Hey, On Tue, Jul 27, 2010 at 7:09 AM, JohnC <[email protected]> wrote: > Please do not take this the wrong way, I am sure you have taken your > best efforts to write bug-free code and do testing, but the fact > remains that the new code base is only 3 month mature.
No problem John, it's a totally valid concern. > Can you provide any insight if and why the above may not be true? Am I > incorrect in the above analysis? But yes, your analysis is not quite correct and is missing some points. The code in github has been pushed recently, but I started working on the 0.9 version in late 2007. Thanks to some early-adapters, who happen to be big users of Cecil, we've been able to exercise the code base when they did the porting effort from 0.6 to 0.9. I waited as long as possible to release 0.9, to be sure to have something at least as good as 0.6. Of course we had small regressions and bugs, but 0.9.3 and the upcoming 0.9.4 are really stable. As a side note, 0.9 is backed by a large test suite, making us confident when we land patches. At some point, I was a bit unwilling to land big patches in 0.6 as I had no idea what I was breaking or not (ô the joy of bearing with youth mistakes). Also note that 0.9 has been tested with all the problematics assemblies I've collected with 0.6, and works just fine, if not better. 0.9 is now used in a diversity of commercial products and open source projects, including compilers, and as of today, we have no known case of assemblies that can't be read or written. I, for one, am still in the process of porting large portions of code from 0.6 to 0.9, and except for the API breakage detailed in the migration page of the wiki, everything is going smoothly. The API breakage can indeed be really annoying. But it really does clean a lot of dark spots in 0.6, so I'm happy I did it. -- Jb Evain <[email protected]> -- -- mono-cecil
