develooper Front page | perl.perl5.porters | Postings from September 2014

Re: Roadmap/plan for Perl 5?

Thread Previous | Thread Next
sawyer x
September 2, 2014 22:44
Re: Roadmap/plan for Perl 5?
Message ID:
On Wed, Sep 3, 2014 at 12:18 AM, Leon Timmermans <> wrote:

> On Tue, Sep 2, 2014 at 5:42 PM, sawyer x <> wrote:
>> I know this is considered a naughty word in some places, but I am
>> wondering if there is a roadmap or a plan for the continued development of
>> the Perl 5 language.
>> I'm not trying to step on toes or to upset anyone. I'm also not trying to
>> usurp any position. I'm genuinely interested in any long-term plans that
>> exist for the language.
> I think that, due to the nature of the project, roadmaps are largely
> personal.

I don't see how Perl, as a language, is a project inherently different than
others that have roadmaps.

I have plans of where I want to take perl, I know others have their own
> (usually orthogonal) plans, but often timescales are not well-defined. And
> nothing is explicitly written down.

Timescales come later, if at all. Even the "by version X" would be too
restrictive to begin with. Having "this is where we're going" is in and of
itself invaluable.

> Ultimately, the direction of development of perl is made by those who do,
> not by those who talk.

This sounds vaguely like "talk is cheap". I'm not suggesting not doing. I'm
suggesting having clear goals. This doesn't prevent anyone from working,
nor does it convert actions to talk. In fact it allowed to reduce talk and
increase focused efforts.

> * Is there a list of features that had been decided to be added to core?
> There are some directions in which I suspect general consensus on it being
> a good idea. Them actually happening is dependent on someone standing up
> and getting it done.

That's good. Would those be part of some direction the project needs to go?
Is that sorted and listed anywhere?

>> * I know there are complaints about the tests. Is there a plan on how to
>> improve them?
> I'm not aware of any large issue in particular.

One glaring problem the TODO mentions is that B isn't covered, that there's
a coverage problem in general, and that it's still using the
custom-generated TAP. I also spoke to people and heard the problem with the
testing suite a random pile of test files.

These are 4 big issues already.

>>  * Are there any long-term changes desired? I know CoW went in some time
>> ago. Anything like that?
> It's not hard to come up with ideas. Implementing them? Well volunteered!

It's not just about ideas. Long-term changes hints of a direction. I'm not
asking whether you can come up with ideas. :)

>>  * Is the documentation where we want it? Is someone herding those
>> efforts?
> The documentation is of varying quality. I think only perlfaq is being
> herded currently.

Is anyone in charge of documentation in general?

>> * I remember Yves Orton mentioning his grief with the XS interface. Are
>> there plans for formalizing and solidifying how to handle some of the warts
>> he mentioned?
> Again, that only takes a person to write some patches

Again, this is not about the actual work. *What* patches? To *what* end?
*What* will they address? Is that agreed upon with others?

When I spoke to Yves, there were some big ideas floating around. It's more
than "just send the patches".

I think much of the bike-shedding is done because there is no clear
direction. If I send huge patches now to address this, I would still have
to go through "Why are you doing this? What are you trying to achieve?" -
Would it be good to know *before* patches are submitted? Wouldn't be nice
to discuss what such patches will do and how?

>>  * Do we know how we want Perl 5 to look like 5-10 years from now?
> I know where I want it to be, but I wouldn't expect consensus on all of
> that.

Hours of discussions can save weeks of work - as long as they don't lead to
bike-shedding and rage. :)

Thread Previous | Thread Next Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at | Group listing | About