Why I Upgraded to UR 5.0

[mikeg]

By now it's obvious that key features/enhancements we'd like to see didn't make it into the last major version upgrade. I just don't think re-posting complaints, lengthy diatribes about the decline of software development, threats to leave UR, predictions of it's demise if expectations aren't met, etc. are tactics that will get us anywhere regardless of how valid or heartfelt they may be.

It's not that I disagree with everything you and others are saying. In fact, I used Info Select years ago, got tired of all my requests being ignored/rejected and dumped it as soon as I discovered Ultra Recall. In my view there were/are important differences with Ultra Recall. First, not just a few, but many of our requests have been implemented over the years. Other suggestions were at least noted and, in some cases, discussed and added to the road map. Last, but not least, there's still nothing better out there that I've found.

That's why I've tried to start a different kind of discussion. Not another invitation to complain, but an effort to reconcile how UR could arrive at another major version with more attention on Windows 8 compatibility than some of the most anticipated enhancements. My assessment is that UR is a great product in search of a sustainable market and expanded user base. From this perspective, kinook may be fighting for all of us by first waging a longer term strategic battle for profitability.

Customers sometimes engage in magical thinking that unseen elves should be working day and night for free or less than a livable income for years on end to satisfy their every wish and demand. With products that are relatively simple to develop and support, the labor of love model actually works sometimes--adding fuel to the fantasy. However, as product complexity and support demands scale up, a steady revenue stream is increasingly essential to keep teams (large and small) going and the electricity on.

We don't want to see UR go the way of Info Select and similar products. But what can any individual do about it? Probably not much, but I do believe strongly that piling on and going negative won't help at all. If nothing else, a little more understanding and patience is helpful and warranted IMO.

BTW, I'm predicting greater success for Windows 8 than some seem to expect. As for the painful desktop paradigm shift, I think there will be compromises/accommodations. In any case, the real show is mobile and Windows 8 will be a bridge for greater app integration across desktop and mobile platforms. Getting on a bridge to the biggest growth sector is a smart move. Maybe someday we'll have some form of UR on all our devices. At least the idea cheers me up. Go 5.0!

[schferk]

Mikeg, I don't have the slightest intention to argue with you, I just want to clarify.


First, we must distinguish between "features" and "homework", "homework" being demands that should go without saying since they are met by almost all contenders, but not by UR. Here, any sustaining need for catching up becomes "unforgivable" in the long run,

which means UR has - as TheBrain has, it seems - extraordinary features like indexing of external (!) pdf's or - as has got MyInfo, too, lately - really good (or in the case of UR, outstanding) cloning / hoisting (whilst cloning or tagging could also be considered traps since you'll never get your material out of such a splendid application once you'll have heavily relied upon these features there) -

and then such "primitive" annoyances (like the one I mentioned) that you won't find in most contenders, even the most minor (and sometimes even free) ones: Such inconsistency in quality within one program - from supreme to ridiculously poor - is laughable to put it mildly... AND puts off many otherwise would-be new (paying) users, too, whilst being a constant means of deception and irritation for the existing customer base.

Thus, from a commercial pov, leaving things as they are, on such "homework to do" field, might be qualified as sheer ignorance (if not arrogance).


Then, we must clearly see, "where UR is really great?", and that's undeniably in its robustness in handling real big data, i.e. piles of (not thousands, but) many thousands of items, often with .jpg's or with millions of characters - but then, the absence of a stable (instead of an ever-expanding-recollapsing) main tree makes it almost impossible to safely (let alone smoothly) manage such amounts of data, from a customer / user pov (hence my - necessary - going back to much lesser sw, relying on the Windows file system (incl. clones of whole parts, big or tiny) in order to get such a stable main tree / "super tree" of which I was and am in ABSOLUTE need.

Which is to say, UR, with its technical strength, "lures" us into this theoretically splendid system, but then leaves us alone with the "rising unmanagability" of our phletoric stuff once the number of our items isn't 10k or 30k anymore, but exceeds the 100k (and by far, in my case). Which is to say, there where UR could be, theoretically, outstanding, unparalleled, it literally betrays us in the seamy side of everyday use, for those of us who really depend on the strength of such an extraordinary system (whilst in MyInfo, e.g., I never dared putting all my things in one big database to begin with, as I then indeed did, in UR).

And, if one entity / developer is ABLE to code such a thing

(the difficulty laying in the DISPLAYED super tree being left UN-changed by any other "purely cosmetic" changes in subtrees (= expanding, collapsing, further hoisting), BUT showing technical changes there (= renames, moves, new creations, deletions), and, of course, any "manual" changes within the super tree itself (= cosmetic or technical), AND, at the same time, indeed reflecting all these changes within the underlying REAL super tree - and from a coding difficulty pov, it doesn't make any difference if that "stable super tree" is shown in tab 1 of many, or within an extra pane, of course),

it's Kinook (or perhaps the men behind TheBrain but who employ their sw mastering on other interests), so it's all the more so disappointing that Kinook don't do anything about this where their coding excellence would produce a product that would be unequalled, from a technical pov AND from the user experience - 130,000 items of knowledge or interest, anyone, AND a perfect system to handle that all in one place?

(And of course, I acknowledge that the charm of "all in one place" is big enough in order for me to put back my stuff into such a splendid system, since here, any renamings / dividings-ups / movings / different-re-clusterings of parts of such a system would be incomparably easier than they are in my current system with its cutting up of my data in diffferent physical files then re-combined by file system means.)


Then, we have Kinook's two-pier business, the programming environment for programmers, AND UltraRecall, and yes, of course I think they could do like thpse publishing houses do, cross-subventioning - we speak of 2 to 3 man-weeks (!) here - ok, make it a month, will all the debugging -, for such outstanding programmers as Kinooks' that is = Kyle himself I presume.

And again, the very first such IM sw on the market being able to SUCCESSFULLY handle big data, and for a - low! - 3-digit price, that would be quickly known by LOTSA people who are in need of such a system - but which has never ever been delivered, now that pc's have been around for some 30 years now (which is sheer incredible but has to do with the intellectual demands of such a groundbreaking feature).

(TheBrain is particularly bad with this; they show you monster maps, but try to manager such a map, let alone USE it - no wonder they heavily rely upon their search functionality... which, more than a decennium after the demise of Folio, does NOT offer the underlying semantic assistance system necessary though if you make people rely on searching.)


Just since you mention Win8 and "Winslates", please allow for my briefly re-mentioning the importance of response times on rather weak systems, optimized for long battery run (and minimized for heat), which constitutes another argument for my wanting better index (updating) management; and let me very briefly remind the importance of future stylus support.


It has never been my intention to "draw off" users from this superior but heavily neglected (whilst not abandoned) program, but in the end, financial and utilitarian considerations alone - "will any pampering measure create sufficient return satisfying me?" - decidedly can't be it.

Just have a look at a splendid feature of Treepad. Whilst any other (?) such program, at the very most, offers a static tree for web export, Treepad offers a living one, i.e. one that, in the web, you can freely operate, reading-wise, as you do with outliner trees on your desktop, i.e. you expand, re-collapse... It's a JOY to see, and to operate, such a superior toy that can become an equally superior attraction of your site, just by the joy visitors of your site feel when they manipulate your extraordinary tree - few of them will ever have experienced such a thing anywhere else.

Which means, excellence in programming also is question of honor: Show your programming excellence to the world, have your "products" / "children" splendent, even when for biological reproduction purposes (= here, for maintaining business), it'd be sufficient that they were ugly and dumb.


And as for our "patience" = willingness to pay for future goodies and joys, that would be greatly enhanced by a revised roadmap, with approximative dates within an added-on timeline.

[schferk]

Allow for some more musings about UltraRecall and the competition.

I

a)

One aspect is, UR allows for "putting all your stuff into one main applic, or reference it from there", which means, you search for pdf's and such - but which stay external instead of blowing up your UR database endlessly - from within UR, which means, from exactly that context within your UR "tree" where those external file "belong", let's call this "from their most natural context" - it's all about "having your stuff grouped together", of course, the basic problem of any IM.

b)

The other alternative has been described by me here, working from the file system, and having your "trees" there, as well as your "external" stuff; you group your stuff within your folder / subfolder system, including your "working machine" prime applic, and you'll have (an) external search tool(s) in order to find things, in your "external" stuff (pdf, Word, Excel, MindManager files, whatever), as well as in the files - multiple here, not one big one - of your main applic.

The disadvantage here is that your external search tools don't give you your search results fine-grained to the "item" as do the internal search facilities within alternative a, which can be cumbersome. On the other hand, with my current inferior main applic AO, I get, from one external (and free) search tool any search item within its respective context, even with accented characters (if I bother to input them in the 4 special chars form that the tool demands for such special chars - a little bit ridiculous but can be automated by a tiny AHK scriipt), and there are even two file managers (both German, both paid, both I own anyway) that allow for finding search terms with special chars by entering them the strict normal way (but they only give the file name, not the context). The first-mentioned special tool even allows for Boolean search (limited in the free version, anything you like in the paid version).

The same facts apply for external searching within UR files, but of course, here you don't need external search tools except for searching different UR files at the same time, but given that UR searches are index-driven, you better employ an AHK script for doing consecutive internal UR searches within your different UR files, instead of waiting a minute or so for the results of your external tools. So, the instant-avalability of your UR search results is another advantage of a system like UR over an IM file system set-up relying on external search tools that often don't deliver immediate search results for the special files of your main applic - for your "external" files like pdf, Excel, etc., the problem is much lesser since the standard (indexing) search tools index these files with no problems.

c)

Within both above-mentioned systems - reference your external files within your main applic or cut up even your main applic into multiple files -, you'll get problem where to store the external files, i.e. all in one big folder, or fine-grained within a folder-substructure, perhaps with several levels of subordination. The "one big folder" quickly becomes un-manageable from itself, and any occurence of not immediate referencing of such a file within your main application will result in a more or less "LOST" such file, i.e. accessible, by chance, by searching, but not accessible anymore by taxonomic means, i.e. by browsing a file structure or by browsing the corresponding context within your main applic.

d)

So many users, at some time, get away from the "one big folder" for external / referenced files, and try to double a structure of subordination within their main applic (big outliner file or multiple outliner files) AND in their file system. Both solutions amount to almost incredible sync problems, since in your main applic, you always shuffle around things, and rightly so, it's MEANT FOR BEING PLASTIC, that's why it enhances your thinking.

Thus, I don't have to tell you that any 1.1 doubling effort between your "storage machine" and your file structure for referenced files is bound to fail within minutes, if you ever try to establish such a duplicity, and any effort in trying to automate this is bound to fail, too, not for technical reasons, i.e. because your scripting capabilities perhaps are stretched to their limits, but for the simple reason that even if you try to sync these structures, more or less automatized, you won't overcome the problem that your external files, most of the time, will span MORE than just one fine-grained item / group of items within your main applic - because these external files are RESULTS of your work, so automatically spanning several such "think / detail / material" clusters in your main applic = "work machine", or because they are IMPORTS, and of course, the respective authors of these had thought and written in other contexts and context clusters, as you do, within your "writing machine".

e)

Thus, do NOT try do store such external files in multiple instances in order to overcome the problem detailed above, since having TEN OR MORE such .lnk files or whatever (the Windows system offering more than one technical solutions to the cloning problem) of an external file would be as mad, as having 10 or so clones of an item of a subtree within your main applic (the seemingly ONLY exception to this statement would be to clone specific subtrees of legal material or previous projects, to NEW "cases" or projects, in order to serve as a a QUARRY there, within your main applic, and your external files, but even then, you would not use clones, but COPIES!). And as for the synching of external files or clones or copies of them whenever you move an item or regroup a bunch of items within your "workspace", don't even think of it: Technically, of course, all this would be possible, but it would take 90 p.c. of your time to sync your things, there wouldn't be any time left in order to PRODUCE, or even to gather, some material.

From this point of view, it's simply dumb to ask for better means of synching on a fine-grained level, and the solution lies within manual synching, but on an intermediate level, and to have indicators there which say, "attention, if you move this, have a look into the respective file folder structure".

...

[schferk]

...

f)

Hence my short description what I do here, or better, what I've done here, lately, in order to get these things straight. It was two long working days hard labour, but it was well worth the effort, and I'm positive to have found the definite answer to this problem, independently of your choice of a big-file system as UR is, or of a distributed system as I'm using at this very moment.

1. Have a global inbox, but without any try of structure, just an inbox from which you export things as soon as possible.

2. Don't export, from there, by scripting, to hundreds of target files of hundreds of sub-structures - technically, UR would allow for this, but you would get response time problems. Have intermediate sub-structures (which in my case are intermediate files that are always loaded and open, and which in your case (UR) would be LIGHT first-level subtrees, within the very first positions of your global tree, and that "only" serve for holding intermediate structures that double the highest levels of your "real", fine-grained structure.

3. Have macros / scripts in order to access the respective "other" item / tree heading (i.e. from your intermediate structure to the "real" sub-tree and vice versa) by just ONE F-key (or by shift-enter or whatever suits you).

4. Do your thinking and actual conceptual work within this "unnecessary", super-imposed intermediate structure. Access your "real", fine-grained structure only if you must look up details, or for real editing work, etc. - Try to "EXTRACT / DIGEST" any intermediate result / thought / aspect / whatever from your "bunch of material" into this doubled intermediate structure - which also means have the least possible direct access means = links, clones, etc., from your intermediate = working space into those "depths" of your material: Except in special cases, don't clone detailed items from there here, but write a sentence or two here, summarizing those details from there.

Of course, what I'm trying to convey here, is a means of avoiding / minimizing the "lost in hyperspace" phenomenon which will quickly get you even within your own parallel system (which is no hyperspace in the scrict sense of the term, but just a tree grown much too big - and the same applies to distributed systems like mine, the technical details of the possible cutting up of the too big a tree doesn't mean anything on the conceptual described level).

5. In fact, what I'm also trying to establish here, is not only "a better view", "a better point of view" in order to not get lost within the details (but only access them for what they are for: referencing, usually, and only in exceptional situations, development WITHIN those details), but especially and in the very first place, AN INTERMEDIATE LEVEL OF RESULTS - FROM WHICH TO WORK FURTHER ON. (People who will have followed my writings in the askSam, MyInfo, outlinersoftware, and UltraRecall fori, know that I'm after both aspects, the organizational, AND the thinking-enhancement one.)

6. Now coming back to our external files (be they self-produced or imported stuff): Synch your folder structure by the applyable (!) main items within your intermediate = working structure. There is another aspect: Since for working on your stuff, and for adapting stuff to YOUR way of organizing, of thinking, of doing things, you split external things into different context as they are presented by those above-mentioned external "authors", just have external things grouped into the "common denominator", which are hopefully represented by your main headings within this intermediate structure.

For important things, even go so far, if needed, to adapt your main headings to enclose such material, i.e. don't be too "original" here, but follow the ways "your" material is commonly being presented by third parties, in order to have a viable main structure; this advice coming from somebody like me, be sure this is good advice - I'm perfectly conscious of the outrage of this coming from me.

7. Also, don't be afraid of some "comment" items there: Why not have an otherwise empty item but which title says, "for this aspect of the subject, also see item xyz" - this will bring an immediate solution to all these problems "car insurance under car or under insurance?" - in all my system with over 100k of items, I only get perhaps about 200 such "SEE items", all sub-structures together (and I spoke about the exportability problem, set up by technical cloning which hence I avoid).

8. As said before, you do NOT need a corresponding sub-folder (i.e. for external files) for ANY main entry in YOUR system, BUT vice versa it's not: For any external folder, create such an item, even if it doesn't contain much - if there is nothing for you to have about it, in your "internal" system, you will have badly created your external-stuff folder in the first place.

9. So, what IS your intermediate, and your external folder structure? In my case, it's about 10 files (business 1, business 2, private stuff, law, tax, social security, "general stuff" including organizational things with scope over several of my prime groups, and more), with each about 6 to 25 sub-files (which in the "private" group, would be spouse, girlfriend 1, girlfriend 2 (yes, that was a joke), health, sports, literature, music, other pastimes, recipies, drinks, our pet (see the following!) - so, in my "big folder", I got 10 corresponding sub-folders, with each about 5 to 22 sub-folders on that further level: As said before, I do NOT create empty sub-folders, just for the "pleasure" of there being a 1:1 synchronicity even when there will never be any external file within the corresponding sub-folder: I just create those file-system sub-folders whenever the very first file to be stuffed into there comes up, or whenever I think there might be files to be put into there some day.

10. As for the "pet" example, I'll have a folder "allmystuff/private/pet" - but, you see, I will NOT go so far as to have sub-folders there, for pet's health, pet's rise, pet's vaccinations, and so on, no sir, I'm perfectly happy with "pet", and then, there are some perhaps 15 external files, pdf's, invoices as pdf's, whatever, and whenever I want to access some of those files, I do NOT do an "enter" within my "private" file when the selected line there is ".Pet" - this "enter" there would immediately open my AO file (= your UR sub-tree) "pet" (by scripting - any "enter" when a line beginning with a dot is selected, opens the according file, would "open" - why not automatically in a new tab? - your corresponding UR sub-tree), BUT I do a SHIFT-"enter" (i.e. on such a line beginning with a dot), and my system "knows" I want to open, in my preferred file manager, the "pet" subfolder - and it does it for me (I even have got a rather elaborate script that's able to just OPEN that folder / tab in my file manager whenever it's already on a tab there, or to create a NEW tab there in order to show the contents of that "pet" sub-folder, if up to now I did not yet open that sub-folder on any tab there - all's fully automatized. And yes, lately I got a second, identical screen, and thus, I get my AO stuff in front of me, and the respective "external", additional files listed on my secondary screen, and while I'm writing this, it comes to my mind that of course, a simple ENTER on a line with a dot should do BOTH, open the corresponding AO file, AND open the corresponding sub-folder for any additional stuff. (Or do it by control-enter in order to not automatically replace any such additional stuff on your second screen when you don't need it.)

11. So you see, the fact that my "pet" sub-folder does contain some 15 perhaps rather disparate external files, doesn't bother me in the slightest, and this way, by doing the synching on that intermediate level (i.e. my AO file level only, i.e. NOT WITHIN my further sub-grouping within these separate AO files), I don't have any such problems by external files spanning several of my subjects there, since I do NOT have the ambition to fine-grain external files beyond reason. In your (UR) case, you would deliberately refrain from creating sub-folders BEYOND your SECOND-LEVEL MAIN HEADINGS: d:\allmystufforwhatever\private\pet, and not beyond this.

...

[schferk]

...

12. If really an external file spans two such second-level main headings of your stuff, put it into it's "natural" context = sub-folder, and create a copy or clone in its secondary context - but also, if the file is called "afile.pdf", rename it "afile(inPetsAndAssurances).pdf", so that on any further moving / changes in grouping will easily remember if you do must do anything about your current storing of this file or not; the same goes for any file names that you want to preserve in their original, but that ain't "speaking" enough: just rename them "originalfilename(RealContent).extension", and you have your cake and eat it though: Preservation of the original name AND knowing what the file is about, without having to open it, a week, a month, or years later.

13. So, in your intermediate UR structure (= within your UR "private" structure, you must KNOW which entries there are links to "specific" (not files in your case, but) detailed substructes, and which entries are just items there, in order to remind you of something, or substructures of current work there, or anything else. My solution, as said, is that "initial dot" encoding, which is checked by any , macro I might trigger, and processed accordingly. For other special entries, you could use initial semicoli, or whatever char you will not use as leading char in normal contexts.

(In my case, it's not only the leading dot, but also, since I use my very special encoding system, not "Pet.ao", but "pp.ao", for "Private - Pet", my macros also check for the very first SPACE in such special "dot leading" entries, i.e. my entries here would not be ".pp.ao", but would be ".pp.ao - Pet" in order to remind me of the meaning of such entries whenever I don't access them but rarely. - But the opened file would be "pp.ao", and the opened sub-folder would be "allmythings\p\p" - that's one of the beauties of my system: The very first char (after the dot) "tells" my script which is the name of the intermediate folder...)

14. Also, you should know which such entry is "just" a special file / subtree, and which are the ones where you also have got external stuff, i.e. which ones also correspond to a sub-folder. In my system, I simply use bold formatting of these last entries (and indeed, most of these ARE bold-formatted, rare are my AO files without a corresponding sub-folder for external material); since UR does not allow for formatting entries, just use special flags.

15. You will certainly not be astonished if I say that of course, UR should allow for such an intermediate structure, e.g. in tab 1, that does OPEN such "." items in further tabs, but that does NOT FOLLOW any expansion / collapsing you'll do within these further tabs, and which it unfortunately currently always DOES.

II

a)

UR pretends, the search of the ultimate IM system has come to an end, with your coming across UR; that's wrong: UR will indeed be the ultimate such system the day it'll integrate such an intermediate system as described here, including the demand set up in I f) 15 (AND when the "homework" as defined by me in my posts above will have been done, that is). It's simply necessary that an IM systems offer a superstructure, and it's worthwile to have EXPLAINED its use to you, finally.

b)

UR is the ONLY such system (within this group of systems that don't cost many thousands of dollars) of my knowledge that offers a viable transition from "just one lone single user" to effective group work. That's why I hope UR will come to see that the fulfilling of my demands will push UR into a class of its own, justifying any development work I'm asking for.

That's why I'm writing here, and not in an MI forum, e.g. As for the competition, there's IQ, which lately, in the outlinersoftware forum, got exactly that kind of negative impression that I'd got without first buying, and that always had put me off from that sw not living up to its promises. NO such promise is fulfilled anywhere, for the time being, and certainly not by TheBrain, far from it.

UR is closest, and with some efforts, could become what a program like AS had been in the Nineties, something really great.

That's a challenge kinook would be very well advised to take up.

I'm willing to give advice, contacting me is easy:

frjansen@yahoo.de (fr is for Fred).

And if it's worthwile, I'm willing to share my scripts and to develop further ones.

Getting my stuff back into UR would be a thing of some minutes (getting it back into AO was the problem, not the other way round) - but as it currently is, I'm not updating to version 5, even though I'd be more than pleased to do so in the future, when UR becomes as outstanding as it should have been a long time ago.

[schferk]

16. Whenever you get too many external files in such a sub-folder, consider dividing your UR subtree as well, to get another sub-folder, too. Or, if these files are rather similar, just numerous, consider getting "divider files", empty .txt files that you name "ez_____.txt", etc., to make separator lines for your files, grouping your files by renaming not by "originalname(content/coding).suffix", but by content/coding.originalname.suffix. And yes, why not doing sub-folders of an additional sub-level in SOME cases, if too many similar files splatter around.

Just avoid, really avoid, a deep file structure mimicking your item structure. Have a rather "high" item level down to which you "synch" your AO / UR / similar material and your external files, and then do NOT try to synch any further down.

(To be continued with other initial omissions or YOUR ideas, ok?!)

[schferk]

17. The inbox thing. As said, you have a simple inbox, no subdivision. Then - and I totally missed this in the above description, any dot item within my about 10 main files, i.e. that "second level", but "second level" not as files, but as special entries within those about 10 first-level groups, also act, not only as triggers for the respective AO files / UR sub-structures, and for the respective sub-folders for external files, but also as TARGETS for my distribution of any item from the inbox, which means, I press one key, then I get 1500 msec in order to enter the target (in our above example: pp for pet), then my system puts the current item (together with any sub-items it might have), from my inbox or from any other place within my system, as a sub-item of such a dot item (and re-collapses the according dot item, of course), and goes back to where focus where before this move - the whole process doesn't even take one sec.

This way, I have one general inbox, my file "0.ao", and about 150 specific inboxes, those dot entries within my about 10 first-level "group" files / your respective dot entries being sub-entries of your about 10 or so first-level main / grouping items; it goes without saying that whenever I get many similar new things / web clippings / whatever, to be set up as new items in similar places of my AO system, I directly put them under such a dot item, instead of putting them into the general inbox. And, in order to not only have the general inbox, and the rather specific dot items like ".pp.ao", I not only have got, within my "p.ao" file, all these ".pa.ao", ".pb.ao", ".pp.ao" entries as described, but I also have there a ".p.ao" entry, which obviously does NOT serve to trigger the display of the "p.ao" file since it is the p.ao. file itself (but which serves to trigger the corresponding "p" sub-folder, of course), but which serves as a perfect TARGET for any item I want to file away not yet under the specific sub-group (= into "pp.ao", in our example), but just under the more general "private" group, in this example. The purpose of this intermediate "default" target is, of course, to put many similar things into the right "super group" group at least, when I don't have the time to reflect upon the according specific group into which it belongs. As an example, I've got a super group "c" for "computing" (incl. programming, net, and many things, and of course, ci = information management, cf = file management, etc.), and whenever I don't have time to decide if some items goes into cf or into ci, I put it into just c. Afterwards, when there are too many sub-items within c, that are awaiting finer sorting, I then remember which details go into cf, which one go into ci, and I sort these items beneath ".c.ao" within my file "c.ao", into those more specific file groups ".cf.ao" and "ci.ao", with the same sort command which I use from my "0.ao" file to sort items from there, be it in a specific or in a more "default" way, depending on my time "to do it right" or just in an approximative way: In ANY case, this system allows for emptying the general inbox in speed time, and since I WORK within these about 10 intermediate structures, most of the time, and not within the specific "pp.ao" files, any new items being sorted into these ".p.ao" or more specific ".pp.ao" grouping items, are at my fingertips when working, which entitles me to effectively work, with an empty general empty inbox anytime, but with having any new material at my immediate disposal whenever I do real work, even when I didn't have had the time to really sort out my new material, into the various really specific files ("pp.ao" / UR substructures (= the real UR "pp" substructure, instead of the ".pp" context within your UR "p" substructure)).

Remember, just one F-key brings me from this ".pp" context right to the "pp" file / tab, and vice versa, and you'd do the same within UR. Also, this way of working - working not only / not predominantly within the specific "pp" context, but within the broader "p" context, does greatly improve your work results, since the utmost specific are at your fingertips, and at the same time, any OTHER material you'd be well advised to use for your writing / drawing up, is at your fingertips, too.

Unfortunately, with the current state of affairs with UR, it is NOT possible to have these advantages to their fullest, because the ever-waving expanding and re-collapsing of your intermediate sub-tree, whenever diving into details of your specific details sub-tree, will create chaos in your intermediate part ( in my examples, the "c" or the "p" part) of your UR system. Hence the necessity, for me, to flee the UR system in its current state of development, into a system where the division of these different levels of specificity - the "c" vs. the "ci" or "cf" level - into different files AVOIDS chaos in the "c" tree when you work upon your "ci" or your "cf" tree.

Hence, the absolute necessity for UR to amend this problem of "waves" created by display movements on deeper levels, in order for UR to become a professional IM system.