Plan for Jason Swearingen
by Jason Swearingen · 08/22/2005 (3:12 pm) · 6 comments
I am working on a C# wrapper of the t2d C++ engine. It is meant as a replacement for Torquescript, but can be used in paralell with C++ and TorqueScript code if the developer chooses.
I figured that since a few people have been regularly inquiring as to the status, I would start posting a .plan for everyone (and save me some copy/paste emails!)
Current functionality:
T2D.NET Core Features (290hrs)
- 95% t2d feature parity (need to validate torque gui works properly)
- TorqueScript / C++ / C# interop
Current work areas:
Verify torque.gui work properly. (5hrs)
Split tutorial/sample into it's own stand-alone mod. (5hrs) (currently it is baked into the basicTutorial)
Future work areas:
Port T2D.NET to next version of T2D (5hrs) (will happen when next t2d version is released)
Port all T2D sample programs to t2d.net (10hrs) (will do this after next t2d version is released)
Add .NET Forms Gui support
- Base functionality (10hrs)
- T2D Specific SDK (10hrs)
Add code security measures
- Strong Named Assemblies (15hrs)
- Cryptographically secure resources (10hrs)
- Hacking countermeasures (obfuscation, etc) (20hrs)
Add .NET Scripting support
- Basic functionality (10hrs)
- Restrict T2D functionality to Signed Assemblies only (5hrs)
Write t2d.net TechDemo/Sample
- using all t2d.net functionality (15hrs)
Finalize documentation
- Fill in any missing intelisense documentation (5hrs)
- add sections for each major api (5hrs)
- add additional explanations for every api (10hrs)
I will post updates as each feature is completed.
I figured that since a few people have been regularly inquiring as to the status, I would start posting a .plan for everyone (and save me some copy/paste emails!)
Current functionality:
T2D.NET Core Features (290hrs)
- 95% t2d feature parity (need to validate torque gui works properly)
- TorqueScript / C++ / C# interop
Current work areas:
Verify torque.gui work properly. (5hrs)
Split tutorial/sample into it's own stand-alone mod. (5hrs) (currently it is baked into the basicTutorial)
Future work areas:
Port T2D.NET to next version of T2D (5hrs) (will happen when next t2d version is released)
Port all T2D sample programs to t2d.net (10hrs) (will do this after next t2d version is released)
Add .NET Forms Gui support
- Base functionality (10hrs)
- T2D Specific SDK (10hrs)
Add code security measures
- Strong Named Assemblies (15hrs)
- Cryptographically secure resources (10hrs)
- Hacking countermeasures (obfuscation, etc) (20hrs)
Add .NET Scripting support
- Basic functionality (10hrs)
- Restrict T2D functionality to Signed Assemblies only (5hrs)
Write t2d.net TechDemo/Sample
- using all t2d.net functionality (15hrs)
Finalize documentation
- Fill in any missing intelisense documentation (5hrs)
- add
- add additional explanations for every api (10hrs)
I will post updates as each feature is completed.
About the author
#2
08/23/2005 (10:53 am)
Darren - Don't forget to port it to BASIC while you're at it. :P
#3
08/23/2005 (12:56 pm)
@Darren: fyi, anything written in .NET is usable by any .net language (C#, C++CLI, VB.NET, J#, IronPython, etc)
#4
they're still probably faster than TScript callbacks...
You you should email me an svn login so I can post patches. :-b
08/23/2005 (3:09 pm)
I tend to use more AoP cascading signals, but the delegates in .Net are still too expensive.they're still probably faster than TScript callbacks...
You you should email me an svn login so I can post patches. :-b
#5
Also, if you are interested in speed of t2d.net VS torquescript, I didnt run perf numbers on this yet, but Aaron Wieland did and he mentioned that t2d.net seems to be aprox 50% faster than torquescript, which i can belive.
BUT, obviously if multiplatform is your cup of tea, this isnt for you... unless you can get this working with Mono (quite possible)
08/23/2005 (3:58 pm)
@Nanomid: fyi .NET2.0 delegates are about 2x faster than .NET1.1 delegates (and about 50% the speed of direct function calls) which is pretty damn good IMO.Also, if you are interested in speed of t2d.net VS torquescript, I didnt run perf numbers on this yet, but Aaron Wieland did and he mentioned that t2d.net seems to be aprox 50% faster than torquescript, which i can belive.
BUT, obviously if multiplatform is your cup of tea, this isnt for you... unless you can get this working with Mono (quite possible)
#6
For what it's worth, I re-implemented the tutorial in Smalltalk, and got a framerate just as good as with TorqueScript. And I suspect that my Smalltalk/DLL implementation has significantly more overhead than T2D.NET. Of course, the tutorial is very simple, but I was still surprised by the results.
08/27/2005 (2:08 am)
Actually, I haven't even used T2D.NET yet (that would require installing both Visual Studio 2005 Express Editions for both C#and C++, and recompiling the engine with the latter -- something I tried and failed to do previously). My tests were done in Smalltalk, using T2D as a DLL, so they aren't directly applicable to T2D.NET. They did suggest that using the scripting interface is significantly slower than calling the macro-generated functions directly, as T2D.NET does, however. I'm not sure, though, how similar calling Con::execute() for each method invocation is to executing compiled TorqueScript.For what it's worth, I re-implemented the tutorial in Smalltalk, and got a framerate just as good as with TorqueScript. And I suspect that my Smalltalk/DLL implementation has significantly more overhead than T2D.NET. Of course, the tutorial is very simple, but I was still surprised by the results.
Torque Owner Darren Stuart