Robert Pogson

One man, closing all the windows.

Monthly Archives / January 2012

  • Jan 31 / 2012
  • 6

M$, It’s Just Not Happening

M$ has recently advised users of XP that “It takes 18-24 months to plan for and deploy a new operating system” in order to hurry people up to migrate to “7″ and reminding them that end-of-support is in about 800 days.

It’s not going to happen, M$. About 30% of PCs are running XP and many of them are a bit old. To buy 450 million new PCs to replace them, in 800 days would need 500K machines per day, about 45 million per quarter. The world is only shipping 90 million PCs per quarter and many are getting GNU/Linux or MacOS. Don’t hold your breath expecting a 50% pop in revenues the next few quarters. M$ has been selling 50 million licences for “7″ per quarter but that includes consumer, business, replacements and new purchases. The replacement part is not the whole ball of wax.

The world has a huge inventory of used but solid and reliable PCs running XP and other OS and they are not going to be suddenly installed of “7″ or “8″ or replaced with a new machine. They will run until they die, in four or five more years, and may well be replaced by tablets or netbooks or any other kind of PC other than Wintel. People will re-install XP and use anti-malware products from other suppliers to keep it alive. Get used to it, M$. The free ride is ending.

According to W3Schools, in the last year, “7″ share grew from 29.1% to 46.1% (+17%) while XP lost only 12.7% and Vista lost 3.9%. A lot of XP machines are going elsewhere. Perhaps as many as 4-5% of PCs are going to GNU/Linux annually. Vista, you see, is much closer to “7″ than is XP.

I like to check the numbers. Suppose you have a five year old PC running XP and you want something faster. Are you going to spend $hundreds on a new PC with “7″ or are you going to install Debian GNU/Linux for a few dollars’ worth of your time and laugh at M$? It’s good to have choice and enough will accept this choice to keep M$’s wet dreams from becoming reality.

  • Jan 30 / 2012
  • 5

Oracle Shoots Foot, Repeatedly, in Oracle v Google

It’s absolutely amazing that Oracle, having some of the most expensive lawyers on the planet, cannot even follow the rules of evidence and procedure. Oracle cannot even provide expert witness to many of its patent claims. The simple thing required by the judge, they have not done:
” On January 6, 2012, Oracle will provide an identification, for each of the 26 asserted claims, of each Oracle product, Oracle-licensed product, Sun product, or Sun-licensed product (“Oracle Products”) that practice or have practiced the claim. Oracle will also identify the fact witnesses who possess information supporting Oracle’s contentions that the Oracle Products practice or have practiced the asserted claims, as well as provide a summary of testimony Oracle intends to elicit at trial from those witnesses regarding those Oracle Products’ practice of the claims. Oracle will also provide source code citations and/or other documentation supporting Oracle’s contentions that the Oracle Products practice the asserted claims.”

Instead Oracle tried another snowjob having prevented its witnesses from informing Google of many things in depositions and so disqualified them from testifying on some matters at trial as simple as how Oracle’s products practise the claimed patents. This all goes to damages that are surely not capable of reaching $1billion let alone the $billions Oracle demanded. Oracle was ordered to produce citations of source code and did not. The judge is likely to stomp all over Oracle for defying his order.

Typical of Oracle’s response is this:
“1. The ’104 Patent

Google’s first objection to Oracle’s identification of products that practice the ’104 patent is that it cannot respond because Oracle supposedly failed “to actually cite any source code or documentation” reflecting how the patent is practiced. (Google Response at 3.) Google repeats this same rote objection for each of the six patents-in-suit. (See id. at 6, 9, 11, 12, 13.) The objection is baseless as to the ’104 patent and as to the other five patents as well. Oracle identified specific source code files for each of the products it claims practices the six patents. (See, e.g., Oracle Response at 2-12.) These are not “a laundry list of names of various source code files” as Google claims. (Google Response at 1.) To the contrary, Oracle has identified specific source code files relevant to each product or product version — typically less than five files — that illustrate where and how the patent is practiced. (See, e.g., Oracle Response at 2, 12.) If Google felt these responses were somehow deficient, it could have raised that with Oracle before its response was due. But it never did. In fact, although it had a full two weeks to respond, Google’s counsel waited until three days before Google’s response was due to ask for copies of only some of the referenced source code files (which had all been produced last year as part of the parties’ procedure for producing confidential source code). Google did not make a good faith effort to prepare its response or to meet and confer during this two-week period, and if it did not have enough time to review the files to verify Oracle’s position, it has only itself to blame.”

A non-expert naming a file as practising a claim of a patent won’t cut it IMHO. I doubt the judge will be impressed by Oracle’s argument when there’s nothing in the experts’ reports to support the idea that Oracle practised the patents. Google argues that Oracle’s experts are not expert as they had no intimate knowledge during deposition.

see Oracle v. Google – Patent Marking – Closing the Gap

  • Jan 30 / 2012
  • 14

Good Years for

I started this blog back in 2007, more or less for fun. Later, I got more serious as I found many popular forums on the web quite biased against FLOSS. Hard work pays off. Thanks to all who contributed.

The data points are statistics for the site and the smooth curve is an exponential growth curve for 261% per annum. This month is the best month ever and there is still a day left. This is beginning to look like full-time occupation. I hope I have some time left for gardening, hunting and fishing in 2012.

  • Jan 30 / 2012
  • 24

US Government is a Website Vandal

The US government in its haste to support big businesses seized a bunch of uploading sites. They claim big business loses money due to uploading copyright material to servers. Tons of data is at risk of being deleted. Certainly some of that data is an illegal copy but much is not and folks who have used uploading sites in good faith are now threatened with losing their data with no due process. The sticking points are that MegaUpload’s funds have been seized as well and they cannot pay the bills for storage and users cannot access their own data to move it.

I can see a lot of lawsuits in the future and liability for taxpayers who may have to pay the bills. I can see people all over the world refusing to store any data on any server in US jurisdiction. This is yet another sign that the USA is going down the technological drain. The world does not need the bureaucracy of the US messing up IT.

see Feds: Megaupload User Data Could Be Gone Thursday


Megaupload users face data deletion US prosecutors warn

  • Jan 30 / 2012
  • 47

Oh Boy! Tilera Servers Out in March

“the ramp for the Tilera chips has been pretty steep, with over 80 engagements with system and network equipment vendors of all colors and stripes, and 20 design wins where the company has committed to use a Tile processor”

Wow! A Tile processor uses a bunch of RISC CPUs on a chip in a mesh. They have 64bit processing and 40bit addressing. The idea is to get close to one processor per thread so that fewer context switches and massive parallelism will get a lot of throughput at lower cost than x86 with SMP. For servers this makes a lot of sense and because they are optimized for Linux and have tools, porting is trivial. Lots of software that runs on GNU/Linux will be able to move quickly to servers running these things. Sampling is happening and production will happen in March. 2012 will be even more interesting than Android/Linux v world.

Of course, there will be particular applications where Tilera offers no advantage but there are lots that will. Anything on a busy website should be helped. Super-computing is a natural. Big data too… I would bet that as this matures, Google and other big players will slurp them up. That could really dent the bottom line of Intel/AMD which means price/performance could improve a lot for us ordinary mortals who do one thing at a time.

I would not be surprised if this technology eventually gets to the desktop. Most desktops don’t need this but terminal servers may be natural. Many of us have ~100 processes running. A hundred users might benefit greatly having one processor per process.

  • Jan 30 / 2012
  • 2

Weather in UK

The weather office in UK has quit providing widgets for GNU/Linux desktops thanks to Adobe dropping support for AIR in GNU/Linux. There is a workaround. On my PC, running Debian GNU/Linux, there is an app called “metar”:
metar -d eggp
EGGP 301220Z 11008KT 8000 FEW020 04/M00 Q1027
Station : EGGP
Day : 30
Time : 12:20 UTC
Wind direction: 110 (ESE)
Wind speed : 8 KT
Wind gust : 8 KT
Visibility : 8000 M
Temperature : 4 C
Dewpoint : 0 C
Pressure : 1027 hPa
Clouds : FEW at 2000 ft
Phenomena :

It’s trivial to extract from that whatever you want:
metar -d eggp|grep ^Temp
Temperature : 4 C

That’s for Liverpool, but many other airports publish METAR data.

It’s trivial to periodically refresh that information in a window or title bar in GNU/Linux, making your own widget. I have mine on the menubar which I access with a single click but I could also make it update automatically in a virtual desktop, say, using crontab. Either way, I don’t need to rely on Adobe for weather info. Isn’t FLOSS grand? Independence from unreliable software suppliers is priceless.

  • Jan 29 / 2012
  • 0

Strategy Analytics: Consumers are increasingly buying tablets in preference to netbooks and even entry-level notebooks or desktops

In Q4 2011, Strategy Analytics found that 26.8 million tablet PCs shipped with these distributions:

iOS 57.6%
Android 39.1%




26.8 million is 150% more than the same quarter last year so stay tuned for more growth and more slippage by M$ in the PC market. “Others” includes BlackBerry, WebOS and MeeGO, I suppose. M$ is thick with that bunch… Android/Linux is gradually overtaking iOS. I predict they will be even within a few months.

Share stolen from Wintel? 26.8 million is about 30% of x86 desktop/notebook shipments so about one in five tablets is taking the place of a desktop/notebook. Expect continuing no/low/negative growth, M$. Some people prefer small cheap computers.

  • Jan 28 / 2012
  • 17

Ubuntu’s HUD: Why It’s A Terrible Idea

I just read another article proclaiming why HUD (Head Up Display) is a great idea. The gist of it is that

  • HUD learns on the job improving its performance with use.
  • HUD allows you to search for menu items by hitting Alt + search terms
  • HUD is faster and easier to use/learn.

In fighter aircraft the idea of HUD was to allow a pilot to see important stuff while looking through the windscreen for important stuff allowing intricate operations without taking the eye off either. That‘s a good thing. Ubuntu’s HUD is not. Continue Reading

  • Jan 27 / 2012
  • 56

It’s True. More People Use Smartphones Than Desktop/Notebook PCs

A survey of 2000 people 18 and over in several developed countries/markets found that 68-86% used desktop/notebook PCs but 76-96% used smartphones. USA was on the low end while UK, France, Germany and Japan were successively higher in both categories.

Besides the bare facts, I conclude

  • 8-10% use smartphones but not desktop/notebook PCs
  • USA, which is on the low end of desktop/notebook usage is on the high end of tablet usage (11%) so tablets are definitely cutting into desktop/notebook usage

This is entirely consistent with M$’s drop in client division revenue. M$ is barely present in smartphone/tablet markets. 10% can have a huge effect on consumer demand. Decimation gets noticed. I expect this will grow and M$ will continue to drop in market share for the next few quarters and “8″ will have little traction. People prefer chocolate bars in their hands rather than vapourware. The acceptability of */Linux on ARM and consumers’ influence on business IT could well open the door to more widespread use of */Linux on desktop/notebook PCs.

  • Jan 27 / 2012
  • 17
Linux in Education, technology

OLPC in Australia

Thanks to reader oiaohm for providing a link to the following video presentation on the One Laptop Per Child implementation in Australia. Key points:

  • remote locations require a programme like OLPC to bring in IT to education
  • focus is on younger students and using IT to teach, not teaching IT
  • just dumping in the technology does not work
  • teachers, schools and communities need to be prepared/gotten on board to bring success
  • students are a huge asset
  • educational results are dramatic
  • it’s GNU/Linux and they provide both Sugar and GNOME

This documentary is about a programme to bring the XO notebooks to hundreds of thousands of students over the next five years. TCO is about $380 with the local school paying only $80 for training/support. $300 is paid by corporate sponsors such as banks/ISPs in Australia. see

Note on the presenter’s T-shirt:“No, I will not fix your computer.”

  • Jan 26 / 2012
  • 50

Geography of GNU/Linux

NetApplications used to restrict geographic reports of web stats to subscribers. Today they will let anyone have the information:

GNU/Linux share for the past month (counting only GNU/Linux, MacOS and that other OS)

Some of those numbers are pretty strange but there are a lot of populous countries with more than 1% share of hits going to GNU/Linux. It’s interesting that USA embargoes Cuba on lots of things but in the above list they are the highest users of GNU/Linux. Maybe they have more in common than they know.

UPDATE I guess they had a misconfiguration. Now they ask for authentication…

  • Jan 26 / 2012
  • 34

Opportunities Lost

One of the things I read about M$ is that M$ facilitated wide use of PCs and so has been a blessing over the years. While it is true that other OS was cheaper than UNIX licences back in the day (~$1000), the licences still cost too much to bridge the Digital Divide.

We can see this clearly now that Android/Linux on ARM is allowing other technologies into the market. Costs per unit can be under $100, several times less than Wintel. The number of people using IT could thus double within a few years, far greater growth in numbers than is typical using only that other OS in IT.

M$ of course mostly cares about its total revenue, and not about units shipped, but imagine M$ had adopted GNU/Linux a decade or more ago and encouraged ARM. They could have cut their licensing/support fees in half, reduced expenses and had a decade with treble rate of growth, giving them even larger revenue than they have now. They locked themselves in to lower profits at the same time that they locked in users to M$’s way of doing things. They now have to raise licensing fees to maintain profits on a decreasing share of the market which will hasten their demise.

At the same time, a billion people, more or less, were denied IT by M$’s short-sighted behaviour.

The score: In a decade of error,

  • 10 billion person-years of computing was lost,
  • $100 billion in profits was lost by M$ alone,
  • billions were kept in poverty years longer than they should have,
  • Earth was polluted/raped by the material wasted/used in PCs replaced every few years, and
  • the world spent $billions more fighting the malware and bloat and re-re-reboots of that other OS.

So, rather than congratulating Bill G and Co. for their success, we should pity them for their failures and regret having given them any business at all. This is essentially what will happen in 2012 as M$ loses more market share, comes late to market with that other OS on ARM, and */Linux and ARM bring IT to the next billion people.

In 2012, the installed base of that other OS on x86 PCs could well fall to 70% or less and units shipped with that other OS could fall to less than 50% at the same time that Android/Linux smartphones ship 300 million units and tablets ship 200 million units and GNU/Linux ends up on 100 million more desktops/notebooks.