Hi Tiago,

I’m the product dev manager for Connotea. I hear what you are saying about the speed. The reason the service is slow is probably twofold. We calculate privacy each time a request is made, and we don’t do caching that well at the moment.

We have an extra box that we are trying to fit in to help speed things up and we are looking at the speed issue as well. We basically have a three step plan for solving the problem:

1. Add more hardware now

2. Do some back-end work to improve responsiveness

3. Experiment with moving to Ec2.

I’m hoping that 1 will be done in the next couple of weeks. The other two fixes will take longer as we are working on some other stuff first.

I know that it sucks at the moment, but it will get better.

Your point about the DOI import/export is also important. You are right and it should just work. We are re-desiging the db tables in order to fix buggotea. One of the things about conntoea is that an entry is stored once in the db and bookmarks are mirrors of this entry. We have been having issues where the post item is defined by the uri, and there are multiple instances of the same paper represented in the DB, each having a different uri. If someone creates a post item, anyone bookmarking the same entity will inherit the meta-data accosicated with the same post. We can go through the db and re-collect citation information if we have a better parser for getting this data. When we roll out the fix to the duplicate paper multiple uri issue then we will do this as a matter of course, and I will make sure that we fix the problem that you have been having with your doi-format not getting sufficient citation information for you. This should happen within the next two weeks, hopefully, or soon after that.

Short answer: sorry… you’re listed now!

Long answer:

http://www.ghastlyfop.com/blog/2007/10/connotea-postgenomic-complaints.html