sisyphus

That build by niXman of 12.2.0 (x86_64-12.2.0-release-posix-seh-rt_v10-rev1) also eradicates the slowness I reported in #114. @brechtsanders, are you able to pinpoint the difference that accounts for this ? Cheers, Rob

I've built this -t-a-k:dont-convert-inaccurate-NV-to-IV branch under a few different config options and found it to fix the 2 test cases presented in this PR (and tested in t/op/numconvert.t.) Furthermore, I've...

The following example (from the OP) puzzled me: ``` >perl -le "sub foo { $_[0] + 1 } print foo(1e+18); print foo(1e+18); print foo(1e+18)" 1e+18 1000000000000000001 1000000000000000001 ``` I couldn't...

> any response to the idea of using SVp_IOK Sounds sane to me, and I'm happy enough to test any proposed implementation. I have quite a bit of code that...

Update: As of today (blead at commit 31e18734071), the slow test is still just as slow. I timed it to take 2 minutes 55 seconds. In the original post (above),...

> Do you have enough knowledge of Visual Studio to easily try its profiling feature No - sorry. BTW, in case I haven't already mentioned it, perl built using VS2022...

I think it would be polite to have the script output something at intervals, just to demonstrate that it hasn't hung. Or at least a (windows-specific ?) warning that "this...

``` >perl -wle "print join ' ', map sqrt, 199999982358225, 199999982358226;" 14142135 14142135 ``` Perl's print() function rounds the given values to 15 significant decimal digits - and at that...

It does indeed fix #21697. LGTM.

Does the proposed removal of the %setlocale_failed part of the condition weaken the tests on the non-problematic systems ? If so, the %setlocale_failed could instead be replaced with (%setlocale_failed &&...