bug#23871: 25.1.50; Undo unexpectedly leads to blank buffer

[Phillip Lord]

Markus Triska <triska@metalevel.at> writes:

> phillip.lord@russet.org.uk (Phillip Lord) writes:
>
>> I'm amazed at how you spot all of these!
>
> Thanks to your extremely nice work on these issues, I am encouraged to
> file more of them. I am very glad that these issues are being corrected!
>
> I have found more issues with undo for which I don't have reproducible
> cases yet, and I can tell you more about them if you are interested.
> They may depend on the exact timing of process output and user input.
>
>> Can we start a new bug for this? We've been discussing about 4 different
>> bugs on this thread, and IIUC, Eli's fixed this one.
>
> I have filed #23903 for this, let us continue the discussion there.
> It's also great that you are including these snippets as actual test
> cases that are run automatically!


Yeah, it's not the best form of testing, but it works whether nothing
else does, and quite a few of your bugs have been subtle. While I like
these, I need to try and make some simpler more unit test like ones as
well. These functional tests are good for telling you something is
wrong, but not good for telling you what is wrong.

Please feel free to submit bugs in the same form.

Phil