This is hardly rapid progress. It shows that LyX is a slowly evolving beast. Indeed, if only one person is committing many changes quickly, then I expect the code to become more buggy not less. Wouldn't you?
I think the number of bugs is very hard to predict nomatter what procedure you have. I am happy that others want to review my work, but I believe that running and trying the software should have priority over long discussions about this or that.
Also, I think that progress is a function of how many obstacles and overhead you add to the process.
I'm not saying that what you propose is wrong or even that my suggestions are better. I'm saying that we should be cautious, that's all. Two people are saying _WIN32 is the wrong macro, so it's unreasonable of you to reject their opinions out of hand. Especially when some of these arguments are backed up with evidence contrary to your own assertions.
I do not accept your evidence. Compile the code with my patch with mingw and show me the errors. When I see that, I accept your evidence.
To help me test the program and find errors, could someone please send the userguide in the latest version?
Thanks, Asger
