% of all PCs are never opened in the field during their entire working life. What's important is drivers for what's shipping right now from major hardware vendors.
"Luckily, Windows more or less stopped being a moving target recently." Haven't looked at what Microsoft wants developers to do for Vista apps, have you? There's a big push by Microsoft to get developers using Microsoft-only technologies embedded in Vista, ones you can't run under Wine because they require non-redistributable DLLS.
"To attract enough non-technical end users to make the hardware vendors care about us, we need Linux to come preinstalled on PCs in a configuration that just works." Finally, the right answer. But that's a political and legal problem. Vendors don't offer Linux preloads because Microsoft penalizes them if they do, and Ashcroft's Justice Department rolled over on keeping Microsoft from doing that.
This essay is aimed at making Linux fanatics happy. What it should be aimed at is making low end desktops for office use cheaper. Push on Leonovo to offer something comparable to Red Flag Linux (which they preload for Chinese consumption) for export. Push on WalMart to sell it. The standard low-end business desktop should become Linux. Your call center people don't need Windows.
This hits Microsoft where it hurts - price pressure. Microsoft wants to charge more for Vista than for XP, and that could be derailed.