README contents from the repo I've just created:

tumblr-pics-downloader

It will download all the jpg, jpeg, png, gif files to your destination folder.
It's very fast because of parallelization.

The downloader is based on the David Shaw's Tumblr_scrape.
Check his code out and compare to mine to see what the changes are.

Differences from the original version:

• speed. I didn't measure it, but using 10 threads makes it approximately 8 times faster. Further increasing of threads didn't bring anything, but you can test it yourself.
• number of supported picture formats is also extended from |jpg and png| to |jpg, jpeg, png and gif|. One can add more formats easily by changing regexp in 2 places.
• code refactored a bit (if it became better or not is for you to judge)

Example of usage:

go to the command prompt and type something like this:

ruby tumblr_pics_downloader.rb http://best-of-tumblr.tumblr.com

Some of the things I am still going to change include:

• add other websites so that you can get pictures not only from tumblr, but let's say, also from mmm-tasty.ru. If only I were a regexp pro it would be already done :)
• use trollop instead of parsing arguments manually (for such a simple script maybe there's no need, just wanna play with it a bit)
• save text if necessary (still thinking if I will ever need it)

If you like the idea and want me to further improve the code let me know by pressing the button.

Any suggestions are welcome
Especially critic from ruby programmers
(how else I am going to learn without your feedback?)