1.0.29

• add quarterly branches
• Wrap lines too long for the screen in the pre-formatted commit messages. #80
• Better dynamics for the logo #85

Welcome to IPv6

Indicate requests from IPv6 clients by including a logo

While here, bring back the bug.

sidebar for backends page; timestamp for vuxml

Provide a sidebar for the backends page.

Fixes #69

When linking to ports, strip any suffixes such as @py27 or :configure

This allows the link to work, but the text will retain the original value.

Better opt-in / opt-out

Make the page useful when the user is not logged in.

When this page was originally constructed, there was no external link to it.
Hence, it would display "why are you here?" if you were not logged in.

Let's change that.

Don't pull back more data than you need.

This change improved the query from 30.2s to 2.3s.

Display links to sanity test failures from the sanity test failure page

I mucked up on the previous release. fixing it here.

Show link to sanity test failure when displaying multiple

Start caching sanity test failures.

With all the fallout from the introduction of flavours, we have many more sanity test failure,
some with very long lists.

We never cached STF before, but now we do.

Since STF are neither ports nor files, but they relate only to commits, I created a new cache
classe in classes/cache-general.php

While here, classes/cache-port.php & classes/cache-file.php were slightly altered to use a
class constant, much like classes/cache-general.php

Code related to packages was removed from www/sanity_test_failures.php - I suspect
it was never used and was left over from a copy/paste situation.

I attempted to avoid file manipulation by using filter_var() in www/sanity_test_failures.php
when accepting incoming message_id values.

I see now that I am caching all results, not just successful reads from the DB. Could
someone fill up the cache via that? Yes, I'm sure they could.

Cache sanity test failures

Start caching sanity test failures.

With all the fallout from the introduction of flavours, we have many more sanity test failure,
some with very long lists.

We never cached STF before, but now we do.

Since STF are neither ports nor files, but they relate only to commits, I created a new cache
classe in classes/cache-general.php

While here, classes/cache-port.php & classes/cache-file.php were slightly altered to use a
class constant, much like classes/cache-general.php

Code related to packages was removed from www/sanity_test_failures.php - I suspect
it was never used and was left over from a copy/paste situation.

I attempted to avoid file manipulation by using filter_var() in www/sanity_test_failures.php
when accepting incoming message_id values.

I see now that I am caching all results, not just successful reads from the DB. Could
someone fill up the cache via that? Yes, I'm sure they could.

Don't always show pkg-plist

1.0.20

we do not need to include rewrite/missing-port.php here.