Mikester's Blog

Xselerator

I have been really into XML/XSLT this year.  I am particularly interested in using XSL to generate source code and I've been very successful in this regard.  The bible on this subject, as far as I am concerned, is Code Generation in Microsoft .NET, by Kathleen Dollard.  This is where I first encountered this approach and I must say there is a lot to like about doing code generation this way.

Xselerator

I've used a number of tools to edit XML and I have been very frustrated because I feel they should be much better.  I've actually used Visual Studio, Delphi, and Dreamweaver, to name a few, and while they are better than nothing, I figured I could develop one that was superior.  The rub is I have had no interest in doing so.  Anyway, I discovered Xselerator the other day and it does what I really wanted an XML editor to do and at a reasonable price.  I downloaded and installed the 14 day trial and on day one, I ordered the product.  There is an annoying little nag screen that takes away all the joy of using the product and I knew I wanted it, so I purchased it.

The product essentially has a left pane and a right pane.  The left pane is your XML document and the right pane is your XSL document.  You can have multiple documents, each indicated by a notebook tab.  Press F9 and the transformation takes place before your eyes and the result appears in a popup window.  Very slick.  It has intellisense-like functionality, tag completion, puts the cursor in the right place, etc.  It is a very nice product and is something that will speed your development along.   That's good news.

Borland to sell off IDEs

Yesterday was an interesting day for Borlanders.  I have a love/hate relationship with Borland.  I feel the company really knows how to disappoint and thrill its customers.  The management disappoints.  The products excite us - usually.

The history of Borland is a fascinating one.  The first product of Borland's that I ever purchased and used was Turbo C 2.0 back in the summer of 1989.  I made a decision to learn what I needed to produce commerical products.  I discovered that most serious developers used the C programming language.   Well I am still on that quest to learn how to develop commercial products.  I had heard of Turbo Pascal some years earlier and I almost got that compiler instead, but opted for the C compiler.  I sure don't regret having done that.  It was easy to go from C to Pascal and C is the basis of so many other languages (C++, C#, Java, PHP).

I loved Borland's products.  I used their C, C++ and Pascal compilers.  I also used some of their other products such as QuattroPro and SideKick.  I loved SideKick - it was so simple and elegant and the first real kickass PIM.  It was of course a DOS program which ran as a TSR (memory resident).

Well Delphi was another product that transformed the development environment.  I have used it since the 1995 version 1.0 release.

Here was one revelation I had after having done C/C++ development.  I purchased TurboPascal for a particular project.  I learned enough Pascal to get by.  However, I was overwhelmed by the compilation speed.  It was blazingly fast and left any C/C++ compiler in the dust.  Delphi was no different.  I realized this was a superior product in so many ways.

I have developed with Delphi ever since.  I don't do C or C++ anymore.  I do use C# (which was architected by the very same Anders Heilsburg that brought us TurboPascal and Delphi).  C# is similarly elegant.  But for high performing Win32 applications, Delphi is the way to go for me.  My only concern has been the viability of Borland.  The management has done some blunderous things - monumentally blunderous in my opinion.  They price their products in a way that drives away business.  I would have enjoyed playing around with Kylix.  I saw the introductory pricing and said to myself it wasn't worth it.  We as developers were very interested.  But this was new territory and I was not interested in exploring it (hobby mode explore) at the price they wanted.  So I concentrated my efforts on doing plain vanilla Delphi programming.

I upgraded a year ago to D2005 - a disaster.  I tried using it a number of times and it lays there dormant on my machine.  It was unusable.  I also got it without SA (software assurance) because they wanted an arm and leg for the Pro Version SA at the time (they since reverted their software assurance policy).  I upgraded, gave the company hard-earned dollars for what was a worthless product in my opinion (I have lots of agreement on this one).  I can only assume that management wanted the revenue from the upgrades and pushed the product out the door before it was ready.  I have not yet upgraded to D2006 - which I hear is wonderful.  I probably will when I have time to do so.

The management had ideas where the company should go back around 1997 and they lost Anders.  Big mistake to let a guy like that go.  They've lost a lot of talent for sure by trying to be something they are not.  Now they are divesting themselves of the products that brought them the most success they have ever had.

So the big news is Borland is selling off the IDE products (Delphi, C++ Builder, etc.), which will be a separate company with the mission of turning out world class development tools.  I am thrilled at the early news about this.  Hopefully it comes to fruition.

I still get excited whenever I fire up Delphi.  It is a truly magnificient product and we all would like to see more great Delphis in the years ahead, including a 64-bit version to give it new life for years to come.