-
-
Notifications
You must be signed in to change notification settings - Fork 372
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
say
optimization breaks Test::Output
#2751
Comments
The module declare an IO::Handle class and do not define the method say |
Can I allow myself a comment on the style? I find the two tags with urgency banners as aggressive, and the term obligatory as extremely rude. |
OK, let's see if we can fix that "wrong code" on the other side. |
Thank you JJ, I'm sorry but I am in a personal condition right now, and even this message is typed on a cellphone, otherwise I would have been glad to help |
@scovit no problem. Working on it right now. Best of luck. |
The exception is in this line:
Really have no idea why. |
Looks like I managed to fix it. A new release has been made. It does not seem to be a Rakudo error, unless you consider as such a LTA error message . Main problem was that it was not using WRITE for the IO::Handle that was defined, and then that said Handle didn't use UTF8 encoding. |
That's not the intent, sorry. When I was the release manager, my understanding was that we shouldn't be breaking modules on a whim. When I see a commit that makes things faster but doesn't have a written justification for breaking things, I file a bug report. Does that make sense? Maybe we should have a written document about it somewhere. Maybe on https://github.com/perl6/problem-solving. |
I don't remember the status of this why it still has a label. |
@AlexDaniel it doesn't seem like too much of a priority right now regardless of the status, so I'll remove the blocker label from it. |
Is this still blocking anything? |
Reading up on this issue and the linked one in Test::Output I'm stumped about why we've not yet reverted the broken change. Breaking modules for some optimizations in Rakudo is not OK unless those modules did something that was clearly wrong. The latter has not been discussed at all, neither here, nor in the other issue. |
$_ := $*OUT;
.print: nqp::concat(x.gist,.nl-out) So I'm unsure how 9770c1a can still be the issue here? |
Is the bug actually still present? Unfortunately neither bug report actually
mentions what was wrong. Many of the links don't work anymore either. That's
why it's so important to always include all the relevant information and e.g.
commit messages like "Fix GH issue 1234" are no use at all.
|
If there was a test failure, there isn't now:
With:
|
Actually, that was due to a workaround, mainly upgrading to 6.d. It's probably still present in 6.c |
It's probably still present in 6.c
So what is "it"?
|
Sorry, the bug. Answering to @jonathanstowe in that comment. |
Right, but I can't find anywhere that says explicitly what that was just that there was a test failure. Maybe I'm not looking hard enough, but a reproduction with the actual error might be useful to save us going round in circles. |
Trying to reproduce it here, lowering the version to 6.c, but no go. I would say it's fixed. Closing. |
Test::Output – Fail, Bisected: 9770c1a
Originally reported on IRC and on raku-community-modules/Test-Output#3.
The text was updated successfully, but these errors were encountered: