Jump to content

ambassadeur

Level 2
  • Posts

    16
  • Joined

  • Last visited

Everything posted by ambassadeur

  1. It is easy for me to imagine all sorts of wonderful solutions, since I don't have to spend time actually doing it. I'm sure Dave's ideas are just as good, but he has things to worry about as he guides this thing along its path, and everything takes time. And I am known for being very long-winded (sigh). When I left HP a couple of months ago, I moved from using MS Office 2003 to the 2007 edition. It has the "ribbons" along the top. I'm still not fully adjusted to it, occasionally failing to find the place where the function I need might be buried. But I nevertheless consider it to be an advance, at least for casual users, as it is very visual. Power users tend to like menu navigation sequences, keystrokes, or even commands. Power users can tag very quickly if tags are searched keystroke by keystroke in real time. For general users, I like the idea of having a 6-8-line-tall ribbon of checkboxed tags with +- symbols to expand & collapse tag hierarchies. It would extend to the right however far and scroll horizontally if it needs to. Keys and buttons to toggle it on or off, expand all, and collapse all. Hard hierarchies (like Colored, Colored.Red, and Colored.Taupe) would have the behavior that checking Colored.Taupe automatically causes Colored to be checked (and conversely, ...). That partial or soft relationship between tags (like my Order may or may not need an item type, but it ideally would have additional attributes/tags), and if the mechanism can expedite that, that would be great; it's sort of like required & optional "attributes" in an "element" in XML. This part of it is at the edge of my UX (user experience) imagination at the moment; I'm not quite sure how that might work. The main thing is that multitagging should be very quick and maximally assisted for both productivity and integrity. The thing I'm working on at the moment is to determine what checks to clear and what checks to check or leave checked after a note has been tagged and until the next one comes up. Perhaps one could define a filter (saved search?) that, when satisfied, causes certain tags to be set, among other things. (shrug) And whether these ideas are shared by others or by Dave, I have no idea at all. Please chime in. For me, I'd like to searches and the actions to take and such to all be scriptable -- like Ruby or Python (I'm a Ruby/JRuby/Groovy man myself, but whatever). My main model is that of research info capture and profiling/categorizing. But Evernote could be useful in many other ways as well. I'd like to use it as an RSS/Atom feed reader, as a blog editor/viewer, and so on and on. Each kind of use will demand certain UX features. I heavily use a tool called PersonalBrain (TheBrain.com). It is a very useful tool that can be leveraged in so many ways that it has a major identity crisis, in my opinion. Things they do for the calendar/to do list folks sometimes ***** up the research/knowledge folks like me. So far, not too badly, though. I think Evernote is a bit like that at the moment. By the way, Dave, I'd really like to see Evernote and PersonalBrain interfaced together. Both are pure Java and have APIs, so it should be readily doable, I would think. But both tools need to evolve a lot on their own, as well. -amb
  2. Like most folks, I've used the subfolder notion for years, in Unix & Windows file systems, in Outlook, and in many other places. More recently, I've come to appreciate tagging and have gone through multiple learning processes with them. Tagging is very popular in modern Web-based collaboration (Web 2.0 or whatever). I've used them with Outlook categories, Shadow Plan & SmartList for Palm OS, Google Mail (Gmail), Google Reader (Greader), Evernote, FireFox bookmarks, etc. Like most folks, over the years, I've evolved a directory tree structure I like for file systems, and I set that up when I get a new machine or a new drive. Also, I've noted similarities to the various bookmark, category (like Outlook), and tag sets that have evolved, and I am now in a mode of aligning them; soon, I will have them fully aligned and will become more rigorous as to how I assign tags to content. After all that, here's what I find: (1) In file systems, I find myself frequently wanting to use (Unix/Linux) symlinks or (Windows) shortcuts to implement exceptions within the otherwise hierarchical structure of file system content. Still, if the tree structure is good enough, the need for them is somewhat rare, but always present. (2) In tag sytems (my use of them is most mature in gmail, so I'll use that for examples), most of my messages are tagged twice or three times; examples include (each tuple are tags for one message): (2a) {order, books}, {order, gear}, {order, clothing}, {order, clothing, john}; (2b) {family}, {colleague}, {family, event}, {event, music}, {colleague, event, conference, IT}; (2c) {career}, {career, opportunity}, {career, education}; and (2d) {community, social}, {community, career}, {community, organization}. (3) Subfolders/subnotebooks by themselves is nowhere near sufficient. (4) Tags by themselves are powerful, especially once mastered. But they exact a price: (4a) It takes me a handful of clicks or keystrokes to get a message tagged in gmail. If I were less OCD, less nerdy, or less intent on having a searchable message store, I would've abandoned it long ago due to productivity loss. (4b) Getting hierarchical structuring done via tagging IS possible; but again, at a cost in productivity (and complexity). (4c) The creative mechanisms I put in place as I refine my tagging are great, but they are not policed. I mistag stuff every day, and I realize it sometimes immediately and sometimes much later. Regardless, I live with knowing there will always be exceptions due to mistagging; my lists of, say, orders are not complete, for example. (5) When developing code, I've learned -- after using SmallTalk & others purely hierarchical (single) inheritance methods, Java single inheritance with multiple interface inheritance, C++ with multiple inheritance, and so on and on -- that there is no "right" answer as to how best to model a problem, a solution, or the world. So.........., here is my proposal............... (1) Let there be subfolders, unless the developers REALLY consider this to be hard or bad in some other way. There's just going to be a level of comfort they will yield, and without them, there will be too many folks who will be put off by or tempted to stray from Evernote. (2) Tagging is really great. Let there be tags. Let there also be a forum specifically for tag structures or usage mechanisms to be used like templates to help users get the most out of them. Evernote can contribute as well as users to this compendium. (3) THE way (in my opinion) to enable productive tagging is to provide (pop-up?) CHECKLISTS of tags. If a list gets too big, SCROLL it (probably HORIZONTALLY, like Windows Explorer List View, as opposed to Details View). Tagging gmail-style is 'way too slow. (4) For rigorous tagging to be feasible for those of us who may NOT be OCD or sufficiently nerdy, let's get a smidgeon creative regarding tags. Mature tag sets tend to have at least partial relationships between tags. An example of mine above is that a message (note) representing an "order" should generally have a tag representing the kind of thing being ordered and possibly who it might be ordered for. Another is that a note of an "event" is pretty barren without also knowing what kind of event it is, and depending on that (like "music" vs "conference"), it might also be annotated with "IT" or "concert" or something. We could implement tag hierarchies this way, if the relationship is a full one (enforceable) rather than partial (user's option). (5) To accomplish #4, I suggest that power tag users have a Settings section wherein they can define relationships between tags. The effect of a relationship is to be prompted with a list of optional "sub" tags. Now, this is arguably icing on the cake; #4 is the main deal. BUT an additional benefit of #5 is that Evernote could then help the user achieve INTEGRITY by ensuring the relationships are maintained or at least not violated. I have spoken. :-) This topic shouldn't degenerate into the subnotebookers vs the taggers. Most of us probably use both right now. If someone gets frustrated and inadvertently snubs the other, let's relax and give each other a bit of a break. -bc
×
×
  • Create New...