|
Another consideration of mine was, how would UR work on a slate, considering my almost 2 GB db? Ok, cut it up into 100 or more db's, so technically it would be feasible... but would it be reasonable? Of course not.
So, as many of yours know - since I related my export adventures here in this forum -, I finally went back to AO, from which I came four years ago.
Now it gets interesting, since at this time, some 6 months after my transition, I've got about 500 AO files; of course, these binary files are much more suited to such file multiplication than single db's: 500 or more UR db's within the same system, when you say, you can't understand why people would have more than a single one? And by the end of the year, I'll certainly have a thousand or so.
I know that out there, there are some advocates of the "single file" theory, preferably within the ASCII format, i.e. without formatting, and they go to great lengths explaining you that outliners just offer something the file system already has on offer for you.
Well, these people are tremendously mistaken. Does anyone imagine you could manage 100 k of single files within a file system, in any manageable way? Storing, yes, but processing? (The advocates of such systems praise various naming conventions by which you then could filter, and yes, they squeeze multiple tags into one single file name, easily...)
In fact, my AO files contain anything from 20 to perhaps 800 items, and if in doubt, I avoid additional sub-levels, inserting separator lines instead: This gives me a good overview of something which "normally" would be an intermediate (but unnecessary) level: Grouping things, clustering things is always preferable to subordinate things (and my cutting up a monster db into up to now 500 single files is another embossing of my underlaying philosophy.
You will observe that I'm here in the vicinity of that old debate "outlining or tagging"?, but than, as with "100 k of single files" vs. "1 monster file with 100 k of items", there's always the right thing to do FAR AWAY from these extremities.
Yes, today I'm against monster files, and I would never come back to them, seeing that their developers are unwillig to supply them with a "super system" that allows for immediate handling of projects, of reference needs, and so on, but continue to force you into their "chaos systems" (cloning and hoisting is very good for a start only, but when you'll have got 2,000 cloned sub-trees, you'll have got an impenetrable jungle, not an information management system of any value anymore).
But this, in no way, will make me desaffect trees, outlines!
I've got two principles in cutting up: Make it self-contained; make it as tiny as possible (and a third one, derived from the two, would then be: and don't bother if one has got only 10 items and another a thousand: don't try to "normalize sizes").
And then, for everything "pm", I use the file system, which offers, as you all know, (sub)directories / (sub)folders, links (= references = "clones"), and, among others, the "comment" attribute.
Then, you have some rather good file commanders, and whilst others can't even display these columns, some of them are able to filter by various "tags", trigger codes, etc., and with an extensive macro / scripting system (and with additional trigger keys, e.g. those of your numeric keyboard, simple, with control, with alt...), you'll realize a really good workflow indeed, and whenever I switch from my browser back to AO (i.e. with all the content I need, within multiple clipboards), my systems brings my back to a screen where AO has three fourths of the screen real estate, whilst a file commander, just showing a simple list of file names together with their "comments", takes the fourth fourth of my screen.
But then, most people out there didn't even replace their Alt-Tab with a simple key, or put their shiftlock key to a better use.
Now compare with my system that's perfectly scaleable, as said before, in both directions: No problems with slate computers, and a big corporation with 10,000 or more staff can rely on such a system easily also.
Nobody can blame me for not having tried with outliners, cf. my contributions for a better MI, than for a better UR - but instead of just complaining, I've got the means to do my own thing. I don't critizise UR for what it is; as said, in ancient times I created my own monster file application. But since I didn't get ANY help or encouragement whatsoever in my tremendous efforts to make other people's monster file applications really good, I stepped back from that concept I today consider obsolete.
It has been today that somebody on the outliner sw forum pretended, in writing, that UR had stopped development, and that brought me here, in order to check for any news on that. There wasn't any, so I seized the opportunity to relate my findings, grosso modo, of these last months without UR now:
I finally got my 3-pane outliner. And you know what? At this very moment, I'm considering begging file commander developers for enhancing their products, since of course my system, as elaborated as it might be in direct comparison with traditional, self contained "outliners" like UR and similar ones, is always a little bit rudimentary as I see it, but even if these file commander developers let me down, as have done outliners' developers, I'll do just more and more scripting, and I'll all get what I need even for the most elaborate standards.
And I even have got divider lines within my file lists there, i.e. I managed to group files within several groups on the same level of subordination, and any combination of data is now available on my fingertips... and ONLY that. I only see those 30 or so outlines I need for my immediate work, and if ever there's missing something, I just do a single further key press, and these parts of my information system get available, too.
I've said it a thousand times within this forum and the other ones: The problem of IM (information management) is not a technical one, but a conceptional one, and not being heard is not necessarily being wrong.
In the outliner sw forum there has been a discussion re granulation of information recently, i.e. how = up to which atomic level information should be cut up (I'd been heavily into these questions 15 years ago, so nothing new for me here). When you, Mark, want to do an outline within an item, it's exactly that problem you're discussing; when MI user (and history prof) "wsp" touts MI (and rightly so) within the outliner sw forum for the former's ability to reference not only items (as UR does) but paragraphs, it's again information granularity he's discussing.
There might be a lot a theoretical possibilities to that, and indeed, on that champ de bataille, I've got my very precise ideas also (just for a start: cut it up into sentences as the atomic level; do 1 sentence 1 line within monster text files (any formatting being encoded by special characters that are resolved accordingly, afterwards); do heavy referencing, cut up the monster text files into as many monster text files as you need for your big corporation... but do all the technical processing on a referential level, processing "atoms", not any more half-way physical combinations, and half-way references to other parts in other physical combinations).
But on a practical, application level, with what we've got today, everybody must search for his individual solution how he will granulate, then re-combine his information bits in order to make them useful for him, and all I can say now, after 6 months' heavy information processing in and out of my new heteroclite, scratch, homemade system:
It works like a charm, and it's the first of my many information systems that does so. But it stands and falls with both components:
Hierarchical systems are a necessity then, provided you hold 'em flat and combine 'em ingeniously.
And yes, kinook could have done this, within UR's borders, not for big corporations, but for all those freelancers and 1 to 5 head business gathering here. But instead, they don't do any super system that would drive UR into another sphere of utility. Cannot help those, but could help myself.
Yes, We Can Do. Individually. Ingeniuosly. Don't be customers, you'd be waiting forever for the manna to come.
|