Re: [boost] [peer review queue tardiness] Cleaning out the Boost review queue

Subject: Re: [boost] [peer review queue tardiness] Cleaning out the Boost review queue
From: Niall Douglas (s_sourceforge_at_[hidden])
Date: 2015-04-02 15:08:06

On 2 Apr 2015 at 11:08, Emil Dotchevski wrote:

> > Both myself and Antony have served as review manager for other
> > libraries since submitting our libraries. The present situation is
> > frustrating, though I'd imagine for Emil it is even worse seeing as
> > he's been waiting a year longer, and yet has been doing all the work
> > a library maintainer does except without the recognition or
> > visibility of being included into Boost official.
> >
> 
> Doesn't this simply mean that there isn't enough interest in the library
> within the Boost community? :)

It would appear so. Another way of saying this is that your library 
is so niche that very few consider themselves competent to review 
manage it. I, for one, wouldn't even try.

> > Whilst peer review is important, it is impractical for very niche libraries
> 
> Should niche libraries be part of Boost?

That's part of the wider debate, definitely. Is quality what we want 
for the Boost brand, or is popularity?

I'm in the former camp, mainly because as the standard library grows 
it is necessarily the case that the low hanging fruit is picked and 
subsequent libraries must be more niche, and less popular. Therefore, 
to grow and evolve Boost I believe should aim for quality, not 
popularity.

> In the case of QVM I like to think
> that a generic quaternion/vector/matrix library is not *that* niche but the
> evidence seems to show that it is. Regardless I don't feel that the Boost
> community owes me a review. :)

I don't think the community does no.

I do think that someone seeking a review manage needs to first give a 
review manage in return. Otherwise it's bad karma.

For the record, when I originally reviewed which library to review 
manage, I did genuinely try to select the longest waiting Boost ready 
library first which was yours. Problem is, I don't really understand 
what makes your library good or bad - I have no experience 
programming such maths on a computer you see, all my experience was 
on paper at university. So I ended up on TypeIndex, a topic I 
understood very well.

Niall

-- 
ned Productions Limited Consulting
http://www.nedproductions.biz/ 
http://ie.linkedin.com/in/nialldouglas/