Coding Assistance Tooling, why so many scattered projects?

Martin B martin.brzenska at googlemail.com
Fri Apr 22 17:41:01 UTC 2022


Hi everyone,

Recently, I wanted to improve my experience with dlang in sublime 
text and tried several solutions. The ones with the most 
potential and which are (imo) futureproof are the LSP based 
solutions "dls" and "serve-d". Unfortunately both are somehow 
incomplete, dls has features that are missing in serve-d and vice 
versa.

I am not here to complain, these are super cool and important 
projects. I would like to speak my mind and learn by feedback: I 
think that code editing assistance is a important key feature 
when it comes to adoption. Maybe it would make sense to move it 
closer to the "dlang core tooling" and handle it in a more 
monolithic aproach. As i understand, right now there are several 
github repos/dub packages, each solving one specific problem 
which are depending on each other and maintained by different 
people... maybe this sounds like KISS and single responsibility 
principle and yada yada yada - but IMO, it seems to me to be 
wastefull on ressources in this case.

I am a webdeveloper, so chances are good that i do not understand 
important keyfacts here, since i am not coding on this level in 
everyday life - but for example: Why are things like libdparse 
even a dub package?

Looking forward to learn :)


More information about the Digitalmars-d mailing list