On Sat, Oct 18, 2008 at 5:18 PM, Louis. Lenders xerox_xerox2000@yahoo.co.uk wrote:
Thank you for your answer.
I agree, that before sending a patch I should write a test first. But the problem is that even a tests are ignored without any explanation. Maybe my tests aren't good, but why anybody just tell what's wrong.
Vitaly claims he gets no feedback for his patches. He knows full well that I give him plenty of feedback. I explained to him what was wrong with his test SHFileOperation patches, yet he repeatedly sent the same patch to the list. I'm not going to repeat the same comment over and over again.
Seems to be common habit on this list. I hope you're not added to the "Julliard ranking s**t list? Yes: REJECT" ( see http://www.winehq.org/?issue=353#WineConf%202008%20Keynote )
Bologna. There is no such list. Every developer on the project has an unspoken trust rating, and every long term developer knows said rating for every other developer, especially Julliard. When you initially start submitting patches, your trust rating is low, understandably since we have no idea of your skill level. As you submit correct patches, your trust rating rises. When you repeatedly send bad patches, your rating drops. It is this trust rating that makes patch reviewers look at a patch from a guy with a low trust rating to say "Oh, another patch from that guy...it's probably not right." Our development model, with such a high bar for entry, is why the Wine project is such a clean, successful project. You can keep bitching about the way it works, or you can do what we all did in the beginning and get over that hump. Start small. Write exhaustive, well-written test cases for many features. Test cases are one of the easiest ways to get new code into the tree. After a while, start submitting tiny fixes that fix broken tests. Through this entire process, you will become a better developer, guaranteed.