[boost] [Removing support for old compilers]

Subject: [boost] [Removing support for old compilers]
From: Robert Ramey (ramey_at_[hidden])
Date: 2013-10-14 12:20:25

I realize that I've asked this before - but I never got a response which I 
found
convincing.  I'll try to ask my question better this time.

The question - What is the value of undertaking this task.

Let's look at the value of according to the type of boost
participant.

a) boost user - all this code in config and the libraries is
internal to the libraries.  After this change the look and
usage of boost will not change at all. - no benefit and
no cost to the boost user.

b) author of a new library - this person can set the
requirements to what ever he wants subject to the
constaint that the code compile and work on any
comforming C++ implementation.  That is, there has never
been a requirement to support non-conforming or
older compilers.  This won't change.  Only authors
who want to support an older compiler would be
affected.  It's safe to say there aren't many authors
in this category.  so no benefit and no cost to
authors of new libraries.

c) boost users on older platforms.  They are to
be stuck with the boost 1.54. I don't know how
big this group is but I'm sure it has some members.
A big loss for this (probably small) group as they
won't benefit from things like bug fixes and enhancements.

d) maintainers of boost libraries.  When I first raised this
question, I was assured that it would be totally transparent
to me and that I would not have to be concerned about it.
Hmmm - well  OK - as long I don't have to wade in and
muck around with ten year old code - (don't even think about
addressing any changes which make old data sets unreadable).
My worst nightmare is having to go back and re-debug
20 thousand lines of code running on 10 different compilers.
Recent postings have made me doubt that one can really
undertake this and guarentee that I won't have to do this.

Soooooo - bottom line:

No benefit to any boost participant
Definite costs to some boost users.
Possible big problems for library maintainers.

So how can this be worth doing?  If one sees real value
in this - how about creating this on a branch or a fork.

Robert Ramey