I firstly analysed Roslyn because of seeming the a-priori-most-appealing-to-me repository (as already explained
), although this experience wasn't too good. Additionally and unlikely what happens with CoreCLR/CoreFX, this code is not too intuitively appealing to someone in my situation (i.e., experienced .NET programmer expecting to easily locate "recognisable parts") as far as it is mainly focused on the C#/VB.NET compilers. In any case, I will most likely contribute to Roslyn in the future.
The aforementioned not-too-positive episode is helpful to understand why I quickly liked CoreCLR so much. This time, I could easily find the code associated with virtually any feature (e.g., all what is contained in the C#/VB.NET
namespace). Having full access to the pure essence of the .NET Framework, my primary programming environment during the last few years! That's why it didn't take me too long to post my two first issues (i.e., some hours after having downloaded the code for the first time and on the next day):
"Problems with thousands separators when parsing strings to numbers"
"Decimal type not able to parse scientific notation"
Both of them are related to the behaviour of the current version of
and, indirectly, to other parse methods of this and different numeric types. Curiously, I didn't realise about any of these issues (i.e., thousands separators behaving unexpectedly and scientific notation not being fully supported) at work, but while answering questions in stackoverflow.com (my SO profile
). In any case, I thought that both behaviours were worth correcting, that's why I created those issues and actively participated in the associated discussions.
After reading most of the comments in the aforementioned threads, anyone should easily understand that I didn't enjoy this experience too much. The fact that both issues were closed and the associated proposals rejected didn't bother me even a bit; on the other hand, I did find the behaviours of some participants completely unacceptable, mainly in the issue #2285. Anyone interested in knowing more about what happened there can take a look at both threads and at the associated external references (IMO all this is a good condensed summary of most of my online self-describing efforts, one of the requirements of the new Attitude 2.0
, a required-but-very-unappealing-to-me work).
In general terms, I am quite happy with various outputs of this curious experience; for example: very clear ideas about what the .NET community implies (i.e., lots of people, not necessarily too knowledgeable or reasonable) or the format of my contributions there (i.e., not-too-controversial-issues + forking + pulling + waiting). Additionally, I got various good advices and met some quite knowledgeable people.