On 2020/03/21 13:20, Neil Bowers wrote:
A good approach is to reply on RT, and include your patch in RT.
The first thing I did was search for issues there. When I saw that
it hadn't been responded to or acknowledged in 6 years, I didn't think
adding my 'me too' case (even with a trivial fix) would be likely to
be seen, thus my post here.
Iâm pretty sure the person who most recently released it has a github repo, so you could also consider going the extra mile and doing a pull request. But probably best be patient, as like most of the world, he probably has other concerns right now.
you could also consider going the extra mile and doing a pull request.
---
I thought by creating a patch against the current version on span as
well
as a new tarball w/the fix was doing that. Sorry.
But probably best be patient, as like most of the world, he probably has other concerns right now.
Well, it's fixed on my machine since I installed the patched version. So time
is not critical. It turns out the module I was trying to use is even more
broken, so fixing this, while useful didn't really enable me to look
at its dependent. Sigh.
Maybe if he has done nothing in a week or two, email him a nudge?
I'm not understanding. If the bug has already been filed and not fixed
or answered for 6 years, why should one expect ...um...
..reading following responses...
Oh. It's you.
So having this fix might increase the chances of you to
responding to the bug.
As I said in the CHANGES log, since we aren't testing division here,
using strings that test as identical should be sufficient for this test.
I'm not entirely sure what all the test was to do, and found some
logical, but questionable behavior, in-as-much as when it shows what
it found, the message shows quotes stripped off, but not in the
"expected text", i.e.
0.667 doesn't compare as equal to
'0.667' in the test. Is that correct? In the test source, I initially
had '0.667' (with a few extra 6's) in both places and it didn't compare
because the source location was dequoted on being read into memory.
That's why I used the "q(0.667)" double quoting, but it seemed a
bit clunky if users are expected to do or need double-quoting (this
feels like a shell script though I'm sure it would be worse).
Anyway, I just needed the test to pass comparing the two
floats - as strings, but having to jump through extra quoting steps
felt a bit less than str8-forward.
I'll copy this other issue into the rt tick.
Cheers,Neil
Ditto ... :-)
