Dropping 5.5.x

[Gabor Szabo]

On Thu, Mar 12, 2009 at 2:17 AM, Michael G Schwern <schwern at pobox.com> wrote:
> Nicholas Clark wrote:
>> Partly because I would like to provide some pressure on the world out there
>> "stuck" on 5.6.whatever that they need to seriously think about moving on.
>> They don't have to *do* anything. They just need to start thinking that
>> they are getting left out in the cold, and whether they consider frostbite,
>> or death by exposure, to be what they want happening in their near future.
>> (Oh, and that if they do nothing, they don't have a distant future full stop)
>
> +1
>
> Part of the reason I dropped 5.5 was I was working in a 5.6 shop.  Their code
> needed some work to upgrade to 5.8.  They were using 5.6 style Unicode.  They
> had some code which depended on hash order.  A few other little things.  All
> the developers knew this, but management never authorized the time to do it.
> Why?  There was no "business reason" and the slow dribble of efficiency lost
> from working with 5.6 was not quantifiable.
>
> What sort of thing were they looking for?  When it was considered to port the
> app to Japanese the improved Unicode support in 5.8 would have been enough,
> but that never happened.
>
> By dropping support for 5.5 I gave every developer stuck using 5.5 a reason to
> upgrade.  They could tell their managers that their version of Perl is
> unsupported and they have to upgrade.  Every time they tried to install a CPAN
> module, it didn't work and they have to spend time patching it, that adds to
> their argument.
>
> We can do the same for 5.6.
>
> (Unsurprisingly, they weren't heavy CPAN users and most of what they had was
> wildly out of date)
>

Here is what I have been trying to say for a while but I think your description
went in a better way.

I think people see the risk and expense of the upgrade only but not that of
staying behind. Partially it is so because you guys spend your valuable time
in supporting ancient versions of perl, partially as they are not heavy CPAN
users anyway so they don't feel the need for advanced features.

AFAIK the Perl community or p5p never claims it is "supporting" a version
of perl but occasionally there is a discussion about stopping to "support" one.

This is due to different levels of "support". The former refers to paid support
the latter to the meaning "it works".

Users - as in companies - either want commercial support and then they go to
ActiveState or maybe some other company who sells them this support but mostly
say that "there is no support" - meaning paid, commercial support - for Perl.

It would be nice if we could give good reasons why to upgrade or if we
could show
at least some of the cost and risk in staying behind.
Declaring on various versions of Perl that it is "not supported any more"
could be a good way.

Gabor