Hey, it’s timotimo again.
Last week, we decided to make building MoarVM with its JIT by default (thanks carlin!). It seems to be pretty stable, but now we’re going to get much more people trying it out.
Now it’s time to look through what else has been done recently:
- lizmat has worked on IO::Handle and IO::Path stuff a lot. It used to be possible to have an IO::Handle that corresponds to a folder or be unopened. That means every single method on the IO::Handle first had to check if it refers to a directory or if it needs to open the file first. That was a major time sink, especially for cases like a while loop that gets individual characters from a file. It’s hard to put into words how much work this has been!
- On top of that, lizmat also sprinkled lots and lots of optimizations all over the path/file related things.
- FROGGS tracked down a major piece of start-up time wastage: on windows, rakudo took about 0.75 seconds to construct $?USAGE, even if it went unused. The new code is also faster than the old one used to be in case you actually do want to see the $?USAGE.
- jnthn implemented the SUPERSEDE and DECLARE cases of EXPORTHOW.
- SUPERSEDE lets you override the behavior of declarators like “class” or “grammar” in the scope your module gets imported into to use different metaclasses
- DECLARE lets you create whole new declarators that act very much like the “class” or “grammar” declarators, like perhaps a “controller” declarator that adds a certain role to all things declared with that directive. (That example is what the initial tests in the spectest suite do)
- Tony-O added a super simple ORM to the ecosystem called “DB::ORM::Quicky”.
Does that list seem not too impressive to you? Don’t worry, things are going to be exciting: This weekend the APW2014 will take place in Austria and it will be followed by two days of hackathon where many Perl 6 developers are going to be on site. Topics that will be covered include The Great List Refactor and there will surely be discussion of what topics and features to move past the “6.0.0” milestone and what we’ll definitely spec/implement before calling that milestone complete.
I am going to be there, too. I will probably post a weekly that includes the things done at the hackathon and the things presented at the APW on tuesday rather than on monday, as monday is the last day of the hackathon.
In somewhat related news, the yapc eu videos from the Vitosha hall (which is where all the Perl 6 related talks were given) have still not been uploaded. Apparently the setup they used to record the talks is suboptimal and requires a lot of extra work or something like that? Sadly, I don’t have insider information for you 😦
Oh and by the way: thanks for all the positive feedback in last week’s comments! I didn’t expect so many people whose names I’ve never seen before to speak up in such a kind manner 🙂
Looking forward to reporting cool stuff next time 🙂