You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
111 of the 188 entries have an empty maintainers field. I opened 40 of them myself, which is why I am writing this.
An entry written by someone who does not maintain the framework is a guess about how that framework is best used. A reader of the board cannot tell whether the person who wrote it knew the framework well, gave it their best, or had a reason not to. I cannot prove any of those things about my own forty either, and that is the point: I compete here, so anyone can say tomorrow that I under-implemented someone else's framework to make mine look better. Since there is always a better way to write something, that accusation will always have something real to point at. I have already had that conversation this week, about an entry I opened.
The fix is not to remove those entries. Coverage is worth having, and most of them are probably fine. It is to say where the number comes from.
Mark an entry with no maintainer as unofficial on the board. A short label and a line in the popup: this implementation was contributed by a third party and has not been reviewed by the project. The reader then knows what they are looking at.
Tell the projects. For each unmaintained entry, an issue on their repository saying that their framework is benchmarked here, where the implementation lives, and that they are welcome to review or replace it and to be listed as maintainer. That converts the label into an invitation rather than a warning, and it is the only path by which those 111 become 0.
None of this costs the board any coverage, and it makes the numbers honest about their provenance, which is the part that is hardest to defend once someone decides to attack it.
111 of the 188 entries have an empty
maintainersfield. I opened 40 of them myself, which is why I am writing this.An entry written by someone who does not maintain the framework is a guess about how that framework is best used. A reader of the board cannot tell whether the person who wrote it knew the framework well, gave it their best, or had a reason not to. I cannot prove any of those things about my own forty either, and that is the point: I compete here, so anyone can say tomorrow that I under-implemented someone else's framework to make mine look better. Since there is always a better way to write something, that accusation will always have something real to point at. I have already had that conversation this week, about an entry I opened.
The fix is not to remove those entries. Coverage is worth having, and most of them are probably fine. It is to say where the number comes from.
None of this costs the board any coverage, and it makes the numbers honest about their provenance, which is the part that is hardest to defend once someone decides to attack it.