Front page | perl.perl5.porters |
Postings from August 2021
Paul, What you say is true in some cases, but for others I disagree with the idea that one can't comment without having first written code using Object::Pad. I feel I already understand the user experience well enough just from the Corinna spec and I don't need to use Object::Pad to understand it. Also using Object::Pad will take an order of magnitude or several more time than commenting hypothetically. I have already committed (per Twitter comment a few weeks ago) to use Object::Pad to build my next real Perl project in (open source), however it will take at minimum a few months if not longer before I can get around to starting that, so commenting from hypothetical only is a very effective way for me to contribute to Corinna betterment now when it is needed most. Also I agree that renaming Object::Pad to Corinna::Prototype per Elvin Asianov's suggestion should boost its visibility significantly. -- Darren Duncan On 2021-08-18 6:54 a.m., Paul "LeoNerd" Evans wrote: > On Wed, 18 Aug 2021 14:30:39 +0100 > "Paul \"LeoNerd\" Evans" <leonerd@leonerd.org.uk> wrote: > >> Also, both of these points are made by reference to myself having >> actually written large amounts of real actual code using Object::Pad. >> [1]. > > Furthermore on this note, I am getting somewhat frustrated in a lot of > these Corinna design discussions, when it feels like everyone else > involved is discussing from a purely hypothetical perspective without > having written any real code. > > Object::Pad exists. > > https://metacpan.org/pod/Object::Pad > > It is an experimental test-ground for all of these Corinna-based ideas. > I began writing it almost two years ago now, precisely for the reason > of giving us a good experimental playground to try out these Corinna > ideas for real in real code (i.e. not just a paper design exercise). > > I often get the impression that nobody else has even seen that it > exists, let alone actually tried writing any real code using it. > metacpan's reverse-depends seems to support this impression: > > https://metacpan.org/module/Object::Pad/requires?size=500 > > At time of writing, there are only three distributions on that list not > written by me. Of those, two of them are just benchmarking it and not > using the code for real. The only other one is written/released by > a $work client of mine - admittedly they're writing a lot more of > that one than I am on it. > > But overall this doesn't give me the impression that almost anyone else > is actually trying to use these things for real. > > If folks are going to get involved in debating the design of Corinna, > can you at the very least be familiar with what already exists and is > actually implemented (thus avoiding any comments on "oh that might be > hard to implement" and the like), and ideally, get a feel for what it > is like to actually use the system in practice. >Thread Previous | Thread Next