Linux: execute a command periodically and watch the result

I am aware of cron and it does a great job when you want it to do something for you weekly, daily or with whatever frequency. But you have to configure it and stop it, and so on and so on. But what if you just want to quickly see something to start happening or the dynamics of something that is already happening and then just move on (for example new files start appearing in the folder). I needed this functionality because a model, that I use, stores the last executed time step in a text file and updates it after each time step (updates, so 'tail -f fname' does not help). What I was doing is executing 'cat fname' million times and frequently until I have found out about the 'watch -n 1 cat fname' (http://en.wikipedia.org/wiki/Watch_(Unix)), which does it for me every second and nicely updates the result. So I decided to write it here for my (and for anyone's) reference.

A python script: example for downloading GlobSnow data (snow extent in this case)

Recently a friend asked me if it is possible to selectively download folders from http servers in linux.. Probably it is possible by using wget or rsync, but I have never succeeded to make them work exactly the way I needed. So I wrote a small script for the task and passed it to him hoping that this might be his first step to learning Python. And below are 2 versions of the same script:

1. The version I have actually given to my friend,
2. The improved a bit more scary-looking version, which is a bit closer to the way I think it should be written.

First, below I show a quick and simple way of downloading files, with a minimal account for possible errors. The next step is to check if the sizes of existing files are the same as the sizes of the remote files and to replace the bad files, if required. Of course, it takes some time to download the data, especially if you need several years. If you work remotely I would suggest using tmux or screen, so your program would continue running even if the ssh session is closed for some reason. But if those are not installed, you still can get away by using nohup as follows:
nohup pyhton download.py >& log.txt &

Cheers and any comments are welcome