I was subscribed to original blog announcing this for a while, and developer after developer asked for more virtical screen resolution. As far as I am aware the project brief was to use existing hardware so that option was never viable.

I gave up following it in the end, because the other ideas - to try and pre-configure software configurations for use seemed crazy in the face of the flexibility of the Ubuntu synaptic package which could acheive a more flexible goal with less work.

I came to the conclusion that this skunkworks was driven more by a marketing goal to seen to be doing something rather than any real desire to meet developers needs.

I've run several SMF forums since I started my main one in 2007 (http://www.melindasbackups.com/forum/) This main also has applications that I have built (chat, airhockey and football) which use the SMF login as a single sign on, and themes which (sort of) match.

I've been into the guts of it, so we have flowplayer video and mp3 players embedded pretty seamlessly into posts (and youtube videos) as well.

We have had a massive number of spammers attempt to join, but we have some extra mods in place to keep them out and so far only one or two out of the several thousand who tried have managed to set up an account that could post. Manual moderation deals with this last few, but its a fairly painless job.

There is still a surrounding static site, with a completely different theme, but we don't really do much with that

That is rubbish. Provided you don't distribute it outside your company then you don't have to make your own software GPL.

Of course if you do want to distribute it in a proprietary sense so you can make profit out of other peoples work then you can't.

I wrote a chat program for fun that addresses the problems. See https://github.com/akc42/MBChat. The software originated as a non secure chat for a fan club (www.melindasbackups.com) that I an the IT director for, but I tried to add security to it as part of a proof of concept for use where people were trying to intercept the comms.

The assumption I made is that all communication between client and server needed to be secure and that there could always be a man in the middle trying to intercept your communication. Since the initial program is downloaded from the server, it is not possible to send the keys needed to encrypt things with the program.

My solution to this was to calculate a RSA key pair on the fly in the browser as the program starts up - and then send the public key to the server and ask it to encypt the key used for communication using this public key. The client then decrypts it using his private key.

Because RSA key pair generation is quite compute intensive - I had to develop a mechanism (which I borrowed from elsewhere) to allow the browser to perform long running calculations without timing out. It does this by breaking things down into small steps and returning to browser for a "tick".

I also spent some time trying to have the server also prove it was the correct server. I came to the conclusion that the only completely foolproof way required each human client to know the "Standard message of the day" and to be able to see it after its been decrypted for the server comms to be sure the server is who you think he is.

I open sourced it when I developed it about two years ago now - but I don't think anyone has cloned it or taken a look.

For a few seconds at exactly 6:am I had up a page at RS that let you order the pi - not express an interest but actually order. I already had an account with them, but by the time I had checked up what my account details were the whole site had gone done.

The press release by Farnell talks about 600 people trying to order a second. If they split the first batch 50/50 then in slightly under 10 seconds Farnell would have sold out. No doubt RS was the same.

So its not surprising that by 1 minute past 6, all you could do is be put in a queue for the next batch.

I fully agree with you. I was the same, so I converted over to KDE4. Then I spent a long time working out how to use facilities like multiple desktops and multiple activities under KDE4. I work in different areas at different times of the day, and the KDE activities approach seemed ideal.

But the practical reality of it was that I just got in a mess.

What I did realise then - how good Gnome3 is. I've gone back to it and now I find the older interfaces limiting.

Things that are really good
1) I run with multiple screens - having my secondary screen remain in place when I switch workspaces is fabulous. I can put an important source document on display and then work in the different other documents each in its own workspace, surrounded by the support applications I need for that part of the work
2) Growing an shrinking Workspaces automatically. As I start getting too complicated (too many windows to work on comfortably in one workspace, I through some stuff to another workspace.
3) Minimising by flicking my wrist (and moving the mouse to the top left hand corner) - and getting a dynamically visual view of applications. I often work and watch MythTV in the evenings, and being able to here something important about to happen on the TV when I am working on an overlaid window, that flick allows me to immediately watch the action on the TV

I do have criticisms though. It was insane to make Suspend the default option closing down - I want power off (I know there is an extension that does this - these aren't yet just available as Debian packages). There seems to be bugs that can hang the closedown process if you cancel it half way through. I can't find my apps very easily - seems to be very particular about having .desktop files in the right place. Often I can install some thing in Debian and I can't find the app.

I'm just turned 60 and wrote my first professional programme (ie I was paid to write it) in 1969. There have been several different themes in this whole article and this seems like a good place to summarise my thoughts on them all

Until about 1976 or so I was a programmer/designer - mostly in assembler language (mostly PDP 11, but lots of other stuff too). About that point in my career I drifted upwards into managerial roles with less of a programming input. Initially I was pretty lousy as a manager, I didn't understand people skills - particularly tending to treat people like robots. I got criticised badly for that in my performance reviews, but I learnt and got better. I think by the end of the 1980s I was considered a good manager. At least I can claim that many of the people I was responsible for in those days are still quite good friends and when we meet socially there is still discussions of team building activities that went on back then. I wasn't at all of them - I was a divisional manager and I had just authorised the expenditure for the project manager to take out his project and a completion of some milestone or other - but we also held occasional divisional level events and these are still remembered too. I had financial targets to meet as well - but I was convinced that such expenditures were trivial with respect to the benefits I got from improved productivity from good morale and from better communications because people knew each other better.

Two other elements of managing seems to me to be important and views that I don't often see expressed here.

1) I felt that to be a good manager I had to understand the technology. It wasn't just at the programming level either (although I did teach myself C and did write the odd bit of software in it), but in terms of frameworks, operating systems, databases, and the software engineering processes. My unit - with my managerial lead - developed an early version of distributed version control which as recently as 3 or 4 years ago I know was still being used on some projects in the company that were nothing to do with me (my junior protégées had grown into senior positions and were using it), and I remember spending a long time understanding the issues surrounding Object Orientation (the mind shift traditional programmers need to design a system round an object concept as opposed to a functional concept is significant and you have to put some quite strong controls in place to ensure a project is successful)

2) All this 80hour week stuff is rubbish and I tried to prevent it. I was convinced then and still am (although my view has modified with respect to those that are doing out of love rather than peer pressure) that if you push people much more than 40 hours per week, except in very short bursts to meet a deadline, then burnout gives you a negative productivity gain

At the beginning of the 1990's I stopped having such close contact with technology at work and continued my learning journey at home. I taught myself how to set up web sites and learnt (initially) Java to make dynamic web sites (using Tapestry), although these days I do most all of it with PHP and Javascript (also self taught), and vastly improved my SQL skills. Also got to learn about Drupal and Wordpress. About 2 years ago I got made redundant, but was in the fortunate financial position that I could pretty well retire if I wanted to. Given I programmed still as a hobby I decided to see what I could do about selling my programming skills again.

I want to tie that into a comment about value (ie cost of a programmer) v experience. Of course on a project you need balance between senior and junior guys. In one year in my career as a manager I hired 27 new graduates into a team of 70 (with a spread of experience). We strained and lost a bit of quality in our development. I think I went to far down the junior balance, but not a long way too far. My most senior guy still actively writing software on a regular basis had had (at that time) about 12 years experience. I think I was gaining little from him over the last 6 of those 12 years on the programming front (his most productive and innovative stuff had been at about 6 years of experience) but he was still immensely useful as a design authority. But I also had younger guys coming through who became gurus in (again at the time) Windows NT as a graphical workstation. About 6 years experience seemed to have been long enough to have gathered sufficient experience to be an expert. (The two have still technically oriented careers, but they are now being paid for skills related to the application of technology to the business rather than as programmers).

Back to me. I have achieved some income - but not a full time role but small projects. I have formed my own company and I sell my skills. I don't charge a vast rate - more applicable to the 6 years of experience rather than the 40 and I can only support myself because I have pension income as well. But I continue to learn and want to. In the past year or so I have become a Microsoft Access expert (including of course Visual Basic) and of the back of that have also learnt Transact SQL (sql server) and Classic .asp. I also developed software in Visual Basic .net. I have generally sold myself through personal contact - I haven't tried hard, but have had zero response to just sending off a CV and covering letter to possible roles - so far everything has come from existing contacts from before who are now in roles where they find my skills useful.

I meant to post that as ME, not as an AC

Unless of course he is playing a double bluff.

It appears the study you refer to says that even with PCI compliance, 30% of companies experienced some form of breach.

That is frightening. Do I missunderstand the figures?

Mine did too - but it did quite a lot of damage to the electronics when doing so, so a replacement power supply did not revive it.

You should use DVD+R for data and refresh every decade or so ago

I just re-read my entire CD archives from 10 years ago (actually between 1998 and mid 2004) and thinned them down and put them on to DVD+R. I could read all of them, although one or two had difficulty with a couple of files. I took the opportunity to thin them down (removing duplicate copies of files) and compressed 51 CDs into 4 DVDs

I also have offsite backups on standard 3.5 inch sata drives of what I would regard as my current "live" data - I have about 200GB covering all the family - music, images, documents and spreadsheets, software I wrote etc etc. A simple USB/sata external chassis for the drive was all I needed to take a backup. I have two sets and attempt to rotate them monthly.

When I went to University in the UK (late 1960s), only a handful of not very prestigious colleges taught the new subject of "Computer Science". I ended up doing Electrical Engineering, because that meant I could go to one of the London colleges (Oxford and Cambridge were out because I couldn't pass Latin - was needed in those days) despite knowing from the age of 11 that I wanted to have a career in computing. Shortly afterwards, it became more important and my brother (9 years younger than me) did a Computer Science course at the same college as I (Imperial College).

In those days, the subject was not very broad - algorithms were immensely important because the libraries that embed these algorithms were not widely available and software packages (as we know them today) did not exist. For some time, my career after university was programming large chunks of what I did in Assembler language (although to be fair, I was writing device drivers for the operating system RSX-11M on a PDP/11 - not mainstream commercial programming) and it wasn't until sometime in the 1980's that the computer languages had progressed far enough to use something like C universally. I remember at the time wanting to get hold of a book on compilers and how they work, because that was the 'in' subject area for modern computer science. Similarly databases, particularly relational databases, was another subject area that was broadening. We hand developed in Assembler some database engines for a large commercial database project I undertook in 1974, and in the early 1980s the company I worked for started to develop a relational database which competed with Oracle. It was only when we had to invest in an SQL parser that the business case required us to sell a 1000 copies that we chickened out - asking the question if that was possible. And if we can all remember it was during the 1980s that GUI's became prominent - the Apple Lisa was the first I experienced, but MS Windows followed soon afterwards. And then again in the late 1980s, Software Engineering, along with all the various methodologies (I remember having fierce debates about which methodology was correct) and Object Oriented programming (again fierce debates over the mindset change needed to use an object oriented language properly). The final point was brought home to me by watching Tron the other day. In the time of that film, programs were written by Users. That was indeed the norm, and it wasn't until the mid 1980s and the explosion of the PC as both an Office and Home Device that users where people who didn't write the programs, but instead went out and bought "Software Packages". It was at this point that a new subject area of "how do I use this program" and "how do I configure this system" became a subject that required teaching to more than just a few specialists.

The point I am trying to make, is that since its inception the field that was historically named computer science has grown to a point where it has too much subject content to be covered completely with a single undergraduate degree. Specialism is required and therefore you would expect different colleges to offer different combinations.