I have fond memories of playing Manic Miner and Jet Set Willy on this machine, as well as writing out computer programs from magazines.  This little box of circuits is probably what launched my interest in computing and inspired me to learn how to do programming.

Notice that the chiclet keyboard has come back into fashion, albeit without the rubbery keys that would sometimes get stuck, or just wouldn’t respond.

Look how things have changed over 30 years and imagine where we’ll be in 30 years time.

Anyway, enough reminiscing… happy birthday!

I am currently working on a Ruby project that uses a lot of regexes on large volumes of text.  It is currently running too slowly, so I decided to try to optimise it by implementing the regex matching code in the D programming language.  D has given me a lot of joy (compared to C or C++)  by making things like string (with Unicode) handling a breeze without taking the performance hit of supposed “productivity” languages.  I painstakingly reimplemented my Ruby functions in D expecting a huge performance boost (actually I expected an order of magnitude performance jump) but instead I was shocked to see that Ruby outperformed my D code by a significant margin.  The Ruby implementation was finished after 80 seconds, whereas the D program required around 280 seconds using the exact same regexes and the exact same input.

I am not being a Ruby fan boy here, actually I will be the first in line to give you a long list of complaints I have with it, so please don’t view this as a “Ruby is awesome, everything else sucks article”.  It is also entirely possible that I am doing something really dumb in my implementation.  I’d like to ask if anybody out there has done any benchmarking with D (especially in the realm of regexes) if they know what might be going on and if there are known performance problems.  I will try to post my code (in Ruby and D) as soon as possible

Just as an aside, the D programming language is a very interesting language and if you haven’t already played with it then you really should.  I can also highly recommend Andrei Alexandrescu’s superb book “The D Programming Language”.  There is, quite simply, no other book of this caliber about the D programming language.

Update 26 March 2012:

Apparently comments weren’t working on my blog, so sorry about that (the captcha plugin was sort of half activated, so I just switched it off until I get around to fixing it properly).  You are now able to comment on this post.  I got a nice email telling me that some people were curious over at the official D forums about my problem.  It seems that my post caught Andrei’s eye, so I thought I should get on with it and post my code and not leave people speculating on the forum.

The problem was as follows:  I was unable to post my work because it contained proprietary code and the data I was using was also somewhat sensitive, so I had to spend some time recreating the problem with “postable” code.  I recreated the problem using different input data and new regexes where I found D to vastly outperform Ruby.  This made my both happy and sad at the same time because I still needed to recreate the problem somehow.

My code loads a file from disk, stores the contents into an array and runs a series of regexes against each line in turn.  The time taken is only measured during the regex phase (so the file loading isn’t counted).  I don’t care if the regexes match or not, but I always run every regex against every line.  The regexes are the same in the Ruby and the D code (unless I’ve made a horrible mistake).

In the D code I am not using Compile Time Regexes, and my Regexes are cached before the test (i.e. I am not  recompiling them on every loop iteration).  Additionally,  I am using the more recent std.regex package and not the deprecated std.regexp package.

I am using Ruby 1.9.2 and the Digital Mars D compiler v2.054

I perform the test twice, once for a more or less random CSV (actually pipe delimited, but that doesn’t really matter) file and secondly for the Complete Works of Shakespeare in text format.  Actually, I stop after just 2000 lines because the performance difference is already obvious.  The regexes used are always the same.  The performance is very much dependent upon which input file I use:

With CSV file input:

D Program: 55,813 ms

Ruby Program: 4851 ms

With Shakespeare text:

D Program: 1973 ms

Ruby Program: 4683 ms

Maybe if somebody could take a look at my code you could tell me if I’m doing something dumb or not.

All the code and test files can be downloaded here: Benchmark.zip

Update 27 March 2012: The more recent v2.058 DMD compiler doesn’t suffer this performance problem anymore.  I just tested it and it now runs significantly faster than Ruby in both scenarios.  Thanks for the support in the D forums!

In the olden days you could pick up a tape cassette containing a Spectrum or Amstrad game for a few pounds (yes, I’m British).  At this time, the gaming industry was in its infancy and the business side of the game was highly underdeveloped.  Games were written by a single (or a very small group) in squalid conditions and these programmers were like rock stars, or mad scientists working alone in their lab (complete with the associated lightning and electrical apparatus).  Some games turned out to be “smash hits” (Monty Mole, Manic Miner, Dynamite Dan to name a few), others were of very poor quality but hey, that’s the price you pay for experimenting.  How things have changed!

More recently creativity in game design went through a really stagnant phase with developers churning out platform games and more recently first person shooters and over the shoulder shooters.  The PC game Doom was an incredible step forward in gaming at the time, but it has spawned (through no fault of its own) a genre that now dominates the gaming industry.  Graphics standards began to rise and this meant larger and larger teams, now approaching the team sizes required to make a movie (actually game making and movie making now have a huge amount in common, but that is the subject for another post).  To fund these endeavors, huge companies such as EA, Ubisoft and Activision are active in video game marketing and promotion.  Big gaming companies don’t like to take big risks with their money and prefer to spend it on formulae that are tried and tested, usually pumping out sequels and variants.  This strategy has resulted in the growth of gaming franchises such as Modern Warfare and Call of Duty.  New games are inevitably tied in with the franchaise so as to guarantee high sales volumes.

Indy gaming on the PC has somewhat of a renaissance in recent years.  With the development of smart phones and tablet PCs where the average Joe has access to the development tools, we’ve seen an enormous growth in the independent gaming scene.  Another avenue for independent game developers has been social media platforms such as Facebook which allow gaming to be integrated with the social experience.  You don’t need hundreds of programmers to write your own games for these platforms and, I suspect, this has drawn many a disillusioned game programmer back out of the closet which they climbed into during the late nineties.

Mark from the Classic Game Room HD explains how Gamers can pick up affordable games for the iPod, which is how it used to be for the Atari 2600.

Zynga has turned, with its partnership with Facebook, into an extremely profitable company.  It develops games for social media web sites and is just one of many growing companies developing for the casual gamer.

On-line distribution now means that the humble PC programmer can start a games company and begin doing what they love creating games.  Steam, the on-line game distribution system, has also started selling independent games.  They have a very large market share in the on-line game distribution market and their catalog of games includes a large and growing number of games from developers that we’re probably unheard of just a few years ago.

Smaller game companies such as Paradox Interactive also stand gain by the reduced distribution costs promised by platforms such as Steam.  Companies such as these can focus their efforts on producing games for their market niche, keeping their fans very happy, and not have to spend millions on packaging and distribution.

With gamers becoming increasingly disenchanted with the large game publishers pushing draconian Digital Rights Management schemes down people throats, market conditions seem perfect for a renaissance of the independent gaming scene that has been ticking along very quietly in the background.  If companies and individuals are able to produce high quality games for a reasonable price and still turn a good profit, this can only be good news for the industry and gamers alike.

As more shoppers move to the Internet for their bargain hunting, we’re left (certainly in England) with quite a bit of vacant high street property. Many stores, even big chains, simply aren’t able to price their products competitively enough to compete with on-line stores that don’t pay inner city rents.

Once the shoppers are on-line, a wealth of information becomes available that helps the prospective buyer make their decisions. Price search engines help to find the best deals without the user even having to move from their seat (let alone hike around town). There are countless sites hosting professional and consumer reviews and opinions to help buyers find the potential pit falls of their chosen product. Other web sites help users find new and interesting products that the shopper perhaps did not even know about.

What does the future of on-line shopping look like?

One benefit to going shopping in the real world, is that you can bring a friend with you. You’ll get a cup of coffee and talk about whether or not those jeans you were looking at were a good deal, or whether they would suit you and, additionally, what else you might buy to go with it. These sorts of activities are currently poorly served by the current format of on-line shopping. Shopping via your web browser can, for some people, feel a bit soulless. Can we ever recreate the “social” aspect of the on-line shopping experience? I think that we probably can get very close!

Social Shopping for the masses!

It turns out that we already do rather a log of socialising on-line anyway.  With social networks like Facebook, Twitter and <insert your favourite regional social network here>, we can communicate at light speed how we feel, what we are doing and whether or not the dog is currently asleep.

A simple web shop that provides Facebook integration does not, however, mean that you have a social shopping site.  A web site must actively support shoppers in their efforts to find products, recommend products to others and to get overall advise and ideas.

Most of all, social shopping must be fun.

Introducing Shelf9

Shelf9.com is a new social shopping platform that aims to bring together groups of like minded people and help them communicate.  A user can find things they like from a very large catalog of products from a variety of on-line stores and brands.  Products can build their own “shelf” with the products that they particularly like, or that work well together.  Building a shelf is done with an easy to use drag and drop editor and, once complete, these shelves can be shared with friends via social networks.

I think that Shelf9 has a lot of potential to help small businesses come into contact with shoppers.  If you have an on-line store, one of your main concerns will be marketing.  If you make your own products, this also includes building up your brand awareness.  To achieve this with Shelf9, you can simply start building shelves with your products (which can easily be added to the platform).  Once added, other users can combine these with other products in their own shelves which they will share with all their friends.  If a shop has high quality products, they will quickly become disseminated throughout the community.

Shoppers get to discover all sorts of smaller brands from shops that they would never have found otherwise.

Shelf9 is completely free for shoppers and small businesses alike and I’d like to take this opportunity to invite everyone to have a go.

I suspect that there are many much more serious systems that would cost considerably more if they became unavailable at Ikea. If the accounting system went down I’m sure it would cause a lot of headaches for them.

While companies are able to save money with the use of thin (and web based) clients, let’s not forget that there are downsides too. As software moves off the terminals and onto central servers and eventually into the “Cloud”, we should start thinking about such problems and preparing for them. If I start my own company, I certainly don’t want to be denied access to essential services because some other company is having technical trouble. That could send a small company to an early grave.

It isn’t that thin clients are bad, it’s just that we’re going to start seeing these sorts of problems more and more if many businesses get their way. Apparently in the future we’ll all be running just a kernel that will give you an Internet connection and all your apps will be streamed to you on demand. A nice idea in theory, but prepare yourself for the server being down!

My Firefox was hopelessly out of date on my desktop so I thought I would upgrade the installed packages. Thinking that this would be less risky than a full blown distro upgrade I pressed ahead. I’ve just spent an hour fixing my graphics settings after my graphics card driver stopped working (or something like that) and I thought I should share the solution because I’m sure others will have the same problem.

I have an NVidia card and upgrading my software packages made it stop working.  I was stuck with a low resolution and all attempts to reactivate the proprietary driver failed.  It would say that it was enabled, but when I tried changing the resolution it would say that I needed to activate it in my X settings. The error dialog suggested that I run “nvidia-settings” to fix the problem.

Running “nvidia-settings” did absolutely nothing, so I had to start scouring the web for help.  I finally found a helpful blog with some helpful advice buried in the comments.

There’s a really helpful too called Envy that solved my problem in a matter or seconds.

To install it do this:

sudo apt-get install envyng-gtk

for Gnome, or

sudo apt-get install envyng-qt

if using KDE/QT.

Then run it.

envyng-gtk

And select ATI or NVIDIA and click Install with Automatic Hardware Detection.

Envy is a really helpful tool

It prompted me to restart and then all my problems were fixed. I’m quite disappointed that upgrading some packages in Ubuntu would cause a problem like this, but I’m happy that there is a community of users who want (and are able) to create helpful tools like this. Thanks to the author!

Agile Software Engineering