Chris's Wiki :: blog/programming/FallibleSemverAndMVS Commentshttps://utcc.utoronto.ca/~cks/space/blog/programming/FallibleSemverAndMVS?atomcommentsDWiki2018-05-21T11:24:10ZRecent comments in Chris's Wiki :: blog/programming/FallibleSemverAndMVS.By Chris Siebenmann on /blog/programming/FallibleSemverAndMVStag:CSpace:blog/programming/FallibleSemverAndMVS:5ca6837b09ef31e177372da7a1554f32f3a2bc38Chris Siebenmann<div class="wikitext"><p>That's one way to put how MVS doesn't advance version numbers by itself.
Another view is that MVS declines to assume that any change is harmless
or good unless you tell it explicitly (by forcing the version number
to advance).</p>
<p>Semver or not, all we really know when a version number increases is
that the version number has increased. By itself, we don't really know
what this means, and we don't really know if the new version will give
our package and our overall system the same results (or better results)
than the old version.</p>
<p>(It would be nice if semver changed this, but it can't. All it can do
is make it more likely that this happens by creating a social agreement
that it should. This is fragile for various reasons, including that
people have different views on whether certain sorts of changes are
semver compatible or not. Is adding explicit error checks for things
that don't work semver compatible, for example?)</p>
</div>2018-05-21T11:24:10ZBy lilydjwg on /blog/programming/FallibleSemverAndMVStag:CSpace:blog/programming/FallibleSemverAndMVS:07cc51bf35d6b8dc78cd87b374c0e6950cc900e1lilydjwg<div class="wikitext"><p>So MVS ignores all later bug and security fixes? This worries me.</p>
</div>2018-05-21T04:02:21Z