Re: rustc panic

["Ian D. Leroux" <idleroux%fastmail.fm@localhost>]

On Wed, 26 Jul 2017 15:59:42 -0600 (MDT) Swift Griggs
<swiftgriggs%gmail.com@localhost> wrote:
> I'm not trying to bag on Rust. I'm just wondering if this weird
> gene-drive for Rust really what's going on here. Ie... people are
> trying to wedge in new language requirements to existing software
> where it really doesn't seem to be a good fit and it darn sure wasn't
> there before.

<counter-rant>

No need for conspiracy theories.  Some people felt that C was a poor
tool for writing a large performance- and security-critical
application, judged (based on the literature) that better tools were
possible, and set out to build such tools.  They were willing to put in
the work, and apparently some people feel that the resulting tools are
enough of an improvement to make it worth implementing /
maintaining (parts of future versions of) firefox with their help.  If
sufficiently many people disagree, they can continue developing the
existing C codebase. Remember that firefox was once phoenix, a fork of
mozilla by people who believed that they could do better than the
mainline project, and were correct in their belief.

I'm not saying you should be interested in Rust.  I'm on the fence,
myself.  But no one is trying to pry C from your cold dead (or even
warm living) hands.  C compilers aren't going anywhere, and even if the
gcc maintainers dropped support for all past C standards there'll
always be old versions (or tcc, or pcc, or ...).  Nor are existing
ANSI-C codebases being deleted from the face of the internet, as far as
I know.  But other people maintaining/updating software have as much
right to choose the tools that suit them as we do.  We can do the work
of adjusting to their tools, or we can live without their contributions.

Of course, it's *nice* when everyone agrees on the same toolset so that
we can more easily piggy-back on each other's work.  And there is a
certain hypocrisy in abandoning cross-platform compatibility once
you're big enough not to need to ride on others' coat-tails, as parts
of the Linux community seem to have done.  But as the lucky beneficiary
of others' work, I don't get to choose the tools they use, and I don't
have a right to insist that they ignore the past 40 years of
programming language research, even if it *would* make my life much
more convenient.  I'm a beggar and can't be a chooser.

I'm reminded of the rants that went around when John Polstra wrote
CVSup in modula-3.  It was an unconventional choice, but it made his
life easier and he was the one writing the program.  Some people liked
CVSup enough to maintain a stripped-down easier-to-port modula-3
compiler just to build it.  Other people liked it enough to reimplement
it in C (as csup). Yet others (NetBSD) just did without.  Those three
options work.  Telling Polstra (or, in the current context, Graydon
Hoare) that he can't use the tool he prefers to solve a problem ...
does not.

</counter-rant>

--
IDL