dmetalious 0 Posted February 8, 2010 Share Posted February 8, 2010 In response to my earlier suggestion for folders, Dave Engberg said: So we certainly appreciate the feedback, but I actually think that it's more important for us to come up with better and better ways to Search and Filter your notes, rather than coming up with more elaborate alternate schemes for you to manually organize notes in to buckets (in addition to the list of notebooks and unlimited hierarchy of up to 10,000 tags per account). Here, I will demonstrate in the simplest terms even a programmer can understand why this is so much nonsense. But first, let me make clear, I love tags. They are very useful. I use them frequently. But they are no replacement for hierarchical directories for one simple reason. Think of the outline for a research paper: Note the hierarchical structure of how the information is organized. We humans have been organizing our thoughts this way for a very long time. Why? Because that is how we think. Categorically. That is how we organize information - in groups, and sub groups and sub sub groups. That is how particles are classified in physics. That is how species are classified in biology. That is how notes are structured when writing a college dissertation - usually anyway. I know. I teach college. Cognitively, categorization is not only how we structure ideas, but how we remember them. Indeed, it is the categorization of knowledge and ideas that provide the handles with which to retrieve the memories of those ideas. And notes are nothing if they are not the augmentation of our own memory structures. So look at the Shakespeare outline. It's got categories(early life, works, later years) and sub-categories like Works>Plays>Tragedies,Comedies, Histories. This formatting encodes information, not just about the data itself, but about relationships to the other pieces of data. These categorical relationships are very important about how we think, understand and most essentially, how we remember. Now flatten it. I. INTRODUCTION - (Brief comment leading into subject matter - Thesis statement on Shakespeare) II. BODY - Shakespeare's Early Life, Marriage, Works, Later Years A. Early life in Stratford 1. Shakespeare's family a. Shakespeare's father b. Shakespeare's mother 2. Shakespeare's marriage a. Life of Anne Hathaway b. Reference in Shakespeare's Poems B. Shakespeare's works 1. Plays a. Tragedies i. Hamlet ii. Romeo and Juliet b. Comedies i. The Tempest ii. Much Ado About Nothing c. Histories i. King John ii. Richard III iii. Henry VIII 2. Sonnets 3. Other poems C. Shakespeare's Later Years 1. Last two plays 2. Retired to Stratford a. Death b. Burial i. Epitaph on his tombstone What you see is the loss of a great deal of information as a result. Information about the organizational structure of information that you cannot encode with tags. "Search and filter" cannot restore that lost information because it was never there in the first place. 2nd Law anyone? So, far from a "redundant feature" as Mr. Engberg asserts, I have demonstrated beyond any doubt that the ability to organize one's notes in a hierarchy is, in fact, quite often a fundamental requirement. Not for everyone. Not for every application. But quite often. Which begs the question, who is your market demographic? Certainly not my college students. I already have enough trouble with their flat, linear thinking. But what of a housewife who snips recipes off the internet. Guess what she wants to do - file them away in categories. Just the way ALL recipe books do. Check it out. Now finally, Mr Engberg makes the following assertion: In retrospect, it seems that simple "flat" organizational schemes are good for finding things in a small set of documents.Browsable organizational metaphors like hierarchies are ok for finding things when your set of documents gets to be medium-sized (e.g. a few thousand). And very large data sets (e.g. all of the notes you might take through your entire life) are better suited to a good Search system. What Mr Engberg fails to comprehend here is that hierarchies become more essential as the amount of data grows because keeping data sets small and manageable is precisely what hierarchies do! Instead of having to rummage through a thousand recipes to find the one for beef stroganoff, you go to beef>sauces. "But why not just do a simple search for beef stroganoff?" you are undoubtedly thinking. Here's why, because it wasn't beef stroganoff. It was something else I can't remember the name of. "What was it...? Oh, here it is, beef bourguignon." You see. This is what those of us who manage gigabytes of data deal with all the time. So there's a pretty big wrench in your theory Dave even if you ONLY see hierarchies in terms of memory retrieval. But as I demonstrated above, there is so much more to it than how to find information. Finally, we're only talking options here. If people want to use tags and search, they can. It will be interesting to see how that works for them. I suspect it will go the way of Google's Desktop which as far as I can tell, is a flop. Most people I know hate the fact that they can't store their Gmails in a directory format like Outlook. The reason why the folder metaphor is persistent is for exactly what I explained above: it emulates the way humans think. Search and tags will grow in prominence, but they will never replace putting your stuff in containers. It's just too easy. Link to comment
This topic is now archived and is closed to further replies.