Published by Robert Pogson July 30th, 2010
in technology.
“Plaintiffs cannot meet their burden with an expert’s general statements about economic theory, and simply throw up their hands when record real-world facts fail to conform to economic theory,” is the quotation from a report filed in the suit against Intel.
This is about Intel paying Dell and other OEMs a “discount” to ignore AMD CPUs. The report to the court likely is about the prices paid. In the absence of sales of AMD CPUs it is hard to demonstrate that consumers/customers paid too much but what of the lack of choice? Combined with Moore’s Law it is hard to prove that Intel’s prices were “too high” or “higher than they would have been” without the discounts. This is silly when you consider the size of the bribes in $billions. If Intel’s prices were not too high, how did Intel imagine they could recoup the payments? Increased volume? Supply and demand do work. If only Intel’s chips are demanded by OEMs then Intel’s prices can be higher. When I look at the price of CPUs I find that Intel charges about $100 more for a chip of comparable performance. When AMD64 came out, AMD did, for a time charge higher prices but that did not last long as the OEMs boycotted AMD. Lack of choice means lack of competition which means higher prices. Also, a customer may prefer AMD but be unable to find them from the usual suppliers with a boycott. The end result of Intel’s discounts would have been the eventual demise of AMD. They had to reorganize, shedding the fabs. If AMD had gone under, Intel could have cleaned up. Do the courts need to see that happen to see that consumers were harmed? AMD was certainly harmed. Don’t they count?
- Robert Pogson
Published by Robert Pogson July 29th, 2010
in technology.
In IT there are problems:
- malware
- energy consumption,
- space,
- networking, and
- maintenance.
For desktop computing no one has had a problem with speed/throughput since the 32bit days of single cores at 135nm. Now 22nm and quad-core are on the horizon with DDR3 2666. This architecture will solve problems for Intel by giving them lower-cost production/more CPUs per wafer and, under-clocked, they will use less power to do the same job for the end-user but no one is likely to see quad-core under-clocked. Are they? These solutions are mostly going to be unnecessary for end-users and will be used by the “partners” to up-sell IT. They will even encourage more feature-bloat from M$.
Servers and a few CPU-intensive niches like graphical engines and games may benefit from the new hardware but where are the thin client chips using 22nm? The world needs those to solve all the real problems, not just window-dressing for the showrooms. 22nm can drive down the cost of thin clients to the point where
-
every monitor will include a thin client at no extra cost to the end-user,
- dedicated thin clients could be the size of a thumb-drive,
- cost per-seat for clients could reach $100
- a thin client could fit easily in most keyboards, and
- every cordless phone would be “smart”.
Instead Wintel will do its best to up-sell and to prevent thin clients from ever becoming mainstream by evangelizing overkill and taxing thin clients as best they can. Fortunately, ARM does not care what Wintel does and ARMed thin clients will take up the slack by next year.
- Robert Pogson
“Schwarzinger finds that schools in Austria are increasingly turning to OpenOffice. The Austrian ministry of Education is supporting this move, paying the schools 10 Euro for every PC that uses this suite of open source office productivity tools.” Wow! and a bunch of schools are using GNU/Linux.
Lower costs, flexibility, and support for open standards would be my guess as to the causes. The one barrier mentioned is that IT workers are still more familiar with that other OS but these moves indicate that is changing. I like the fact that the government is actually paying schools to use OpenOffice.org. The advantage to the government is that they will have future citizens and employees familiar with OpenOffice.org and since the government pays for licensing that other OS and its Office suite, this saves them money. Say that other Office costs the government $50 per PC. By bribing the schools to use OpenOffice.org $10 the government saves $40 per PC not getting Office and increases the proportion of schools using OpenOffice.org. The schools actually get an income to cover the trivial cost of installing OpenOffice.org so they have some slush for their IT programme or whatever is their priority for spending. That’s brilliant.
- Robert Pogson
Published by Robert Pogson July 28th, 2010
in Uncategorized.
Candace Hoeppner has done a lot of work to see that the registration of long-guns ends in Canada and she still is working hard to see that a couple of votes in the House of Commons approve her private member’s bill, C-391. She had a good interview about the process. As usual, keep in touch with your MP to support the bill. She says a number of NDP and possibly one Liberal are likely to vote for the bill but the Liberal leadership is asking all Liberals to oppose the bill. We shall see in September whether the bill goes to the Senate this year.
- Robert Pogson
Published by Robert Pogson July 28th, 2010
in technology.
Yes, Dell has shipped a lot of GNU/Linux but, no, they don’t make it easy.
Google finds five hits for “site:dell.ca ubuntu”, some of which are dead and the “choose OS” thingie does not list ubuntu.
Google finds 37000 hits for “site:dell.com ubuntu”.
Does that make sense to anybody? Is Dell a global corporation with the corporate knowledge that GNU/Linux does well in and outside the USA? Does Dell actually know how to sell stuff? Imagine a car dealer with “Keep Out” signs all around the lot. Image a fish monger with huge signs saying “Our fish stink!”. That’s what Dell is doing with Ubuntu and GNU/Linux.
- Robert Pogson
Published by Robert Pogson July 28th, 2010
in technology.
To me tablets are pills I swallow from time to time or the thingies the UPS drivers have me “sign” on.
This is July, 2010 and in this Year of ARM, the tablets are coming. KMART will sell some. This is a new price-point, too, $150. That makes them affordable by most KMART customers and they should move a lot. I guess KMART never got the memo they should not give shelf-space to GNU/Linux…
This second half of 2010 should be quite hot for such devices. The manufacturer, Augen, makes all kinds of consumer gadgets and is not a “partner” of M$ and so in uninhibited by taxes and exclusive dealing with M$. Their retail partners may be uninhibited too. KMART sells products from M$, too, but may not have gotten the message. Interestingly Augen and KMART push Android and not ARM. Perhaps it is the year of Android, too.
I have been predicting such moves for months. ARMed OEMs will spring out of the woodwork pushing all kinds of ARMed thingies and selling them everywhere at much lower prices than Wintel can compete. By the end of the year normal PCs should be shipping with ARM and GNU/Linux. This is more than a month before school starts. Expect a big sales push now and just before Christmas in North America and Europe. Expect a big sales push in the rest of the world for the next year… Expect M$’s “partners” to be clamouring to get on the ARM/GNU/Linux bandwagon within the year. They will not put up with M$’s drag much longer.
- Robert Pogson
I have had Grade 1s in a computer lab with a bodyguard. It is a frightening experience because you never know what the little ones will do next but you can be sure they will be doing something…
My respect goes out to Helios and friends who gave a 4 day summer workshop for 22 kids. The kids not only got to know the ins and outs of ATX PCs but also installed GNU/Linux and played games on it. I would bet everyone involved learned a lot. 
This shows that GNU/Linux is not only for geeks. If you have some grown-ups in your organization who are reluctant to change, perhaps this example would inspire sufficient effort. The benefits outweigh the costs:
- relative freedom from malware
- relative freedom from anti-malware
- freedom from monopoly, and
- superior performance at lower cost.
- Robert Pogson
Published by Robert Pogson July 25th, 2010
in technology.
M$ says it’s “all-in” with cloud computing and they want to convince business, in particular, that would be a reliable “partner”. Then they do this:“Windows Live shares your Messenger contacts”. In TFA, you will read that “private” means M$ thinks it’s OK to tell everyone with whom you communicate, the list of everyone with whom you communicate! So, now your boss gets to know you are talking with the competition, or lady A knows you are friends with lady B. It sounds like a mine-field.
Social networking has its place in business as in other activities but such openness makes it very difficult to keep confidences. The problem with Live is that M$ took Hotmail and Messenger and plugged them into its social network with single-signon. Businesses used Hotmail and Messenger for business purposes and now they are tangled with non-business stuff. I can see M$ doing this to increase the size of its network but I cannot see business or anyone else for that matter sticking with Live.
Where I work, there are several people locked-in to Live. They think it is the way to do everything on the web and completely confuse Live with the Internet and Live with e-mail and Live with their lives. They are absolutely amazed that you do not have to belong to Live to go places on the Internet. I can understand that. Every new machine coming out of the big-box stores promotes Live even when you fire it up the first time. I remember taking a Vista machine out of the box years ago and finding it insisting that we needed a Hotmail account to do updates… (or something – they make it very complex and they are great salesmen). It’s just another reason to avoid M$. We should not be pressured to become a partner of M$ even if we are foolish enough to use their OS.
Of course, social networking is not “the cloud” but it is part of it and M$’s clouds are dark and dangerous. Their motivation seems to be to extend the monopoly. If your role in life is not to support the monopoly, be careful to avoid entangling yourself deeper with the monopoly. Find organizations who will deal fairly with you. Organizations you can trust. I have trusted Google for a long time. They provide a good service at a great price. I heavily use their Gmail and search and maps and sometimes Trends. I have started using the desktop search function of the Google toolbar and the Google-chrome-browser. Those last two are a bit risky but they work so well and simplify my life so much that I think the risk is more than compensated by performance. I trust Google. I do not trust M$. I have not trusted M$ since I switched away from Hotmail a decade ago when my chief worry about them was spam dominating my in-box. That same year I switched away from using their OS. I have not regretted either decision.
- Robert Pogson
Published by Robert Pogson July 24th, 2010
in technology.
We have two monopolies supporting each other in IT: Wintel. Intel and M$ work together to support each other. M$ keeps cranking out more bloat that requires new hardware that begs for more features/capabilities in the software … Here’s what we can do about it.
Monopoly is not good for us. Monopoly is good for those who have the monopoly, in this case, two powerful corporations with fewer than a million people. We are thousands of millions. We can do more and better whatever the monopolists can do. Monopoly is not good for us because we pay too much for IT and are limited in what we can do with IT because we depend on what the two monopolists do. Then there are their partners. Need application X in 64bit? Nope. Need application Y to run on ARM? Nope. Need application Z to run on another OS? Nope. Need your network to be secure from intruders? Nope. Need an upgrade? Nope. Pay full price and you have to buy version 12.34 first, etc.
Hardware. We can buy ARM, AMD, even Apple. If you are locked into Intel because the stuff you run only runs on that other OS and it only runs on x86 you can change. You can use emulators, virtual machines, and terminal servers. There is no need to have the latest and greatest and most expensive CPU from Intel every few years on every PC so they can idle at a lower CPU utilization.
It all hinges on software. If your software is not platform independent why isn’t it? Computer Science 101 is about logic. As long as you do not build in a dependency on Intel or that other OS you are free of the monopoly. Your dependencies will be some list. Everyone using Wintel will have a similar list of dependencies.
If you count up all the folks who depend on some application it will be millions of users, perhaps hundreds or thousands of millions of users. If each one contributed $20 per instance, they could cooperate and build the most wonderful application that met all their needs in a year or so. The largest such cooperative project is probably the Linux kernel. It was started by one person who was joined by a few dozen others and they were successful in a year or so. That project has thousands of contributors now and it is very healthy. Sun did that with OpenOffice.org. They looked at the cost of a new round of licences for Office and bought the company producing StarOffice and ran with it. It paid them to do that for a few tens of thousands of licences. There is no reason at all that the world cannot produce the software it needs without relying on monopolists. There is no magic in software. If you specify the features, inputs and outputs, any decent programmer can create an application that does what you want given enough time. Given enough programmers the job can be done in less time.
Invest in FLOSS. Free yourself from monopoly. You can start right away by migrating parts of your operation to GNU/Linux and identifying the parts that do not migrate readily and fix the causes of that non-portability. Fix it by finding a FLOSS project that does what you need done or creating one. There are lots of resources on the web. FLOSS is reusable so you do not have to reinvent the wheel. Just use the wheels others have developed and contribute to the world under a Free Software licence.
These days the monopoly seems very healthy because it is taking tens of billions of dollars out of the world’s IT budget annually. That is unstable. If the world invested that amount in their own software and produced platform independent software with the money, monopoly in IT could be a thing of the past in just a few years. The advantages would be many beyond more cost-effective IT: local employment instead of corporate outposts, improved interoperability since everything would work with everything, and no more malware since the world can do a better job of producing secure software without copying vulnerabilities from previous releases.
- Robert Pogson
Published by Robert Pogson July 24th, 2010
in technology.
These days, it seems impossible to find Ubuntu on Dell sites. You can find pages with a checkbox for FreeDos and Linux but they lead to the ever-popular FreeDos and Redhat WS. No more Ubuntu. No explanation…
Could it be that:
- an agreement with Canonical has run out and has not been renewed?
- that the most popular distro of GNU/Linux does not sell on Dell’s chaotic site?
- has M$ increased the payments to downplay Ubuntu?
Since Dell has recently been caught out boycotting AMD and accepting payments for the boycott from Intel, is it not very likely that Dell would boycott Ubuntu upon payments from M$? What do you think?
“Intel made exclusivity payments to Dell in order for Dell to not use CPUs manufactured by its rival — Advance Micro Devices, Inc. (AMD). These exclusivity payments grew from 10 percent of Dell’s operating income in FY 2003 to 38 percent in FY 2006, and peaked at 76 percent in the first quarter of FY 2007. The SEC alleges that Dell Inc., Michael Dell, Rollins, and Schneider failed to disclose the basis for the company’s sharp drop in its operating results in its second quarter of FY 2007 as Intel cut its payments after Dell announced its intention to begin using AMD CPUs. In dollar terms, the reduction in Intel exclusivity payments was equivalent to 75 percent of the decline in Dell’s operating income. Michael Dell, Rollins, and Schneider had been warned in the past that Intel would cut its funding if Dell added AMD as a vendor. Nevertheless, in Dell’s second quarter FY 2007 earnings call, they told investors that the sharp drop in the company’s operating results was attributable to Dell pricing too aggressively in the face of slowing demand and to component costs declining less than expected.”
That sin cost Dell $100 million and Mr. Dell $4 million in penalties. Do you think that Dell may have stopped the Ubuntu listings promptly after this became public out of a deep need to shoot itself in the foot? I believe that, having accepted exclusivity payments from Intel, Dell would accept exclusivity payments from M$.
There’s some history here. A year ago, The Var Guy posted that Dell had cut back on desktop PCs with Ubuntu… Dell recently agreed to push Canonical’s cloudy stuff but today, Canonical is not listed as a “partner”. Dell is not listed as a partner of Canonical.
It is hard to imagine that Dell thinks there is more future in FreeDOS than Ubuntu GNU/Linux.
- Robert Pogson
A guy has a nice report of a construction project, a PC he built for $200 and a bit of his time. It uses an AMD64 X2 CPU and 1 gB of fast RAM on a minimal motherboard. He was under budget and if the construction and installation time cost $50/h, this project cost less than $250. He then benchmarked it against a $300 box with “7″ installed in the factory. He got what he paid for and it is faster in every test compared to that other OS on similar hardware.
This is a fine example of people doing what they can do for themselves instead of paying someone else to do that. FLOSS is like that. Need an OS? Write one. Need an application? Install it. I have done such projects many times. I built servers for Easterville because we could get better servers, for our purposes, than we could buy. We installed Ubuntu GNU/Linux because it did what we needed done better than that other OS. Another advantage of such techniques is that one never needs to hire external support. Just fix it yourself or re-install in a worst-case.
The first PC I ever built on my own was a project like this. I wanted a GNU/Linux terminal server adequate to run a 30-seat lab. Never having done it before made it fun. The students and I researched price/quality/performance on every part and drew up a purchase-order requisition. When the goods arrived we then built the system and installed the software in class. It was an excellent educational project and amazed everyone with its performance. We tried to optimize every bottleneck and every price/performance point. We even graphed prices versus price per MHz for CPUs and price/MB versus memory module size to get the most for the least.
An OEM can always build a system in volume cheaper than an individual but when you add in the price of that other OS, an individual can beat the system.
- Robert Pogson
Published by Robert Pogson July 23rd, 2010
in technology.
Here we are a bit more than half way through 2010, The Year of ARM, and India has announced a $35 ARM tablet… I don’t know how they can build the parts for that price. There’s no room for M$’s cash cow in that. I was thinking the world would go to $100-$200 this year for ARMed devices. It could be subsidized but TFA reads “costs $35 to build”. This is unexpected, but wonderful. If true, India could ramp up production and connect the subcontinent in a few years for much less than Wintel would cost. This could change everything. China will be doing the same, I would bet.
- Robert Pogson
10329
7993
107
2
0
14301 
7198 
7152 
2718 
2028 
1227 
198 
7 
0 
0 
0 
0 