Michael Kleehammer
Michael Kleehammer
I suspected moving the `connect` function would not be popular, but there is something else to think about. I've been wondering whether pyodbc 5 should provide a factory object instead...
Hmmm... Even if it is a bug, it would be easy to work around *with* a comment indicating why we did it even though the specification doesn't require it. (Otherwise...
@jstastny-cz This looks clean, but I'm wary about adding something to the interface without knowing what someone would do with it. I'm not sure how one would make use of...
@keitherskine Is this something we can get from Travis-CI integration? They let us test in different environments - can we get artifacts from those tests?
I forgot about cibuildwheel. Great idea.
The difference is that in 4.0.17 we used `SQLGetDiagRec` but for Unicode error message support, which is a critical item, we switched to `SQLGetDiagRecW` (notice the 'W' on the end...
Driver bugs, or sometimes purposefully ignoring the spec, have been a constant over pyodbc's lifetime and I've tried to work around them where possible. For example, the text-encoding settings are...
I would very much like them to be consistent. I'm just not sure what the right answer is. Is the DB always in UTC and we should therefore convert it?...
I saw the PR - great job. I left a question on one part. This may be the first item for v5 *(#945).
The only two fixes I can think of are: Convert to 00:00. Technically 24:00 *is* 00:00, just on the next day. This could be surprising if there is an associated...