Robert Pogson

In this video clip, he says, “I don’t see it a issue. OK. No matter it’s Android or it’s GNOME or a special Linux. I think you can keep improving the performance. The original eeePC finally still used the Windows XP and now the Windows 7. I think mainly it’s because of the compatibility consideration, mainly because also the pricing. OK. At the beginning I think Windows, at that time, the price difference is bigger and they’re willing to sacrifice and then trying to minimize the gap. So I think it’s always the trade-off between the compatibility. So the performance, I think, is not a real issue.”

(This is my interpretation of the words on the clip. Your interpretation may be different. I deleted some unnecessary stumbling words like, “uh” and “you know”)

If this is the unbiased truth, I think ASUS missed an opportunity. They could have produced units with both software systems and let the market decide. The incremental cost would have been tiny compared to the market value of millions of units. I think they had M$ over a barrel in that at the time ASUS had a monopoly on the netbooks and also M$ was legally prohibited from requiring an exclusive deal. If they had supplied both kinds, consumers would have been able to choose based on the OS on identical units, and they may have sold more units in total because folks who loved GNU/Linux may have been put off buying XP and waited until others sold netbooks with GNU/Linux. They were already tooled-up to supply GNU/Linux so they would not have to add any costs to keep it going. As ASUS is not foolish, my conclusion is that the price of XP must have ended up being negative… to pay for the missed opportunity.

I have just been reading M$’s annual report for 2010. M$’s business plan worked well for them. The PC division received $14 billion from the OEMs. That’s for about 300 million PCs. If consumers had not paid for that other OS, they could have bought about $14000 million/$500 per PC = 28 million more PCs. Then there’s the money consumers would not have needed to spend on malware… and some paid for Office 2010 and “down/up-grade rights”… The bottom line is that because OEMs prop up M$’s monopoly they sold 10-15% fewer PCs. Now, that’s a tax. They made less money because they were M$’s partner. Business are the same. If they spent less on software, they would be able to spend more on hardware. Then there are servers. Would OEMs not want to sell more servers?

OEMs, are you listening? Perhaps not. Too bad. The world will start buying ARM. Are you pushing ARM or letting newcomers eat your lunch? The newcomers are making smart-thingies this year. Next year they will pump out many millions of small, cheap PCs. The world will change whether you change or not. Perhaps you should change to survive.

I love data. Much of my education at university and later life was learning how to select it, collect it, analyse and present it. All the data in the world means little if it is systematically biased. That is why data on the prevalence of GNU/Linux in the universe of PCs is so problematic. There is little good data. The trial of SCOG v Novell gives us some surprising data. The jury pool was 52 people randomly selected. The random selection is of course biased somehow depending on what list(s) the data was collected: property owners, drivers, whatever, but that pool is supposedly fair enough as input for the selection of a jury so it may be acceptable as a population of consumers/end-users.

Judge Stewart asked this question of the assembled candidates:“Ladies and gentlemen, do any of you know what Linux is, L-i-n-u-x? If so, would you please stand”

Here is the raw data coming from that question and my analysis. On previous pages of the transcript, you can find more information about particular candidates identified by number.

24 Would you briefly give your jury number and tell
25 us how you know what it is, what it is that generated your
Case 2:04-cv-00139-TS Document 854-1 Filed 04/19/10 Page 108 of 152
108
1 familiarity with it.
2 JUROR NO. 3: Juror number three, XXXXXX XXXXX. I
3 have not worked with the system. It is an open-coded system
4 that’s an operating system. There are, I know, some basic
5 setups so those users that aren’t programmers can use it. A
6 lot of my friends have actually started using the Linux
7 software instead of Windows. I have very little contact
8 with it myself.
9 THE COURT: Thank you.
10 JUROR NO. 4: Juror number four. Some of the
11 platforms that my employer uses are based on Linux operating
12 systems. We interface with it, but we do not use it
13 directly in my specific business unit.
14 THE COURT: Mr. XXXXXXX.
15 JUROR NO. 14: Juror number 14. My familiarity is
16 somewhat in passing. Working in marketing communications
17 with Web development companies, I work closely with ITs, so
18 I do understand some basic differences between Windows and
19 Linux as it relates to Web development.
20 THE COURT: Thank you.
21 JUROR NO. 16: XXXX XXXXXX, number 16. In 2001, I
22 built a new computer and used Linux.
23 THE COURT: Thank you.
24 JUROR NO. 20: Juror number 20. I know what I’ve
25 read in computer enthusiasts publications. I tried building
Case 2:04-cv-00139-TS Document 854-1 Filed 04/19/10 Page 109 of 152
109
1 a computer based on Linux. It didn’t work very well, so I
2 went back to Windows.
3 THE COURT: Thank you.
4 JUROR NO. 21: Juror 21, I have done some pearl
5 scripting in Linux.
6 THE COURT: You’ve done some what?
7 JUROR NO. 21: Pearl scripting. It’s a program
8 language primarily on Linux.
9 JUROR NO. 26: Juror number 26. I personally do
10 not know a lot about Linux, but by husband and my sons do.
11 We have approximately ten to 15 computers in our home, half
12 of which are Linux based and half of which are Windows
13 based, or others. I look at my book shelves and I see
14 Linux, Linux, Linux.
15 THE COURT: Thank you.
16 JUROR NO. 27: Juror number 27. Both of my
17 parents use Linux a lot, but they are working engineers.
18 THE COURT: You don’t have any personal
19 familiarity with it, though?
20 JUROR NO. 27: No.
21 JUROR NO. 28: Juror number 28. I have used
22 multiple Linux programs in my field. I haven’t currently
23 used them for probably the past five or six years.
24 THE COURT: All right. Thank you.
25 JUROR NO. 41: Juror 41. I have both used and
Case 2:04-cv-00139-TS Document 854-1 Filed 04/19/10 Page 110 of 152
110
1 built software for Linux distributions over the past 15
2 years.
3 THE COURT: Thank you.
4 JUROR NO. 44: Juror 44. I’ve heard of it in
5 passing. Just that it’s another option besides Windows. I
6 thought it was a free operating system.
7 THE COURT: Thank you.
8 JUROR NO. 46: Juror 46. I haven’t personally
9 used it, but I have a lot of friends who are big fans of
10 Linux.
11 THE COURT: Thank you.

Jury-candidates who stood were: 46, 44, 41, 28, 27, 26, 21, 20, 16, 14, 4, and 3

Jury-candidates who had used GNU/Linux were: 41, 28, 21, 20, and 16

So, 12 out of 52 (23%) stood indicating they new of GNU/Linux.

5 out of 52 (9.6%) stood and indicated they had used GNU/Linux.

That’s way more than 1%, people. You can argue that Salt Lake City is full of geeks or that lightning struck the court-house but a lot of people know about GNU/Linux and would use it if they found it on shelves. I expect a few who did not stand had used GNU/Linux but did not know it, say in a netbook or some gadget. A couple even built PCs and tried GNU/Linux…

GNU/Linux is ready. Are you?

I recently received a FUD post about four companies: IBM, RedHat, Google and Oracle. The FUD being sown was that these folks while pretending to be FLOSS supporters were actually crassly fighting open interoperability. The truth is much different.

IBM is a stalwart of the FLOSS community. They generate huge amounts of FLOSS code and install enormous numbers of FLOSS systems globally. The idea of FLOSS is that people, including corporations, should be able to run, examine, modify and to distribute FLOSS. IBM is about the strongest supporter of that in the corporate world.

RedHat is one of the largest contributors to FLOSS on the planet and for its size is probably the largest. GNU/Linux is huge globally but only a few years ago one had to search a bit to find solutions to problems. It was standard operating procedure to add “site:redhat.com” to a search of Google for some error message or configuration problem. They supported K12LTSP which was my preferred solution for schools after 2003. That magic opened my eyes to what was possible using existing hardware.

Google is a glorious example of a profitable business built on FLOSS. Google not only uses FLOSS as it is entitled under Free Software licences but it gives back by encouraging new FLOSS programmers (Summer of Code) and contributing cutting-edge FLOSS projects such as Chrome Browser, Chrome OS and Android. The effectiveness and wide acceptance of just these projects marks Google as a main pillar of FLOSS and why M$’s fudsters pay so much attention. Besides the obviously vigorous projects, Google inspires much best practice in IT such as Hadoop and even care and feeding of hard drives. Google is one of the foremost repositories of knowledge of the server/data-centre.

Oracle is not famous for FLOSS. They are famous for a successful non-free database, proposing thin client computing (which is finally taking off) and acquiring SUN. With the latter acquisition they own the openoffice.org, Java and MySQL projects. They seem to be in the process of shedding OpenSolaris. Some other FLOSS projects are still being sorted out in the SUN acquisition. My opinion of Oracle on FLOSS is still open. They obviously have serious channels to large business and with MySQL and OpenOffice.org seem to be developing channels to small and medium sized business. How they fit in the global picture of FLOSS is not clear to me but they can be an asset of similar effect to the others if they have the corporate will to do so. A big plus for Oracle is that they are the enemy of my enemy… and they push GNU/Linux as a platform for their database and MySQL.

So, bizarre interpretations of software patents and interoperability do not give me much doubt about the sincerity of these players as assets in the FLOSS community. M$, your fudster has failed.

In 2009 I was whining that AMD and Intel need to satisfy the netbook market. Intel brought out the Atom and did well but AMD passed. Now, almost two years later, AMD is coming to the game with a suitable chip. They might earn some thin-client action with that as well.

Some people just ignore me… AMD in its ivory tower just did not see that the netbook was here to stay and gave Intel a huge lead. Where is the AMD of years ago who brought out amd64 ahead of Intel? Perhaps there were other business reasons for passing on netbooks but publicly they said, “We’re ignoring the Netbook phenomenon–just thinking about PC form factors above that form factor.” Well, better late than never.

M$ is asking anyone affected by the latest hot vulnerability in their amateur operating system to call the police. Don’t waste your time. M$ is part of the problem, not the solution. They want you to waste your time so that they can pretend to fix something in the meantime. Instead you should call your IT guys/gals (if you don’t have any, hire some or a consultant/system architect). Give them these instructions:
Get M$ out of my system STAT!

(STAT (statim – immediately))

Ideas that may be proposed include switching to Macs but that is the most expensive commonly used option and also is a system with a lot of top-down nonsense. GNU/Linux is a much better solution. It is a cooperative product of the world designed by people who do not want to invite malware into your PC, not salesmen.

The fastest way to go to GNU/Linux is to migrate files and databases to GNU/Linux servers and to convert the client machines to GNU/Linux via a network installation. Done right all that needs to be done is to set every machine to boot PXE as the first option and set up an LTSP terminal server in each department or office or set up an installation programme to boot with the machines. You are far better off to have a couple of days or a weekend of scheduled downtime than the devastation of the latest malware M$’s trojan invites in. Extremadura, Spain, converted 80K seats over a weekend. You can do it. To speed up network installations, I recommend using apt-cacher-ng to cache any packages downloaded from Debian’s repositories locally so that they can be served at LAN speed instead of web speed.

There is no limit to how bad malware can be. It can range all the way from sending spam e-mail from your machines to selling all customer lists and sabotaging data by rot over a long period of time so that by the time you catch it weeks of work could go down the drain. The worst case is killing your operation through lawsuits charging negligence in allowing disaster to happen when reasonable people know you do not allow malware to run on your systems.

If anyone objects that the migration is impossible or too much effort, retrain them or fire them. If your software is incompatible with GNU/Linux use other software or hire someone to create your own.

Come back to your senses. When the criminals tell you to call the police you know they fear the police less than your attention to your own business. Don’t let M$ tell you how to run your operation.

For ideas about how to migrate to GNU/Linux, check out these resources:

http://www.redbooks.ibm.com/redbooks/pdfs/sg246380.pdf

http://www.redhat.com/migrate/

http://www.novell.com/products/desktop/

http://www.debian.org