vertiginous
improve pik update
The current implementation of pik update is poor.
I'd like to have it work similar to package, and install. Download a zip file, extract, and copy over the existing files. The last step will have to be done from a batch file to avoid stepping on a running process. This shouldn't be too hard to do, since it's how pik use works.
I'll need to figure out a place to host the files. I have a cloudfront cdn set up, but I'm not sure about cost. The other option is rubyforge, but I'll have to sort out how to push the files there, and I'd have to worry about slow replication times.
The final piece is going to be checking for updates. The simplest way would be to just have the files overwrite themselves without checking for updates, but this wastes bandwitdth. I could revive a simpler version of packages.yml, that only contains information about pik versions. If there is a version in that file newer than what is running then update.
