You may have noticed that appypi commands were almost the same as in aptitude , the Debian packages manager. It is not done unpurposely. Actually, you can find the same install, remove, update and upgrade in aptitude. The list command is not available in aptitude but in dpkg The show command is not available in aptitude but in apt-get.
Except for the launchers, everything appypi will create on your disk is confined to the .appypi directory.
For each app installed, any launchable script located in the package is translated by appypi as a launchable script. Any of these file is situated in the user bin directory. It creates it if it does not exist.
You don’t want to query the list of ALL Pypi packages at the installation of any package. That’s why appypi uses a cache of this Pypi list, to be fast at installing several packages in a short interval. This cache is written on disk in a single file, using Python’s pickle functionnalitie
Of course, Pypi packages list grows over time, and appypi should keep it up to date. It is done automatically if the list file is more than 7 days old. At the moment, there is no way of modifying this duration, but it seems reasonable. If this does not suit you, you can manually trigger the update with a simple:
$ appypi update
appypi maintains its own local cache for Pypi packages. It may be redundant with your own pip cache if you set it already (and you should), but this way appypi is kept independent from pip. Then, each single package comes either directly from Pypi or from the appypi local cache.
The annoying thing is that if pip find a package in the cache, it asks the user what to do with it:
The file /home/user/.appypi/appypi_cache/glances-1.5.1.tar.gz exists. (i)gnore, (w)ipe, (b)ackup
There is no way to tell pip what to do programmatically before its version 1.3. For previous versions of pip, you have to manually choose (i)gnore anytime it is asked. Shitty, I know...
In case it is too disturbing for you, you can delete the full cache with:
$ rm -rf ~/.appypi/appypi_cache
As you can imagine, entering in a virtualenv for each command is a bit time-consuming. Your package may therefore be take a little more time to start than usually. It is therefore not recommended to use appypi with particularly time-sensitive usage. For any other use of common packages, appypi is the way to go :-)