Evernote 5 is here

Posted by 22111
Oct 11, 2013 at 12:56 PM

dan7000, I don’t have any knowledge about such big systems but I know there is a tendency towards management information systems in that scale, hence those “views”, and “big data” and its “aggregation” and half-automatted analysis. I know some not-so-big dms, and they all function by virtual folders (which often are not named this way) and clones, and some try to facilitate inter-personal workflow, from one staff to another/“the next” staff member, beyond shifting the item/object manually into the “ToDo” folder of that “next” man or woman; of course, access rights management is an integral part of those systems, which means of those I know, several try elaborate functionality in this respect, to the detriment of other aspects. There are big differences in clarity of functionality there; in Germany, there is ELO Office, originally by Leitz (the lever folders inventor (?)), very often installed, but from what I have seen, one of the worst programs when it comes to intuitivenesss and all.

All those database-based outliners are based on relational databases, so the technical storage of data is not that important, and “views”, if they are not augmented by some form of AI, are nothing more than virtual sub-headings in a tree, or clones to a virtual folder, perhaps half-automatted, by kinds of tag inheritance (thanks to wsp for telling me that in EN, this would have to be done manually) and similar procedures, as in Zoot.

As I said, in the end, almost no one in a big corporation looks at the “big tree” if there is any, and technically, in most such systems, it could be created, similar to AS, from the content of some record fields, and then ordered by the content of some other record field (for example record number); this would not even suppose similar record fields in all of the records, since the automatism could build up this “tree” by some alternative fields. So again, it’s not that different after all.

As said, the advantage of “trees” - in fact, to be precise, of just ordered LISTS, instead of simple tagging - lies in the fact that if I really need those 20 or 30 “siblings”, in tagging this list is unordered, whilst in a “tree” or the like, I can order those siblings and look at or work on them in that order; if you try to get so something similar by search and/or tagging alone, you must do it by entering order number in an additional “record number” record field, but of course, this is done by the outliner for you, behind the scenes, and a no-outliner system could of course display, upon request, a field in which you would order the entries/finds (from searching/tagging) manually, by drag-n-drop, or by alt-arrow keys, and then assign/reassign the respective order numbers for this “view” or this “stored search” or such.

So, again, it’s about finds being automatically listed in some order, and an outliner assures this: You put your stuff into some category, and if your effort is justified, you do some manual sorting there; if it is not, you leave your material there in the original = chronologic gathering order, as it would be for a tagging system.

This easily possible ordering of siblings - as Hugh correctly stated in other words - is the thing outliner preferrers are after, it’s not the big tree that quickly becomes so convoluted in most outliners, since their developers don’t see the necessity to display various sub-trees, most of the time, of if they see it and offer hoisting, they do it in such a way, example Ultra Recall, that your overall view is always mixed up with your detailed view.

Outliner preferrers are NOT into hierarchy, we just need the big tree in order to quickly get access to our detailed “views” - and it goes without saying that for “views”, current cloning implementation, be it of items or of sub-headings = grouped items, is not fast, and not smart enough, I developed this in another thread some days ago: For every “view”, this gathering follows other rules, often very similar, but slightly different, though, and for such minor dissimilarities, it’s manual work and rework - again, I don’t know the details of the real big players in this field.

But it’s not “outliner people vs. tagging and search people” - it’s just that if (you think) you need ordered siblings at the micro level, today’s outliners available for individuals make this easy, whilst today’s tagging/search tools are not of too much help in this respect - not speaking of professional software that perhaps is much better, without even presenting one stable outline; as said, from any relational database, you can create multiple outlines on the fly.

Btw, that’s another missing point in today’s outliners: Technically, they would be able to present other views of the tree, AND the tree as you see it, so they would be able to revert to it, after your sorting by different attributes, but then, those sortings remain stable: You cannot switch views, you just could destroy your outline in its previous form. “Different views” are very rare, for example Bonsai does it.

This is not to be mixed up with flat views by tagging or searching, as in MyInfo. Technically, the “outline” you see in your “outliner” should not be but ONE “view” of multiple possible views, but then, developers do not offer this, except for in AS, which in its current version, you could use as a traditional outliner (if you have a macro automatically fill up some fields for any new item), and for alternative views, too.

It’s a pity in modern outliners there is one structure, by parentage, as “THE outline”, and then, perhaps, “stored searches”, but them giving flat lists of hits only, and no “stored outline by attribute x, then, y, then z”, for example. Except for being immensely practical, this would also finish with the misconception that an “outline” is some very special kind of organizing things: it’s just one view, by “parenting”, among many possible views. Currently, the unnecessary limitations of developers’ efforts too much define the presumed, very special “kind” of organizing things.