No in my experiences it is IT's insistence on security practices with zero thought on how this will impact the end user. A few examples:

I was looking after a Jenkins server for a project we were running, Jenkins was running on the latest version of Tomcat with a Java 6 runtime. However due to customer requirements there was also a Java 5 runtime which as being used to generate the build. IT felt the need to un-install the JDK 5 and upgrade the the machine's Java 6 (to the version with Oracle in its name). The removal didn't update the Java_Home directory causing Tomcat not to work. They decided to do this just as we were starting an Integration & Test phase for a major release. The Jenkins server was linked to me on their records but at no point did they think to mention it to me.

Same Jenkins server, this was running fine and suddenly the builds starting failing. 3 hours of investigation later I find out its because the Jenkins server password has been changed. Never mind the server username was the name of the project (e.g. projectXYZ). IT came up with a new policy which stated all server accounts needed to be > 48 characters and they had changed them all without notifying a single person on the project.

How about when IT decided that in a software house no one needed Admin access, which would be fine except they tried to forbid admin access on projects which were developing software which required Admin access (for a number of reasons). Those projects had to go up to the business director and have him shout at the IT head to fix it.

Or the fact they decided no one should have USB. A great idea except I was working on an embedded project (along with a dozen other projects) which required an unencrypted usb stick to load the software on to the test rig. Once they realised how many people had a problem they tried to limit it. But when your working on a 5 man team only allowing 1 person to transfer files causes you to loose a lot of man hours.

I can think of dozens of other examples, none of them were dictated by upper management. People hate IT because IT doesn't look at how to better help people work.

I was considering power draw and size, if your dealing with an electric scooter an old android phone is going to have an affect on the battery life. A windows Mobile 5/6 phone might not be too bad...

Your right that something like google latitude/foursquare/facebook would be simpler (and you an get it on the N95, 5300, 5800, etc...) the problem is data plans. In the UK Pay as you Go offer either 500Mb free a month if you top up by at least £10 a month or £1 for 25Mb (valid for a day). I'd rather put £10 on a card and have the device start streaming its position once a minute when I ask it to.

Wouldn't it be simpler to get an old Symbian phone and write an app?

A Nokia N95 has GPS and it very low power compared to modern phones (week between charges), you get a cigarette charger hooked up to the battery to keep it charged and then write an application that listens for text messages. Upon recent of the text message he phone would text/email it's number. Then all you would need is a water proof case and a pay as you go sim card.

A quick check on ebay shows them going for £40. I'm getting a Honda CBR 600 RR next week I might do this.

I'm buying my Mum a toaster this Christmas... as a bit of fun. Every time I visit my parents I almost set fire to the kitchen using the old one (timer doesn't work) and there is now a long running joke about it between my Mum and me. That won't be her actual present, just something silly under the tree

I agree this idea does sound like a terrible gift, there seems little thought about the intended person and putting software on it will involve opening the packaging which will make it seem like a used usb stick

Mod this person up, ignoring everything else about Christmas I have always thought it as a time to show family and friends you care about them. Gifts should be thoughtful and hopefully meaningful I would argue that they can be utilitarian or something you would buy yourself.

Last year my parents bought me a Nokia bluetooth stereo headset, I would have bought them myself when my old set had died. They got them because over the year I owned the previous set I would mention about how they are better for online play, the gym, how I kept calling them on the old set and how gutted I was that my old set were dying. I never asked for them for Christmas and they meant a lot because it showed they do listen even when I go off on a technology rant.

I should point out that you shouldn't let money be a factor, one of my little sisters is trying to set a limit no Christmas spending. The rest of the family has been trying to explain that the amount you spend doesn't matter it's more important to get something meaningful. Heck last year I got my Dad a £5 survival kit and it has to be the most used and talked about present he got that year

This is a good idea, I don't understand giving away a usb stick. They tend to be secondary things e.g. giving someone a digital photo frame with some family photo's on it.

The gift of a USB stick with OSS on it seems very generic and not really very thoughtful. I always thought the point of Christmas was to show you care about someone by taking the time to think up a gift/gifts tailored to them. For example one of my little sisters never used to read recreationally and I would tease her about it. This year she hasn't been able to put a book down and is also a huge fan of Harry Potter so I'm getting her the Harry Potter Special Edition book set.

I appreciate people have budgets, heck of all the things I gave my Dad last Christmas the £5 survival kit is one of the things he goes on about the most. It doesn't have to be expensive just bought with that person in mind.

No but I saw this Doctor who movie

Your wrong, if you understand a subject it shouldn't matter how a lecturer teaches it. I don't understand differentiation and integration the way it is taught in UK schools. It took me a lot of effort to work out what things meant and why*, when I went to university the lecturer there did such things another way. I didn't claim it was unfair I wrote things down as he did went back to my notes and figured out what he was doing.

If in your first year you cover integration and differentiation is is unreasonable to expect your second year to reteach integration and differentiation before teaching something like laplace or furrier transforms.

If you don't have the knowledge for a class either you need to learn it in your own time (and possibly find someone who can help you) or you shouldn't be there. Your choosing to study that subject which means you must care about it, if you don't care about it why are you there?

*This is my biggest gripe about STEM teaching, it is only taught one way. A few years back my little sister was being predicted a D/C in her Maths GCSE. My mother called me home to try and help, I spent a single evening going through her Math curriculum book explaining how I understand the subject matter. He teacher remarked on a stunning improvement and she got an A*. I'm not a great teacher I just see and think about Maths in a very different way.

At my university (University of Plymouth) during the first year they would take on 70 people for electrical engineering related degrees and the facility expected more than half to drop out over Christmas. The second and third years would be limited to ~30 people.

The first term would cover all the preliminary knowledge required for the degree, I think the idea was to make sure everyone was at the same level but the pace of it meant it was more a refresher. It was designed to weed out people who wouldn't be able to complete the degree.

While it sounds harsh out of the dozen people I know who dropped out I think only one left because he couldn't keep up. Which was his own fault, myself and others had set up a study group and frequently invited him but he always said no. The others were international students who stopped coming to lectures after the first few weeks.Most of them seemed more interested in partying than learning.

No it needs to fix Picasa Web Albums. I tried to load my entire photo collection in to Picasa Web Albums and came across 2 problems. Firstly local tags don't aways get put online, when they don't appear online the application detags them locally. Secondly it creates duplicate people I have a person called bob, I upload 2 photo's of bob to web albums on my local machine that is one person but on Picasa Web Albums bob now has two person albums. This problem is compounded by the fact you can't merge or delete duplicate people albums. I have no photos in web albums at the moment but my account is stuck with over 400 empty people albums most of which are duplicates. Lastly just because I tag someone in Picasa Web Albums doesn't mean they are tagged in a Google + stream and tagging a person through the Google+ stream does not tag that person in Picasa Web Albums.

The people manager issue has been a known bug for more than a year and Google haven't even acknowledged it, for me the draw of Google+ was circles and Picasa Web Albums. If Picasa Web Albums is broken what is the point?

As a software engineer I take offence to that. I work for a CMMI Level 5 rated business unit, process is everything. First you develop requirements, then design use cases, then outline the high level design (functional area's, etc..) if required you complete the design in UML to a class level (depending on target safety level). This is documented and reviewed by peers, you generally have to justify any design decisions. You also have to specify any external interfaces and detail them so a 3rd party could implement them. Once the design is approved you develop component, system and unit tests which should trace back to requirements. If tests are needed after the code has been written (say new requirements come in) the person who implemented the code can not write the tests. Finally after all this is done you get to implement the code (depending on size multiple people may do the implementing).

We tend to run products through iterative testing process, when we are happy the tests and all project documentation are independently audited. At that point the product is considered "ready". Occasionally if your unlucky you get to see the customer integration and provide technical support.

What would a professional engineer do differently? In my next job is labelled software consultant and my biggest fear is it won't have decent engineering processes.

For me it's more about the manufacturing yields, the article mentions TMSC are struggling with their 40nm production process and this thing is 28nm being released next year. From what I understand TMSC is being used to remove Apples reliance on Samsung, I wouldn't be surprised if this allows Samsung, etc.. to jump ahead as TMSC don't sound ready to mass produce the chip.

Dual core makes sense because of power saving issues, you can have one low clocked core which is enough for basic phone functionality which is turned off when you started using the phone. In this sense I could even understand a triple core chip, you would have one low power core for when the phones not being used, then when it is you can move OS/Background processes to one core and have a 3rd core for running the main process.

Surely a purpose built GPU would give far better gaming improvements than an additional A9 core.

It sounds close to the official number as quick glance at BBC news's website shows. Nothing to do with collusion, people keep talking about all of the CCTV camera's in the UK police have been identifying people through that. Most of the papers have also been printing pictures of rioters in an effort to identify them, there are about half a dozen stories of mothers turning in their kids when they saw the child's photo in a national paper.

This meeting is the higher ups way of looking like they are doing something to daily mail readers, I'm hoping nothing comes of it especially when you realise Twitter & Facebook were used by people to organise clean-ups and identify the rioters.

There has also been a lot of talk about the harshness of the punishments handed down to rioters. The UK doesn't require mobile phones to have Government available GPS tracking like the USA. You can only check-in with Facebook/Google latitude and not twitter.

I've tried to find a middle road, if I come across design flaws/hacked code. I write a series of unit tests for the component and for the individual classes while I write the hack fix. Once I've come across the same class 3/4 times there are usually unit tests covering the class in depth. So when I go to management I can promise that there is a low/no chance of the behaviour changing and point to numerous issues we have had to fix because of the problem.

That is a wonderful idea which doesn't work in practice. I'm currently writing a very simple web application, I'm sticking purely to the XHTML 1.1 and CSS 2.0 specification. Now IE 7\8 have their own idiosyncrasies but they are generally predicatable and well documented. Safari (buggy as anything on my machine) and Chrome seem to strictly follow the standard but have a few CSS deficencies (collapsing borders of Input button types for example). Firefox lives in its own little land while it has a wider range of support some of the CSS fields work strangely and there are a number of display issues. For example I have something like the following:

table style"width:20%;border:1px;"
tr
td
p text /p
/td
td
input type="button" value="test" style="width:100%;"/
/td
/tr
/table

Now in IE6, IE7,IE8,IE9,Chrome & Safari if I were to change the value of the button using javascript the table cell would resize to fit the new value in Firefox 3.5 it doesn't.

Funny thing is when I started the project someone suggested simply writing a, Eclipse RCP, iOS and Andriod application because it would be quicker. Considering the time I have had to spend I think they might have actually been right.