Robert Pogson

One man, closing all the windows.

Monthly Archives / June 2010

  • Jun 25 / 2010
  • 0
technology

The Closed Community of Apple Produces a Lemon

The wonderful ithingy we here about has a major bug. If you hold it in a fairly natural way the 3G signal strength drops greatly. A poorly designed antenna is the problem. Holding the phone loosely seems not to show the problem but if you were stressed in an emergency or while parachuting, you might squeeze it a bit, I figure.

So, while those guys were running about with the prototypes in disguise and not squeezing them they missed a fine testing opportunity. Too bad they sent the cops after Gizmodo or they might have had real-world testing.

Openness works, even in tech.

  • Jun 25 / 2010
  • 2
Linux in Education, Teaching, technology

GNU/Linux is More Than Good Enough in Education

Here’s a blog about one of my favourite subjects, using GNU/Linux in education. The authour has it right that most uses of IT in schools do not require that other OS. I disagree on a few minor points.

I think GIMP is just fine for web development. Adobe has advantages for print but not the screen. Monitors are RGB last time I looked… I think Office 2010 or even Office 2007 have nothing to offer schools they could not obtain using other components of FLOSS such as LAMP with PHP and MySQL.

My high school just did reports using word-processing documents file-shared and write-locked. I don’t think we had a single collision amongs four teachers and all their students’ reports. Some teachers used GNU/Linux and some used that other OS. It all worked. Indeed the simpler interface of FLOSS apps tends to be easier for students to learn which lowers the overhead of introducing them to particular apps.

The authour is right on when he points out that saving money per-seat really pays in education, freeing funds for other things. We should recommend FLOSS for everything in educational IT unless there is a compelling reason. That that other OS is out there is not compelling. We prepare students for the future not to be slaves to M$.

  • Jun 25 / 2010
  • 0
technology

Blog on OS Fanbois

Adrian Kingsley-Hughes gets somethings right and a few things wrong in “They exist now only in the minds of fanbois …”

  • There’s a War Between M$ and Apple – He says there’s not but M$ is at war with everyone, even its partners, by taking the lion’s share of profits in IT. M$ and Apple have sued each other and made things not work for users of the other OS. Remember the blue “e” on MacOS? Where’s Office 2010 for MacOS?Where’s the ribbon?
  • The Year of the Linux Desktop is Coming – He says it’s not and he is right. It is already here becoming widely accepted in 2009 thanks to Vista, netbooks and ARM.
  • Open Source = Secure Code – He says everyone has bugs and he’s right but open source means they get fixed faster and you can check that your bug is fixed before you use the code. That is security. He also neglects the fact that if your code is public, you will make a better effort to make it right. If your work is hidden in a binary, it is easier to shrug off lots of details because no one will know about it.

Not a bad effort as blogs go, but I had to clear up a few things.

  • Jun 25 / 2010
  • 0
Uncategorized

Games in GNU/Linux

The games market is pretty small but vigorous. One supplier of games for that other OS ported a game to GNU/Linux in six weeks and published the web-stats that resulted. They claim their investment in the port was repaid and sales/downloads continue. Their estimate is that 15-21% of their gamers use GNU/Linux and the release of the GNU/Linux port caused a spike larger than that for the MacOS and other OS versions. So much for the 1% number… I think that gaming is fairly OS-independent test of popularity. No doubt some GNU/Linux users bought the game just to wave the flag, but I doubt that. I would not, but then, I am no longer a gamer.

  • Jun 25 / 2010
  • 0
technology

Unravelling US Extra-territorial Law

One of the side-effects of waiting breathlessly for Bilski at SCOTUS is seeing other things that the supremes decide.

A recent opinion slaps down rulings made by lower courts over decades that US securities fraud laws applied globally. There is so much assumption in the culture of the USA that the world should do things the way they do them: film, politics, copyright, software patents…

Here’s hoping the decision on Bilski comes out Monday with similar clarity and kills the evil of software patents.

  • Jun 25 / 2010
  • 0
technology

Intel Hedges Bets

Intel is one half of Wintel but they are not stupid. Everyone can see the writing on the wall:

  • small things work
  • that other OS is getting serious competition from GNU/Linux

The result is that Intel is investing in GNU/Linux, porting Android 2.2 to x86. This will re-open the door to netbooks for GNU/Linux and Java, two of M$’s nemeses. This is also a good plan for Intel to keep Atom vigorous by making the same software available for the ARMed netbook available on Atomized netbooks.

It’s all good. There is choice.

Read the comments on the Reg.

M$ is down to pre-”7″ stock values. So much for “7″ saving the company or Ballmer.

  • Jun 24 / 2010
  • 10
technology

YES !!! BILSKI !!!

(wrote the intro a few days ago. Opinion was given 2010-6-28)
Today, as I had hoped for months now, we have a decision in Bilski, which is about more than patents on methods of doing business. It is about software patents, the patent trolls’ delight. M$ has been using them to extort and inhibit GNU/Linux globally so it is about time they were laid to rest. I wonder whether or not the suckers that paid M$ to leave them alone will ask for their money back. I wonder how big a hit the loss of $billions of value in their software patents portfolio will have on share-price.

The decision is to affirm the lower court’s denial of Bilski’s patent. Unfortunately they also state that the “machine or transformation test” is not sufficient to deny a patent. Business methods are not categorically excluded, either…

Stevens concurred but emphasized that stating that the “machine or transformation test” was not the sole test did not suggest the Court ruled that many patentable ideas could not be identified or ruled out by it. He also stated the reasoning should have been that business methods are not patentable:“These clues all point toward the same conclusion: that petitioners’ claim is not a process, within the meaning of S101, because methods of doing business are not, in themselves, covered by the statute.”

Breyer wrote:“In sum, it is my view, that in reemphasizing that the “machine of transformation test” is not necessarily the sole test of patentability, the Court intends neither to de-emphasize the test’s usefulness nor to suggest that many patentable processes lie beyond its reach.” Scalia agreed with him.

So the Court is ambiguous on software-patents but did not shoot down the test. If a software-patent application is for some algorithm represented by code it is abstract and not patentable IMHO.

M$’s stock was down 1% on early trading but rebounded to down 0.5% on the ruling. I think we have seen the end of M$’s threats now. No one should take them seriously as the Court certainly did not.

You can read the ruling here.

Update: Other views of this opinion are all over the map. Experts keep score on the parts of the opinion:

9 justices agreed that Bilski fails to get a patent.

4 justices opine that patents may be broadened in the “information age”

5 justices opine that the “machine or transformation” test is too narrow and business methods can be patentable

see ScotusBlog

My reading is that software is an abstraction and thus is not patentable. 5 or more justices not agreeing supports this. I think a lot of software patents will have to be chucked. M$ ended down 0.91% on the day, that’s down $18 billion. Whether Bilski is the reason is debatable. Down is still down. Why is a monopoly down? Because the table has been tilted against them by their own actions or changed circumstances. Perhaps it is a little of both.

  • Jun 24 / 2010
  • 18
technology

XP/”7″

“Dell estimated the percentage of commercial PC users who have updated to Microsoft’s Windows 7 as still in single digits.”

That says it all. The most successful version of that other OS ever, is not being adopted by business who have no need of feature bloat and eye-candy that consumers lap up having few choices in retail.

If Dell hangs its hopes on a revival of the Wintel monopoly soon amongst customers who have choices like GNU/Linux and MacOS, they need to develop Plan B immediately. Dell’s competitors will have a shot at the same growth, if it ever comes. Now, Dell needs something to differentiate itself from its competitors like promotion of GNU/Linux as the best and most efficient platform for IT in business. Business is about making money not providing money to M$ and Dell. Business is looking for more cost-effective IT and if Dell does not give it to them, someone else may, perhaps a competitor.

More than 50% of business still used XP and they will continue to use it as long as possible. That could be years yet and GNU/Linux is already being chosen by many. The longer business looks a the cost of replacing all their IT infrastructure to go to “7″, the less likely that will be to happen. They can get like-new performance from most older equipment by re-purposing it as thin clients and only upgrading the servers and network. For all but the smallest of businesses this is obviously the best approach. Do the maths:

  • save $50 on each hard drive not bought,
  • save $50 on gigabytes of RAM not needed,
  • save $100 or more on licenses for software,
  • save $hundreds on lower cost of maintenance,

and multiply by thousands of seats. Even if they go to “7″ on terminal servers, M$ still charges per-seat license fees, eating into all that saving. Dell will lose revenue in the long term by the migration to thin clients but in the short-term, the supplier who gives the customers lower costs will win. Dell needs a larger share of that lowered cost of IT to maintain profitability. They cannot get it by sticking with M$.

  • Jun 24 / 2010
  • 0
technology

Buddies

Do you have friends you love who are squabbling? I do. DevonIT and IBM. I have done business with both of them over the years and they make great products. DevonIT is an important supplier of thin clients. I used their boxes at Easterville and they worked perfectly. Ordering and shipping was very smooth too. IBM made many of the PCs in use here and Lenovo made our newer ones. Our older machines are pushing 10 years of service and only a few have failed. In the 1970s, I bought a tape reel from IBM so I could transfer magnetic tape from a UNIVAC spool to one that would work on our IBM System/360. IBM’s local guy did not know what to do with cash… I was just a starving student at the time, not a corporation.

Anyway, in 2007, IBM, being big in servers, and DevonIT, being big in thin clients that needed servers, began making deals to develop and ship various blade servers and clusters of same to form a pillar of the expanding (slowly…) thin client business. DevonIT contributed development funds and IBM was to produce and distribute the server products. It should have been a deal made in heaven. DevonIT would get great credits in dealing with businesses who will sooner or later need thin clients and IBM would get a bigger piece of thin client servery with a leading supplier of thin clients.

Then came the slowdown and the project fizzled. DevonIT who contributed the funds with borrowed money defaulted on a loan and DevonIT passed the buck to IBM by suing them for a bunch of things. DevonIT was particularly ticked off that IBM had cancelled the project for blade servers while soliciting contributions for the clusters… Ouch!

I hope they will resume being buddies now that thin clients and all the other IT business is picking up, but it is in the courts so anything can happen.

  • Jun 24 / 2010
  • 0
technology

Do You Like Speed?

Speed is probably the most important thing in IT for me. I want what I want when I want it. Newer hardware helps. More efficient algorithms help. Google has made some developments that will help the speed of web browsing, a bit here and a bit there.

They have done some work on the TCP stack, HTTP headers, their browser, of course, and DNS. What has M$ done to help you get things done faster lately? Nothing, except to slow your PC so you rush out and buy another with newer hardware and a new licence fee for them.

Way to go, Google!

  • Jun 23 / 2010
  • 0
Uncategorized

Who’s Driving the Bus?

The bazaar approach to Free Software development is hard to understand/categorize. If you want something done, you either do it yourself or get someone else to do it. When no one is in charge, that gets interesting.

KDE was my first GNU/Linux desktop environment. I still remember the silliness of accidentally clicking on the hole in the gear and nothing would happen… It was utterly solid compared to that other OS I was using at the time. Then there were the library wars and I went with the flow to GNOME. On some older PCs I find in homes and schools, I use XFCE4 because it does what is necessary and not much more.

While I wandered in the forest looking for mushrooms, KDE evolved, or someone turned the wheel of the bus in the direction of a “new desktop paradigm” etc. Stuff I cannot imagine. I naturally clutter any flat stop with stuff so I can see what I am currently using. The old paradigm worked for me. Even the ancient MacOS of the 1970s worked better for me than some of these new ideas. I have no memory, at least not one that works in real time, so I need to see clutter. Eliminating clutter eliminates the usefulness of a desktop for me.

Others, though, are driving the bus of KDE and have chosen to “improve” the desktop. Others, who feel as I do are trying to preserve the look and feel of the 3.5 version. Whether the group doing the work can sustain an independent branch of KDE is a question. KDE is large and complex and the libraries it depends upon changed, causing some of the development of the 4 branch.

Apparently, I am not alone. The site of the developers, http://www.pearsoncomputing.net, was /.ed… Perhaps the group will get enough support to make a second KDE branch work. There are those who love change for the sake of change or for some particular features, but “if it isn’t broken, don’t fix it” applies to software, too.