Some thoughts

[RazvanN]

On Tuesday, 26 November 2019 at 09:11:33 UTC, S.G wrote:
> On Tuesday, 26 November 2019 at 05:05:09 UTC, Jab wrote:
>> On Monday, 25 November 2019 at 16:53:43 UTC, berni44 wrote:
>>> a) Remove bugs
>>
>>
>> I feel this gets brought up a bit, how pull requests sit idle 
>> for years along with bug reports. People working on D work on 
>> whatever they want, fixing bugs and handling pull requests are 
>> just chores that no one wants to do.
>
> Fixing D bugs is a one way process so I don't even recommend 
> doing this seriously, e.g on a daily basis, several hours per 
> days. This will inexorably leads to frustration. Fix bugs if at 
> work your company needs some particular fixes, fix bugs if you 
> need some fixes for your own side-projects, fix bugs if this is 
> required for your studies. But don't fix bugs just for the love 
> of the D programming language.

I have fixed 150+ bugs and closed another 200+ (invalid or fixed) 
in the compiler and I can say that this is a great way to 
understand the innards of the software. Since documentation is 
scarce, fixing bugs is a fun way to understand (and document) 
what happens in either dmd, druntime or phobos and I encourage 
anyone who is interested in having an in depth understanding of 
the D compiler to pursue this endeavor.

Although the list of unresolved bugs is large I'm sure that 
aproximately 40% of them have been fixed in the meantime or are 
invalid, 40% are easy to fix, thus leaving only 20% that are more 
difficult to fix. This estimations are based on the distribution 
of bugs that I have encountered so far.