semicolons in dig output

[Matthew Pounsett]

I had a question from a one of our developers today about semicolons in TXT
records.  It seems he and someone he's working with were getting different
output from dig when querying for TXT records that contained semi-colons.
One person saw them escaped, the other didn't.  It caused some confusion
because they weren't sure what was actually in the data.

As far as I can tell, it looks like dig's behaviour changed somewhere
between 9.8 and 9.9.  In later versions semicolons in quoted text aren't
escaped.

I see several posts from the archives about people being surprised by the
escaping, but nothing that indicates anyone at ISC thought it was a bug.  I
also haven't (yet?) found any release notes about the change in behaviour.

Was this actually a change between BIND 9.8 and 9.9?  Was it deliberate, or
an accident that might be reversed at some point?

Thanks!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.isc.org/pipermail/bind-users/attachments/20161104/c6e134be/attachment.html>