I got some thousand files into a folder structure by exporting part of my items into them, automatically, from a competing software's exporting function.

Now I wanted to import them into PB Prof (Trial), in order to evaluate PB for my purposes (I've got dozens of thousands of items in my current software, so transferring them into PB will be an important step in that, I simply cannot do that by hand).

Now I must see that the folder structure hasn't been maintained in PB's thoughts structure but partly only, and then the notes of my items, i.e. those rtf files within the structure, are listed as attachment files within those thoughts.

In the same way, even for items I have got thoughts within PB, their original notes (i.e. the contents of those rtf files) have not been put into PB's thoughts' notes, but are attached to the thoughts in their rtf file format.

Thus, for several thousand items / notes, I only get under 1,000 thoughts in PB, and for all those notes, I don't get any notes within PB, but several thousand rtf files attached to those under 1,000 thoughts only.

I'd never had imagined that PB's importing from a file structure would not allow for some optioning by the user; my option would then have been, make a PB thought for every such rtf file, and put the rtf text into the thoughts' note fields (being able to contain rtf format it seems, or is is html?), thus, no rtf files attachment whatsoever but perfect integration of their rtf content into PB' notes, in formatted form.

As there is no other practical way to import things from competors into PB (?), I wonder if there's really no way to rightfully import them by importing a file structure - or does PB staff simply do not see the importance of this for would-be migrating users?

I cannot bear the cost of hiring a programmer for doing this work.

Afterthought: Are there ways to better import in html format, again using that third-party file structure in rtf format? I.e. the competitor exports into files, formatted in rtf format; perhaps there's a transitional way from which then PB can import in html format...

but doing a thought for every such file, and putting the files' contents into the notes field of the respective thought, that is.

Please let me say that PB's notes are just wunderful, since just hovering with the mouse over a thought shows the respective (rtf or html, i.e. formatted) text in the notes' field. So, if you always leave that field open, rather big, (vertically) 1/3 of your screen or even on a second monitor, you'd have a tremendously "accessible" information system, without additional mouse clicks needed for this deep-level information display.

But then, first, you must get your information into those fields, and let's say, it'd be 50,000 of them...

Boy, I'm not aware of an easy way to do this. If your other software doesn't export to an .opml file, can it export to a single text file? The reason I'm asking is that you can format text files to import both the thought names and notes (page 213 of the PB6 Guide). If there is a certain character delineation that notes the title and the notes, you could do a find/replace to convert it to the outline format.
You could try e-mailing and see if they have any input/ideas.

P.S. Welcome to the forum.

macOS 10.14.6
TheBrain 11.0.119
Hi zenrain,

Thanks a lot for your answer and welcome.

Doing "one" file should be possible, but I hadn't been aware that PB does correctly import such formatted files, I'll have a look into this and shall report.

For the delimiter character, that should be possible my macro, by putting it in front of every original item (I'm speaking of a double of my original files, of course).

I said "one", since of course, if this way is viable, I could do about 10 files or so, with about 3,000 to 8,000 original items each, or even 20-30 of such files, that's not the real problem in this.

But there's one thing I don't know yet, from your kind answer: My original items are "indented", i.e. are on first, down to fourth or fifth level - would PB able to "understand" this "indenting", and from which trigger? Since you're speaking of a delimiter character, it seems to me that this is precisely the problem with this, and for the export from my original program yet:

How to indicate the indent level of every original item, by export, first? (Okay,my current program is ABLE to to this, macro-wise, since it lets me select all my items by their indent level; thus, I would be able to insert special characters, in order to identify the indent level.

But then, how PB would be able to identify those levels / special characters, and put all my items of any such a file into a new items' / "thoughts" ' hierarchy? That seems impossible to me!

I cannot do with 50,000 items, almost all of them on the same level, i.a. all my current hierarchies lost!

Why would it be necessary to address PB staff by mail? Ain't they reading this forum, and / or don't they answer here?

Yes, PB does honor Tab indenting as a hierarchy.
Thinking a bit more about this though, the text import allows you to reference/attach hyperlinks, but not files themselves, so I don't think that will work after all. Sorry.

The PB staff do read the forums, but I think they scan the issues threads much more frequently than how-to threads. E-mailing support will get you a faster reply.

macOS 10.14.6
TheBrain 11.0.119
Thank you very much for this follow-up, zenrain, and I see what you mean: How-to's are answered by power users, whilst suggestions are considered by the staff, and that makes perfectly sense.

I'd like to say that I'm aware of some shortcomings of PB, e.g. the relative impossibility to introduce "order" in your things, "order" meaning 1, 2, 3... - of course, you can manually order steps by making the first element of the thoughts numbers, from 1 to ..., but then, whenever you introduce some step into that range, you either will have to rename all the following steps manually, or you must insert some "12a", "22a/b/c" or whatever, and that does not make a neat impression on your mind...

And all's about thinking enhancement, though, thus that's an important consideration.

There are other deficiencies of PB...

but then, for some years now, and using at this moment the very best of traditional outliners (in the Win world, that is), Ultra Recall, and in its 3-pane variety (which I found immediately out when trying it, whilst neither the developers nor any of long-time users of that program were aware of its (accidential) possibilities), I've never put PB entirely down, since I'm sure that it has tremendous potential to enhance your thinking, even when up to now, this potential has not been put into work...

which is to say, there must be some fault in the execution of the brilliant basic ideas that led to the creation of this program, and once clearly identified, those faults could be overcome, and the program could become a real winner, making its users real winners - since word would spread of that.

But PB staff seems to have never really been interested in enhancing PB's importing capabilities, and this is paradoxical since they insist in their public speak on PB's tremendous capabilities of handling monster files; well, if people like me who have 50,000 or more real items to put to trial within PB, could do that, they would not only become some more customers, but especially, they would become valuable sources of insight into the needs of users with that amount of real data...

Please note that for databases, response times, etc. are at stake; thus, dummy data will do, for most of your data; 1,000 real records might be needed, and then 49,000 dummy records of various kinds, in order to set up the haystack in which then you'll search your needles.

But for an information management system, that sort of approach is almost useless, except for response time, etc., of course: If you want to optimize an information processing system, you'll need real data, in which you're doing real work all the time, and then you'll have real experience for large amounts of data, and their interaction.

For there is a profound difference between most (not all, but most) data in databases, and information you put in information systems: Most "normal" data is not interrelated, in re-use, but for statistic means, i.e. you'll have perhaps 50,000 customers in your database, but each customer is to be treated separately, except for statistical analysis, which means that all those records, again, will be treated in a similar way (or in a million similar ways), one record after another: That's work modern processors are made for.

On the other hand, 50,000 information data items are INTERRELATED in many ways, and to let us view those interrelations (that we must establish manually, though, up to now) is the strength of PB...

And its weakness, I might say, since as soon as you'll have a monster file - which I cannot build up in PB with my data, for the absence of adequate import facilities -, you need MULTIPLE VIEW INTO that monster file, and PB has not come through with something valuable in this respect yet, it seems.

BTW, there's a very interesting thread here, currently, discussing PB thoughts, data types and tags, where some poster contributes thoughts that cannot be overvalued.

But then, it clearly shows that PB's INHERENT qualities in this respect are simply not yet implemented; data types, tags and whatever you want are just superpositions in any outliner like PB is, they are ersatz for something better, not yet the real thing.

For being perfect, PB should have some (pre-defined) "intelligence" built into it, allowing for PARTLY VIEWS, perhaps even with "false-loose-ends", indicating pathways to other partly views, but GIVING A CONSISTENT VIEW of the things you need to consider; PB in its current state does NOT do this yet.

I'm not giving away secrets here when stating that ZOOT is trying to develop this approach; but they're far better at "automatic tagging", as I see it, than they are in "automatic display" of the inherent choices of this "tagging" - I have the very strong feeling that it's PB that has the utmost potential...

but then, potential on its own is not worth a lot; just see the demise of askSam, lately - some weeks ago, they even buried their forum which did too much harm to their remaining sales indeed -, and that program had not only had tremendous potential in the eighties, but even some years ago, had introduced the splendid idea to build up outlines (i.e. flat outlines, unfortunately, i.e. one item could not be present in more than one place of the tree, but for a beginning, it was splendid as such) of various field contents (and only to some indentation levels, but again, for a beginning, it was outstanding) - and now look at askSam 7, out there for some 3 years now, and they seem to have come to their end.

Even now, I know rather well what an almost perfect PB would be, but with my stuff put into it, I could be much more specific about it, and I would not have to give virtual examples I'd need to make up, but I could speak of my actual experience with PB (and its current deficiencies)...

and, let's put it bluntly, I'd be much more interested in contributing to the optimization of a program that I could really use, waiting for better things to come, than in giving my 2c to some software that intrigues me, yes, but that's out of my reach: in the same way, I much respect some Mac software and would like it to be developed even further, in the right direction, but I would not my time in considering detailed options for that possible development: I don't own a Mac, and I never will, since I (can) do some programming in the Win world, but not for the Mac, so I stay with the (perhaps lesser) world in which I can do some things I need to do, while in Mac elysium, I'd be helpless.

But again, please allow me to put this perfectly clear: The 3-dimensional representation of information bits, in half-automated and CONSISTENT VIEWS, is the secret to real thinking enhancement by an information manager (and the roots of askSam's tree-building-on-the-fly were rather near it (so I'm insisting on askSam here), notwithstanding its poor execution on first try that could have been greatly enhanced afterwards but unfortunately never were, in spite of my insisting and encouraging, and that of others) - and I'm certain PB is not far away from this target... but does not do enough, currently, to make the last step.

More precisely and on a more technical level, what I'd do within the development of PB, I'd concentrate not so much on the nodes, the thoughts, and their categorization (i.e. what this concept of data types and tags does), but more on the vectors, the relationships between far-distant step-parents and their step-children, and between far-distant step-siblings, step-cousins, and so on:

I'd concentrate on the automatic / semi-automatic establishment of such relationships, yes, but first and over all other, I'd concentrate on the visual representation of SUB-NETS within the net, depending on such special relationships, excluding most, but retaining those you'd want to see.

But first, people like me should be enabled to put their real data into PB in order to give much better advice, from within the lot.

And I'll send PB staff an email in order to draw their attention to this thread and its problems / possibilities. ;-)

But then, PB staff, perhaps you could dwarf this input window even further: from 5 short lines into some 3,2... or why not ONE line of just 12 characters, like in the very first electronic typewriters, back in 1980? Just joking, but then, it really is underwhelming... whilst the search function of this forum, or more precisely the graphic representation of the results it gives, is a splendid thing I like very much.

Yes, I can see how the inability to import your data could hinder your testing of the product. Especially with large data collections.

For custom sorting, if you have time please vote for that feature in UserVoice.

macOS 10.14.6
TheBrain 11.0.119
Hi zenrain,

Done, I gave it 3 votes so there are 17 votes left for me to attribute to other things; I welcome your idea to give a real link since on my own, I would not have found it; in general, it's a very good idea to facilitate thinks for people when you want them to do things. Thus, well done, and successfully done.

As for the PB staff, I got a service ticket...

But then, for things-not-easy, we must allow for some response time; that as well is a valuable advice in management; a constructive answer after a week outweights by far a so-so answer on-the-spot.

But afterall, I know from experience that out there, there are a lot of software developers that are more of less advice-resistant, if not outright advice-proof, and I would NOT be happy if that was the case with PB staff; as of my experience, a lot of developers, even having done VERY good work up to then, are dollar-minded and preach themselves, "with all that, I've got 80 p.c. of my 'natural' market - why should I double my efforts in order to please a lousy 20 p.c. of them much more, then?".

Thus, it's not the 80/20 "law", but then, that law is flexible in its percentages... ;-)

Of course, what those dummies do not see, is that their "natural" market would double, triple, rise tenfold, IF they did that additional effort - completely NEW markets would suddenly appear, their competitors (= in a conservative, "as-is" market that doesn't seem expandable to THEM) doesn't even suspect to exist, out there.

I don't like Steve Jobs - the a**h*** factor is very high in dealing with him, that's for certain -, but he's the man who has perfectly understood what I say here, who puts it in practice more - indefinitely more - than any other person in the computer industry and way beyond, and who gets rewarded for that in a righteously obscene way : He makes that 50 P.C. MORE EFFORT... but his revenue stream did not double, it did explode into stratospheres.

PB staff did never do that additional effort, and, as I hadn't been aware before but have just been learning by my voting for the numbering / sorting function on your behalf, they even seem to have CRIPPLED their software at some moment, since without that ugly suggestion, that "please give us BACK that function" wouldn't make any sense.

I'm Fred in, I'm schferk (= "Schweine und Ferkel", = big pigs and little piggies) in's UltraRecall, I'm Fred X in (see my advice this thursday = in 2 days from today, on Axialis Icon Workshop Prof.), and I'm Fred in - oops, they finally got rid of their very much too encumbering forum, showing in every detail their intellectual and business suicide; - I know exactly what I'm speaking about when I'm discussing things in my ever-lasting search of excellence.

(In the last moments of the nineties, I developed and marketed a cascading outlinining software where not only clones were a matter of course, but in which users could freely decide on the content of multiple, cascading item lists of freely-to-be-determined indentation levels, and where "history" functions rated most highly, when in today's programs, "immediate access" functions are undervalued, to say the least.

Unfortunately, I'm not a programmer, so I did this most splendid thing with Allen's (oh yeah, ex-MS's Allen's, believe it or not) Toolbook 6.1, so it never went stable, since the underlying programming language wasn't, even when my programming there was faultless - and, folks, I can prove it.)

This being said, my search for the ultimate thinking-enhancement thing has became compulsive, since I had been near it on my own means there, and I'm searching for software developers that have the financial means to experiment - PB's staff has those, because of their corporate biz -; so, all this is reduced to that question, who, of those professionals out there, is able to show some of the intelligence that Steve Jobs shows to us every day of his life.

And no, I don't own an iPad, iPhone, iWhatever, or even a Mac - but thats because I'm stuck with Windows since the little programming I'm able to do, I can do it in the Win world, but that's the lesser world, and PB staff, that, among other things, have the intelligence to do PB on BOTH platforms... so, why not "hire" a little bit insight where it could make 'em really big.

"Hiring" meaning that credits must be given, and that developments must be made (after discussion that is, and that goes without saying, but it could NOT be "please give me this, please give me that", as in, you see?); under these conditions, PB could become just right outstanding, within reasonable time, and for free, as I'm concerned.

Or make your thing, and never get through; some people LOVE to stay in their comfortable niche.

Let's pray PB staff don't belong there but envision bigger than that.

In fact, they say Apple is the most valuable incorporation in the it world (software / hardware at least)? Did it ever occur to anybody that this tremendous success did NOT come with some consolidation in the software world where nil, mediocre and sometimes even acceptable software is splattered all around the place, but where not a single software is really outstanding, every one of them clinging as hell to their deficiencies and even to their conceptual impossibilities to break through?

Software developments during the last ten years have been almost imperceptible, except for the database refinements of google, facebook and their likes; for the individual user and even for the corporate user, this last decade has been lost; just compare what's here today, with what has already been there in the second millennium, and you must agree with my conclusion.

PB is one of those software with the potential to give us something really new; it's up to them to accept the ferment I'm offering, in order for that to realize.

Something on the opposite of service tickets, then.


Thanks for the post.  Lots to absorb here and I've passed your thoughts on to a few others at TheBrain as well.  Also, thank you for voting on some features at  Our engineers do visit this site and the votes do make an impact on future builds.

In a nutshell, you are looking to import a large quantity of content from one application into PB - correct?  We do have several import options in PB.  It seems that you have already experimented with the folder import and know its capabilities and limitations.  You may be interested in sampling the XML import as well.  I've attached a file that outlines this feature.  Please take a look and let us know if it is something that may be of use to you.

Larger, database imports cannot be facilitated in PersonalBrain and would require our enterprise product, BrainEKP.

Thank you,
Hi Matt,


Ok, I did (for a minor tree, not for one containing thousands of items; in fact, 600), in UR, "Expand All", then I selected the expanded items.

I did a control-c. Then, in PB, I did EDIT-PASTE THOUGHTS. Of course, this didn't do anything, but then, I wasn't really expecting something useful to come out from this childish try.

Then, I let that UR tree selected in whole. I did, in UR, FILE-EXPORT-XML(OML), and then, I unselected all elements, in the dialog, to be included into the export file, except for "details(rtf)", i.e. the content of my text fields - of course, I could have deselected that element altogether, but bringing the tree, and the tree only, without any contents, into PB, wouldn't be of any interest to me. (I also optioned for "export child items (recursively)", of course, this being the aim of my tries; note that I even unselected the icons to be exported.)

Export in UR went fine, but afterwards, I had to add the ".xml" suffix by hand to the naked filename, "brain1", perhaps that was my fault, UR works very smoothly in any circumstances as I can assure, it's the most professional program out there, among traditional outliners (which PB is not which is the reason it intrigues me that much).

Then, I opened another new file in PB and optioned for FILE-IMPORT-BRAIN XML. Almost immediately, I got the message, "Completed pasting thoughts. 0 created. 0 modified." Then, I got the message, "You have to wait for this process to end" or something, and indeed, I needed to use the Windows Task Manager in order to stop PB working; thus, not a "crash", but sort of a "freeze".

New try in UR, with Export to xml/oml, but some special items less, only ordinary text items (mostly plain text, some with some rtf formatting, nothing special), like last time, but this time, I specified, to file "c:\brain2.oml". Export without problems. Afterwards, I manually changed the suffix from "oml" to "opml" since in PB's import function, there's an "opml" but not an "oml" format.

I opened PB anew, and opened a new file there. Then, I did FILE-IMPORT-OPML.

Intermediate result: No crashing or similar behavior within the very first seconds, but my harddisk has been constantly working now for more than 45 minutes, but without any acknowledgement from PB's site it's working on something, the menus seem to work normally.

So I even succeed in tagging the source thought, and add child thoughts, etc., manually, as if there weren't any import process running.

But my hard disk continues its unusual work, so perhaps PB has got some separate importing / transponding process component working independantly from the main program? Whatever, I close down PB in order for my harddisk to stop working...

Well, with PB closed down, my harddisk won't go to sleep but continues its work...

Anyway, new try: In UR, I just export some 15 (= 3 indent levels including the first one), instead of 600, items into the oml format which I name ".opml"; in PB new program start, new file.

As before, and after another 20 minutes' waiting time (while I'm writing this that is) for 15 simple UR-exported oml or opml items to be imported, I give up.


Since you speak of alternative import formats for the corporate versions, please let me say that I perfectly understand that you want to charge corporate users quite a lot of money for corporate use, i.e. distributed-architecture, collaborative uses, but that then, it's beyond my comprehension why allowing (if that is) corporate users import of their original material, where those same materials (?) cannot be imported into PB.

As for EXporting, I could perhaps understand, since that would be hindering people from departing you... at this very moment, "the whole web" (= the cognoscienti, that are) is laughing about InfoSelect's owner's not having withdrawn important exporting features, but about his naïveté to believe that anybody out there will believe him that this was upon users' requests!

But for IMporting your stuff into a program, and IF the technical foundations are right there - in your corporate versions, it seems -, it just seems to be quite another proof of my developers' niche theory, they simply don't want to outgrow a select userbase ?!


I've said it elsewhere, I'd be happy to acknowledge any claim to have gotten that idea before I put it into practice, but in about 1997/1998, I thought very well to be the first one to have got that idea, and even today, many years later, I don't have it seen put into practice anywhere:

My cascading lists, back then, were perfectly smoothly to be used, since I had a 3-button-mouse back then... as everybody else, that is... but thought, well, if my lines in all those lists / panes were divided up into 3 parts (= you can make it 4 or more, but 3 then is perfectly useable even for rather narrow panes), let's say the first 25 p.c., the last 15 p.c., and the whole rest (= from 26 to 84 p.c. of the pane's width), you'll have NINE possibilities to trigger commands with your mouse, not only three!

And with that repartition, 25, middle, 15, you won't have any difficulty to set your mouse clicks without even looking, and without getting any unwanted, erroneous command. So, formatting, copying, transferring of items was more than smooth, as was the decision into which of my multiple cascading lists (= some with text pane, some as intermediate lists only, without the text contents being displayed yet) a clicked-on item was to be expanded (in fact, I only had one virtual tree (and with automatic unwanted-recursion checking), of which the screen representation then was cut into those multiple panes, etc., question of not having any indentation but BETWEEN panes, NOT WITHIN panes, sort of mandatory hoisting if you want...

This way, a normal click "in the middle part" expanded the item into the "main" list (= with contents display), a normal click in the last 15 p.c. of the line expanded the item into the "next" one, and so on, in cascade (and of course, items with children, i.e. "expandable" items, were automatically formatted other than items without such children), whilst a normal click on the first 20 or 25 p.c. just set that item as the parent of the current pane, i.e. its content / list was replaced by the children of the clicked-on item.

Right-clicks in those 3 parts did different things, among others, I didn't had ONE content menu, I had THREE of them, for entirely different sets of purposes, formatting, displaying the item's children in one of several pre-defined additional "fixed" panes, and so on - you easily imagine my ease of use of that system, with having at hand, at any given moment, exactly those ToDoLists I wanted to see - and there were plenty of forth-back commands between all those elements, and many more.


I'm not saying here that my system described here-above would be helpful in PB; in fact, it cannot be implemented but onto "fixed" pane widths (that are variable as such (but identical from one item in a list to the next one, independently of its name length), since we're speaking of percentages, but we're speaking of a given line length fractioned into percentages, whereas in PB's screen's items, you'd have to position the mouse upon percentage fractions of the individual item's name lengths, and that wouldn't do).

But I'm saying that I am able to do a lot of valuable things THAT WORK, and of which nobody else (?) seems to even think about (and in fact, even 9 different mouse clicks weren't enough for me then, so I divided my lines up into FOUR parts, but then, memorization problems set in...), and many developers seem to cannot see good ideas coming from the outside of their family bunch, and thus, nowhere REAL excellence = excellence up to the last highest steps that make all the difference, is to be found, whilst I cannot program the perfect software on my own means...


Of course, in traditional outliners, I come close to a good approach, even without the help of the developers.

In askSam, everything was more than awkward, and buggy, and it went from bad to worse.

In ActionOutline, nothing was possible since I didn't see the possibilities yet file groups and macro systems on top of a program can bring.

For MyInfo - which I touted and tried to help out of its mediocrity without any common sense -, I finally constructed a macro system I could have used in an identical way for AO, and without profiting from any feature MI has "above" AO; a lot of (necessary) little things were done upon my request, but nothing really useful that would have make MI stand up out of the crowd o many.

Ultra Recall had intrigued me for a long time, but its idiosyncrasies had always put me away from really trying it; being fed up with MI's lack of potential, I played around with UR, here and there...

and then I discovered its children pane could be used as an intermediate pane, making UR a 3-pane outliner (= a crippled one, since there is currently no good way to show up if such children there have children of their own or not), and from that moment on, it only took some minutes I put all my stuff into UR.


But of course, UR having a very fine cloning feature (whilst AO hasn't got any, and MI a very rudimentary one), I put all my stuff into one monster file... and UR worked perfectly with my 50,000-plus items.

But, might I say of course, then problems... in my HEAD began!

In fact, I fiddled around with hoisting, with favorites, with heaviest cloning all over the "place"... and to put it bluntly, my memory and my "intellectual ordering capacities" weren't up to par with all these 50,000-plus items, especially since there ain't but 50 (!) "favorites" for one file (= be it of 50 items or of 50,000!), and my begging for allowing for more wasn't even answered.

(UR developers always say no to me - and to most wants of fellow users, but then, PB's ordering feature, having been there up to version 3 and being gone since?! ;-) - when I ask for some feature, and even pretended they didn't understand what I say since English is a foreign language for me as they rightly intercepted once; thus I amused myself to use pidgin English there for my very first posts there, in order to appear the most dumbo possible, but quickly lost interest, not in posting there, but in posting there in a way to appear as idiotic as some people there pretend, but indeed, my alias there was motivated by my impression, "pearls before swine".)

Thus, I was STUCK with the INTERNAL means of UR, since a sphere / are / field level just containing 50 items, without clones there, and no serious grouping, is simply not sufficient.

But as soon as you sacrifice that super level, being made possible by favorites, you are lost, since your main tree MUSN'T be the main level, it must be the first INDENTED level, the project / area level, in such a monster file.

(There are ways, with use of the hoisting feature, to organize something simili-indented with the tab function, but then, as with so many things else, UR developers refuse to introduce something almost any other classical outliner has, a title line above the text / contents field, so I refrain from using tabs there, since that's the only way your (one and only then) tab is long enough in order to serve as such a title line even for longer titles.)


Thus, I ended up with applying my macro system upon UR that I had, in a similar way, used already upon MI: I cut up my UR monster file in many files, and even more and more; at this point, I'm into about 90 different files, and there will be more than 100 soon, that's for sure.

I had to sacrify the master level for this, since if its right that you can use 50 favorites for each file, those favorites ain't useable as a master level for more than one file, so there was no way.

Instead, I use manually-maintained (!) menus, per HK (Hot Keyboard), with an additional keyboard (but you can do the same with Alt-x/y/z, of course), as my master level and as my first level under this master level.

Thus, KP (keypad key) 1 brings me a "general" menu, with several sub-menus, and under every such sub-menu, I put (manually) full names for my various files, and of course, identical files can manually be put then into several sub-menus of such a menu, or even (and often), into sub-menus of several different such menus; I help myself with putting an asterisk after the name in case of such double or multiple use of a file.

The actual file names are one- or two-digit only, such as b.urd or cl.urd, this being because in the "open windows" menu list in UR, it's they that are displayed, and if I have open dozens of such files, it's a question of possible length of that list on my screen... or so I thought.

In fact, I could put there up to 50 of my 1- or 2-digit file(name)s, but then, I quickly discovered that UR has another limitation, there's only 25 of such file names displayed, be there enough room to display more or not.


In order to quickly access individual files, I put then an ampersand in my HK menu entries before the character to which I want to attribute the entry, and then, my macro goes, "KP1 a/b/.../z", i.e. TWO KEYS, without any Enter or other, opens every wanted file for me.

Now I have EIGHT of such menus, with various sub-menus, put on 8 different keys on my additional keypad (Cherry 4700, that is; I'll buy something better soon, but then, with 128 keys or more, even with color encoding of the keys, it's quite impossible to memorize, so some 30 or 40 keys WITH SCOPE, as have the 21 keys of my Cherry, might be best, scope meaning that the same two (and color-encoded) keys, in every program, go back and forth a tab, e.g., thanks to HK's interception of the standard key combinations sent out by the Cherry (= Alt-Shift-a...z; attention here, don't make it Shift-Alt-a...z and avoid triples (= Alt-Shift-Control), both in order to avoid problems with many programs)).

(And to say it all, if HK's system variable functionality were more reliable - but it isn't; the developer hasn't answered my asking for amending; the program hasn't been updated for years; but it has some good points I cannot find elsewhere, so I accept that deficiency for the time being -, that scope could even be varied WITHIN any such program, depending on the keys pressed / functions triggered just before... - in fact, that's the reason because I need some 60 keys or so, on mid-term, whilst I'd much prefer the 21 Cherry keys, with perfectly reliable scope by system variables, beyond just HK's (reliable) scope by program caption.)

These 8 different menus, then, are in relation with different aspects of my personal life, with different businesses, with different programming adventures and with many legal dispositions.


And, of course, this system, where "Standards" and "Projects" are a little bit mixed up (but only just so that I don't lose my clear view of all of them), is done in order for NOT having to mix up things on levels 3 and deeper, i.e. anywhere in UR: There, files must be self-contained, be their big or tiny ones.

E.g., I do some progamming for some business, so that business has a menu with several sub-menus (= all in HK, as said, since UR doesn't offer such a "master" functionality, as explained above), but also, my relevant programming efforts, including scripts, database details, data mining details, and so on, are listed in a sub-menu, or even, scattered among different such sub-menus, where they belong.

On the other hand, those software and programming files are also listed under the appropriate headings = submenus in my Comp menu.

In the same way, legal sources are listed in my Legal menu, under various sub-menus - and then, if I hadn't brought them together anywhere, how could I check them regularly for accuracy / changings having intervened, by their web sources? - but also, these legal resources are listed under every sub-menu everywhere, businesses, projects, current affairs... - where I need them.


Here, things become really complicated, but first, let me make an insert: To HK submenus, i.e. subheadings in those menus, I have also attributed shortcuts = single characters, but for another purpose: They are groups. Thus, whilst "KP7 x" (= 2 keys pressed) will load / open / go to "Law X", e.g., "KP z" (= again just 2 keys pressed) will load, e.g. a GROUP of laws, Law X, Law G, Law Z and more of them, whatever.

In order to have some order in these things, I do the following: The subheading, when triggered, loads all the files it is subheading to, i.e. all its virtual children, whatever they may be.

This way, you must sometimes multiply your groups = subheadings, but then, it is perfectly possible to group several files in any other way, i.e. to make just a special entry (= perhaps beginning with a special character, or in uppercase letters) which, when triggered, does not load just one file, but several of them, e.g. all the remaining ones within that group / subheading under which that special entry is filed.


It goes without saying that the manual maintenance of such a complicated, in multiple ways interwoven system is awkward, time-consuming and prone to multiple errors, so it should be included INTO any serious such information management system. But for the time being, and being IN ABSOLUTE NEED of such SEVERAL MASTER LEVELS, before then, in (UR's or whatever competitor's) software's tree, the DETAILS start to flow upon you, I haven't got any other chance than to do it manually, since that's the price for GREATLY ENHANCING the usability AND usefulness of any such program, and as having said, UR is one of the very best... not the least because of its intermediate children pane.

That intermediate pane gives you an additional level because it functions this way: Provided that all options are set as they should be set (which is kind of a problem in UR, or in fact one minor of its multiple problems in learning to use it as to your very best), you can scroll within the main tree (of which the first level is supposed to be the first level overall, when in my case, as explained, it's already level 3 or 4, so imagine you have that main tree "open" to some second level, then, in my case, the additional pane will show you level 4 or 5...) -

and without that main tree changing by any means by that, just displaying the children of these items in that additional pane; you can do a scroll in the main pane by arrows, or you can even click upon an item with your mouse, then proceed with the arrow keys, all this immediately brings the children of that item into the children pane.

And it goes further than that: In the children pane, you scroll and / or click (again, when all options are set right), and you see those items' contents (e.g. their texts), and all this does NOT make any change to the main tree (nor does navigation in the children's tree make any changes there); it isn't but by DOUBLE CLICK that you can, on purpose, "go elsewhere" if I might say so.

This behavior is as smooth as it can get and is very near my "splendid navigation" I had created in the past, in my ToolBook tries - except for the fact that in UR there ain't those HIGH LEVEL navigational facilities any serious user is just in need of, on top of that smooth navigation further down.

But you understand now why I left MI's developments.


And now for that complication I promised you.

I had adopted that fractionizing of my data into files which I than grouped into various "packages" / "projects" / "standard clusters" / whatever, since I had to overcome, at the time, MI's (real or just unfoundedly feared) deficiencies: In fact, its cloning feature didn't allow for correct monster-file-wide cloning of parts of the bunch since any (?) changes you made after cloning to your cloned material (or to your clones) weren't then automatically transferred into the clones (or vice versa), let alone for the awkward way of doing the cloning. And for the perhaps just feared deficiencies, to say it all, I never dared to build up a monster file in MI, just having not enough confidence in its stability (but then, perhaps wrongfully so).

But by no means do I pretend that the splitting up of your base material, or of your working material, into FILES is somewhat natural: IT IS NOT!

But what can we do?

As said before, in MI it was technical reasons that withheld me from putting it all into one file; in my own program years ago, I had built up one single file indeed, but then, I had less than 10,000 items at the time, AND I had ALL those indentation levels I had dreamt of, at my fingertips, integrated into the program (and therefore automatically updated after every imaginable operation there), and so my vision of my base and my working material there was perfectly clear at every given moment, notwithstanding the fact that I did very heavy cloning, of single items here, of groups of items there, just as it seemed necessary and / or useful in every instance.

And I said it before, my own monster file in UR was overwhelming to me, when in fact I had put all my efforts into align it to my needs -

it isn't but now, with my having fractioned all my things again, that I get a clear vision of my things again, those "high levels" holding it all together, but allow me working on just some grouped things I want to work on, all other things being withheld from my vision... but being there, waiting in the background, in order for my possible needing to integrate them...


And this integration of "things out there" into the things you're working on, isn't but very poorly supported whenever you put things into different files.

For project 1, you need article 12, 245, 1247 of a given law, and a dozen more articles of 5 or 6 other laws, especially, and in a way you must work on them, and some others just for reference purposes; in project 2, it's again several standard articles, and again some various special things you need to integrate; holding out SOME recurring "standards" can be often useful, but then, the variations are endless... so what to do?

Let's have a (very far, very poorly informed) look upon Google and other real monster databases, with not 50,000 items as in my case, but with a billion of that.

I know nothing of their systems, but if I have just a fractional bit well understood their functioning, they put a decisive differenciation between references / referrers, and data; they distribute the data any which way they can, for physical storage, but they do tremendous work upon the referrers, the pointers, you name 'em.


Why not store your data in files, in some "natural standard grouping" entities, as long as your pointers allow you to have instant access to anything, anywhere?

At this time, UR even allows links to specific items in ANOTHER database, but it's more than awkward in use, so that it certainly cannot be useful except for exceptional, rare use conditions (whilst all linking / cloning WITHIN a database in UR is of the finest programming quality anywhere to be found), and so, poor me, I do what I started doing in MI a year of so ago:

I cut up my information into every tinier chunks, while desperately trying to preserve common sense in doing this.

Have an example. For rarely used bits, special law articles, I've got the law in a file (with indication as to the last check date from official sources), within the "Laws" KP, of course, but then, I just copy that given article as a special item, there (whilst the law itself is just another item, or parts of it are fractioned into different items). (The same with groups of articles: Instead of copying them directly from their source file to the target file, I replicate them in the source file, as sort of a reminder, and copying them from there only.)

From there, I copy that given article, as such, into my current project where I need it to be inserted.

In the law file, when checking for updates, I check by electronic means (well, using Beyond Compare), with macros (and thus, even for fractioned laws, I always upheld an integral version in order for Beyond Compare to its work in that), comparing the new, raw version with my UR-integrated, formatted, and perhaps even annotated version, no problem with that.


Then, since I see then where there are changes (that will go into my stored versions), I also see which ones of the "extracted" parts / articles are affected by this. Up to now, I used MI's "Search over all files in a given directory" function (which was awfully slow in spite of all those indexes there, on my comps); now I must adopt one of those comp spider programs in order to find all copies made into other UR files of law articles affected by changes; UR files are searchable by those programs, but with MI and with UR, it implied and implies a LOT of awful manual work.

And there's more to it: If you are maintaining (electronic) "manuscripts" in which those articles are of an issue, you must UPDATE your target files (and that's why I do all this, in my case), in the according way; if you had and have legal projects in which those articles are cited / of importance, you must decide IF you must update or not.

Very fortunately, and with a minimum of logical thinking, there is a solution to this: You must put your "manuscripts" in one sub-directory, your "current legal (or current legally affected) affairs" into another one, and your past legal stuff in a third one, and you must employ your comp spider to the former sub-directories only...


Similar consideration apply to all affairs where citations / usage of parts of reference material perhaps needs updating; this can be technical stuff e.g., but also, anything where your reference material is updated by new information or new work on that, while this updating is not automatically transferred / transferrable into your target material / in your target "files".

Thus, as a first step (in thinking, that is), it seems that SMOOTH INTER-FILE REFERENCING is the answer, where references would automatically / half-automatically update themselves, i.e. where on any changes on the original material at least, i.e. in our above-mentioned example, the law (resource) file(s), to any target of those links, everywhere, would be sent a sort of a message "original has been changed!" - more than just this automated "message" (which, if received by an archived file, will never be treated in any further way) could then be imagined to be done by automatic means, by applicating rules, e.g. "files representing this and that quality", e.g. not yet closed, in that sub-directory, having a certain code value in a certain field, or whatever.


It seems evident that for every professional 21st Century use, information must be ATOMIZED, in the legal word up to the part of the article (and that's what those monster databases out there seem to do)... but then, just the atomization, without an automated / "natural" agglomeration ("leaving things together that belong together", e.g. the multiple articles of a division of a law, and those divisions"), will NOT be helpful.

Thus, again, it does NOT seem to be important in which way those 500 articles of a law are stored,

BUT it is important that ANY SINGLE ONE of them can be addressed and incorporated everywhere (link, clone, copy, depending on your needs in that case; in fact, the difference between a clone and a copy is just the automization of the update (= from the user's point of view that is, from a technical point of view, a clone should be a reference, and a copy a double)), and without any technical difficulties then,

AND it is important that NATURAL AND SYNTHETIC GROUPS of them can be addressed and treated in exactly the same way: natural groups, i.e. law x part y division z; and synthetic groups: law x articles 3, 70, 620, 11.


So much for the reference material; now for the targets. Oops! As we all might have seen again and again for now, ANY try, casual or serious to any degree, to differenciate between reference and target / working material, has failed, and for cause:

In the law world, things seem to be rather easy for a start: Most material can indeed be sorted into the reference material, or into the working material group, but then, as soon as any commentary material, of sources outside or inside of your corporation, gets added, things become confused already.

And for any projects, ok, a filed project will then start to become reference material for new projects? Why only then? Even a living project and any tiny or larger parts of it can become valuable reference materials for other projects... even having started earlier than that source! And this applies to every matter in the whole big world, any fact is result and raw material at the same time, everything is entangled, and I have brought my law example for the only reason that in that mind, referencing was uni-directional at least, so that my line of thought wasn't distracted by the actual non-linearity of the processes to handle there.

This notwithstanding, in your actual work, you are searching for references to include in something you build up, so for almost every given situation, there isn't such a mess in handling your work, but the underlying information system must work in any direction in order for you to have some ease of management.

It's another application of what I said elsewhere, "In order for your work to be a little bit easy, the (unseen) information system's programming has to be complicated like hell - and the other way round." - in fact, many a times, I saw programs that were meant to "put it simple", but that in fact were only lazily (and cheaply) programmed, but did NOTHING to help users do their work.


Thus, it seems to me that we need, is not a monster file with contents: What we really need, is a big file - and which, in a big corporation, will become a monster file in itself, indeed -, that contains all those references, all these groupings, all those links, pointers, cluster maker strings and arrays, that allow us on-the-spot and prefigurable sortings, clusterings, adjunctions whereever and in whichever complexity we may need them.

Finally, it's not interesting if article 45 of law x is, physically, near articles 44 and 46 of the same law, BUT it's important that from anywhere where we need it, the 45, we have access to it, AND that we have also access to them, to the group, be it 44 to 46, or whatever, natural or scattered, and that natural groups (...files, you see, but not physica files anymore but files in your mind) ain't but cut up for some purposes, and notwithstanding stay uncut.

Finally, by referencing (and cutting up and regrouping references instead of physical files), we could get to an unseen liberty in information handling = INFORMATION REARRANGEMENT that will finally HELP US IN THINKING.

For the time being, ANY program, even those that rightly promise to enhance your way of doing things - and indeed, where would I be with 50,000 items on paper instead of being them in an ice database like UR? -, GET IN YOUR WAY AND DON'T LIVE UP... perhaps not to their promises since they even didn't promise what you promised yourself from their use...


But you KNOW it's there. You KNOW and bet your soul on it that the technical possibilities are there, outright... and those people who have such grandiose means... are just letting go, instead of working on those possibilities day and night?

Out of world-wide excellence, except for those net phenomenons (Google, Facebook, to name a few), it's just the iPad we are entitled to?


In fact, an information management system must be a SERVER, must establish OFFERINGS, instead of withholding relevant information from you... and at the same time, it must semi-automatically HIDE all nuisances from you, all interference.

There are 3 levels, as I see it, those things that you want to be before your eye, those that wouldn't do but bother you, and they must be entirely discarded, and those in-between, where you're looking... or even where you're not even looking 'cause you didn't think yet of it / them.

By inherent definition, those might-be things are, technically sorted in the last group, the "unwanted" one, thus, in that group, there must be sort of semi-automatic offerings of clusters of materials whenever you need them, and that let you quickly decide if parts of them are worthy of consideration or not, in this context, and then "quickly retreat", graphically-wise.

I insist, all those materials that are "discarded" in the current context, must be discarded in a way that they don't make you think, "could there be something I overlooked?"

It must be - become - the RICHES group, by way of that intermediary swelling, not the SPOIL group.


When I discard things that belong in (too lose) a way together, into another physical file, in UR, instead of just doing a hoist, it's because I want it REALLY discarded from my current task, and if I put it then into one of my sub-menus, it's because there, it doesn't bother me any more, but is AVAILABLE to my consideration anew when I WILL decide upon this - I think this argument is one of the secrets here. And then again, I look upon my (rather ugly) HK menus, and it gives me ideas; and any such file is discarded by one single key pressing again, incl. its sub-trees (as explained above, and by my key control-F4 on my Cherry).

Instant access to CLUSTERS of things - thus my problems with the search approach, and thus my not mourning MI's "search over all files" (except for the very special law updating thing) -, and instant discard, but, if applicable, with instant retaining of an / some element(s) of what you've just been seen, is it... would it be. (see below)


The Brain - all's there, all's connected, most of which is NOT available, and many, many things are much TOO prominent there; I'm speaking of the brain, not of The Brain, of course.

And that's it: Of course, The Brain has to strive to emulate AN IDEAL BRAIN, not those brains of ourselves we almost all are forced to endure.

PB's search functionality tries to overcome, but that's not enough.

In fact, PB does not do too much, at this time, in order to bring us that extra space into our thinking world, but it would be and unkind and outright false to pretend that it wasn't all there, in embryo.

All things being well considered, I'll I'm asking for is for PB to go three-dimensional, which at this moment it is not. I know those 3-dimensional representations, there are developments in some universities; there are not usable for any practical purpose (except perhaps for weather forecasting or what do I know).

But I'm asking PB for cutting us out every 2-dimensional slice out of such a system, by our choice of the moment, the cutting out being provisional, and then being REFINED by our further choices, be they partly manual, partly semi-automatic (= applying of filters, removing of others...) or partly standardized / stored, with variants, combinations, cutting out of subgroups...

PB is one of those few programs that could start to PROFIT from today's pc's possibilities -

and to make us profit from them.


I'm aware that current AI "developments" go into restriction, not widening, of your choices / experiences / informations, e.g. the infamous ongoings with Facebook's presentation automatisms, and I'm aware that the masses like to be led in that direction, to thinking uniformity - but there is big money to be made from individuals and tiny enterprises once words spread there's something pretty and intriguing AND totally exceptional, delivering outstanding results.

So another decade has been spoilt, and it's getting worse. But where could be hope of trying to catch up a little bit if not with The Brain? Who has been willing to go that far? But don't stop short of the target, there's way to go, I could join the venue for some steps together on that last but so far seemingly impregnable mile.

(Now, after some more hours, not a single UR item appearing in my brain3 plex yet, but that could be amended, easy. Even inferior programs manage to import from UR.)


Newsletter Signup  Newsletter        Visit TheBrain Blog   Blog       Follow us on Twitter   Twitter       Like Us on Facebook   Facebook         Watch Us on Youtube  YouTube       

TheBrain Mind Map & Mindmapping Software     Download TheBrain Mind Mapping Software