Welcome the channel on the development of Cro, a set of libraries for building reactive distributed systems, lovingly crafted to take advantage of all the Raku Programming Language has to offer (cro.services). This channel is being logged for historical purposes.
Set by lizmat on 24 May 2021.
08:19 patrickb joined
rba The virtual server with most of the *.raku.org/*.perl6.org websites will be down for maintenance today 8pm CEST for approximately an hour... 13:57
Geth cro-http/master: 4 commits pushed by (Jonathan Worthington)++ 15:43
cro-webapp/templates-from-resources: a5c54bf441 | (Jonathan Worthington)++ | 6 files
Refactor template library test

Use a META6.json to handle the lib/resources more "officially". We'll want to write a similar such test for the new template location handling that includes resources, so moving this under a subdirectory for consistency.
cro-webapp/templates-from-resources: 5856b58210 | (Jonathan Worthington)++ | 7 files
Implement templates from resources

This builds upon the recent router plugins and `resources-from` features in Cro::HTTP to also provide a way to serve templates from a module's resources. Furthermore, it makes `template-location` inside of a `route` block only apply to the routes into that block and those included into it. This is a potentially breaking change, however also an increase in consistency. A `template-location` outside of any `route` block still works, and will function as adding the location to a set of global template locations.
cro-webapp: jnthn++ created pull request #41:
Support templates from resources, refine template-location semantics
jnthnwrthngtn There was also a release of IO::Socket::Async::SSL today that addressed some occasional failures in the Cro::HTTP tests (of note, t/http2-client.t). 15:50
16:13 patrickb left
japhb When you make a correctness fix in a dependency, do you set minimum version requirements on that dependency? If not, why not? (I'm curious because I've started to do that in my own modules, but it's not clear to me if that is a net win for all users; for example, what if a version bump included a fix for one bug but created another? Should I force my users to prioritize the bug that caused problems for me?) 16:59
jnthnwrthngtn I haven't tended to so far, though github.com/croservices/release/issues/2 means we'll do it across the core Cro modules from the next release 17:00
I'm tempted to do the same with Cro::TLS against IO::Socket::Async::SSL too, given there's quite a few fixes there recently that make things more stable 17:02
Altreus japhb: I tend to be of the philosophy of worse the devil you know 17:07
A known fix is a bigger win than an unknown breakage is a loss, on the theory that everything is reasonably well-tested and therefore egregious issues are not going to factor in 17:08
japhb That's a decent point, especially if a release is a bugfix-only release (meaning that there are no features slipped in at the same time). 17:09
Kindof argues for our community in general getting better about semantic versioning, so that it's highly likely to be safe to bump a dependency if only the patch level changed. 17:10
Many of us (me included) are still treating sub-1.0 release numbering as "Does this change seem likely to break things or not?" as opposed to "Strictly separating functionality versions from bugfix-only versions" 17:11
But it might be worth shifting that towards the latter, recognizing that most of our entire ecosystem claims pre-1.0 versions. 17:12
Altreus There's always the modulo of "Is the author scared of committing to 1.0 or is it actually in flux?"
DBIx::Class for example
japhb Yeah, true.
Altreus (Assuming perl5 examples work here)
japhb I started with "Larry languages" back in the Perl 4 days, so sure. :-) 17:13
Altreus > DBIx-Class-0.082842
japhb Oh sheesh, I hadn't realized that there's TWO leading zeroes there. 17:14
Altreus This was already a mainstay of large Perl projects when I started my first job in 2008
So I'm usually happy to pretend the first 0 is a 1
japhb Yeah, agreed.
Altreus But yeah the general idea of semantic versioning sits well with me, even if it doesn't literally mean x.y.z 17:15
19:08 hankache joined 20:26 hankache left
Geth cro-webapp/templates-from-resources: a85c7465fe | (Jonathan Worthington)++ | .github/workflows/ci.yml
Be sure to test against Cro::HTTP HEAD
japhb Oh, now 23:16
*that's* a really useful idiom
Didn't even know zef could install directly from a github zip 23:18