|
Boost : |
From: Robert Ramey (ramey_at_[hidden])
Date: 2023-02-24 18:04:51
On 2/22/23 3:10 PM, Glen Fernandes via Boost-announce wrote:
> Furthermore, as Boost release managers we support authors exercising
> this freedom, and do not intend to impose any restrictions on what
> language support they must maintain.
I believe that traditionally, new boost libraries have been required to
compile and pass tests using that the latest version of C++. I don't
believe that similar requirements on pre-existing boost libraries as C++
evolves ... or am I wrong about this? If a library stops working
because it depends upon something that has been deprecated, should the
library be suspended from being in boost? I gather not. I'm not
advocating for this. I think the stated policy is correct as it stands,
but that leaves a few questions unanswered. Generally, library
maintainers want their code to survive and usually maintenance is not a
huge time suck. If no maintainer is sufficiently interested to follow
this policy or if the library is hard to maintain, maybe it's time it
should be set aside.
Robert Ramey
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk