Slashdot Mirror
IBM Releases Desktop Linux Presentation
An anonymous reader writes "DesktopLinux.com, in coordination with the Desktop Linux Consortium, is making select presentations from Monday's groundbreaking Desktop Linux conference at Boston University's Corporate Education Center available. Sessions from the well-received program included talks from key companies and open source projects bringing Desktop Linux into the enterprise. The first presentation available is from IBM's Sam Docknevich, Linux and Grid Services Executive for IBM Global Services and is titled "Open Source Desktop - Directions for today... and Tomorrow". His presentation discusses IBM's push into the Linux desktop market, an initiative from inside Big Blue."
1) You can't buy a IBM Thinkpad unless it comes with Windows. That 'old "Microsoft tax"
2) IBM can't be bothered to support FreeBSD on their laptops. Public case in point - the use of Type 165 for the partition that held the backup info. Private case - IBM staffer claimed they'd help with a USB implementation issue on one type of Thinkpad. (The USB doesn't work at all with FreeBSD and the only way Linux works is if you force the probe order in some wonky way.)
3) Many of the new style Thinkpads come with the Intel wireless - the one only supported under Windows.
I'll believe IBM cares about Open Source when they address the 3 above. Otherwise its the swapping of one corporate master for another.
IBM can make far more money selling services. PowerPC really offers no tangible advantage for the average desktop system over x86's descendants.
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
That linux isnt for the desktop... Who should i belive.. big red,or big blue...
( as i sit here submitting from Konq running on FBSD 4.9 )
---- Booth was a patriot ----
http://www.desktoplinux.com/files/article003/sld01 3.html
Once people switch from the mindset that there are things in the red circle to the red circle is shrinking and only will get smaller over time is a key to understanding the impending disaster Linux has turned out to be for Microsoft.
A better version of the slide would be to show the red circle shrinking over time down to nothing.
far better security
Not necessarily true. I know it's an old retort, but trust me - once Linux gains appreciable market share on the desktop, the virus and trojan writers and script kiddies will descend, and the exploits - and they're sure to exist - will be found.
Sure, it'll be harder for them than it is with Windows, but not impossible.
far easier maintanance
Rubbish. Far easier for you to maintain, perhaps, and I'd even be willing to agree that proportionally, there are more clueless Windows "pretend-admins" than there are Linux ones, but a properly skilled Windows admin is as effective at their job as a properly skilled Linux one is at theirs.
If you're talking about patches and updates, well, a few months ago I ran Windows Update and Mandrake's update application one after the other, both on more-or-less unpatched installs. The Mandrake one had about 10 times the amount (by byte count) of updates that I selected (let alone available) than XP did.
Of course, you get far more applications with a Linux distro than you do with Windows - but that hardly matters. It all still has to be maintained, and we're talking about desktops here, not servers; desktop installs tend to be rather less selective as to what gets installed. Even if the number of critical/security updates is lower for Linux, it takes someone time to read through the list and work out what needs to be applied and what can wait.
It's official. Most of you are morons.
In my experience 1/2 of linux users that I know IRL switch thier browser identifier to IE. Konqueror(and other browsers) should make that option automatically disabled when visiting google. It's too bad these people don't just enable it for specific sites that require IE, but most of them are busy and "just can't be bothered" with it, so they stay identified as IE all the time.
True genius is grasping a situation like a peice of fruit, and peircing it just right so that it drains dry.
Gentoo's advantage all depends on how heavily you optimize with default CFLAGS. I use the following on my Athlon XP:
CFLAGS="-march=athlon-xp -m3dnow -msse -mfpmath=sse -mmmx -O3 -pipe -fforce-addr -fomit-frame-pointer -funroll-loops -frerun-cse-after-loop -frerun-loop-opt -falign-functions=4 -maccumulate-outgoing-args -ffast-math -fprefetch-loop-arrays"
I haven't benchmarked this personally, but I can tell you that, qualitatively, Gentoo "feels" significantly faster than Mandrake or Debian on the same machine, and leaves Win2K in the dust.
I noticed that, in the article, the following CFLAGS were used in the test:
CFLAGS="-march=pentium3 -pipe -O3"
Now, even ignoring all the various hacks tacked on to the end of my CFLAGS line, there are some VERY important flags that the benchmarkers seem to have left out. For one, on a 2 GHz Celeron, march should be set to pentium4, so that gcc will optimize for the newer celeron core and use sse2, among other things. Using mfpmath=sse also yeilds a very significant performance benefit, as it optimizes ALL floating-point calculations for the sse SIMD instructions.
Anyway, if we had a different set of CFLAGS, one could definitely expect sse/sse2 optimizations to have yeilded a GREAT performance increase in the GIMP test, while the various memory and loop optimizations from my CFLAGS would almost certainly have edged Gentoo ahead in the gnumeric and kernel compile benchmarks.
Anonymous Luddite: "What do you think of the dehumanizing effects of the Internet?"
Andy Grove: "Not Much."
The IBM and Novell connection just seem stronger now with slide 17.