From: Boost-users [mailto:boost-users-bounces@lists.boost.org] On Behalf Of degski via Boost-users
Sent: 10 May 2017 14:55
To: boost-users@lists.boost.org
Cc: degski
Subject: Re: [Boost-users] Cutting headers

On 10 May 2017 at 11:36, Paul A. Bristow via Boost-users <boost-users@lists.boost.org> wrote:

The interdependence of Boost libraries isn't a 'problem' - it may be an inconvenience, but is an unavoidable consequence of missing
built-in language features and Boost's invaluable working for multiple platforms, targets and compilers and standard versions.

In my opinion, supporting things like VC7.1 in boost is actually doing the world a dis-service. The CRT's that come with VC7.1 (f.e.) have severe, known (to black-hat hackers) security issues AND ARE NOT SUPPORTED BY Microsoft.

Microsoft is clear about what is and what isn't supported. Boost should not support more than that. If one doesn't adopt this approach, the backlog (of mess from the past) will never be cleared and windows will forever be characterised as THE unsafe platform, which I think is un-warranted.

When will boost stop supporting VC7.1 (and all the other out-dated stuff, VC8 etc etc)?

Well I think that it is, slowly, by adding new libraries that say C++11, or even C++17 only.

Some reasons for not trying to actively strip out C++very_old handling code are

1 Not gratuitously mucking up those who are stuck maintaining C++98 systems but want to use the latest Boost version (surely a good idea?).  If they have security issues, *they* should fix them; current Boost will still work.

2 It isn’t good use of limited peoples time to do this.

3 There is a very big risk of collateral damage.  If it ain’t broke, don’t fix it.

In short, evolution, not revolution - that usually means bloodshed ;-)

HTH

Paul

---

Paul A. Bristow

Prizet Farmhouse

Kendal UK LA8 8AB

+44 (0) 1539 561830