[RANT] Should we try to keep compatibility with old perl5s?

[Shlomi Fish]

Hi all!

This post is a little flamebait, so please try to keep the discussion civil.

Anyway, after reading the discussion in this public github issue, and following
some of the links (especially
https://szabgab.com/what-does--if-it-aint-broke-dont-fix-it--really-mean.html ),
do you think I was being unreasonable, or should I as a CPAN
author/maintainer/adopter accommodate for people running old perl5s, in my case
5.10.x and below:

https://github.com/shlomif/perl-XML-SemanticDiff/issues/3

This reminds me of what chromatic wrote here -
https://www.nntp.perl.org/group/perl.perl5.porters/2008/09/msg140206.html :

«
This is why we can't have nice things.
»

Any comments or opinions? I think I'll relax by watching a nice and fun video.

-- Shlomi

--
-----------------------------------------------------------------
Shlomi Fish http://www.shlomifish.org/
http://is.gd/htwEXQ - Integrating GNU Guile into GNU coreutils

Joel’s Generalisation: If it happens to you, it happens to everybody.
(Or: It’s never only you.)
— Based on http://www.joelonsoftware.com/news/20020402.html

Please reply to list if it's a mailing list post - http://shlom.in/reply .

[Arthur Corliss]

On Sat, 11 Aug 2018, Shlomi Fish wrote:

> Hi all!
>
> This post is a little flamebait, so please try to keep the discussion civil.
>
> Anyway, after reading the discussion in this public github issue, and following
> some of the links (especially
> https://szabgab.com/what-does--if-it-aint-broke-dont-fix-it--really-mean.html ),
> do you think I was being unreasonable, or should I as a CPAN
> author/maintainer/adopter accommodate for people running old perl5s, in my case
> 5.10.x and below:
>
> https://github.com/shlomif/perl-XML-SemanticDiff/issues/3
>
> This reminds me of what chromatic wrote here -
> https://www.nntp.perl.org/group/perl.perl5.porters/2008/09/msg140206.html :
>

> ?

> This is why we can't have nice things.

> ?

>
> Any comments or opinions? I think I'll relax by watching a nice and fun video.

I think this begs a question: how many developers are actually testing
with those older versions? From a purely pragmatic perspective I'd think
devs should only officially support down to revs they're actively
testing, but at the same time staying cognizant of the oldest perl revs
shipped as part of non-EOL'd Unices, etc.

Personally, I'm still supporting 5.008003, but I occasionally consider
whether newer syntactic sugars might be worth a jump. I have to admit I
quite supporting 5.006005 just because it was getting tedious having to
maintain my own patches just to compile and install it.

Ideally, whatever your choice, a dev shipping code for the benefit of the
community shouldn't be badgered for not wanting to take on the extra
maintenance efforts. At the same time, said dev shouldn't be surprised if
wider use of the same contributions are limited until the broader community
catches up.

Do what you want, dude. We might not all make the same decisions, but we
all get it.

--Arthur Corliss
Live Free or Die

[Linda A. Walsh]

Shlomi Fish wrote:
> Hi all!

> do you think I was being unreasonable, or should I as a CPAN
> author/maintainer/adopter accommodate for people running old perl5s, in my case 5.10.x and below:
>
>

Going backwards it becomes more difficult to support an old version,
along with
older versions having fewer users.

I tested some of my modules back to 5.8. I didn't test any on 5.6, as
it was
just too hard to build. If there was a clamor to support some of my modules
on 5.6, I might ask the reporter to do testing and try to see if I can
fix something.

Of course this implies only using a 'core perl' that worked on 5.6.x to
develop
extensions in -- another thing that might cause problems.

Another issue -- how many years ago was the release?

If it is old enough, it might be argued that it is hard to support some
things
on SW so old, it may have scores of unpatched bugs.

Really though, I can't imagine that *one* answer would work for every
situation.

OTOH, if the user wants to add support for the old version, you might be
willing to merge the changes into your source?

Just thinking...

my 4 cents (used to be 2 -- inflation hit).

-l

[Andy Lester]

On Aug 17, 2018, at 4:41 AM, Shlomi Fish <shl...@shlomifish.org> wrote:

I see - in that case I suppose they can afford to either pay me, say 100 USD
per dist per year, so I'll support older versions of perl there - and for
everyone, or alternatively do the backporting work themselves. And in both
cases, stop complaining.

I suggest that trying to optimize away complaining is a losing battle.  You will never make everyone happy.  Don’t optimize for whining.

[Shlomi Fish]

Good advice, but see the first item of
https://www.joelonsoftware.com/2007/02/19/seven-steps-to-remarkable-customer-service/
and the story at the bottom of http://linuxmafia.com/faq/Devtools/ides.html .

--
-----------------------------------------------------------------
Shlomi Fish http://www.shlomifish.org/

http://www.shlomifish.org/humour/bits/Can-I-SCO-Now/ - “Can I SCO Now?”

What is is. Perceive It. Integrate it. Act on it. Idealize it.
— Leonard Peikoff

[Rocco Caputo]

On Aug 17, 2018, at 10:26, Shlomi Fish <shl...@shlomifish.org> wrote:
>
> On Fri, 17 Aug 2018 08:19:26 -0500
> Andy Lester <an...@petdance.com> wrote:
>>
>> I suggest that trying to optimize away complaining is a losing battle. You
>> will never make everyone happy.
>

> Good advice, but see the first item of
> https://www.joelonsoftware.com/2007/02/19/seven-steps-to-remarkable-customer-service/
> and the story at the bottom of http://linuxmafia.com/faq/Devtools/ides.html .

"But when you think a little harder you can usually find a deeper solution: a way to prevent this particular problem from ever happening again."

I agree that complaining is a significant problem. I look forward to hearing your final solution for it.

--
Rocco Caputo <rca...@pobox.com>

[Shlomi Fish]

Hi Rocco,

by fixing bugs/quirks/etc you can prevent further complaints about the same
issues. But people like to complain, including to a large extent I as well. See
https://shlomif.livejournal.com/39215.html (BTW, I recall finding an LKML post
where Torvalds did say this, and it was a reply to a complaint about him using
https://en.wikipedia.org/wiki/BitKeeper before its gratis version was
discontinued, which in turn prompted creating git and other VCSes, but he
wasn't the first to say the quote.)

> --
> Rocco Caputo <rca...@pobox.com>

--
-----------------------------------------------------------------
Shlomi Fish http://www.shlomifish.org/

Understand what Open Source is - http://shlom.in/oss-fs

Chuck Norris can become root on OpenBSD. By using nothing but /bin/echo.
(By: ZadYree and Shlomi Fish)
— http://www.shlomifish.org/humour/bits/facts/Chuck-Norris/

[Shlomi Fish]

Hi Linda!

On Sun, 12 Aug 2018 04:46:24 -0700
"Linda A. Walsh" <l...@tlinx.org> wrote:

> Shlomi Fish wrote:
> > Hi all!
> > do you think I was being unreasonable, or should I as a CPAN
> > author/maintainer/adopter accommodate for people running old perl5s, in my
> > case 5.10.x and below:
> >
> >
> Going backwards it becomes more difficult to support an old version,
> along with
> older versions having fewer users.
>
> I tested some of my modules back to 5.8. I didn't test any on 5.6, as
> it was
> just too hard to build. If there was a clamor to support some of my modules
> on 5.6, I might ask the reporter to do testing and try to see if I can
> fix something.
>

5.6 has become painful to build indeed.

> Of course this implies only using a 'core perl' that worked on 5.6.x to
> develop
> extensions in -- another thing that might cause problems.
>
> Another issue -- how many years ago was the release?
>

see http://perldoc.perl.org/perlhist.html .

> If it is old enough, it might be argued that it is hard to support some
> things
> on SW so old, it may have scores of unpatched bugs.
>

These bugs often don't get in the way of running the module and the tests.

> Really though, I can't imagine that *one* answer would work for every
> situation.
>

yes.

> OTOH, if the user wants to add support for the old version, you might be
> willing to merge the changes into your source?
>

see my replies on https://github.com/shlomif/perl-XML-SemanticDiff/issues/3 .

> Just thinking...
>
> my 4 cents (used to be 2 -- inflation hit).
>

heh, Israel stopped minting 1 and 5 agoroth (see
https://en.wikipedia.org/wiki/Israeli_new_shekel ) coins because they costed
too much to produce.

> -l

--
-----------------------------------------------------------------
Shlomi Fish http://www.shlomifish.org/

What Makes Software Apps High Quality - http://shlom.in/sw-quality

I often wonder why I hang out with so many people who are so pedantic. And
then I remember — because they are so pedantic.
— an Israeli Perl Monger