Bump minimum Python version to 3.5.4

[rodrigc]

Scope and purpose

From https://twistedmatrix.com/pipermail/twisted-python/2021-February/065465.html , suggested by @richvdh

As it happens, it also fails (for an import of typing.Deque) on Python 
3.5.3, as used by Debian oldstable.

Craig, what is your intention here? I think it's ok to drop support for 
these ancient versions of Python 3.5, but please could you make sure 
that python_requires gets set to 3.5.10 if that's what's been tested 
against?

The minimum version we are testing against in CI is 3.5.4 on Azure.

So in the interest of getting this release out, I will bump the minimum version to 3.5.4

Contributor Checklist:

• The associated ticket in Trac is here: https://twistedmatrix.com/trac/ticket/10099
• I ran tox -e lint to format my patch to meet the Twisted Coding Standard
• I have created a newsfragment in src/twisted/newsfragments/ (see: News files)
• I have updated the automated tests.
• I have submitted the associated Trac ticket for review by adding the word review to the keywords field in Trac, and putting a link to this PR in the comment; it shows up in https://twisted.reviews/ now.

[richvdh]

the change looks good to me.

It looks like the CI is unhappy about something, but I can't imagine how this would cause a failure on python 3.7, so perhaps it is unrelated.

[adiroiban]

Unfortunately there is still a communication gap between Twisted dev.

I am flooded with GitHub notifications ... so I hang on #twisted-dev to catch up with Twisted development.

I have reviewed and approved #1524 and it's now merged.
After that I saw this PR.

For now, I think that the final release note should include. Python 3.5.4 is the minimum supported version of Python.

I guess that this will be the the last release to support python 3.5... so there will be less pain in the future.

[rodrigc]

@adiroiban , it seems like you and @graingert don't pay attention to communication done on the Twisted mailing list.
That's unfortunate.

@glyph @exarkun

[adiroiban]

I agree.

I did the review as Thomas pinged me over IRC and ask to review it as it was a blocker.

I did the review without checking the mailing list or the other reviews.
I have only checked GitHub notifications, and #1525 was not in the GitHub notification list.
Only later I checked Twisted PR lists and saw it and added a comment.

My understanding is that #1524 was created before this PR... and in #1524 I haven't see any reference to this ticket.

My suggestion is that instead of using the review tag, to also use the GitHub PR review request mechanism.
This is how I do reviews for klein, treq, pydoctor, ldaptor

I don't know how to coordinate between Americas and EMEA timezones... without having peoples on IRC.

I understand that this incident is just reducing more the collaboration and trust between me and Craig.
I have not ignored this PR on purpose and there was malice.
I am sorry for this incident.

But I think that we can look at the good part and see that this issue is not fixed.

Regards

[rodrigc]

@adiroiban can you answer the question that I CC:'d you on here:

https://twistedmatrix.com/pipermail/twisted-python/2021-February/065460.html

[richvdh]

I think this can be closed as a duplicate of #1524.

[rodrigc]

Duplicate of #1524