Version 3.2 (xx.2015)

have a config-home were logging/configs take place
more simplification (just one cache implementation, no db)
started writing tests - but now abandoned (I would need some non-copyright streams for this)
captcha support
removed/added animeloads
tried different gui systems: qt,tk,gtk - all to complicated
use gooey as gui system - it just starts from my argparse config
commandline:
- limit download speed
- resume downloads when no arguments
- interactive mode to select stream files
use yapsy as plugin manager
- could again remove some of my own code - yapsy seems ok
use urlresolver api for getting flash-download url
- this is a bigger change as it makes this tool a bit less fragile and I don't need to put so much time on developing these things anymore
unified config-setting/loading, commandline and gui (via gooey)

Version 3 (16.05.2014)

python3 compatibility
removed broken streams and not used library
some new streams and pages
static code analyze with landscape.io
test automation with travis-ci
simplification and cleanup in all parts of the code
removed many threads
renaming tools to flashget and having everything more like other python projects
the config file is now ini format and can be in the home directory
have a setup.py for installing

stream_extract: implement putlocker
improved cache interface
added new caches (Leveldb, HypterTable, KyotoCacheComp (=Kyotocabinet with compression turned on))
urlmanager will look if it received binary data and won't cache it
http library improvements (header parsing, error handling)
nicely close the program and join all threads
own extension manager
extension manager for pages and streams
direct download of streams via commandline
streams are objects now too
support https

after so long time another version increment is needed change in version numbering cause of simplicity

allow disabling of curses interface
improved http library with many bugfixes
(re-)implement animeloads, eliteanimes, movie-loads, kinox.to
extracting all data from sites and store it into a mysql database
improved extract-interface
allow multithreaded extract-all through specifying start and amount
replaced my own cache implementation with kyotocabinet cause filesystem got too slow
dropped idea of client-server style program

again no particular reason for this, i just think this version is stable (:

generic xvid implementation + some supported pages for it
my http-implementation
- bugfix: correctly get the port from a link
- bugfix: keywords (like content-length) in http-protcol aren't casesensitive
- using a dns-cache
added crypt-it container.. but currently they don't realy work.. you just can display the links inside a container
first idea + some code about making this program in a client-server style to support later more guis (but currently doesn't work)
many other smaller things, just use git cherry -v 7e46d271

there is no particular reason why exactly this revision got version 1, it's just that i can add a version with the commandline-option

cause this is the first version, i will mostly list features here:

tools for easy adding new commandline options
own http-class which implements keep-alive and which is (compared to other python http-libs) very small
classes to handle the webpages, where flashfiles can be found. supported pages:
- anime-loads, animekiwi, anime-junkies, youtube, kino.to
classes to handle the flv-streams, which can extract a direct-download link. supported streams:
- veoh, eatlime, megavideo, hdweb, sevenload, youtube, imeem, hdshare, zeec, plain (direct dl-link is inside html)
class for logging + config-params to filter some logging stuff not very advanced yet
class for creating a gui in ncurses not very advanced yet + has some bugs with the scrolling in those logs but has already some nice features and is, i think, quite fast