I just got my EEE PC

I finally got my EEE PC an I'm typing this post on it. I'm far a fan. It's not perfect, but the most important part is that it's right for the price. I managed to wangle it on Salary Sacrifice which means it totaled somewhere in the realm of $300 AU. It does 90% of what our old $2500 laptop used to do and a few things extra. It's so light, in fact is scary how light it is. I'm scared of breaking it which throwing it across the room to my wife. I have massive hands so writing on the keyboard is a little cramped, but I can touch type on it without too much trouble. I think with practice it would be easy, but I have a wireless keyboard setup so I'm using that mostly. I have it hooked up to my LCD TV in the main room so the screen real estate is massive (40' instead of 7') and it scales well. Shame my Full HD TV has a few troubles with higher resolutions due to the frequency. Only problem I've had so far is that Pidgin doesn't work. The installed version won't connect to MSN or ICQ and I'm not sure why. It could be the version, it could be the network settings but it's dissappointing that it isn't working out of the box. I've crashed it once or twice as well, but no more than I would have a windows system considering that I'm trying to tweak it. If anyone ever reads these things, I highly recommend getting one. They're so cheap that if you think it sucks you'll lose almost nothing. They're great fun, and you'll really love showing off the voice command program to your friends.

.NET Reporting

I've had a chance now to look into reporting for .NET and it's the first time I've been dissappointed. The general reporting tools aren't much chop but lucky for us they've included Crystal Reports from Busiess Objects. This isn't necessarily a bad thing, in fact this is a standard reporting tool and most of my developers are having no trouble getting the hang of the new version. The application I've been developing is all but complete and while I haven't taken the best development approaches due to prototyping I'm very pleased with the resulting code and application. What has dissappointed me is something I would have thought should be trivial, though it probably shouldn't come as a suprise. Deployment. Just deploying your web application isn't enough. Installing the crystal reports distributable isn't enough either and neither are the things you might try after that. I had to go searching the internet for an hour to find that I had to look through the install files to locate some files and copy them into the installation web application root for it to work. There was no documentation on this, no clear indication of what to do. I suppose with most environment's I've worked with there have been problems like these so I shouldn't be too dissappointed.

Developing in .NET: Initial Impressions

Well I'm a C++ backgroung programmer who initially started developing in VB 6 many years ago and moved into C++ shortly after, which was a welcome relief. I've avoided Microsoft technology every since, so developing in .NET is a big step for me. Why not try Java or Ruby? Well I would have liked to, but in the team I'm currently working with here .NET code and Microsoft technology is already heavily embedded so I could try to introduce a new language that would take a long time for developers to learn (and many of them do not want to) or I could learn a new language myself. I decided it was a good oppertunity to try something new. So with absolutely no .NET training I fired up a project, a website project. I created a page, added some code behind the page. Fired it up. Within 10 minutes of starting my first .NET project I have a login form that processes my username. So far so good, but I had heard this about .NET that it was very quick to get simple code in place. So I read up on Master pages, create a framework for the interface. Then I extract the code behind into a nice class structure with an interface into our reporting environment. I add a database structure and interface into DB2 and get one of my team to extract some authentication code (in VB, where as I'm writing C#) into a re-usable class. One week later, roughly 18 hours into development and I have a dynamic reporting application that only needs the final touches and interface to be beautified. I have to say that I'm fairly impressed at the speed at which I can create such an application. I spent 3.5 weeks creating exactly the same application in Java and I had an API to work with for the reporting component and had a good knowledge of Java before starting the project where as this is the first time I've programmed in C#. I still don't consider myself an avatar of .NET. I don't agree with many of the development principals that they use and the Microsoft restrictions that exist still bug me. But for its purpose, .NET appears after my first impressions to be a pleasure to develop good clean code in.

Running Ubuntu

Well I thought I'd finally give Ubuntu a shot. Good idea, I had a spare drive to install it on and figured I could set it up without impacting on my Windows setup. I'm afraid of Vista, haven't heard anything good about it yet and when I do get a new laptop i would like to have the option of something other than Microsoft's DRMware. Anyway, I found a distribution on my ISP's freezone and downloaded it at ridiculous speed. Extracted the ISO onto a CD and here I am. 2 hours is what it took me to install. Would have been one hour except that Ubuntu seems to have trouble reading the DNS servers from my router. I had to set them manually. If anyone loads Linux only to find they can ping but can't browse in Firefox, go to your router's home page (usually http://10.1.1.1 or similar) and copy the DNS servers into the network settings of Ubuntu. I have to admit if I had no computer knowledge then I would NOT have ANY IDEA what I just said. But hey, everyone has a child somewhere in the neighborhood that can help them out right? So far so good.

Borland Reporting Tools

I have developed a lot of code in Borland tools and I've got to say that I'm dissappointed with the reporting tools available for the environment. I suppose I'm too much of an optomist to have expected too much from Borland (now Code Gear) but the environments they provide always show so much potential to be great (but more often end up falling far short of the mark). In our upgrade to BDS2006 some time ago we lost the ability to use Quick Reports. Rave reports was the solution provided now, though why I'm not sure. Rave Reports offers nothing new and opens up a number of holes every time you take a step tempting you to fall in. I'm a C++ developer at heart, but if you decide that Rave is for you then you're going to be hamstrung. Specifically database components will not work due to compilation issues and an inability to find object files. I'm sure with a few days work and searching around for help I may have found a solution, but I think that software development is at a stage that if it doesn't work out of the box then it isn't worth the effort. For anyone who happens to google this post due to the same error, turn back. If you have another option that I would highly recommend it over rave. Specifically why not try writing a web interface reporting engine using Perl, Ruby, PHP or even Java and plug that into your BDS app using a web page interface component as I have now done. The error is:

No DataLink drivers have been loaded. Use the DataLinkADO.RDV driver for the Rave server/IDE or include the unit RvDLADO for applications.

Not a particularly helpful error message, though at first glance it looks like you might be able to solve it. I had no such luck and I recommend you don't bother.

‫‬‭‮‪‫‬‭‮҉Crazy Character

Ok so I stumbled across this crazy character. It's a string of unicode commands that reverses the text that you enter in a browser. I couldn't get it to work in word or other windows applications, but in a browser it does just fine. Plase the character into text somewhere and everything you type afterwards will be refersed. Example: ‫‬‭‮‪‫‬‭‮҉Backwards Text is Funny! ‫‬‭‮

Is There a Future in C++ Development?

I've been a developer for some time and I may as well start this blog with something that's been interesting me for the last few weeks. My background is in C and C++ having come through university while it was still very popular and promoting it throughout my career in various positions. I now find myself in a position that is quite heavily involved in using C++ to solve business application problems on multiple platforms. Company direction here in recent years is aiming directly away from C and C++ for future projects, which is not surprising. The industry appears to be moving forwards with .NET and J2EE platforms with far greater intensity than C and C++. However having developed in all of these technologies before I find they all have their own particular advantage. So, does C++ have a role in today's Software Development future? A quick research into the matter finds many articles that point out that C++ is still very valuable and will be around for many years to come. Conversely if you browse developer forums the flame-wars that exist when discussing the future of C++ burn brightly with very few people sitting on the fence and quite a large majority opposing C++ and it's future as a language. Developers, in particular the vocal ones, appear to choose their language and defend it vigilantly in a similar fashion to gamers defending their chosen gaming platform. This of course makes sense, and both instance are for the same reason: Survival. Developers who have spent many years learning a particular language stake their career on the future of the language they follow in a similar fashion to the gamer who spends hundreds of dollars on their chosen gaming console. It is interesting to me that even after years of professional experience and learning that a Software Developer's attitude can be so easily compared to that of a pre-pubescent teenager squabbling over which gaming console is better. C++ maintains many arguable advantages over the other languages included in this article just as both .NET and Java maintain their own arguable advantages over C++ and each other. Once all the flames die down after arguing which one can do what better, what it all boils down to is that there is no task that you cannot do in any of the three environments. Some tasks will take longer in one language than the other, some tasks will run faster and more efficiently in one language over another, however in today's modern environment of computer power is this as important as it once was? So if the differentiation of each language and platform can be reduced to programmer preference and correct language for each implementation, does C++ have a future? Now I will express some more subjective opinion. I believe that it does, and a strong one. The one thing that C++ has going for it over the other languages, aside from the physical differences in syntax. That is that it compiles to native objects allowing you to control any aspect of the operating system that you like. Java and .NET allow you to use anything within their framework, but nothing more. This allows you a greater level of control making C++ far more effective for software that diverges from the operating system standards and requires performance levels far greater than regular business applications. This brings particular attention to games, for which C++ is well renowned already. It also makes C++ and appropriate choice for server applications with high processing volumes and large requirements for efficient processing and manipulation of data. So does C++ have a bright future? Only time will tell. These are just the opinions of a single developer who has grown his career with the language.