On another note:
I have had a few people comment about converting new site to www.unix.com versus community.unix.com, so let me remind everyone:
This site has over 200K man pages and none of the man pages migrate to the new forum (anytime soon, we may write a plugin for this sometime in the future, however).
So, this site has a lot of reference material which is very important, including man pages and code.
In some instance, the migration of legacy posts with code from here to discourse has results in instances of "code mangling" where, for example, the original vB4 to Discourse migration script changed new lines chars "\n" to hard line breaks in code fragements.
Our moderator (and Poster of the Year 2019 as well), Scrutinizer is working on perfecting this Ruby script and when he is done; we will test this on a staging site and then update all the posts on the new site.
Even when the new site reaches "perfection" (or 99.9..., whatever that means), this site has a tremendous amount of important reference material (man pages, original code, etc).
So, eventually we will turn this site into a reference site and keep the reference site available. However, this "conversion to a reference site" is going to happen slower than we originally thought because of some parsing errors we found (we can all thank Scrutinizer for this, looking carefully at code) in the original Discourse Ruby vbulletin.rb script.
Frankly, after working 14 days in a row, 12 hours a day, on this migration, I was disappointed to learn of the errors in the original vbulletin.rb script; but I'm glad we found them and we all owe a debt of gratitude to Scrutinizer for working on modifying the preprocessing routine of the original Discourse vbulletin.rb migration script.
Thank you Scrutinizer!!
Regarding updating the posts:
After we have fully tested the new preprocessing routine on a staging site, we will update the posts on the discourse site. Discourse keeps both the raw (preprocessed) and cooked (post-processed) text for each post in the DB. We will update the raw posts and recook them when it is time to do the rebaking.
Because of this hard line break issue for the "\n" char in code fragments, Scrutinizer is looking at more enhancements to preprocessing at the same time.