Marc, This is certainly disheartening, after all the help and advice you have given me with EN 2.2. I haven't bothered with this beta yet; after seeing that it appears Evernote has completely changed directions and is becoming primarily a Web-based, "Shared-content" type application, I immediately decided that I would just stick with EN 2.2 for as long as possible. Jim Jim, Let us keep perspective, please. This is, after all, early Beta code. We do not yet have anything even approaching the (eventual) version of EN 3 that is to fully replace EN 2.2. My method for "Beta Testing" EverNote code, as I have had the privilege of doing since the early EN 1.* code, is to simply USE it hard, in everyday use, with the full expectation that all the documented features work and that all the features are documented. I pound the living s**t out of EverNote, every day. When I get some new Beta code, I make a careful backup of my working EN databases and then load the new code. I understand that if I run into a total disaster, I may lose a few days worth of notes. So be it (such is the life of a Beta tester). I should also talk about what I do not do when I test Beta code. I have a long history of architecting, designing, writing, testing, and applying code. In particular, I have worked on "Tiger Teams", teams of people whose goal in life is to break code (and not to be gentle or polite about it). One of the things that formal software testers do is to map out a long list of features and capabilities, and then design a set of tests to stress and break those features and capabilities. I don't do that with EverNote Beta code. Instead, when I am beating on EverNote Beta code, I simply USE EverNote. I use a broad range of EverNote's features, with special stress on what I call "The Essence of EverNote": self-assigning categories; a hierarchical category tree; and the category intersection panel. I also bring notes into EverNote by nearly all the documented methods, and I output notes in nearly every possible fashion (including exporting partial or whole databases for consumption by other EverNote users). On an average day, I add at least 25 new clippings to EverNote, and frequently as many as 100. I also regularly prune (delete) notes that I decide I do not need or that I think are superseded by other notes. So I churn my databases quite a bit. I also edit at least 100 notes per day, usually more. On any given day (I've counted) I send out at least several dozen EverNote notes as e-mail messages, and frequently as many as 100. I get requests for information all the time, and frequently I have some related information sitting in one or more of my EverNote databases. Just about weekly I get requests for someone for information about topic XYZ, and then we start corresponding and it turns out that they also are seeking information about topics ABC and PDQ. So I send out even more notes. All this traffic usually results in people requesting "related" information, so I fire up the EN Classic (2.2) Category Intersection Panel and go looking to see whether I actually have any such related data. Frequently I discover that I do, so I start building many new Automatic Keyword Categories to tag that data in ways I had not thought of previously. I start building new category trees to hold the new categories, indexed into the data (notes, mainly web clippings) to organize what I have. Then I fire up some focused Google searches on the web and on my hard drive to see what else I can find. That probably sounds pretty abstract, so let me give you a real-life concrete example from the past few weeks. I was asked to provide whatever material I had on fMRI brain scans to a psycho-therapist (PhD) in Minneapolis, my old stomping ground. I have known this therapist since I was a teenager, and she is a close friend of my family. My mother (also a PhD psycho-therapist) was having lunch with her friend, and pulled out a color printout of an EverNote note I had sent her. The EverNote printout had pictures of fMRI brain scans, showing which region of the brain light up when used in a specific fashion. The friend grabbed the printout of the EverNote note out of mother's hand and demanded to know where to get such information. Within a few hours the Supreme Mother In Chief (I bet I'm not the only one in the world who has a Supreme Mother In Chief ) had tasked me to pass along whatever material I have on fMRI brain scans and behavior. I had several hundred EverNote notes on the subject (with text, graphics, source URLs, and related links). I began dispatching EverNote notes via e-mail. Within a few weeks the therapist began passing along additional requests for related information. One request was for the brain region which controls the sucking reflex in human infants. I had no idea whether I had anything about that, so I started running searches on my EverNote databases and probing the Category Intersection Panel. I built a few new Automatic Keyword Categories to see what I had and what I could pass along. I ran Google Desktop Search on both my multiple EverNote databases and on the files on my hard drive. Within a few hours I had dug up the material she was looking for. The requests for information and relationships kept pouring in. It turns out that she works with Minneapolis and St. Paul parents who have adopted orphans from Russian orphanages. There is absolutely no documentation of what happened to these children before their adoption and transport to the US. Some of these kids (from toddlers to teenagers) exhibit bizarre developmental disorders. So the therapist is looking for information on brain scans showing whether these kids have physical brain damage (trauma, neglect, abuse, or whatever) or whether they have properly functioning neurology and only have to be treated behaviorally. I was able to pull out an article talking about the specific fMRI results for children who had been neglected or abused. There are good fMRI brain scanning resources at the University of Minnesota (my Alma Mater), but the neurologists are not necessarily trained in behavior or childhood developmental disorders and don't necessarily know what to look for. I think you see where I am leading. EverNote Classic (EN 2.2) is without peer for this sort of a retroactive data mining and fishing expedition. I can find information in my own extensive EN databases, and I can mine my databases for information and relationships that I did not even know I had (the "Essence of EverNote"). It usually takes me just a few hours to find what I have, find what I did not know I had, and to start tracing out through the URLs in the materials I have to find the material that I did not have. All of the newly found material goes into my EN Classic (EN 2.2) databases. As I go along I usually just export the necessary material into a new, special-purpose EN database, and then work with that database. Sorry for the long real-life "How I Use EverNote 2.2 to Help Make the World a Better Place" story, let us now get back to the discussion of beta testing EN Lite (EN 3 beta). If I could, I would just move my working EN Classic databases into EN3b and continue doing what I do (using EverNote hard every day, all day). But I can't. Not won't, can't. EN Lite has database size limitations (one of my EN Classic databases is up over 500MB these days). EN Lite only has 32 (or so) "Saved Searches", whereas I need hundreds in any one database, even in my highly focused special-purpose databases. EN Classic lets me build a stand-alone special-purpose database, and then see only the category tree for that one database. EN Lite puts all my notes into a single database, and does not (yet) let me look just at one notebook's notes and only at the tags associated with that one notebook. I have security concerns about EN's web servers, as some of my databases contain privileged information. So, EN Lite has neat new multi-platform capabilities (which I think are great) and neat new capture tools on my SmartPhone. EN3b is, in my opinion, a technology demonstrator to show what the new EN 3 can do. I think that is great, particularly given all the platforms the EN Development Team is addressing. They need lots of help to beat on the EN software on all the different platforms, and to test how the different platforms play together. I am happy to assist with that aspect of the EN3b Beta Test effort (and I will continue to do so). But what I REALLY want to do is to slam my EN Classic databases, with 10,000+ notes and with 1,000+ self-assigning categories (EN Classic's Automatic Keyword Categories, EN Lite's Saved Searches) and hammer away in the EN Lite code. But I can't. So I'm frustrated. I want to take the shiny new Formula One race car with the green paint job and the elephant logo decals out for a qualifying lap on the Formula One race track, but all I can use right now is a Toyota Prius with bald tires. The Prius gets great gas milage and can run on many different kinds of race tracks, but it can't yet do what I want to do: open the throttle and start turning laps at 200+ mph. So I am chomping at the bit for the far more fully-capable version of EN 3,the full EN 2.2 replacement, so I can finally stop calling it EverNote Lite and start calling it EverNote.