Alex Pooley's Blog

Overnight, a couple of generous readers left six new SpyFu promo codes in comments on my previous SpyFu post.

I know from my logs that people are interested in these codes. Here’s the six posted last night, and I’ve also added my “private” SpyFu discount code to the list otherwise it will go to waste.

These codes expire in about 10 days (9 days now I guess).

• Invite Code 1: SNKY-AWXZ1
• Invite Code 2: MKWV-JLWU2
• Invite Code 3: E9EW-GSQ53
• Invite Code 1: RVRY-SQID1
• Invite Code 2: HEYK-38BH2
• Invite Code 3: LFSS-NLXA3

and my private code *drum roll* …

• 1EFF-QIE50

JPA is a new standardized Java persistence system that takes the best parts of the current persistence technologies. Recently while using JPA, I had noticed that large transactions took increasingly longer to process. It turned out that by default, JPA cross checks a pending query with all managed entities to ensure that the query is up to date. To disable this default behaviour and shoulder the object management yourself, you can do the following:


EntityManagerFactory emf = Persistence.createEntityManagerFactory("YOUR_PU");
EntityManager em = emf.createEntityManager();
em.setFlushMode(FlushModeType.COMMIT); // disable auto-management


A concern among many SEO and SEM experts is that search engines may penalize duplicate content (web pages ripped from PLR, open publishing sites). After learning about the internals of search engine algorithms I suspect dup content is not an issue for any search engine.

There are three generic information retrieval techniques: boolean, probabilistic, and vector space models. Vector space models appear to be the current hot topic, specifically with latent semantic indexing (LSI) leading the charge. With vector space models you project the data in to a set of co-ordinates and then use the co-ordinates as a metric to compare the data against itself. Duplicate content is projected to the same co-ordinate. In general, similar content is projected close to similar content, and away from dissimilar content.

Given that duplicate content will be located at the same co-ordinate, a search engine would simply continue to rank the content by other criteria such as page rank, domain authority, etc. So I guess if someone is ripping your content and they are considered more authoritative - you have a problem; but I suspect this would be a very unusual situation.

[Also, I had read on Matt Cutt’s blog that you should not be concerned about dup content though I can’t find the blog post.]