# Knowledge base¶

## A single per-tag Feed¶

'/my/feed': {'view': 'feed', 'if': lambda e: 'whatever' in e.tags}


To have a feed for each tag, use the newish Atom/RSS per tag view (a configuration example is in your conf.py).

## Performance Tweaks¶

Markdown instead of reStructuredText as markup language might be faster. Using a native Markdown compiler such as Discount is even faster. Another important factor is the typography-filter (disabled by default) which consumes about 40% of the whole compilation process. If you don’t care about web typography, disable this feature gives you a huge performance boost.

A short list from slow filters (slowest to less slower):

1. Typography
2. Acronyms
3. reStructuredText
4. Hyphenation

Though reStructuredText is not that slow, it takes about 300 ms just to initialize on import. Typography as well as Acronyms and Hyphenation are limited by their underlying library, namely HTMLParser and re.