April 23, 2003

Attention RSS Aggregator Developers: User Choice Is A Good Thing

Dave says "Aggregators should not organize news by where items came from, just present the new stuff in reverse chronologic order."

I say, "Sure, but what about letting me choose?"

I prefer to use Newzcrawler in a subject oriented manner, and a source oriented manner depending on what my need-of-the-moment is. Chronological orientation is useful as well and one of the things I miss about the Radio aggregator.

Each orientation has its merits. Subject matter orientation enables my research, source orientation gives me a better sense of "relative quality that can be expected based on past experience" and chronological is a good way to spend a lazy Sunday morning getting caught up.

Problem is, most software developers are intent on telling me what the right orientation is instead of providing me with the flexibility to do what I want.

Posted by ross at 11:17 PM | Comments (1) | TrackBack

Googlediving: How to Visit Google's Last Result

Bret asks how one can navigate to the last Google listing for any particular search.

The answer is simple, but not automated. Hacking the URL is the only way to do it.

First, you have to get yourself a hackable URL. Do a search for whatever you want, and scroll down to the bottom and click on any of the result pages in the Google results navigation (you know, the one that looks like this "GOOOOOOOOOOOOOOOOOOOOGLE"). The resultant page has a slightly different URL than the page you were just on - a URL that you can mess with fairly easily.

In this slightly different URL, you should find a key that says "start=x" where X is a positive whole number. This number will always be less than the number of non-similar entries on a given subject and (I would guess) be equally divisible by the number of entries per page that you have chosen to display - the default is ten.

Changing this number in your browser address bar will allow you to "quickly" jump to specific results pages. For instance, changing it to 100 in a result set of 200 entries, will get you halfway into the results...(yes, obviously ;)

The trick lies in guessing where the last page is. Bret noted that there were 950,000+ entries for ICANN, so presumably changing this number to 949,990 would do the trick, but natch - that just brings up an empty result set.

I decided to see if this was just a big number problem and dumped 999 in. Again, nothing. Scaling back to 500 got me something again - so I split the difference and went up to 750. No results again. I dropped back to 700, got results and then decided to increment by 10 until nothing.

In this case, the magic number is 720 - there are 727 unique entries in the GoogleDB for "ICANN" - the last of which being the Webopedia "Who's Who" bio's that start with "C" - includes Vint Cerf, ICANN's current chairman.

Personally, I can't wait to see what name the community gives to this particular activity. I vote for "Googlediving" - it needs a better last page.

Posted by ross at 07:13 AM | Comments (0) | TrackBack