Slashdot Mirror
CNet Article On 2.4 Kernel
jho writes "This C|Net article talks about how the Linux 2.4 kernel, armed with Firewire, PnP and USB support, will be better equiped to tackle the desktop market. It's a intresting read as far as how Linux is being pushed to the mainstream. Have a look. "
At the time I post this, they're about 34 replys. I want to first start by saying clearly "I LOVE LINUX"
Now for the flaming part:
Why is it that the Linux community, for the most part can't accept a little press awareness or a little humor? I am a member of MDLUG (Metro Detroit Linux Users Group) and recently someone posted a funny site that was sarcatic to "Micorsoft" (as he put it), and the group went mad about it. Slam after slam was posted about how STUPID this guy was and I thought his site was hilarious (sorry I don't have the URL handy). I laughed my ass off.
Now CNET is paying attention to Kernel 2.4... so what, big deal. The direction that Linux goes is controlled by what the Linux community wants and what the developers are interested in. God bless the people with that kind of knowledge.
The point of the whole CNET article was that things like USB and Firewire were being worked on, who knows if it will actually be implemented or how stable it will be. As of right now Linux is still geared at people who are computer literate, but I think thats soon to change, especially if things like USB are implemented.
Enough said... now will the uptight people in the Linux community please take a minute to pull the underware out of the crack of their ass' and just be glad that Linux is getting press time? Linux is serious but even Linus said it should be a little fun, and comments like most of these are worthless attempts to slam people that don't know as much as you do.
IMHO, it's just as bad as M$ taking advantage of people's ignorance by hiding the fact that Win98 is version 4.1 and Win95 is version 4.0, how many people realized that? Not many I'll bet, atleast until they spent $89 on it.
Impossible? Never is anything impossible... The thing is why would we want to have Winmodem support? So it can make Linux as crappy as Windows? PNP support doesn't mean Winmodems man :) Just support for the devices so we don't have to do ISAPNP config files hehe
Well, you can tell linux is going mainstream when the press starts commenting on the latest linux vaporware. 2.4 isn't out yet. It doesn't have all the features listed in the article in the kernel source. Reminds me of how the press treats a certain Redmond-based company - "w2k will have xyzzy feature - blowing away the competition!".
Okay, mark me down now - I've spoken blasphemy against the holy os...
--
I notice that among some other things, the new kernel will have at least some support for Winmodems. We all know that they're crap, but it's good news anyway, because lots of newbies don't know that. There's a lot of scorn among Linux users for anyone whose skills are anything below Guru. This is going to have to stop. A lot of people use Winmodems, and anything we do to make it easier for people with low-end hardware that the guy at Circuit City told them would be fine is great -- GNOME, Winmodem support, wheel mouse support, popular programs e.g. WordPerfect & Netscape. So this is a Good Thing (TM).
Switch the . and the @ to email me.
Well, because you want more people to be able to run Linux, of course.
Anyway, it's not broken. It just has offloaded more of the work to software, rather than hardware. It's not any more broken than software DVD decoding compared to hardware DVD decoding is.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
Yes, the kernel will have more functionality. Sure, this is good for the OS, but that only tackles half of the problem of entering the desktop market. Linux really needs a somewhat standardized set of distros to sit on top of it, and a lean, mean GUI before it can hit the desktop market with any real force. I hesitate to say that we have such a userfriendly and joe public ready distro in anything we see about today.
"What is now proved was once only imagin'd"
"What is now proved was once only imagin'd"
William Blake, The Marriage of Heaven and Hell
As it happens, I just bought a USB scanner today (HP 4200c). I was thinking of getting a parallel port one, but the USB scanners were 2-3 times faster in all the benchmarks I've seen.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
-- Give him Head? Be a Beacon?
-- Give him Head? Be a Beacon? :P)
(If you can't figure out how to E-Mail me, Don't.
I don't know too much about USB or firewire. Are such devices for the iMac compatible with PCs?
and slightly back on topic, i thought java was oak. it was called oak for awhile, but then they realized there was another language called oak so they renamed it to java. *shrug* thats what i heard
Full quote here
You were saying?
Well, if you want Linux to become a desktop OS, as many people do, you must support Winmodems. An increasing number of PCs come with them. Nearly all sub-$800 PCs come with them. If you have no winmodem support, that's a huge chunk of the market that is not going to use Linux, no matter how good the rest of your OS is.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
I don't see your logic. Virtually all sub-$800 PCs come with winmodems. They do so even though Linux doesn't support them. They will continue to do so even if Linux continues to not support them. The only difference is that there will be a large percentage of users who cannot use Linux, and will stick to win95 instead, since it lets their modem work.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
Mac (and LinuxPPC) user here: What the friggin' hell is a Winmodem?
Wordnik, a dictionary project which aims to collect
However, the first time the power flicks out and now the thing has INODE blocks all of the place the grandma who bought a gateway will be lost. Specially when it loses some config file (what's w/ Mandrake 6.0 and the kdmrc file??? I need a UPS.)
Erm...
I hope you feel stupid right now. 2.4 will have journaling code in the form of ext3fs, which is being finished up right now by Steven Tweedie. This means no more long fsck's.
As soon as a file is changed (due to write() or mmap()'ed file access, or anything else) this info will be written to a conveniently located scratch area on the disk. If the power fails, and it comes back up, the fs notices there is stuff in the scratch area and does the modifications if they haven't been done yet.
And don't let anyone tell you NT has journalling, it doesn't. It has half journaling, which means that the metadata is journaled, but nothing else. So basically, your still screwed if the power failed when your 5 gig database is left in an inconsistent state.
Well, NT's file/print sharing code came right from OS/2 (according to older NT documentation). The "Server" service even used to be called "LanMan Server" in NT 3.x. NTFS is also based on HPFS.
However, the NT Kernel and HAL stuff looks nothing like anything that's in OS/2. The folks who designed this worked at DEC on VMS, so the design is influnced by VMS, but that's not the same thing as "based on".
--
Business. Numbers. Money. People. Computer World.
Of course, it does bear mentioning that NT is based on a lot of work done on VMS which was more stable than the Unix systems of its time, which in turn was based on the work done on PDP-11 systems. This work on the PDP-11s was also where some of the ideas that begat Unix started, so "I'm older and more mature" arguments don't really work in this context.
Does the same argument apply to languages like BASIC and FORTRAN which are all older than C, or possibly even ALGOL (not sure on that one).
A lot of the point about relatively new technology (including Unix, VMS and languages) is that you really can't tell what is going to be a success and what is a failure until you have the benefit of decades of hindsight. Who knows - BeOS may be the system we all end up using in a hundred years time, after all it does have a very good architecture. Open source systems may end up with a difference between ideology and implementation and crumble, the capitalist society we live in may fall to anarchy etc.
Remember than when Unix was invented, no one 'knew' it was a success. They all have to start somewhere.
Fear: When you see B8 00 4C CD 21 and know what it means
Oh, come on, surely you have a better memory than that. Before Windows 2000 came even close to seeing the light of day, it was hyped as the ultimate replacement for any operating system that ever existed, and I include VM/CMS, VMS, RSTS, etc :-).
I don't see anything wrong with the same happening for Linux. Rampant speculation is just a part of being popular. Think of it as comparable to the Star Wars hype.
Incidentally, I had to buy a car and wound up talking to a Realtor(R) car owner. When she learned that I was a computer guy, she vented her spleen about the new Windows-based computer systems in her office freezing up all the time. It was interesting because she'd already absorbed the anti-Windows perspective from the media, and was ready to suggest that MLS switch back to their old system or put together something, anything, else.
I think the public may be wising up. Frankly, I never thought they would, but to see this happen is refreshing to say the least.
D
----
It's possible to write drivers for anything. It's not possible to write a single driver for all winmodems, since there are so many different types. Further, it's not possible (or at least very hard) to write drivers for winmodems without information from the manufacturer, since the interface is not documented nor standardardized.
But, it's quite possible to make the rest of the code "friendly" to winmodem drivers by modularizing the modem code and writing skeletal drivers, which makes the task of writing a winmodem driver as easy as possible.
That's the wonderful thing about linux: instead of 'features' like an animated paperclip 'assistant', the people developing linux are answering only to thir own needs and things that there is a demand for. There are no marketdroids, no billionaire empire moguls commanding from on high (yet), and no single point of development. These are all Good Things.
With this kind of a development system, I'm not at all surprised that there's no PnP (yet). I hate Plug-n-Pray. Firewire and USB? Bring 'em on, these I can use! However, I'm not yet confident enough to hack my own mods to a kernel, so all I can do is let it be known that I would like these features included in linux, please. Then an amazing thing happens: Enough people want something, and a very generous person dedicates a chunk of his or her life to staring at a glowing screen, and a few kernels later, things work.
Know what, folks? This is a great time to be alive.
"...America's great minds of today, teaching America's great minds of tomorrow. Poor bastards." -- A Beautiful Min
There are currently firewire implementations that aren't in the main development kernel. They're not vapor and they're certainly being worked on.
Man walked on the moon in 1969, the year before the dawn of the Unix epoch. Must we discard the aging space travel technologies pioneered during and prior to those days and start anew?
The first block-structured language, ALGOL, is now about 40 years old. That software technology is still alive in the latest languages -- Java can trace its roots directly back to ALGOL.
Firearms are now about 500 years old -- and still rule the world in a very literal sense. Further, all the modern pistols I've seen are mere refinements and variations on a 90+ year old design.
It is not the arbitrary incident of when a technology is originally developed that determines its importance now and in the future. The solid stuff lasts and is refined and improved. Like it or not, the works of Ken Thompson and Dennis Ritchie were so fundamentally sound that new variations of Unix are just hitting their stride now, three decades later.
That kind of quality is something that fans of Bill Gates' young empire, now pushing its third OS family, can only dream about.
Geeky modern art T-shirts
Winmodems should go the same way...
<^>_<(ô ô)>_<^>
If the even numbers are stable, and the odds are devel, what happened to 1.4.x?
Linus decided there had been enough changes to warrant a jump all the way to 2.0
Maybe 2.3. will be 3.0.x
Perhaps, but I doubt it. Linux said he wanted to experiment with faster iterations. In other words, he wants fewer new features going into each release, which will allow more time to stabilize the code while it's still in development. He's also been good about not making huge version number jumps too quickly, and when he makes them it's for a damn good reason (wasn't 2.0 the first stable kernel to use ELF binaries?)
2.3 isn't 2.4's beta. At most, it's the alpha, more likely the pre-alpha. 2.4.0-pre will be the beta of 2.4 (just like 2.2.0-pre1 through -pre4 were the betas of 2.2). So, no, Linux 2.4 is not in beta yet.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
Other old computing technologies -
- Windowing GUIs
- Mouse
- Ethernet(!)
- Object oriented programming
- C programming language
And speaking of old computing technology, have you checked to see what's running under Win98 lately?General technologies older than 30 years -
- Telephone
- Internal combustion engine
- Jet aircraft
- Light bulb
- Electricity
- Radio and television
- Indoor plumbing
Can I assume you don't lower yourself to using these old, outmoded technologies either?It's interesting, Joe Pranevich's piece about what's new in Linux 2.4 at www.linuxtoday.org didn't make it to
I mean, the time when it instantly was "News for Nerd", whenever mainstream press (or something remotely resembling it) just mentioned Linux in an article is long gone. IMO Joe's article is much more NfN. Stm., than the C|Net one.
</slightly OT>
<extremely OT>
Sorry, but I've always wondered: Why do C|Net have that pipe in their name (or is it just the logo)? What is it supposed to mean?
</extremely OT>
Please alter my pants as fashion dictates.
Well, it that case, might as well ditch X. We don't want the type of people too ignorant to use a command line to use Linux. Ditch RPM and DEB, we don't want people too ignorant to compile their own software using Linux either.
As for nonwinmodems, the cheapest I've found is a US Robotics 33.6 for $50. The cheapest 56.6 nonwinmodem I've found is $95, compared to $20 for a winmodem. Most purchasers of sub-$800 PCs buy them because of the price, not because they have a preference as to what type of modem it has. As long as it gets them online, and saves them money, it's fine with them. I don't see a problem with that.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10