Subject: Re: [boost] [review queue] What to do about the library review queue?
From: Niall Douglas (s_sourceforge_at_[hidden])
Date: 2017-03-15 16:16:30
On 15/03/2017 11:19, Vinnie Falco via Boost wrote:
> On Wed, Mar 15, 2017 at 4:53 AM, Niall Douglas via Boost
> <boost_at_[hidden]> wrote:
>> There are libraries in
>> the Boost main distro with open bug count nine years old. That's
>> unacceptable in any set of libraries claiming to have quality.
>> Last time I looked in detail (before the corporate sponsorship effort),
>> about half of Boost libraries had maintenance issues, but there was a
>> hardcore 20% that is damaging the Boost brand and those need to go.
> I've been using Boost for years now and I didn't even know about these
> libraries with maintenance issues. And knowing about those issues now
> does not deter me from continuing to use Boost - I will simply avoid
> the libraries with problems and continue to use the good ones. Can you
> provide some evidence that this problem has "damaged" the Boost brand?
The poor folk who have locked themselves into the unmaintained
Boost.Iostreams generally come to regret it. Additionally it should
never have passed peer review, its design is fundamentally flawed.
There are increasing problems with Boost.ASIO as it gets ever further
away from ASIO.
Boost.Test before develop branch finally got merged to mainline was a
real problem. Lots of wasted time for a lot of folk over many years.
Some feel that the lack of cmake based build and test is a brand
damaging problem. I'm less convinced of that personally, but then I
generally only use the header only Boost libraries which don't suffer
from being misbuilt by package repo maintainers etc.
Some Boost libraries have muxed in support for the C++ 11 STL, others
have not. That is causing big problems for some Boost users. I'd
personally throw all the C++ 11 STL unhelpful libraries into deprecation
by this stage.
I'm not subscribed on stackoverflow to other Boost libraries, but as a
general rule, frustration and pain expressed there is a good indicator
of ongoing brand damage. Again, without an active policy making
political leadership, there is nobody to identify brand damage, and even
if identified, no means of addressing it. If we at least could segment
the bad libraries into a "bad Boost" distro, it would be a huge improvement.
-- ned Productions Limited Consulting http://www.nedproductions.biz/ http://ie.linkedin.com/in/nialldouglas/
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk