a remark about 'stability'

[Rainer Weikusat]

,----
| I suspect that the CPAN-hostile environment which still promotes
| CGI.pm as the state of the art in Perl overlaps strongly with the
| older-is-better-stability-over-all enterprise Linux distribution
| group, of which Red Hat for example won't provide Perl 5.20 for
| several years,
`----

I also suspect that the fraction of people who have existing code they
don't want to see broken because of 'political attempts' to promote
this or that not-yet-abandoned wepp fremmwuerg strongly overlaps with
the fraction of people who have existing code they ....

But you, dear modernists, can do better than outlined above: Just
break more stuff which always used to work more rapidly and you might
manage to make Perl 5 go the way of Perl6^Wthe dodo: Nobody uses it
because it isn't fit for use anymore. Imagine that you can probably
get 'RedHat' to never ship Perl 5.20 and onwards at all because
none of their customers wants it.

 

[Rainer Weikusat]

,----
| I suspect that the CPAN-hostile environment which still promotes
| CGI.pm as the state of the art in Perl overlaps strongly with the
| older-is-better-stability-over-all enterprise Linux distribution
| group, of which Red Hat for example won't provide Perl 5.20 for
| several years,
`----

I also suspect that the fraction of people who have existing code they
don't want to see broken because of 'political attempts' to promote
this or that not-yet-abandoned wepp fremmwuerg strongly overlaps with
the fraction of people who have existing code they ....

To supplant this with some 'real-world data': Among the things I need
to deal with at the moment are two packages making up a 'system
management GUI' I'm supposed to maintain and extend in order to
provide a more complete set of features. These consist of 7,404 lines
of code based on CGI.pm (5,553 Perl, 1,043 sh, 808 C) and even if I
would agree with 'strange' opinion statements a la 'in order to use
the CGI protocol, you have to know the CGI protocol' (Really? Who da
thunk ...), considering what amounts to 18 lines of Apache
configuration 'hard to deploy' (Someone confusing his ignorance with
'objective technical difficulties' OR someone hoping to confuse
ignorant others?) or assuming that "it hasn't changed much since 1998"
would constitute a quality judgement (the stove I'm using today also
"hasn't changed much" compared to stoves I used to use ten years ago
-- clearly, the fact that it works instead of being 'under
construction' is abhorrent), there would be no way in heaven or hell
to convince my boss that rewriting these seven-and-a-half thousand
lines of code in a currently more fashionable way in order for them to
become more ... fashionable would be a sensible way to use company
work time.

 

[Rainer Weikusat]

Oh, do shut up. Nobody cares about your paranoid ranting.

That's exactly the kind of 'intelligent discussion' of possibly
far-reaching core perl API changes which has happened in this
particular area so far: I hate it. Whoever disagrees with me doesn't
count. Ergo: It is, for all practical purposes, unused, and can be
removed.

RedHat won't provide 'Perl 5.20' for several years because they have
existing customers who use the current code and don't fancy 'a la mode
redevelopments' of software they've presumably been using for years.
Changing working code because tools go in and out of fashion
with this or that (vocal minority) 'developer community' doesn't make
any sense from a business standpoint. And J. Random Internet pseudonym
can rant and rave all he wants about that, it won't matter.

Insulting me is a lot easier but even this doesn't really accomplish
anyhing: It just highlights the lack of arguments in favor of the
change.

 

[Rainer Weikusat]

That's exactly the kind of 'intelligent discussion' of possibly
far-reaching core perl API changes which has happened in this
particular area so far: I hate it. Whoever disagrees with me doesn't
count. Ergo: It is, for all practical purposes, unused, and can be
removed.

Also, responding to obviously unfounded user concerns (assuming that
they are 'obviously unfounded' which is not obvious to me) by
accusing the people who voiced them of 'paranoid ranting' (which is
code language for "I know something you don't know but I won't tell
you because if I did, I couldn't slam you as ignorant imbecile in
public anymore") is exactly the kind of 'welcoming empathy' which
generates trust towards a 'developer community': They do whatever they
want. If this hurst or discomforts someone, he's obvioulsy one of 'the
wrong kind of people' and free to take all his silly problems on a
long walk over a short plank.