User talk:AW/guideToRobocode
- [View source↑]
- [History↑]
Contents
Thread title | Replies | Last modified |
---|---|---|
Still working on it | 1 | 03:08, 16 May 2013 |
Still haven't decided on a robot | 8 | 22:13, 9 March 2013 |
Corrections | 0 | 23:18, 7 March 2013 |
Just wanted to let you know that I'm still working on this, I've just been pretty busy with finals and then an internship. Expect updates this weekend.
I thought I would start some of the background information since it doesn't rely on a specific robot. Any suggestions so far? (Obviously it should have some pictures to explain things) Would videos be better?
Perhaps a vote for which robot I should use?
My only suggestion so far is that I'd probably leave the 30+ bullet points of Robocode rule details / formulas for later, or in an appendix. It's going to be boring to a newbie, and without any context, they probably won't retain that info anyway. A few sentences of overview about robot/bullet capabilities would be easier to read and draw them in more. You can link to specific rules in the appendix from the main text as you go, if needed.
As for which robot, I think it's really hard to find any objective answer. I doubt anybody's that familiar with multiple top bots' full source code. And even if some folks are, it's a tiny data set to draw conclusions from. I still say that since you're so much more familiar with Gilgalad, it would take a really compelling reason for me to vote for any other bot. I do think Diamond is pretty well designed, and judging from Skotty's incredibly detailed write-ups of XanderCat, it sounds like XanderCat is too. I don't know that DrussGT is a bad choice, either, besides occasional comments from Skilgannon about how messy the code is.
Great start, though I agree with Voidious that it might not be best to put all the rules together right at the top.
It would make sense to use DrussGT, as it is the highest ranking 1v1 robot. But really, the most important thing is how well you understand the code, and I suppose Gilgalad would win that.
Perhaps, instead of jumping into a bot with thousands of lines of code, you could start with walk-throughs of a few sample bots and/or Super Sample bots, just to give the reader a basic idea of a robot's code. After all, I know you're writing this guide for experienced programmers, but you have to leave something for the noobs. :)
If it will go into the code, leave DrussGT out of it for now, I'm at the beginning of a fairly big refactor and cleanup. I'm also in the process of writing up a 1-vs-1/Understanding DrussGT page, in the style of Melee Strategy/Understanding Coriantumr, so I'd rather not have redundant info.
If the robot chosen is designed in a sufficiently modular way, maybe it would be good to have simpler variants of the robot used in early sections to explain things, while keeping the same robot framework used in the late sections.
As far as which robot, I haven't been browsing the code of many robots lately but I feel like Diamond is one of the more cleanly designed ones for the most part, however for the purposes of this type of guide has the drawback of having a lot of melee-related code. DrussGT is well-coded, though the way in which it is (last I checked) less modular than some others may make it not as easy to organize how it's explained in the guide. I haven't yet looked at Gilgalad and XanderCat and some others though I think I will take a look shortly out of curiousity in any case. In any case though, I think the most important thing for the choice of robot is how clearly the one writing the guide understands it :)
If you are writing the guide I definitely suggest going with the robot you understand best. Perhaps include examples from other robots which do about the same thing as you are explaining to give a look at other implementations, but those shouldn't be the default.
Yeah. Now that you and Skilgannon are writing guides, I've decided to stick with Gilgalad.