Linux Can’t Kill Windows

April 21st, 2005

What follows is something I wrote on my personal blog. But after posting it, I realized it may be more appropriate for E Minor.


So Krishna thinks that this chap Ian makes very valid points when he says ‘Linux can’t kill Windows.’ I couldn’t disagree more. Not with the can’t kill part but with the valid points part.

I am going to quote liberally from Ian’s post but I suggest you read it first in its entirety ‘cos if you do that, I suspect you’ll agree with me in disagreeing with Ian and Krishna without having to read what follows.

Operating systems are this centuries (sic) infrastructure. It’s simply too inefficient to have multiple choices in this area, the world naturally gravitates towards a monopoly.

Infrastructure implies monopoly? Telecommunication, transportation and banking. All infrastructure industries where I see healthy cut-throat competition.

You wouldn’t want there to be 5 bridges right next to each other going over a river or 4 train lines right next to each other going from the same place to the same place.

But I could ask PowerBuilders to build a bridge on this part of the river and BuildOMatics to build one on another part of the river and still expect that traffic will be able to ply on these bridges and the connecting roads reliably, no? Operating Systems are infrastructure.

Businesses aren’t going to pay for training and support of multiple operating systems in their organizations and it’s all about business.

No, but my business can use Linux and yours can use Windows and ensure they work together based on standards, right? Just like phones and banks? Operating Systems are really infrastructure!

Let’s say Red Hat takes off and all the sudden they become what is on every users desktop in every organization in the country. What have the businesses accomplished? They’ve traded in one monopoly for another. You don’t think Red Hat would start charging like they were a monopoly? I do.

It’s very unlikely we’ll see a Redhat or any other Linux based monopoly emerge. The GPL ensures that. (Linux geeks: cf. RHEL and CentOS.) Further, this again assumes that a monopoly is the only viable option.

So why should businesses take that risky path of swapping out one OS for another which may only be slightly better?

Businesses take calculated risks. No one is asking companies to throw out their existing systems to install Linux. It’s for them to evaluate their requirements, calculate costs (including risk associated costs), map out the future and then decide if making a switch to Linux makes business sense.

Agreed that changing infrastructural elements carries higher risks. Nevertheless, an agile company will make even fundamental changes if it means more profits and a better future outlook.

Until there is something to compete with Office nobody is going to switch out of MS products. There’s simply too much productivity on the line to mess with switching to another Office like solution. By the way Open Office and the like are not the answer. MS has 40 BILLION dollars sitting around. You’re not ever going to catch them in the business software market by trying to copy them. The only thing that can unseat Office is a revolutionary new product, like the Spreadsheet was 20 years ago.

Why aren’t OpenOffice and the like unlikely to partake a chunk of the business productivity software marketshare? If I find a new solution that delivers identical functionality at a fraction of the current solution’s cost, I’d be crazy not to switch! So delivering required functionality is a viable way to gain market share.

And the current thinking in office apps is approaching a dead-end, something Microsoft’s billions haven’t been able to fix. Companies are happy with their Office 97/XP investments and see no reason to upgrade to Office 2003 and beyond. They already have all the functionality they need from the current set of productivity tools. So if OpenOffice can deliver that functionality, it gets that market!

Of course, if a revolutionary idea comes along, then it’ll be easier to dent Office’s monopoly. Doesn’t mean the current approach is unlikely to yield any results!

The whole idea is that Operating Systems and other basic software are becoming infrastructure, something Ian so rightly gets. But with the emergence of an infrastructure come standards and interoperability. So you get an ecosystem of competitive entities, not a monopoly.

Now going back to the can’t kill part; I believe ten years later we won’t even be talking about operating systems since they’ll be, well, infrastructure. :-)

Instead, someone will be writing ‘Google can’t kill Microsoft’ and I’ll be refuting that. ;-)

In News Today

April 1st, 2005

  • GNOME announces new branding strategy (G12 and not GNOME 2.12) and a new pay-per-use subscription model. More here.
  • Adeptiva, a Singapore based computing solutions provider, has decided to quit the GNU/Linux business claiming that it has become increasingly difficult trying to make 1970’s technology work.
  • KDE Dot reports that the KDE project has chosen to deploy BitKeeper for its source code repository.
  • Abiword forked - announcing Spatial Abiword for the GNOME desktop.
  • Davix - a free Linux-like kernel for x86-64 systems.
  • Paris Hilton hired by the OSDL as Exposure Executive for Linux.

Finally, the release!

March 28th, 2005

With the release of the new KDE User Guide, the audio cd ripping guide I wrote some time ago, the one which got a highly commended note by the judges of the KDE Docs competition, finally sees light of the day! Props to me!

Now to go and check if that guide still works as-is for KDE 3.4!

PS: You’ve just read the most hyperlink dense post on E Minor ;-)

Luminocity - I want it - videos!

March 25th, 2005

Seth Nickell has posted a bunch of cool videos demoing Luminocity, a GL based windowing/composition manager. To sum up, this is the Quartz equivalent for Linux. These videos look so good that I no longer feel the need to get comfortable with E17 just for eye candy!

Here’s a thought… what happens when you run this GL based compositing manager on a GL accelerated X server?

Update: Seth answers that question.

Adobe Reader 7 for the impatient

March 16th, 2005

It isn’t official yet but you can get Reader 7 via Adobe’s ftp server.

Luis Villa has some screenshots but they all point out to what he thinks is wrong with the program and don’t highlight any of the good stuff. I agree in part with his views: support for CUPS and printer detection should have been there. 1

Luis: It should use gconf to store and read preferences so that it can be remotely configured and managed by tools designed to read and write gconf keys…

I disagree. Adobe has to cater to those who don’t use GNOME too. Tying it too closely with GNOME technologies means they will be limiting themselves to a fraction of an already tiny market.

This point applies to much of what Luis has written earlier when he’s asked ISVs to write GNOME applications. Not Linux, mind you - GNOME apps. ISVs have to balance their resources. If they are going to write an app for Linux, they will try and make it as desktop agnostic as possible. Integrating it too much with either GNOME or KDE does not make financial sense!

The way forward is to get more integration stuff on freedesktop.org and get ISVs to confirm to those standards.

Luis: Appears to use gtk 2.2 instead of gtk 2.4 or 2.6, each of which would allow access to the newer, improved file selector, and improved toolbar and menu APIs.

The problem here is that open source software usually follows the release often philosophy with new features being added rapidly and bug reporting / QA being done by users. ISVs typically can’t afford to do that. They’ve got to do internal QA and support a release for longer cycles. This means that adapting to a fast changing (GTK) API is going to be difficult. Add to this the fact that documentation for open source technologies doesn’t generally keep up with the latest codebase changes and you can see why an ISV would prefer older technology to the cutting edge.

Frederico has more to say on the API issue.

What do you have to say?


  1. An easy workaround for apps that don’t detect printers is to get them to print to kprinter. Non-KDE users can install just kde-base to get kprinter.

MPEG-4 support in GStreamer

March 15th, 2005

Ronald Bultje writes about some initial MPEG-4 muxing support in GStreamer:

  • Writing tags to music downloaded from iTMS (yes, they’re untagged!).
  • Capture MPEG-4 video/audio in a MPEG-4 container using Cupid.
  • Transcoding my Ogg/Vorbis files to MPEG-4 so I can transfer them to my iPod.

Great news! I hope this leads to a more general foo.{avi|mpeg|mov} -> foo.mp4 solution. This is something that I badly need! I’ve been struggling to get mpeg4 video on to my phone but have had little success so far.

  • Mencoder doesn’t write an MPEG-4 container.
  • FFMPEG always creates files with terrible A/V sync. Is there a trick to this?
  • mpeg4ip fails to compile!

So please, give me an MP4 creator for Linux!

Easy Gentoo Install

March 13th, 2005

Found on bash.org:

<@insomnia> it only takes three commands to install Gentoo
<@insomnia> cfdisk /dev/hda && mkfs.xfs /dev/hda1 && mount /dev/hda1 /mnt/gentoo/ && chroot /mnt/gentoo/ && env-update && . /etc/profile && emerge sync && cd /usr/portage && scripts/bootsrap.sh && emerge system && emerge vim && vi /etc/fstab && emerge gentoo-dev-sources && cd /usr/src/linux && make menuconfig && make install modules_install && emerge gnome mozilla-firefox openoffice && emerge grub && cp /boot/grub/grub.conf.sample /boot/grub/grub.conf && vi /boot/grub/grub.conf && grub && init 6
<@insomnia> that’s the first one

Nero Linux

March 12th, 2005

The Nero people have launched a Linux version of their popular CD/DVD burning software.

While they are certainly best in class on Windows, I don’t see any reason for K3b users to switch to this because there’s nothing new by way of functionality. Maybe GNOME users who don’t want to install the kde-base libraries just to run K3b would go for it? Then again, it’s an ugly GTK1 app which no one would want to be seen running on their desktops in 2005. ;-)

Overall, I am happy with this announcement ‘cos more software for my platform of choice is always a good thing.

Find that leak

March 7th, 2005

If you are the kind of cowboy who loves hunting down runaway RAM, you’ll be happy to hear that Novell and OSNews have announced a set of bounties to help fix memory bloat in GNOME. Wanted dead or alive posters at the MemoryReduction wiki.

(Our earlier related story.)

A tale of the G and the Cow

March 5th, 2005

Some beautiful and excellently done wallpapers for Gentoo lovers.

(via jimmac)