The code for the issues tracker was updated to be python 2/3 
bilingual. Also the client_nonce was added to multiple templates
for javascript code. However they were never deployed to the host.

Need to do this. Current files to update:

    website/issues/detectors/messagesummary.py
    website/issues/detectors/newissuecopy.py
    website/issues/detectors/nosyreaction.py
    website/issues/detectors/no_texthtml.py
    website/issues/detectors/patches.py
    website/issues/detectors/severityauditor.py
    website/issues/detectors/statusauditor.py
    website/issues/detectors/userauditor.py
    website/issues/extensions/local_replace.py
    website/issues/extensions/spambayes.py
    website/issues/extensions/timestamp.py

    website/issues/html/_generic.calendar.html
    website/issues/html/_generic.help.html
    website/issues/html/_generic.help-list.html
    website/issues/html/_generic.help-submit.html
    website/issues/html/user.help.html
    website/issues/html/user.help-search.html
    website/issues/html/user.item.html

also need to update the config.ini in mercurial from the current
active one. Also need to generate updated active config.ini using
roundup-admin updateconfig.

To restart and activate .py changes run:

 sudo systemctl status roundup-roundup

deployed tonight also included rss feed detector from issue2551085.

These changes also include drag and drop of files and paste of images.

spambayes failed on restart. no xmlprc_ in roundup.anypy.
Reverted that set of imports and reference to xmlrpc_ changed to use
xmlrpclib as in 1.6.1 release. xmlrpc_ is a 2.0.0 python3 thingy.

other detectors/extensions and html files deployed without
incident.

Changed config.ini to generate html4 rather than xhtml code.

Server restarted and seems tobe running fine. Setting to pending.

Couple of additional fixes to the rss generator. Hardcoded +0000 timezone 
offset for timestamps. Also changed from author to dc:creator from
dublin core since author requires email address and we don't want to 
display those.

Autodiscovery still doesn't work for feedbro but it looks correct AFAICT.
rss feed validates correctly.

Autodiscovery must be an issue with feedbro, feeder finds it fine
given the url of an issue. Closing.