February 2008 - Posts
I'm straying into dangerous territory today and saying that I think our House of Lords is an important institution that provides a counterbalance to the House of Commons. Lord Benjamin Mancroft has created a furore today (I was just listening to him on the Jeremy Vine show on BBC Radio 2) by criticising the standards of healthcare he was given when hospitalised in Bath recently. He singled out the nurses as being 'grubby' and 'unprofessional' amongst other things (read the story at the BBC News web site). Whether he was right or not, I don't know and I'm not taking sides in that argument. However, he said one thing that struck me. I can't remember his quote word for word but it was along the lines of:
Most people don't have the platform that I have to raise these issues. I don't care what people think of me, but I needed to raise these issues on behalf of the people of Bath. I could not have done this as an MP in the House of Commons because I would have to worry about getting voted in again.
With the main political parties having just a gnats whisker between them in terms of policies and prime ministers becoming ever more presidential, our House of Lords is more important than ever. The government clearly finds this reality check inconvenient, therefore the moves towards abolishing the Lords as we know it and replacing it with an elected body. An elected House of Lords will presumably be voted upon by the same voters electing the government, the likely outcome will be a House that agrees with the government all the time. It will therefore cease to provide the important counterbalance that it does today and will become irrelevant. We cannot afford to lose our second House. It is important that the House of Lords is appointed and not elected, precisely so that the Lords do not need to worry about votes and can feel free to raise these important issues and force them into the public arena.
I attended a seminar a few days ago that explored current and emerging uses of technology and communications in business. The event was hosted by Rhondda Cynon Taff Business Club. There was a presentation by Light Age Communications on what's currently available for businesses in technology land and a talk on Web 2.0 and social networking by Mike John (research student) and Dr. Geneen Stubbs both from the University of Glamorgan.
To set one apparent controversy straight, having lived and worked in Silicon Valley, California for four years where such things are invented before breakfast, I think I'm safe in reporting that the correct pronunciation of Web 2.0 is "Web two point oh" - but it really doesn't matter how you say it, it's the thought that counts.
If you're interested in Web 2.0 or social networking, then Mike's blog On a Hill will be of interest. He has an article On Twitter, Hillary, Barack and Paul that has me thinking about how useful all this stuff is. I must admit that I haven't tried twitter, but I have yet to be convinced that applications of this genre have any useful purpose. I'm almost certain their popularity has more to do with feeding the egos of bored teens who imagine their friends need to know exactly what they are doing at all times. One day, I hope, they'll all wake up and realise there is more to life than what happens on the screen of your mobile phone. When I was a teenager, I made models, played in the forest, built rope bridges, dammed rivers and went off on backpacking expeditions around the Isle of Wight. I never felt remotely upset that some of my friends didn't know what I was up to every moment of the day.
I've tried Facebook and I feel pretty much the same way about that, too. In fairness, Mike pointed out in his talk that in tiny writing tucked away at the bottom of the page is a link to "business services" that merits further investigation, so I'm reserving judgement on Facebook for a little while longer while I try that out.
Some Web 2.0 stuff is pretty cool, though. Wikipedia is the classic example (here is my one and only entry: Llantwit Fardre). I refer to Wikipedia all the time and regard it as one of the great accomplishments of the Internet age. Some will say that information found on Wikipedia should not be relied upon, because "we don't know who has written it". However, I would say the same argument applies to books. Certainly, the author's name is on the cover (as it is on a Wikipedia article) but that is meaningless unless we actually know that person. I think it can be argued that in many cases Wikipedia is likely to be more reliable because it is arrived at by concensus.
Poignantly, during Mike's talk, he tried to show a page from Facebook only to be greeted with a big red "Access Denied" message from the firewall at Valleys Innovation Centre. Apparently, Facebook is not considered suitable for viewing during work time. I take a dim view of that sort of thing and I've made my views on blocking access to the internet clear. Local authorities are often on the trailing edge where technology is concerned and they are often the last to comprehend and respond to technology trends. It is therefore somewhat ironic they they are the very people charged with the job of promoting technology to the local business community.
Remember Clippy? Remember Office XP? It seems like the biggest selling point of Office XP is that Microsoft took away Clippy.
Those office assistants were a pretty neat idea at the time, but are now regarded scornfully. How times change. I was cleaning up my hard drive today and stumbled across the story of Clippy's demise. I thought you might enjoy reminiscing so I've attached the first episode. I have a couple more laying around if anyone is interested...
Oddly enough, I still regularly use a piece of software that uses an agent character - Merlin - to great effect. ACP (Astronomer's Control Panel) controls my observatory. I often leave it on 'auto-pilot' and while I am watching TV, Merlin chatters away in the background keeping me informed about what is happening.
I have taken on a new customer this week and they have an interesting business that deals with energy (gas and electricity), measuring consumption with smart meters and ultimately using less of it to reduce your carbon footprint and save money. They showed me a neat little box that measures the power being used by any device. They demonstrated this on their Dell desktop computers, which drew about 130 Watts in normal use. We then shut the computer down and I was amazed to discover that they were still drawing over 30 Watts. That adds up to a few tens of pounds per year to run a computer without even booting it up. For a single PC, that might not seem too bad but for a large organisation, multiply that by hundreds or thousands and you can see that quickly adds up to a lot of money for nothing.
I seem to remember someone from Microsoft saying that Windows Vista would use new power-saving features to bring power consumption down to about 1-2% of that required for normal operation when the computer is sleeping. That is certainly not the case on those Dell PCs, which were AMD Athlon CPUs and pre-installed with Windows Vista Business OEM. More like 30%, an order of magnitude greater than claimed.
I tried the same test on a "TiGra Networks Information Worker Desktop PC" and was pleased to discover the results were better. Lower power consumption of about 85 Watts during normal operation, despite having dual cores (Intel E6600), a fast 1066 MHz front side bus, 4Gb of RAM and a beefy add-on 3D graphics accelerator. When placed into hybrid sleep (S5 suspend), I recorded 7 Watts of consumption. Better than the Dells but still almost 10% of the normal power, nowhere near the claimed 1-2%. This does serve to confirm my own conviction that Windows Vista needs the right hardware to perform properly and it also shows that quality components do make a difference - we use Intel motherboards and CPUs and all other components are brand name and WHQL tested for maximum reliability and compatibility.
With energy prices soaring, it is easy to see how businesses waste a lot of money powering devices they are not even using. If you are interested in reducing your energy consumption and/or carbon footprint, check these guys out.
I have a few sites that are now totally Windows Vista Business and SBS 2003 where the users run as Standard User. One consequence of this is that printers have to be deployed by group policy, because standard users don't have permission to install print drivers. Windows Vista provides a mechanism to enable this, once you figure out that you need to update your active directory schema to get it to work properly. One particular site has two HP Colour LaserJet 2600N printers attached by Ethernet to print queues on the server. The printers are deployed using Windows Vista Printer Deployment (i.e. by group policy). This site has been having a problem where users could print but their jobs would just stick in the print queue and never print. The jobs could not even be deleted - until the Print Spooler service on the server was restarted, whereupon the job would print.
I've worked around this problem for some time now by giving a trusted user a script to restart the print spooler, but it was a messy hack and we were both pretty fed up with things not working like they were supposed to.
I opened a support incident with HP and they made an attempt to resolve the issue, had me try all the obvious but futile things like reinstalling the driver (like I hadn't tried all that already!). Surprise, no fix. Eventually, they informed me it was a "known issue with small business server" and as evidence they cited various other reports of exactly the same thing happening from various online sources. Flawed logic aside, I clearly wasn't going to get anywhere with HP so I opened a case with Microsoft instead. The attitude there was much better and I felt that Microsoft was really determined to stick with the issue until it was resolved - which it eventually was after debugging the print spooler and various other shenanigans.
So here's the scoop. HP's printer driver is broken. The HP driver installs a custom Print Processor that maps to ZIMFPrint.dll, whereas most printers use the built-in winprint print processor that maps to localspl.dll - the Microsoft engineer changed the print processor back to the Windows built-in one (as shown below in the screen shot) and everything works as it should. Now why the hell does HP insist on installing non-standard broken software when the built-in component works perfectly well?
Microsoft 1 : HP nil. This was a nasty problem that I would never have figured out without Microsoft PSS. Oh, and guess what - the very professional, well-spoken engineer I dealt with was based in Bangalore. I've seen a consistent negative attitude towards PSS engineers based in India especially from our American cousins. I just can't reconcile that with my own experience because I have consistently had my issues resolved by professional, well-spoken Indian engineers. I'm beginning to wonder which end of the phone line the problem is on.
[Update: Vik from PSS has reviewed my article and pointed out that I made two mistakes. First, he is located in Bangalore, not New Delhi (sorry Vik). Second, he asked me to point out that WinPrint provides basic features and so some advanced HP-specific features might not be available after using the above workaround]
I saw this on Sean Daniel's blog - classic. Have you ever wondered what Microsoft people get up to while they are waiting for their LiveMeeting to start?
If you are developing ASCOM drivers or applications in Visual Studio (i.e. you are generating .NET code), there's a compatibility issue you should be aware of that could prevent your code from running on 64-bit systems.
When you create a new project in Visual Studio, it assumes you want to generate both 32-bit and 64-bit code for maximum compatibility. The target architecture is set to "Any CPU" by default. Most of the time this is what you want but for ASCOM this causes a compatibility problem. There is a workaround but first some background information.
The Problem
The reason is this: some of the ASCOM platform components were generated with tools that predate 64-bit CPUs and consequently they contain only 32-bit code. The .Net Framework does not currently allow 64-bit applications to call into 32-bit COM objects. So, applications and drivers compiled using the default target architecture will run fine on 32-bit systems, but when they are run on a 64-bit system they will run as native 64-bit applications and will then fail when they try to bind to any of the 32-bit ASCOM platform objects.
Replicating the Problem
If you are using a 64-bit system you can replicate this condition with the following trivial C# code snippet:
using System;
using ASCOM.Helper;
namespace ASCOM_64bit_Test
{
class Program
{
static void Main(string[] args)
{
Chooser c = new Chooser();
c.Choose(String.Empty);
}
}
}
Create a new C# console application and paste in the above code. Then add a reference to the ASCOM Helper object:
Now run your project. Here is the result you will see:
The Workaround
Now, right click your project, go to Properties, Build tab. Change the Platform target from the default to 'x86'.
Now run your project again, and it will work.
Thanks to Peter Simpson (author of Conform) for his help in confirming this problem.
Update: I asked Mike Ormond from the MSDN team to peer-review my article as I wasn't 100% sure of my facts. Mike's response is as follows:
Well it’s not really an area of expertise for me but as far as I know you are correct. Essentially, a 64-bit process cannot load a 32-bit dll into its address space. By far the easiest option is to force the build target to be 32-bit as you describe. There are other approaches though which take advantage of the fact that 64-bit and 32-bit processes can communicate with each other even if they can’t coexist in the same process space. Eg this article in DNJ Online describes this type of approach: http://dnjonline.com/article.aspx?ID=jun07_access3264
It seems there is an instant messenger virus at large at the moment. The virus sends messages similar to these (the URLs are deliberately obfuscated):
"Hot or Not? hxxp://mymsngallery.my.funpic de/viewimage.php?youremail@someplace.com"
or
"this really looks like you hxxp://mymsngallery.my.funpic de/viewimage.php?youremail@someplace.com"
Opening the page actually calls up an executable file that infects your computer with malware.
If you need further advice on preventing and/or recovering from an instant messenger virus, you can find advice over on the TiGra Networks blog.
I'm not a religious person; I consider myself spiritual but I think religion does more harm than good. Nevertheless, I have much respect for Dr. Rowan Williams, Archbishop of Canterbury and head of the Church of England. He must be wondering right now what he has done to deserve calls for his resignation even from within his own ranks. The bigots have their knives out, but hang in there, Dr. Williams. I respect your courage to speak out, to challenge the established order and to stimulate debate. May you continue to do so.
I haven't read the full text of Dr. Williams' recent speech and, I suspect, neither have most of his critics. This isn't the first time the Archbishop has suggested there might be a need for the state to become more caring by reabsorbing some aspects of faith. I remember when he first took office as Archbishop of Canterbury and he gave the Dimbleby Lecture, his theme was more or less along those lines. This time, he's made a controversial speech suggesting that the legal system could benefit from being more inclusive and trying to accommodate some aspects of a religion other than his own. That in itself is noteworthy and shows he is a radical thinker. So before we all rush to defend our respective corners, whether we are religious or not, whether we are Christian, Moslem, or another faith, whether we agree with Dr. Williams or disagree with him, lets please show him the respect he is due - which I think he has earned - and at least try to understand his reasoning and enter into a sensible debate with him. The idea that he should resign for challenging our prejudices is just ridiculous.
From the Archbishop's web site:
In his lecture, the Archbishop sought carefully to explore the limits of a unitary and secular legal system in the presence of an increasingly plural (including religiously plural) society and to see how such a unitary system might be able to accommodate religious claims. Behind this is the underlying principle that Christians cannot claim exceptions from a secular unitary system on religious grounds (for instance in situations where Christian doctors might not be compelled to perform abortions), if they are not willing to consider how a unitary system can accommodate other religious consciences. In doing so the Archbishop was not suggesting the introduction of parallel legal jurisdictions, but exploring ways in which reasonable accommodation might be made within existing arrangements for religious conscience.
A classic example of what Dr. Williams is alluding to is the refusal of Jehovah's Witnesses to accept blood transfusions. There was a case a while back where a child of Jehovah's Witnesses needed a life-saving transfusion and the parents withheld consent. They were demonised in the press and of course the state overruled them using the justification that life must be saved at all costs - but who was right? Is the state justified in overruling matters of faith? Its not clear cut. We need to have these discussions and to feel comfortable expressing our views openly - that is what it means to live in a democratic free country. In the climate of political correctness society is too quick to condemn those with dissenting views. In the end, that diminishes the freedom of everyone.
The 2008 launch wave is gaining momentum. Microsoft announced a few hours ago that both Windows Vista Service Pack 1 and Windows Server 2008 have been released to manufacturing.
Windows Vista service pack 1 will likely impact most users in the short term. At TiGra Networks, we've been actively involved in testing pre-release versions of Service Pack 1 on both 32- and 64-bit systems and we're pleased with the improved stability and performance. We'll be upgrading immediately and will be recommending the same to our customers running Windows Vista. No doubt there will be a lot of information circulating in the press over the next few days and not all of it will be accurate or fair, so we'll try to keep an eye on what people are saying and present our real-world experience here on the blog.
You can expect to see Service Pack 1 on Windows Update and on the Microsoft Download Centre in mid-March, with deployment by Automatic Updates beginning mid-April. TiGra Networks will deploy Service Pack 1 for our customers as part of the normal automatic update cycle.
Windows Server 2008 is the first of a slew of exciting product launches slated for 2008. Visual Studio 2008 is already shipping and we can still look forward to SQL Server 2008 with its new data types supporting binary documents and geographical data types. The implications here are obvious for applications like SharePoint with its document libraries and Graphical Information Systems. Later in the year, of primary interest to me and TiGra Networks' customers, we can look forward to Small Business Server 2008 codename "Cougar". We are currently beta testing Cougar under non-disclosure agreement and can't discuss it publicly yet, but I think I can safely say it is going to be as popular as its predecessor. Later in the year TiGra Networks will be hosting a series of "IT Surgeries" for local businesses and we'll be showing off SBS 2008 "Cougar" as soon as we possibly can.
Chris Rue has a stinging appraisal of Groove over at the funcave.
"Groove has to be the worst software I’ve ever used in my life. [it] has the most idiotic user interface seen since cc:Mail. [...] To top it off, [...] this thing totally sucks at handling the group part of collaboration. Which is the only part of collaboration that makes it collaborative, to be quite honest."
Ouch! Let 'em have it, Chris! I do actually think Groove does some useful stuff, at least for me, it does. However, what really spoiled the party for me was the fact that Microsoft couldn't even be bothered to finish it properly (Folder sharing doesn't work on 64-bit systems, which means one of the neatest things Groove did - synchronising my IE favourites and other files across all my PCs - doesn't work). Microsoft even failed to address that issue in Service Pack 1 of Office 2007.
I do agree with Chris that Groove has a long way to go, but we should know by now, Microsoft never seems to get anything right on the first attempt. Meanwhile, there are a few things that Groove does do really well that I can't think of an alternative solution for. The fact is, Groove Networks was acquired by Microsoft and Groove is currently a bit of a square peg in a round hole within the Office System. Lets hope that, once Microsoft figures out how Groove fits into their collaboration strategy, they will make a better stab at it in Office "14". If Microsoft wants us to "Get into the groove" then It's a case of "boy, you've got to prove your love to me". Am I showing my age now? Never mind, Chris P will appreciate that.
SanDisk is now up to 72Gb with the SSD SATA 5000 - directly equivalent to what many servers have in the form of SAS or SCSI drives, which come in 74Gb capacity. However, imagine how many of these little cards (5mm thick x 1.8 inches wide) you can fit into the space of a standard 3.5" hard drive! They have 2 million hours MTTF, power saving modes when not actually being accessed, no moving parts, no latency, negligible access time and performs more than two orders of magnitude faster than any magnetic hard drive and an average file access time of 0.1 milliseconds (100 microseconds).
These little beauties are going to make _so_ much difference to performance, storage capacity and power consumption. Only trouble is, despite press releases and glossy brochures, they seem to be vapourware at the moment...
What do you suppose would happen if, say, a printer manufacturer produced a new, whizzy printer - the WhizzyPrinter 2000, but didn't bother to supply a driver with it? How do you think that would affect sales of the printer? What if the company PR department said, "we don't need to make a driver. Our printer is so good, that our customers will want to just buy it and it will not matter that it doesn't work with their particular word processing software". Do you think that customers would rush to buy that product? I doubt it. You might pull this off if you are Steve Jobs, but not if your product is a printer and there are dozens of competing printers on the market each with their own drivers that make it easy for application software to work with those printers right out of the box.
To release a new printer without a driver would be crazy. It would be planning to fail; It just would not happen. The Chicken and Egg Syndrome would ensure that the new driverless printer would fail stone dead. The only way to bring out a new printer and stand any chance of success is to have it instantly work with all the software already available. That is exactly why Microsoft Windows, Apple OSX, Linux and pretty much all other operating systems provide an architecture where each printer manufacturer provides their own driver which just "plugs in" to the operating system. It is just common sense that the best person to produce a printer driver is the printer manufacturer. It is a basic reality that the printer must work with existing software.
We have a similar technology in the world of astronomy. It is called ASCOM - the Astronomy Component Object Model. ASCOM sets out the standards whereby equipment makers can provide a driver that simply "plugs in" and most astronomy applications will instantly work with it. The user simply selects their particular make and model from a list of installed drivers, just like a printer. This is a great model for equipment vendors because it allows them to get around Chicken and Egg Syndrome, ensuring their equipment instantly works with existing applications and giving them a fair chance of gaining market share.
Yet, most astronomy equipment makers today do not provide ASCOM drivers with their equipment. Surprisingly, they just release their equipment into the wild and just hope that by some stroke of luck, it will gain acceptance and market share. Sometimes, they include a hastily cobbled-together control panel application that facilitates basic interactive control and, in an act of sheer arrogance, expect all users to ditch their existing software and switch over to theirs. Sometimes they are lucky, there are a lot of "old fashioned" astronomers out there who still like to work all their equipment by hand, but they are getting scarcer. Most astronomers these days are children of the Internet generation and expect some degree of automation and certainly a level of sophistication. Robotic equipment such as telescope mounts, observatory buildings, digital CCD cameras, focusers and filter wheels are the norm. It is de-rigueur to drive all this equipment with software such as ACP, TheSky, Starry Night, MaxIm DL CCD, Cartes du Ceil, CCD Autopilot and others. Astronomers simply expect their equipment to work with the software they already have.
Why then are equipment vendors not stepping up to the plate and providing high quality ASCOM device drivers with even their premium equipment? Could it be that they just don't understand the competitive edge this will give them? Could it be they just don't have the skills in-house? Or maybe they lack the vision to design automation into their products? Equipment makers are currently staking the success of their products on the willingness of hobbyist astronomy enthusiasts to adopt their equipment and somehow make it work with their software. They are producing equipment with RS232 serial interfaces when the world uses Ethernet. They produce imaging devices without frame buffers that rely on PC software to perform real-time download operations. They design hardware with under-powered and under-resourced PIC chips and 8031 microcontrollers that were appropriate 20 years ago but much less so today - when they could be designing in modern processors with rapid development tools, LCD displays, USB & Ethernet interfaces and other modern technologies.
The astronomy market is ripe for a step change in attitudes. Consumers are becoming frustrated with under-specified hardware and poor software support. There is room for a vendor of quality equipment that really understands the implications of Chicken and Egg Syndrome on their product's chances of success. Is that vendor going to be you, or your competitor?
TiGra Astronomy undertakes development of commissioned ASCOM drivers and firmware for all types of astronomy equipment. We have produced drivers for some of the most popular devices, and our drivers power the largest operational optical telescope in the British Isles. Our software is produced to highest quality standards using industry best practices, the latest design and development tools and we have over 20 years experience in producing firmware for embedded devices. If you don't have the skills in-house to develop your own high-quality automation systems, then why not work with us right from the design phase to make your product a world-beater?
I saw this article in Microsoft's UK Security Newsletter (which you can subscribe to online). Ed Gibson, chief security advisor, relates the following story:
The owner of a '.com' web domain name is telephonically contacted by someone claiming to work for a domain registration service. The caller advises the owner that someone else operating a similar business is set to register the '.net', '.org', and '.co.uk' domains of the same web address. The caller said he can stop these registrations, and therefore possible dilution of the owner's business, but the owner must buy the domains now - with a credit card. The owner agrees to pay the asking price of $200 for each.
The owner is feeling quite proud of herself . . . until another telephone call a couple of days later. The caller tells the owner that her credit card did not go through; could she please confirm all the details one more time to ensure her domains are captured in her name. Without thinking, she confirms her details.
Days later she finds out that several thousand pounds have been charged against her credit card.
The moral of this story: the owner should have simply registered the domains herself, if at all.
I wanted to repeat the information here because this is a scam I have actually seen happen to one of my customers. Fortunately, they did the smart thing and rang me for advice and I was able to prevent any damage being done, but I bet lots of people fall for this, because it sounds very plausible if - like most non-IT people - you don't understand the intricacies of the domain naming system.
Never commit to purchasing something like this on the spot. If it is a genuine offer, it will be valid in an hour, or a day. This scam relies on putting you under pressure to nudge you into making a bad decision.
So let's pick this scam apart.
- When someone goes to a domain name registrar and asks to register a domain, generally that process is completed online and automatically. The registrar would probably not have a chance to intervene.
- Assuming the registrar did have the ability to intervene, then by contacting the applicants competitors, the registrar would likely be breaching a number of data protection laws. A reputable company would simply not expose itself to legal action in this reckless manner.
- If the domain names were genuinely available, then you could just register them with another registrar at the normal rate, probably for under £10/year. There would be no need to spend a large sum to assure the registration.
What to Do
If someone contacts you out of the blue and puts pressure on you to make an instant decision to purchase something, I suggest the following:
- Tell the salesperson that you need to get authorisation from your boss, partner or co-director to make the purchase and ask for a name and telephone number where you can call them back. Keep a record of the company name, salesperson name and phone number. If they will not give you a number, end the conversation there and then.
- If they give you a number, call it back and see if it is a genuine number and is answered by the company and/or person they claim to be calling from. If not, <click, brrrrrr> (sound of telephone hanging up).
- Even if they pass those tests, Stop. Think. Cool off. Have a cup of coffee. Now call your technology consultant - or me, or another Microsoft Small Business Specialist - and ask for advice.
If this scam worries you into registering all the domain name variants for your business, then OK, by all means go ahead and register them, or ask your technology consultant to do it on your behalf. Do it at the prevailing rate, though - don't pay orders of magnitude over the odds. As of today, 1st February 2008, it costs typically under £10 per name per year.