Surfulater PrevGen on bits, again

Posted by 22111
Nov 21, 2013 at 05:19 PM

So it is there again, 40$, and in light of what I had said some weeks here, I seriously considered buying at that price, and so I had another look into the developer’s blog, and from here, my decision was negative, again - and I had considered that program at least half a dozen times, over the years.

In donationcoder.com, I once explained to a tagging afficionado why tagging is inferior to outlining, and I will not redevelop the arguments here. Just some keywords of the discussion:
- superiority as primal means of classification
- no problem with tagging as a secondary means of classification
- not that much against tagging, but very much against taggins-as-it-is-today, i.e. without a (tags) tree (most of the time), without dop-down lists from which to choose your tag(s), and so on

Now there is no doubt that Surfulater’s developer is one of the really smart guys within the business, and in his blog, (most recent article, ““On Organizing Content”, in Surfulater-NextGen”, Posted on August 14, 2013 by nevf”), his development is straightforward, and you see he’s willing to delve into any (also, technical) depth of the matter it will take to deliver a state-of-the-art product; whoever being in search of conceptual exellence highly appreciates such an approach.

But then, Neville does not only do away with all the usual functional/usage problems of tagging (and if I understand him well, he even tries to present (mainly? additionally, I hope) those tags you MIGHT want to tag the item with, which, if it is well done, will be tremendously useful, but then, he clearly states that tagging, INSTEAD of contextual outlining, will be the paradigm of his software, so there will be no need to beg for any “dual” system. And for me, this means that any development in NextGen will be useless for me as a possible PrevGen user, since I NEED ordered context (see below), thus will not be able to “upgrade” some day.

And PrevGen, as it is, has problems with öäüéàè and all this, and when you reside in Europe, you will necessarily have many web pages from Continental sources, and they all contain such non-ASCII chars; also, there is no functionality to even mildly “scrape”, which means that for a homepage with some 12 immediate “children”, you will have to store all this manually if you need to store these pages; perhaps with a little scripting, again…

So now for the a “double” argument where I am contradicting tag proponents. In that article, Neville says that putting one subject into different contexts is “cumbersome” or something. His argument is similar mine, above, with tagging: It’s simply the CURRENT STATE of doing such things, or more precisely, of HAVING TO DO such things, i.e. cloning of items, that is really cumbersome, whilst OF COURSE it is possible to make this functionality available “at your fingertips”: Just as Neville is doing it now with tagging!

Which means that of course, a state-of-the-art outline should PRESENT you possible “parenting targets” “under” which then you would (also) categorize your item in question, and this from at least three pov’s (and I very much hope Neville will do the same for his tag proposals):
- current context: other items here: to which other contexts some of them were linked, too?
- history, i.e. targets you will have used lately
- standards: ToDo’s, but also standard “inboxes” for the upper levels of your (context or tag) tree (and of course, trom there, those clones/tags should be easily movable to more precise locations in those contexts (or tag sub-hierarchies, which would mean that “general” tags should be easily replaceable by more precise tags)
And I detailed before that such things should be made accessible by keyboard, too, for example by automatic on-the-fly (and not-persistant) numbering of the list entries in such additional (pop-down or durably-displayed) list panes; and of course, there should be functionality for processing “combinations”, combinations of items, and combinations of targets/tags (and combinations of those combined).

Now so much for the “cloning is cumbersome whilst tagging is so easy” argument - in both cases, it’s all in the execution, i.e. it’s all in the understanding and the smartness of the developer.

Where I really do miss a conceptual category that (current) tagging will not be able to provide, is, as I often said, context, and I’m even willing to conceive that a perfect tagging system could display lists of items “tagged similar”, or “tagged similar in this respect, and then with these other tags, too”, and your software could provide many such alternative sub-trees, in a row, or even in several columns displayed concurrently, on a large screen, without need to scroll; in fact, I often have MISSED such a “third dimension” in outliners, and tab loyauts (cf. Ultra Recall and multiple hoisting) are not a valid replacement, all the less so since they currently only display hoisted subtrees, which means that you will have to have created many a clone there, in order for those alternative tabs/lists displaying really “alternative views”, whilst ideally, such alternative views SHOULD CREATE THEMSELVES, according to your (multiple) criteria choices.

So, in the end, at least “primal context” is available within traditional outlining, whilst this, and “alternative context”, too, could be brought to systems relying on tagging, but what is missing in such “tag context”, is order. Let alone hierarchy, just order = sequence.

I know that even this feature could be implemented, and in different contexts, the same item could have different order “numbers” - technically, it’s all easy, it’s just a matter of the design of the underlying relational db, and of course, whenever you get displayed some list, you could manually re-order your items therein, and this new order would be stored somewhere, in order for you to get the list in the form you will have rearranged it another time whenever you will “reopen” that list.

But in this further “reopening” lies the psychological problem: “All” these views (most of them, anyway) are ephemeral, “nothing concrete, just there”. As I said before, I put my stuff into my outliner files, in general inboxes, in more specific ones, and again in very specific ones, or even exactly where I would like it to be, and most of the time, I only refine my “placement choices” when then I need something, in its context: Then, I sometimes have to re-order all the surrounding stuff for hours, and in the end, every one of 200 items I have got there, will be “in perfect order”, in THAT sense that the whole of it constitutes a “complete picture”, a “perfect/rounded gestalt”, or whatever you might call it, and it’s then only that I can both base optimized (not “perfect”, just optimized) decisions on it, or quickly use it within other contexts for which it constitutes an “element”.

And in almost every instance, without any exception I could be aware of, this grouping, sub-grouping AND ORDERING is the necessary factor for my thinking to put those details “into perspective”, “into relation”: The core of my thinking process lies here, in ORDERING things, and, as I said before, sometimes it’s quite long lists (but grouped with separator lines), not that much 7/7/7 cascaded hierarchies.

Now when the material is not of such big importance to my work, I just “browse” some 20, 30 or 60 items, but even then, either they are already pre-ordered in some way, or I now quickly pre-order them when browsing them, and in fact, I only get some “view” of it after this reordering which accompanies my browsing, and so, I “go thru” these items at least two times, onces in order to decide upon their RELATIVE POSITION within their (sub-) list, and another time in order to make my “real” mind up from them - in practice, this is an iterative process, of course, I read here, then move it some positions down in its siblings’ list, then see the next item is “not right here”, and put it at the end of the list (or even send it to an alternative inbox), read some other, without moving them yet, but in the end, the order of all those some dozen items will be much more “logical” than it willl have been before my “passing under review” this little bit of my knowledge.

Now you could pretend that my way of “thinking” is different from yours, but is it, technically? As we all know - and I developed this elsewhere -, and depending on your relative intelligence, you only can preserve some 4, 5, 6 (for real smart people, even more) “core ideas” within your immediate memory, your “thinking space”, and it should be logical that indeed, there’s much advantage in “good grouping” your core elements when you try to “get them all together” in your mind. Now this “grouping” can be made by sub-tagging, or, in an outliner, by sub-headings, but this is not the same thing, since it will cut up your list, into disparate sub-groups, whilst my “adjoining” elements will NOT make explode your list into separate entities.

In the end, it would certainly be possible to entirely replace smart outlining by smart tagging, but Neville isn’t near that, and my outlining is as “smart” as I can get it at this moment, with my technical know-how, and with overlaying external scripts on pre-fetched, dumb software, but with this system, I at least have preserved and enhanced some basic outlining advantage tagging will have way to recoup. And even then, having become on par, it will have to prove its possible advantages I don’t see at this moment yet. Again, we’re speaking of conceptual advantages over really smart outlining, and not of biased line-ups where technical details are designed and executed really dumb in one system and state-of-the-art in the other, to completely blur if not override the comparison of systems.