velmarin wrote:
I encourage all of you Windows 7, Visual C + +, and Intel Compiler
over my dead body...
Linux + bash/make/gcc + a good IDE (or VIM or Emacs and gdb if you're a real pro)
I have never been able to use make comfortably. I find the syntax is awkward. Things clicked for me when I started to use Rake, which is Make with Ruby language incorporated. It is extremely powerful and easy to use. Since I have most of my scripts in Ruby, that was very natural. I use gedit (I only need color coded syntax, but developers lately decided to screw their search... arggggghhh), ruby scripts, Git, xxdiff, and for the C language I use GCC, Intel, and Mingw64 compilers (cross compiles to windows). Let's not forget grep and all the small linux tools. Gedit complements very well with Git, when you are jumping from one branch to the other. It detects what file is new and it asks you to re-reload when you click on the tab.
velmarin wrote:
I encourage all of you Windows 7, Visual C + +, and Intel Compiler
over my dead body...
Linux + bash/make/gcc + a good IDE (or VIM or Emacs and gdb if you're a real pro)
I have never been able to use make comfortably. I find the syntax is awkward. Things clicked for me when I started to use Rake, which is Make with Ruby language incorporated. It is extremely powerful and easy to use. Since I have most of my scripts in Ruby, that was very natural. I use gedit (I only need color coded syntax, but developers lately decided to screw their search... arggggghhh), ruby scripts, Git, xxdiff, and for the C language I use GCC, Intel, and Mingw64 compilers (cross compiles to windows). Let's not forget grep and all the small linux tools. Gedit complements very well with Git, when you are jumping from one branch to the other. It detects what file is new and it asks you to re-reload when you click on the tab.
Miguel
I am a huge fan of Ruby, perhaps I will give Rake a try myself.
Capital punishment would be more effective as a preventive measure if it were administered prior to the crime.
Visual Studio 2012. Great for C# .Net development. Excellent debugging and performance measurement tools. Good tool if you're willing to take a slight performance hit for the rapid development that's possible with an IDE focused on object oriented design.
Admittedly, not the best choice for a chess engine where code performance is king. But a good choice for GUI development. Now that I've written a decent chess engine (on par with TSCP, strength-wise), I'd like to write a chess GUI. Visual Studio will simplify that task.
emadsen wrote:Visual Studio 2012. Great for C# .Net development. Excellent debugging and performance measurement tools. Good tool if you're willing to take a slight performance hit for the rapid development that's possible with an IDE focused on object oriented design.
Admittedly, not the best choice for a chess engine where code performance is king. But a good choice for GUI development. Now that I've written a decent chess engine (on par with TSCP, strength-wise), I'd like to write a chess GUI. Visual Studio will simplify that task.
Hope you produce something interesting that runs on Linux via mono.
Don
Capital punishment would be more effective as a preventive measure if it were administered prior to the crime.
Hope you produce something interesting that runs on Linux via mono.
I'll steer clear of WPF & Metro user interfaces. I have no experience with Mono but I'd imagine it has better support for traditional Windows forms than the newer user interfaces.
I don't know why people like so must to have debugging capabilities. I hardly ever use them. I admit I used then when starting to code the move generator, but when you have a stable move generator and a stable search, with good assets (the best preventive debugging tool), you should don't need it. The times I am in the need to debug I only do same printf and with that I am able to fix. Also you must take into account that chess engines are same Times very difficult to track in a debugger for those so many recursions. and stack.
Also a good symmetry check function to test the evaluation is one of the best preventive measures
I don't consider myself a very goods programming, but trying to avoid the debugger let to better programming preventive habits, and if you don't have, the better, that you learn the hard way.
Kempelen wrote:I don't know why people like so must to have debugging capabilities. I hardly ever use them. I admit I used then when starting to code the move generator, but when you have a stable move generator and a stable search, with good assets (the best preventive debugging tool), you should don't need it. The times I am in the need to debug I only do same printf and with that I am able to fix. Also you must take into account that chess engines are same Times very difficult to track in a debugger for those so many recursions. and stack.
Also a good symmetry check function to test the evaluation is one of the best preventive measures
I don't consider myself a very goods programming, but trying to avoid the debugger let to better programming preventive habits, and if you don't have, the better, that you learn the hard way.
I use gdb once in a while - but I do believe debuggers can make you lazy. I grew up not using them and developed a knack for finding bugs quickly and have a test as you go style. I make good use of printf statements for debugging.
debuggers really can save you some time but I only pull it out when I get to the point that I'm scratching my head - which is rare.
Capital punishment would be more effective as a preventive measure if it were administered prior to the crime.