Slashdot Mirror
Where Does Linux Go From Here?
With the success that Linux is currently enjoying Linux.com (also owned by SourceForge, Inc) asks the question, where do we go from here? With such a high level of success and greater corporate participation (on both the consumer and provider fronts) will the spirit of freedom and idealism remain true or will the ever-present corporate bottom line eventually take over? "Linux is surrounded by proprietary IT firms. Some of them view Linux as a profit maker, others as a threat to their profits. Both sides represent a challenge for Linux in holding to its ideals of freedom and openess. The first large IT firm to really grok Linux was IBM. It has a long and mutually beneficial association with Linux, Apache, and other FOSS projects. The company has learned the language and the mores of the FOSS world, and has made significant code contributions as part of those projects along the way."
Can I be the first to say that it is going on the deskytop! Oh well, may be next year...
Will code for new sig.
will the spirit of freedom and idealism remain true or will the ever-present corporate bottom line eventually take over?
How much do we have to worry that something will "take Linux over"? No matter what corporations do, they'll always have to release the source code, which means people can always fork it. Wasn't that the point?
There are two fields where Linux is lacking compared to Apple and Microsoft: How easy it is to screw things up and games. Of course driver support is important, but that is driven by demand of the market, not demand of the developers, so I consider the previous two reasons of higher importance when discussing how best to expand Linux in the market.
Now I know ideally we should all be intelligent enough to be able to operate Linux without screwing something up, and if we do be able to fix it. But the layman is not and will not have our technical ability, however simple the task may be. Since Linux does not have technical support often in the same way Apple and Microsoft do, users are driven away for fear of an inoperable computer. They would rather have a computer that works 50% of the time than 25% of the time. As far as business use for Linux, obviously they have the resources to be able to have any problems fixed and prevented, but personal users can not do that.
As far as games, Tux Racer does not cut it. Email and web browsing of course are workhorse reasons for having a PC, but you can do that on your cell phone nowadays. Honestly, game development seems to be in a bit of a catch 22 in the same way that driver support is a problem. Investors need to see profitability in the market, so they want to see market demand. However market demand isn't rising because there isn't enough of a reason to switch to Linux when you can't play the hottest new games on it. Of course games do get ported, however initial release of games for Linux I think is vital to bring the average computer user into the fold of open source.
Just my two cents.
How many articles are we going to have on the same topic? Just a bunch of nonsensical ramblings about "corporations" and "freedom" with about 0 substance. You don't like the direction Linux is taking in terms of "corporate influence", then fork it, end of story.
Monstar L
Your comment is valid but misplaced. SGI uses Linux on its Altix line of hardware and their add on software makes it aware of the specialized hardware within Linux.
The main problem that your pointing out is a problem with x86 hardware. Windows and every other OS that runs on x86 have the same problems. The problem stems from there being no standardization and no evolution of the technology. Essentially all of these 50K x86 servers are just big beefy PC's with built in redundancy and some extra monitoring options and are essentially an evolution of the original IBM product. The vendors are moving in the right direction though and Linux is supporting this hardware as they need to. For example the IPMI interfaces and BMC interfaces have direct access with ipmitool where you can read chassis and sensor information. Window's cannot natively talk to the IPMI interface without add on drivers. Linux supports MANY of the raid controllers from within its kernel unlike windows that has to have drivers to even see the drives.
Im a big Solairs fan myself because it truly is an integrated hardware and software platform. The big issue is cost. Thats why a number of companies in the industry are willing to put up with the shortcomings of x86 hardware and related OS's. Its a big trend now for people to go with the cheap mantra when in actuality it costs them more in the long haul in downtime and data loss.
Even further into the cheap cheap cheap mantra is the concept of VM's which were abandoned in the 1970's because they never worked and because of the political battles between departments for resources of the mainframe. This is being looked at again for the concept of energy savings and hardware savings but virtual machines really equal virtual performance. Many companies have the concept that you can oversubcribe the resource of one machine and serve many users and "hopefully" not all of them will need resources at the same time. The Administrator of a VM box is under the pressure from that many more users when there is a hardware failure. It amazes me that companies put all their eggs in one basket on cheap commodity hardware that is known for failure. Loose a system planar or other low level non redundant piece of hardware and suddenly you have hundreds of servers down instead of just one. And of course because we are cheapskates we have to call up and scream at our vendor to get out ASAP because we were to cheap to have a spare kit and an Admin onsite and or capable of switching out hardware.
You get what you pay for one way or another and most people are obsessed with the price tag rather than the true costs. Buy the bottom of the line machine and spend your time down on the phone to India.
I really, really, really want to support Linux. However, frankly, I just don't have the time to hassle with it. I have made 4 endeavors in the past. And currently have a 5th endeavor for my wife. If I were to give Linux a score grade it would "C-".
I know that's not what a lot of you want to hear. But it's the truth. I don't want to spend several days trying to get a 802.11g wifi card working. I don't want to have to use some install manager or try to figure out how to get some script to run from the terminal in order to install an application. I simply want to be able to click and launch it, and have it install. Sadly, driver & software installation hurdles plague Linux. (In fact, these were the same issues that plagued Linux when I tried it repeatedly in the late 90's early millenials.)
I will say, it's improved quite a bit. At least in video card support apparently. But the truth of the matter is, I'd take XP & OS X over Linux. And that's because I'm anti-Linux or don't support Linux. Far from it, I wouldn't have tried it for my wife's (non-critical use) machine. So please guys....focus on these issues. (And don't say "Linux isn't really for the masses." Because everyone else keeps trying to push it that way. And that is the slated goal of many.)
Best of luck all...
- The Saj
Linux is never really going to "take off". That's because the whole Linux/FOSS model is anathema to what it takes for a desktop OS to really take off. What Linux needs to "take off" is a single, easy-to-use viable distro for the public. Which is not going to happen, because Linux and FOSS are all about having lots of choices and having everyone customize it for themselves. Sorry, not going to work. Software developers don't want to worry about working around the differences (whether real or perceived) in umpteen dozen different distributions (and variations of those distributions), most of which have significantly less than .1% of total desktop market-share. Heck even Ubuntu (currently the distro that most nearly meets these requirements) has at least three variations.
There is no -1 Disagree mod. Slashdot.org/faq defines mod options. USE IT.
I love Linux as a concept: An open-source, free as in beer, free as in speech, tweakable operating system offered and supported by multiple vendors. But Linux as a reality is an hodge-podge of incomplete applications spread across multiple subtly-incompatible distributions.
Moments ago, I read the following thread on the Rapidsvn mailing list. Rapidsvn is a very nice front-end for the Subversion version control system. I've compiled it, made changes to it - it's quite nice. I like it especially since it works on Linux, Mac, and PC -- all three are OSs I use to some degree. So the following is not a dig on this particular project. It is one example of something that happens a million times every day:
(P.S. I chopped the thread for brevity to make my point) Hi, I have downloaded rapidsvn 0.94. I am trying to install on SLED 10sp1. I enter
at the command prompt. I get a lots of messages and finally:
checking for APR... not found
configure: error: APR is required. Try --with-apr-config.
I tried...[various things] but got the same error message. I installed all the available APR's for
listed listed as version 1.2.2-13.2
Any ideas how to install rapidsvn -- I really want a gui interface on
linux similar to tortoisesvn on windows.
[various responses about apr-config, apu-config, downloading pre-built binaries, etc. but no solution] So we have a fairly simple GUI program, with no crazy dependencies. This application is not available in binary form for this distro, and since there are many major Linux distros and you never know what will happen if you install an RPM from another one. You can't compile it from source without a CS degree, and you need gigs of development libraries to do it.
This is the Linux I know, and it is why I have Linux on that other partition so I can boot it up now and then and see what the state of Linux is. But so far, it's always stuff like this. The challenge with Linux isn't learning the UI or thinking differently or anything. It's just getting stuff installed and getting it to work properly. I've never gotten a Linux distro up to the productivity of either my Mac or my Windows PC. I've maybe gotten 80% of the way, but with 500% of the effort. It's just not worth it.
Hopefully WGA will put a stop to this.
Change is certain; progress is not obligatory.
I'm not so sure. I think the real next stop is actually handheld devices, be they cellphones or tablets. Not only that, but I'm willing to bet most people won't even know or care that these devices are running linux. The only people who have ever cared about what they run on their desktops is A) Geeks and/or B) Fanboys.
Who, in their right mind, would devote thousands of development hours cobbling something together, then cast it into the wind where basement developers use "what they want, and [get] rid of what they don't?"
Red Hat, Canonical, Novell... hell, even Microsoft have a few open source projects kicking around.
I write bullshit
There's something missing from the article "Where does Linux go from here?"
The users. There's virtually no mention of them. There's talk about companies who are connected with Linux, about the technology, about the freedom of open software. But of the actual users there's only one passing reference.
If you want to advance Linux, start thinking about the users - their needs, their desires, their problems, and so on. To begin anywhere else is to neglect the most important part of the equation, and Linux will remain a "system for nerds" forever.
And so far as "Where does Linux go from here", send it to rewrite.
Send the list to canonical, with a copy to the kernel team, and post on every major linuxforum. It might actually get you somewhere. I would have modded you +111 insightful if I had modpoints left.
Since Microsoft has a monopoly on the desktop (yes they do, the court ruled so) there isn't much in the way of options.
#1. Develop your own system, keep it proprietary and hope that Microsoft doesn't see enough value in taking it from you.
#2. Support Microsoft's system and hope you can:
2a. compete with everyone else doing the same
2b. make enough profit to survive, but not enough so that Microsoft moves into the market itself.
#3. Go Open Source / Free Software and try to get your system enough marketshare that you can turn a profit, somehow.
#4. Give up on the computer industry and close your shop.
Sun has realized that #3 is the only option short of just giving up. At least they have something marketable - their expertise in the systems that they designed and that they built.
Sigh...
Wrong on ALL accounts.
There IS a standard way for ALL Linuxes to monitor EVERYTHING. Smartd+snmp will do the trick and it works just about the same for them all.
As for the rest, vendors create their measly, unoperable, stupid shitty stuff to supposedly "monitor" things. They are all really pretty badly made, but its "their" way.
That being said, "Linux" as an abstract entity has no room in corpoland, you need to start thinking RHEL or SLES, and there you will see: all RHEL is monitored the same way, all SLES is monitored the same way. If you wanna use the crap monitoring stuff dell, hp and ibm attempt to push down your throat, by all means, do it, but dont come crying to us.
NO SIG