So, for those of you who don't know, I uh... took first place in the AIRS Xtreme Challenge today with my Google Co-op.  I don't like to brag but I figured I'd post something on the subject.  Thanks for all the kind words from everyone and a big congratulations to all the runners up including Amybeth Hale, Chris Hayes, Terri Coligan and also a big round of applause to anyone else who took part in the competition! 

The one thing I will say is that I have my eyes on that SourceCon Challenge next and so you can all expect some fierce competition!

Happy Searching^^

So, first of all I would like to make it known that your truly has currently placed in the top 4 finalists in the AIRS Xtreme Games. Winners will be announced today at the AIRS free webinar which you can sign up for here if you haven't already! Also a Big Congratulations to my close friend and colleague, Amybeth Hale who is also in the top 4!! Oh, and Kudos too all the other applicants and finalists also, I'd give individuals kudos, but I don't know who you are...

Now in light of this I was asked to provide a working demonstration of my entry which as I'm sure you all have guessed is in fact the Google Co-op. Well, I was asked to provide this yesterday and with the Finals being announced today I made it a priority to clean up and present some of my latest development which brought some new problems to light.

As you all know I've spent a lot of my time lately focusing on the recruiting side of matters and less on the sourcing so my projects have slowed and I haven't made any real releases over the last month or so. It was during this time that the folks at Google released a bunch of new features to the co-op, made some changes to their own engine, and added some code to play with also. Well having to put together something for this event put me on the spot and I spent all night tackling what I thought was a glitch in my code, when in reality it's a bug that Google is not currently aware of.

When you place an excess of 50 key terms in background Google ignores words longer than 6 characters. Initially I thought it was an issue with quotations, however I was proven wrong through some heavy duty bug testing last night. When you put "software developer" into a CSE with an excess of 50 key terms embedded, you get results dictated by the embedded words only and it ignores "software developer". This is a big problem. So having re-written all my code and tested, tested, and performed more testing, I found that if the word is longer than 6 characters, it ignores everything. As a temporary solution I did find that if you put "-" hyphens between words instead of using quotes... the search returns correctly! So.... Software-Developer works just fine... pretty odd huh?

I'm going to do a file a full bug report with Google today, but I suspect it has something to do with their change in the wild card operator. In any event here I thought I was crazy but it turned out I am crazy and Google has a bug.

SO... What's this Beta thing you're wondering? Ok, so this is something I've been working on for a while and just got around to putting up a presentable model, but guess what... Google has another bug... Anywho, in collaboration with a very good friend of mine who is a much better developer than myself, we put together a CSE/API combo that includes clipping!

What's clipping? Clipping is the ability to create a separate table within the window your browsing in and copy links you want or don't want over to it for later use. How is this useful? Well, let's take a look and the CSE itself.  (LINK REMOVED)

So the first thing you want to is through in a search like this:

java more:nynj

Now what this beta engine does is perform the same search you'd run in myco-op just using the tabs to apply the refinements.  What you'll notice are the two boxes to the left of your results "keepers" and "for removal".  Clicking on either box will switch between those links you want to keep and those you want to get rid of.  Once you perform a search and find a site you like or dislike, you click on the link that says "copy" below the site of your choosing.  This will add the link to the "keepers" or "for removal" columns.  Now if it's a keeper, once completed, the title page will store in your cookies and come up whenever re-visited.  Until we've finish that, we've put a small icon in the top right corners of the boxes which looks like a small filing cabinet with a plus sign.  By clicking on that icon, a new page will open providing a list of the sites you've added.
If it's a site you don't like and it's going to the removal side, clicking the icon provides you a text based list of the URL which you've chosen to discard.  At this point you bring up the Google Marker and exclude in bulk adding to your excluded site annotations. I'm going to leave this beta engine up for display for a few days, but not more than that as it's still a work in progress and I'm not a huge fan of presenting incomplete work.  Also, Google has another bug within the AJAX API I believe as the "more results" area seems to recycle back to the default co-op page and not within the same location.  Could be me, could be Google, hopefully we'll get to the bottom of it soon.  

Until next time, Happy Searching!

Cheers^^


Edit:  I've removed the link to my CSE for the moment.  My apologies, I'll return it once I fix a MAJOR bug...  

Jim Stroud gets down and dirty for the answers everyone has been looking for! Check it out!

 

On this day (not sure how many years ago) the recruiting/sourcing world was blessed with he who would be called Jim Stroud. To which we now rejoice!