Reini Urban

• About: Working at cPanel on cperl, B::C (the perl-compiler), parrot, B::Generate, cygwin perl and more guts, keeping the system alive.

Recent Actions

• Posted Overview of current maintainer fails to Reini Urban

  schmorp recently complained about a maintainer fail in Math::BigFloat, and since this is not an isolated incident of the last years, I try to give a summary of all the maintainer fails…

• Posted Storable security problems and overlarge data to Reini Urban

  Some people might be aware that we (not p5p) fixed Storable already for the most egregious security problems with Storable, with public CVE, explanation and metasploit module, but those bugs are still in Storable as on CPAN and in the latest perl5.24.

  Now a new problem arose on the horizon…

• Commented on The removal of the lexical topic feature in 5.24
  RGS: You are right, that the adoption rate of my $_ and the _ prototype was a problem. I would explain that with the experimental burden. Nobody is using any experimental p5p feature at all (besides say), esp. since all...
• Commented on The removal of the lexical topic feature in 5.24
  Flavio Poletti: You are right. The CoC sidebar was only written to explain the background why this removal suddenly went through. Without 2nd deprecation. With those new CoC rules technical criticism on wrong decisions is not possible anymore, they are...
• Posted The removal of the lexical topic feature in 5.24 to Reini Urban

  Now that p5p removed the broken lexical topic feature in 5.24 it's happened, and how cperl fixed that problem.

  In 2013 Rafael Garcia-Suarez posted to p5p a plea to ="http://www.nntp.perl.org/group/perl.perl5.porters/2013…

• Posted YAML patches done to Reini Urban

  I've finished most of the needed YAML work, esp. for YAML::XS.

  libyaml upstream has now a patch with the new options NonStrict and IndentlessMap https://github.com/yaml/libyaml/pull/8

  YAM…

• Commented on On YAML and YAML::XS inconsistencies
  And with the latest commit in https://github.com/rurban/yaml-libyaml-pm/commits/cperl-core YAML::XS is back again in the modern ages. Add native DumpFile, LoadFile support filename in error messages. Add dumper options: Indent, BestWidth, Canonical, Unicode, Encoding, LineBreak, OpenEnded (kept defaults) Add loader option: NonStrict,...
• Commented on On YAML and YAML::XS inconsistencies
  Fixed the first part $YAML::XS::IndentlessMap with https://github.com/rurban/yaml-libyaml-pm/commit/7f1af737489f11cdd9f7139fe989c0f4d4fbc004 Now only $YAML::XS::NonStrict is needed to pass the validator tests, and YAML.pm needs to accept YAML::XS IndentlessMap....
• Posted More on YAML, syck looks much better to Reini Urban

  In my last YAML post I said libsyck is not maintained anymore.…

• Commented on On YAML and YAML::XS inconsistencies
  toml is great for your private config files, but we need to solve YAML mess for perl5 core. Given that libyaml is broken and YAML.pm is broken, we are only left with YAML::Syck and YAML::Tiny (CPAN::Meta::YAML). Next blog post about...
• Posted On YAML and YAML::XS inconsistencies to Reini Urban

  Personally I'd prefer YAML over JSON for local config data anytime. YAML is readable and writable. It's better than .ini, .json and .xml.

  …
• Commented on Fixed 5.22 problems during my compiler port
  You should all have not to deal with p5p anymore, as they are not able to maintain their product, are making the product worse and worse, are not able to provide basic new features which are outlined since 2002, are...
• Posted Memory savings with -fcow to Reini Urban

  B::C has now better support for copy-on-write (COW) strings with about 6%

  The perl5.18 implementation for COW strings is totally broken as it uses the COW REFCNT field within the string. You cannot ever come to a true successful copy-on-write COW scheme. Y…

• Posted Memory savings with cperl and AvSTATIC to Reini Urban

  B::C and cperl has now proper support for copy-on-grow (COG) and

  COG means that the array of SV* pointers is allocated by the compiler statically, not dynamically, and that the cperl runtime creates a new array whenever the array is extended (copy-on-grow).

  …

• Commented on Perl 5 Porters Mailing List Summary: December 8th-13th
  Do not despair about RJBS's wrong comment that no one is working on signature "aliases" (i.e. call by ref), as with $_[n]: I not only implemented all those features they were discussing recently, which is the discussion following my RFC...
• Commented on Fixed 5.22 problems during my compiler port
  After waiting 12 years and seeing parrot going down with the same people problems convinced me that you need to call the bad apples the bad apples. Revoke their commit access immediately, as their work is very harmful. Drive them...
• Posted Fixed 5.22 problems during my compiler port to Reini Urban

  I uncovered and fixed many 5.22 problems with was busy to port the 3 compilers B::C, B::CC and B::Bytecode to 5.22.

  As I said in my interview it's my belief t…

• Commented on cperl-5.22.1 released
  Versioning is close to perl5 proper. Since I added so many optimizations on top of 5.23.0 I went for 5.22.1c. There are no API breakages. The next stable releases will be 5.22.2c (with windows support and we will see what...
• Posted cperl-5.22.1 released to Reini Urban

  https://github.com/perl11/cperl/releases/tag/cperl-5.22.1

  The name cperl stands for a perl with classes, types, compiler support, or just a company-friendly perl, but currently it's on…

• Posted gdb-dashboard to Reini Urban

  https://github.com/cyrus-and/gdb-dashboard

  wget https://raw.githubusercontent.com/cyrus-and/gdb-dashboard/master/.gdbinit -O .gdb-dashboard

  sed -i 's,python Dashboard.start(),#python Dashboard.start(),' .gdb-dashboard

  joe .gdbinit
  source .gdb-dashboard
  pytho…

• Posted How I made my core testsuite 2x faster to Reini Urban

  There is more than one way to make perl5 twice as fast, but this is what I did today. I fixed it on one machine.

  My Macbook Air gives constantly better results in my hash function benchmarks than my big Linux Desktop PC, because it ha…

• Commented on [BREAKING] [GPW 2015] German train drivers' union calls 5-day strike starting Tuesday
  The other option, bus, is by far cheaper (e.g. 7 Euro from Berlin, 22 from Munich) and faster, so it's not that dramatic. See https://shop.flixbus.com/...
• Commented on On Dave Mitchell calling B and B::C a failed experiment
  This is not a personal attack. It is a defense against public attacks against your core infrastructure. I'm naming the two bad apples here, because they are the only ones who started the personal attacks so far, which led to...
• Posted On Dave Mitchell calling B and B::C a failed experiment to Reini Urban

  While being blocked from p5p I had to read a new outragious statement by a porter. Who is in reality not a porter like me, just some developer who happens to have no idea what he is talking about.

  Re: OP_SIG…

• Commented on On OP_SIGNATURE
  Having now implemented all my plans with the old slow signatures, I have to take back some of the criticsm above. OP_SIGNATURE is in fact a really good idea. I'm still not too keen about the separation of "Too many|few...
• Posted On OP_SIGNATURE to Reini Urban

  Since it is not possible to write p5p criticism to the mailing list, I'll have to do it in my blog. @p5p: think over your guidelines. I don't believe that stuff like that needs to be blogged.

  DaveM now introduced a new OP_SIGNATURE which assigns run-time args according to the compiled sign…

• Commented on Analysing CPAN Testers' Reports
  45 minutes via webscraping? My script gets the reports in 2 minutes, with some yaml caching, and files it under t/reports/$version https://github.com/rurban/perl-compiler/blob/master/t/download-reports...
• Commented on Today’s bit of black perl
  That's know forever. Changing @_ elements directly is destructive. shift @_ creates a copy. With methods destroying $self (i.e. $_[0]) is of course evil....
• Posted A little warning to EUMM and shell-script users to Reini Urban

  I sometimes need to write shell-script test scripts and not perl, to be able to test perl scripts, without interfering with perl, and also for performance and easier IO reasons.

  In order to find out with which perl this distro was built, we need to parse the generated Makefile.

  Rec…

• Posted The sad story of pseudohash criticism to Reini Urban

  I just had to endure MJD’s horrible pseudohash explanation at the Pittsburgh Workshop.

Comment Threads

• dly commented on Fixed 5.22 problems during my compiler port

  I agree with you having incompetent coworkers/whatever-fellow-open-source-developers-are-called is a big problem. As someone on the outside looking in, this situation really bugs me. Your work on cperl seems pretty awesome, but cperl would be even more awesome if it didn't exist, and instead you actually cooperated with p5p to make perl itself more awesome. Just as you've chosen to point out their incompetence; instead, you could choose to bury the hatchet and cooperate with them. Replace personal attacks with
  

  p5p has spoken: by banning you repeatedly for…

• Yary commented on Fixed 5.22 problems during my compiler port

  ++dly Thanks for the good description of mentoring vs. badgering. Alas, transferring skills is another layer of talent and effort on top of employing those skills. At the moment this split seems healthy for all, though I share your hope that these projects can eventually successfully combine forces.

• Marek commented on On YAML and YAML::XS inconsistencies

  Is this improvement available on CPAN, too?

• Matt S Trout (mst) commented on The removal of the lexical topic feature in 5.24

  The _ prototype failed to be adopted long before the experimental pragma arrived on the scene, much though I guess that doesn't fit the desired narrative.

  I find it fascinating that in spite of reini's assertions to me on Twitter earlier that he always focuses on technical aspects first, once again the actual technical questions in this thread have been entirely ignored by reini in favour of further long form personal attacks. One might postulate that this tendency to insult and ignore attempts to engage in technical argument on the merits is a more parsimonious explanation of his ex…

• Aaron Crane commented on The removal of the lexical topic feature in 5.24

  Reini,

  You have been told before that personal attacks are not welcome here. I am no longer willing to provide a platform for you to publish them, and as of today, I am therefore disabling your user account here. This is a permanent action; it will be in place for as long as I am directly involved with blogs.perl.org.

  To be clear, this is not censorship: there are plenty of other places where people can blog without relying on blogs.perl.org. In addition, I will, if you choose, provide a 301 redirect from your existing posts here to another site. If you wish to take me up on …