Wed 11 Mar 2009
Thing 15: Rollyo
Posted by johnofjack and tagged 23 things, bad implementation, broken, broken things, good idea, rollyo, whatever, wth.
I rolled one. It didn’t work. It was intended to be an image searching roll, and at first it had four sites defined: flickr, altavista images, google images, and yahoo images.
The first time I tried it, it pulled results only from Flickr. I deleted Flickr from the URL list and tried again. Then it pulled results only from Altavista, but they weren’t photos. The text alongside the box for URL entry said that you should use only top level domains, which would be why the Altavista part didn’t pull photos (the URL was www.altavista.com/image/default rather than images.altavista.com ). So I deleted that URL and tried again (supposedly searching Yahoo Images and Google Images), and it told me there were no results.
The search term in each case was “Watchmen.” I am certain that both Google Images and Yahoo Images have results for that term, and indeed a quick trip to the sites confirms that they do (and quite a lot of them, for obvious reasons).
So then I decided the site was broken, and decided also that I should move on.
My roll (restored to three URLs, pulling results only from one, and therefore still broken) is here.
The site is a great idea, and it could be useful in a library setting–very useful, in fact–if only it worked.

March 20th, 2009 at 10:40 am
I was also frustrated by the way Rollyo behaved. Maybe someday it will get tweaked enough to be able to search sub-pages within a website. Since it’s basically free, I guess we’re getting what we paid for.
A.
March 31st, 2009 at 7:55 am
I agree. That top level domain thing limits what you can do. I wanted to add the YALSA blog to my searchroll, but Rollyo will only search wikis.ala.org, which includes all ALA wikis.