06-22-26
Version 4.0.0
Replace deprecated array('u') decoded sequence buffers with plain Python
character lists. This keeps decoding pure Python, removes Python 3.14
deprecation warnings, and changes the return type of longs_to_char_array.

Make TWOBYTE_TABLE lazy to avoid import-time memory overhead, and use bisect
when applying N-blocks to slices.

04-17-12
Some modifications to cmdline_reader and download module so that they can be run
as functions too. added more flexible twobit_reader which can write to files too

03-01-12
Fixed super() syntax to remove warning

02-29-12
Added ability to run as a script, e.g.
python -m twobitreader example.2bit < example.bed

Added download module to fetch genomes from UCSC

Bugfixes: max/min in get_slice (thought that was fixed)
Updated version to 2.0
Downgraded status to Beta

02-10-2012
Minor bugfixes
Updated version to 1.05
