Fwd: TeX - ! Extra alignment tab has been changed to \cr

[Carlos]

I provided the only test file in the whole thread that demonstrates
> the error message under discussion and the error seems entirely
> appropriate in this case..
>


The reason why I find the error message far from informative is because as
long as it prevents compiling the file successfully is non-informative. And
the reason as to why I find the implementation not to be correct, is sipmly
because even by having an informative file as you said earlier - then that
compilation is also read and compiled accordingly. You might call me
Stubborn, but what good does it do to read an informative message if the
implementation is incorrect. I'd rather read a non-informative message as
long as the compilation matches somewhow the outcome.

I mean (looking sideways now in astonishment), let's have a scenario in
where
you gradually start modifying macros and definitions around to break the
entire
TeX infrastructure on the planet with informative messages during
compilation (
I know it sounds far-fetched, but this is just an exapmle for crying out
loud),
and on the other hand, these informative messages are deemed quite
informative
and with good intention,  especially coming from someone like you who's been
involved with the project for such a long time. Then, shortly after these
modifications take effect, a few people start asking you what those
modifications you committed were all about, and you start telling them that
you'd like to educate the public at large about the nooks and crannies of
TeX
in general. I'm sure enough though, that even by taking into account that
you're a mathematician at heart, but so are a great many users who in turn,
will start asking you further questions about the changes, so your
informative
end goal is not going to be taking too serious for long, before your
faculties
that drove you to do such a thing are called into question in the process.
And
unless, unless everyone else follows course, and loses the faculties at the
same pace than you have
- an unlikely event though - your informative end goal is out the window.

This is akin - in a silly way though - to be stranded on a desert with no
goods
whatsoever and coming across a sign that reads Water Ahead. The message
reads
that there must be water somewhere. The message is thus an informative
message.
Then one keeps walking and walking and after so many kilometers there's
another
sign that reads pretty much along the same lines than the previous one. It
turns out that the information may have been correct all along, but the
distance to travel to get to where the water is at, wasn't specified, so by
the
time one gets anywhere close to where the water is, one is completely
drained
and dehydrated. Then the information is as irrelevant as not being able to
find
water in the first place, even though the message was nonetheless correct to
point out that the water was somewhere. So yes, most people will choose
water
over a sign, regardless how informative that sign is.


On Tue, Jul 9, 2019 at 2:08 AM David Carlisle <d.p.carlisle at gmail.com>
wrote:

> On Mon, 8 Jul 2019 at 22:12, Carlos <linguafalsa at gmail.com> wrote:
>
> > And you or Mittlebach or Karl for that matter could get a hold of Knuth
> and ask him
>
> I could not possibly do that after all these messages I still have no
> idea what you think is a bug. You have provided no test file that
> shows a problem nor said what behaviour you want to change, only
> stated that you find the existing error message confusing.
>
> I provided the only test file in the whole thread that demonstrates
> the error message under discussion and the error seems entirely
> appropriate in this case..
>
> If you want to make a bug report supply a test file
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://tug.org/pipermail/texhax/attachments/20190709/7a259b6d/attachment.html>