[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: texinfo-6.4.91 pretest
From: |
Gavin Smith |
Subject: |
Re: texinfo-6.4.91 pretest |
Date: |
Mon, 28 Aug 2017 21:47:23 +0100 |
User-agent: |
Mutt/1.5.23 (2014-03-12) |
On Sun, Aug 27, 2017 at 10:32:28PM -0600, Assaf Gordon wrote:
> Hello Gavin,
>
>
> On Sun, Aug 27, 2017 at 11:58 AM, Gavin Smith <address@hidden> wrote:
> > https://alpha.gnu.org/gnu/texinfo/texinfo-6.4.91.tar.xz
>
> The following 5 tests fail on several systems:
> FAIL: t/c-u-m-x-scroll-forward.sh
> FAIL: t/inc-sea-forward.sh
> FAIL: t/inc-sea-forward-nonregex.sh
> FAIL: t/inc-sea-insensitive.sh
> FAIL: t/inc-sea-history.sh
>
> Eg. on CentOS 7, Fedora 25, Alpine Linux, Mac OS X.
> From cursory look it seems it doesn't fail on Debian/Ubuntu systems.
Thank you for testing this.
I tried these tests on a Fedora system I have access to, and reproduced
these failures.
My theory is that it is because there are characters being used in those
tests that have special meanings to the terminal, e.g. C-u for kill
line. I suspect that due to changes in timing, these bytes are being
sent to the Info program before the special meanings of these bytes are
turned off.
I found a bug in the 'pseudotty' program used in the Info tests that
meant that the code there to turn off the special interpretation might
not work. After this, all the tests passed.
I will look at the logs you've posted to see if there is anything else.