Dagfinn Ilmari Mannsåker
Dagfinn Ilmari Mannsåker
@demerphq The `Compress-Raw-{Zlib,Bzip2}` changes need to be released to CPAN and imported to blead before this can be merged.
A simpler example: ``` $ mkdir foo; touch foo/bær; $ perl -MMojo::File=path -E 'my $dir = "foo"; for $upg (0,1) { if ($upg) { print "upgraded: "; utf8::upgrade($dir) } else...
This also breaks when the trace exporter is configured via the environment, rather than using arguments to the `NodeSDK` constructor, because the `propagator` option to the [`traceProvider.register()` call](https://github.com/open-telemetry/opentelemetry-js/blob/main/experimental/packages/opentelemetry-sdk-node/src/sdk.ts#L249-L255) comes from...
> > I thought we could use this in public headers because C++? > > Ugh; it's `mg_vtable.h`, indeed. Does that still count? Even if we can't use it there,...
> The `DOINIT` code doesn't appear usable outside of the executable/libperl.so since the magic functions are marked hidden. Looking more closely at things, `DOINIT` is only defined in `INTERN.h`, which...
I just had a thought (and pushed a commit to do it, for discussion): should we respect `pg_bool_tf` and bind native booleans as `t`/`f` instead of `1`/`0` when it's set?
The [script that generates the JSON files](https://github.com/endoflife-date/endoflife.date/blob/master/_plugins/create-json-files.rb#L61-L66) replaces slashes with dashes in the output filenames, so the correct file name for `releng/14.0` is https://endoflife.date/api/freebsd/releng-14.0.json This should be documented in [the...
I just noticed that the github actions only test the latest minor of each major, and `hv_fetchs` was backported to 5.8.9. I've tested manually that this (and [an equivalent change](https://github.com/bucardo/dbdpg/commit/a59d964c011332d8fce943f682c04179f5801e22)...
> @iabyn > > > Looking at this proposal compared to my earlier 2019 signatures proposal (http://nntp.perl.org/group/perl.perl5.porters/256679): > > My proposal allowed _optional_ positional parameters as long as all named...
Instead of repeating that logic at every callsite, why not move it into `SvPV_shrink_to_cur` or a new wrapper macro/function?