Development

Wesley Shields (wxs@) wrote in to point out that the commits for Mk/bsd.sites.mk where not showing up correctly. Not sure if you know about this or not but it appears that (at least) Mk/bsd.sites.mk is not being properly updated on freshports since August 5. I compared these two URLS: http://www.freshports.org/Mk/bsd.sites.mk http://dev.freshports.org/Mk/bsd.sites.mk Yep. The full commits

FreshPorts has a web page caching issue for all non-port commits. This also affects commits in the ports tree that do not touch a port. e.g. Mk, Tools, etc. I hope to fix it later this week. dev does not have this problem. beta & www do.

I noticed last night that the list of ‘Latest Vulnerabilities’ on dev and beta differ from that found on production. Specifically, prod lists entries for: php5 Jul 23 php52 Jul 23 php53 Jul 23 rubygem-activemodel* As previously noted, the new vuxml processing code is on the dev and beta servers, but not yet in production.

This is a partial list of things that had to change for svn. I’m basing this on my cvs logs and memory. ~/.procmailrc – make sure we process the incoming email from the new mailing list mailing list subscription – subscribe to the new commit messages adjust verifyport.pm to not fetch directories (which appear in

Can we produce a src only news feed? I think so. FreshSource, our sister website, runs off the same database instance as FreshPorts. It merely uses a different web interface. It has a new feed, but it covers all commits, not just src and ports. The SQL used to generate that news feed is: SELECT

I have identified a lingering SVN issue, relating to tag expansion when accessing the repository via the web interface. The VUXML page is not displaying the details of the last processed vuln.xml. It should show something like (this example from pre-SVN): Revision: 1.2752 Date: 2012/07/11 Time: 01:47:55 Committer: swills This problem is related to non-expanding

With the recent move of the FreeBSD ports tree to SVN from CVS, there are some subtle changes on FreshPorts. Early this morning, I noticed a commit which demonstrated a side effect that I had not anticipated. The issue isn’t obvious from the above link, but I noticed it when the commit failed to properly