1)

From my research, text-based outliners don't have clones (automatically updated "copies" of items in the tree; in the same tree=the same file) - none of them has, in fact.

And then, database-based outliners SOMETIMES have clones, but not all of them have.

Those that have (sometimes in their best versions only):

Ultra Recall, MyInfo, The Brain, InfoQube, ConnectedText, MyBase, Zoot
Surfulater, ADM, NoteCase, Whizfolders, InfoSelect
Chandler, KeyNote NF, OneNote, Leo, Treepad
Smereka TreeProjects (?)

Those that don't, in spite of being database-based (but perhaps that's my misconception in some cases):

RightNote (SQLite! has something weird, wsp called it a "virtual notes feature", but I checked and checked again: NO clones! For its other qualities, it's a shame it has no clones, and will probably not introduce them, because of this other feature by which they errouneously think they can avoid them)
TreeDB Notes, Tree Notes, Evernote
MyNotesKeeper (but has "bookmarks", within the same tree)

Corrections and additions would be welcome.

(As said, NO program (of my knowledge) offers clones maintained in different databases (even within the same directory, so in theory at least, updating should be possible). This means cloning items will blow up your database, and you'll get in troube when you try to cut off some part of the database to make it separate database: Another obstacle in your trying to separate the things you work on, from your reference material.)

2)

It should be said that some (again, database-based only?) outliners provide cross-referencing to individual items, even in other databases.

This is an obscure feature of Ultra Recall: You can set a reference to an item within another database.

In MyInfo, you can set references to items within the same database only (?), but also to particular paragraphs of them.

I don't know of other outliners providing linking to particular items, let alone to particular items in a different database.

Some additional info available here? I think such links could replace clones in many instances, and especially so for links to different databases. (But it should be done in a better way than UR does it: As soon as you didn't do it for one week or so, you'll be completely lost, it's really complicated to do.)

Btw, in TheBrain, it does NOT seem to be possible to link to an item in another "brain", judging by a post to this regard.

Anyway, this seems to be a subject of real interest, if you're serious about information management.

Off the top of my head (I'll follow up if I think pf more with the help of daylight):

(1) Brainstorm, "namesakes", text-based, apparently it was David Tebbut (Brainstorm's inventor) who came up with the iea before Dawe Winer, but the latter beat him to it with implementation.

(2) Evernote, Surfulater

I can't recall the exact context of my remarks about Rightnote now, but I think what I said was that RN allows "virtual notes," an attractive feature apparently inherited from Keynote. The virtual notes are not clones or cross-references. You put a sort of placeholder for an external file (say a text file or a PDF) in your tree, and from that point onward RN treats it almost as if it's a part of your database -- you can do searches and alter that file from within Rightnote -- but the virtual note is not actually absorbed into the database and does not increase its size. It's a clever idea.

MyInfo does cross-referencing at the paragraph level (why don't other programs?), but cross-references won't work across "topics" (i.e. databases). Its clones work very well, but again they are restricted to a single database.

Evernote abandoned cross-references when it moved into the cloud in 2008 -- along with a lot of other strong features -- but fairly recently, I think within the past year, cross-referencing has triumphantly returned. Since EN puts all of your stuff into a single database, you don't have to worry about the potential problems of cross-references from one database to another. I must say that the reappearance of cross-references at last has made Evernote usable for me again, because I depend upon them for organizing and outlining.

Bill

WSP, I'm not sure I did NOT mix up those "virtual items" you spoke of, and what I remember from RightNote. I last time checked it out some months ago (cannot trial such things anytime, without having a valid image, since by trialling in my running system, I would not have any chance to trial again, later on, even if I just install for one day), and did not find clones (as you say, in MI 6.0 now, clones are very well implemented, as well as in UR, for example), but something "weird" - did not take notes about it, which I should have.

Now, I try to remember from their web page, and it's this thing, I think, "Any child notes of a folder note automatically get tagged with the note’s caption. This allows you to organize and tag at the same time.", together with the functionality around auto-tagging, and I even remember them saying, on bits, something like - I hope I don't falsify their opinion here - "our tagging system comes so advanced that we don't even need clones anymore".

Which is not right, of course, since outliner users don't generally want the children tagged with their parent's name, but perhaps a clone of the parent itself, in some different content, and then he will access those children by selecting the cloned parent.


As we both say, MI cannot cross-reference to different databases. But even if there is no internal functionality for it, intended by the developer of such software, it might always be possible to access individual items within such a database from the outside, meaning by some "open database abc" command line command, and then put an identifier behind this general command (and since this would be a command line command, it should also be possible from within a database of the same kind, meaning there should be some "trick" for this, even for MI, and if there is, it should be possible to trigger such a command from within any other program, and also from within MI, as an example); I suppose that in the end, for most of database-based outliners, this should be possible, with a little help from the developer, or with having a good look into the used database.


Surfulater: Since surfulater is a web-grasping specialist, this feature opens up some additional use for it. As said by me in another thread, most people too much try to enter whole web sites into their outliner, from my point of view, blocking, by this wish, too much of the respective developers' developing effort in their trying to implement correct processing/translation of the last developments in web sites' technology.

But why not have category and sub-category items in a "specialist" like Surfulater, putting all your web sites into that, instead of your outliner, and make the corresponding category and sub-category items in your outliner links to the corresponding Surfulater items?

So the very next question would be, is it possible to have such links into Surfulator's immediate competitor, WebResearch (from macropool.de), too? If not, this would make Surfulater (at least in its current desktop shape) THE outstanding application for your web sites collection, with direct access from your respective outliner topics.

I believe WhizFolders has a good linking system. It seems pretty comprehensive.

For an example, a UR item would be called by this command, from ANY application that allows links in general, for example web links:

ur://C:/UR/i.urd?item=3019,1000

And there is a command in UR that shows the needed identifier, for your current item. So once you know how the outliner accesses its database, a similar scheme should work for any database-based outliner.

Even for people doing their work in MS Word, or within a 1-pane outliner, or any other text processor, this should be of high interest, since they could store their reference material in an outliner, and continue to do their work within the application of their choice; the same, as mentioned, for their web sites, stored in Surfulater.

And even to Evernote... Well, I would have thought EN is really well (as wsp explained here some weeks ago) for enterin info into your system, but then, that most info there will ultimately be stored in your outliner or such, so the usefulness of cross-referencing EN content is not immediately obvious. But then, if it's easy to do, you could leave your reference material within MULTIPLE EN files, with just the indentation levels they allow (is it 2 or 3?), and just start another EN file, where in your outliner you would have started another indentation level, not available in EN.

In such a way, you could even avoid the transfer process between EN and your traditional outliner, just do one EN file for any new broad subject, then do cross-referencing between those files, and between your "main application", be it EN, be it Word or something else, and your numerous EN reference files.

(If there is such a possibility, with their subscription model, that is. But anything should be welcome that is able to break up your 10,000 items data monster, I think.)

"I believe WhizFolders has a good linking system. It seems pretty comprehensive."

Please forgive me if we share this concept, but we are always speaking of links to OTHER databases, right? Not of internal linking only, but of linking, from the outside (for example one other file of that outliner, or a third-party application) to a particular item within one other file/database of that outliner?

If you only mention internal linking capabilities (which are far from useless either), please say so.

The same applies to Surfulater and EN: We are well speaking of external linking to individual items, are we not?

22111 wrote:

The same applies to Surfulater and EN: We are well speaking of external
linking to individual items, are we not?

....................................................................................................

- For Surfulater: yes. You can create links as Plain or HTML. Here's some info from the excellent help:

Plain: sulkb://kb=MyKnowledge,Fid=5040,Rid=5068
HTML: Atlantica Blended Wing Body

Handling moved and deleted articles

External Links specify the knowledge base name, the folder the article is in and the article itself. If an article is moved from the folder it was in when an external link is created, to a different folder, Surfulater won't find it in the original folder. When this happens, Surfulater searches the knowledge base for the article and displays it, along with a pop-up tip informing you that it has moved. You can update the external link in the source document to correct this if you want. If the article has been deleted, a message is displayed indicating the article was not found.

Locating the Knowledge Base

External links include the name of the target Knowledge Base, however they do not include the name of the folder the KB lives in. This enables KB's to be moved to a different drive or folder and links still work.

In order for Surfulater to locate a KB specified in an external link, it must be in one of the two known locations: View|Preferences|Folder for Knowledge Bases and 'Send To' folders or My Documents\My Surfulater, or already be open in Surfulater.

....................................................................................................

For Evernote: Yes. However, as noted by Bill/WSP, EN only allows one database (per user account). So one cannot link from one database to another, but of course an EN database can have as many notebooks as you'd like, which for me is just as well.

check this thread for linking to an item within a database:

http://www.outlinersoftware.com/topics/viewt/4706

1)

Jim, thank you very much of the link which I hadn't been aware of.

Alexander, EN not bad, Surfulater perfect, I'm much more interested in this program now than I previously had been!

Brainstorm: As Alexander says, possible also (I never touched Brainstorm, since it's plain text only; I had assumed that the very first practical implementation of clones had been within "HyperCard", but that was simply my first encounter with that concept)


Now for the content of that OS thread:

For UR: described above; the identifier of the respective item can be copied from the attributes pane. And also, Alexander described it like this (so this "obscure UR command" is known better than I had thought here):

"Select the item you want and click (on the menu bar) Item / Copy Item Command-Line; you may have to fully expand the Item menu the first time. There’s also keyboard shortcut Ctrl-Shift-I. You can then paste the result to the program from which you will be linking from. In UR, you can also create a Windows desktop shortcut to the item through the same menu."


For Connected Text: It's possible even from the outside, as Dr Andus points out at page 2.


TheBrain was positively mentioned there, but "from brain to brain" (and thus, to do it "from the outside") does NOT seem to be possible, or simply the posters here did not find the "trick" to do it, and the developers did not share the info:

http://forums.thebrain.com/post/how-to-link-to-another-brain-6076028


OneNote: Possible even from the outside, as Steven Zeoli points out.


Surfulater again: On page 2, Dr Andus gives the details for it: perfect!


"Mindjet, EssentialPim and GemX do-Organizer can do that with their own protocols" ???
I don't know about EssentialPim and GemX-Do (which is a "very colored" program to put it mildly), but for MindJet, I'm in doubt, since I tried in vain (but then assuming, perhaps too soon, it wasn't possible); what MindJet can do, is linking to other files in general (but not necessarily to specific items in other MindJet files).

What I CAN say is, EssentialPim is buggy, and the developer doesn't answer mails asking about those bugs even when described in detail, and there is no forum or such, so EP is one of those program where there is no help but the one the developer is willing to give out; for a perspective, UR's "help" might appear very cold, but whenever a bug is brought to their attention, they see to it immediately, sort of "question of honor" to them.


Zoot: Details on page 3: perfect! on top, the details, and then Dr Andus makes a comparison, and then Alexander (pages 3 and 4) gives a hint what to do for UR's absolute links.


So much for my thinking I had brought up a new subject! Sorry!


2)

As for relative vs. absolute links, I think as soon as you do the link from the outside, not from within one outliner file to another outliner file of the same outliner program, this problem will always present itself? Or do I think too short her?

I could think of not activating those links directly, but intercepting them by a macro, which puts the content of a variable into the link path, but this would mean of course that your "system" only runs in the very specific environment you create for yourself.


3)

Yesterday, I said, "outliner users don’t generally want the children tagged with their parent’s name, but perhaps a clone of the parent itself, in some different content, and then he will access those children by selecting the cloned parent".

Well, I think that's right, in general, but there is an additional problem to referencing to / cloning a sub-heading within a different context:

In many circumstances, you would NOT want to have ALL those children put into the second context, but only some of them. Let's say you have a sub-heading with 12 children; you clone the sub-heading. In its original/main context, all these children are useful, "in their right place". But in this secondary context, only the children 3, 4, 6, 8 and 10 will be of any meaning, and in a possible third context, it's only the children 2, 4, 5, 7 and again 8 and 10 will be useful.

Nobody ever has found a valid solution to this problem; sometimes, I cut up then this 12-item sub-tree into two sub-trees, for "of general interest" and "of interest only here, but not in the second context I need the other sub-group, too", but this is far from elegant, a lot of fuss, and certainly not a valid approach if this occurs often.

On the other hand, an automatic tagging system like in RightNote would than tag all 12 children identically, and then, you would de-tag some of them, which even creates more chaos than my approach described above.

Selecting just some of your items here, in order to do individual cloning to that second context, does not seem to be smart either, since in many cases you will add additional items, later on, to the principal location, and what about those if they there, you cannot see which items there you will have cloned to a second location (let alone a third one), from there, and which ones are unique there?

In UR at least, the clone, AND the original item are assigned a special icon, indicating it's cloned state, so this is a big help here, but only if you clone individual items from this sub-heading; if you clone the sub-heading instead, there is no such indication for individual items then for their respective usefulness within that second context, so when doing work there, you always have to check for their content - you could do something about their respective titles, of course.

In this respect, it is worth mentioning that in MyInfo and some time ago at least, additions/changes to the child items of such a cloned sub-heading were NOT replicated to the second location of the cloned sub-heading, when in UR, this replication of sub-tree is realized in all circumstances.

I don't know to what extend other outliner users face these problems in their practical work, but I regularly do. Let me give an example: You have lots of (legal) rulings, sorted in groups and sub-groups; then, you have legal cases. Now, in many such cases, a whole group of rulings apply, but with the exception of some of them, exactly as in the above example.

If only 2 or 3 out of 12 or 15 rulings apply, you would clone those individually, but if 6 or 7 out of 12 apply, you would be tempted to clone the respective sub-heading, and then you have the above-described problem that some of its content will NOT apply.


Perhaps these considerations will give way to a better understand why outlining PLUS TAGGING might be a very good idea for some uses, or outlining plus simili-tagging by putting keywords into "attributes".


If somebody has some ideas/experience on this, I would be very thankful for him to share them.

Two ideas:

Even when the "original" item gets a "clone" icon, as in UR, that item does not indicate "cloned to location a" or "cloned to location b", which is a problem whenever you clone standard items, several times, to different target locations, so what is missing here, is some function that within a sub-heading, will show the respective cloning target locations in a table or such, but since they might be multiple, it would perhaps be preferable to have a function, not starting from the "origin", but from your respective target, and which shows, wherever you "are" within your material, "this item/heading has been cloned to the target of interest", this being some heading/sub-heading of its own. This means, a function would, when triggered upon a certain heading anywhere, then check for and visually indicate any other item/heading, wherever you are navigating, if that particular heading/item has been cloned to that heading and its subtree, where you triggered the command, for example by displaying those specifically-cloned items with a colored background.

And also, there would be an additional function that, within such a setting "selected target for gathering elements", a simple click/key pressing would clone any other item/heading (together with its sub-items) to that particular target, meaning in this technical situation of your outliner, you would not have to identify the target again and again for such cloning.


Second idea: It seems obvious many problems arise from the fact that outliners today mostly don't display two items at the same time, and even those that do, will not display but the content pane of that second item, but not its "core" identifiers: no subtrees, and no title-as-target, to which you could "send" things; this latter functionality, though, would be rather easy to implement, in comparison to "full" display of a second item.

If you had two subtrees, side-to-side, one where you are navigation in order to gather things, and the other displaying your target to which you are collecting things, instead of having to switch back and forth for this, all these tasks described would be much less cumbersome. But as said, this could be facilitated by a function "set this item as collecting target", and then making the outliner indicate visually which items "are already cloned to that target", and which are not, and this would resolve the above-mentioned problem of not knowing, by a simple "is clone(d)" icon, if some item has been really cloned already to your current target, or to some other location within your big data.

I think this would be one (of certainly several) valid solution to the problems stated above.

In another thread some days ago I said, often, it's the missing coding capabilities of developers that is responsible for missing functionality, but here we have clearly a case of first having to develop how to realize a function best, and then the actual coding would not be so difficult!

It goes without saying that within your file system (physical as well as on your pc), the problem is similar: You have grouped reference material, be them reference material from the beginning or just old projects from which parts then (should) become reference material for new projects, and again and again, only sub-groups of files (and then, of content within those files) will "apply", when other content or other files within those sub-groups / sub-folders will not be relevant. The only conceptual difference to what is within your outliner is the possibility, in your outliner, to better fine-grain those different contents, meaning you cut them up into more pieces, in order to make one bit apply and another one not, when in files, there will be more disparate content, in most cases, than in one outliner item (ideally).

Technically, to replicate such "sub-groups of originally connected items" on the file system level, is possible even if you try to do it all within the file system, for example by .lnk files, but perhaps there is a real interest instead of doing it by database links, which means by a document management solution but which should have such "gathering" functionality for fast building up new compounds from spread, "old"/standard data.

For physical files, some makers offer "thin folders", which at first sight, is a very good idea to better realize such multiple combinations, but then, most things do have some "natural context", in which it's very handy to "have it all together" in sort of a "standard combination", from which you deviate only scarcely, depending on the respective context, and this invalidates this "thin folder paradigm" (I'm not speaking of separate customer files here, of course, but of files to be combined in order to be useful).

So we have a disruption from physical files to pc, but it seems that except for real expensive software, our pc tools don't make enough advantage of their additional possibilities-over-paper yet.

I once considered cloning a real paradigm shift, and it is, considering paper files, photocopies, and such, but their current technical realization is too cumbersome yet in order to become really helpful for "project work", which is NOT about "putting fome file into several categories" only, but which is of partial combinations of single items, and of sub-groups of "connected" items, in endless variants, so I hope some ideas above, or additional ideas posted here, could find their way in some outliner or another.

I cannot resist to give this citation:

from http://takingnotenow.blogspot.be/ from "Blumenberg on Luhmann, Part II":

"Blumenberg had to re-order his cards several times, depending on the purposes he had at a certain time. Luhmann never did. Whether that was an advantage may be doubted because repeated interaction with the material is more like to stimulate revisions and growth over time. While Luhmann's work often is repetitious and not very elegantly formulated, Blumenberg's work is much more polished. However, both their texts reveal how they were put together, i.e. that they are accretions of more basic units."

You must know that Luhmann did his index cards the same way most library archives are stored: chronologically, meaning books/cards 3, 105, 2058 and 19386 belong together, in a certain way, but you would never know without the index, and then, no "natural context", but gathering one-by-one. For books delivered to a counter, that's the right system, but I assume if you have 85 such index cards, "held together" just by an index, and spread over thousands of physical cards, you will probably not gather all them but do some triage, hoping you have the content of the remaining 45 sufficiently memorized, taking out 40, and having to restore them all afterwards, one by one, being lots of work in itself, even without those other 85 you judge (from memory) "not so important here".

The same applies to gathering material to work on/with within your computer: If it's too cumbersome to "gather it all", you will probably not do it, by this leaving out important elements. So the computer/software should help you in gathering all important / probably relevant material in every instance, instead of your having to do "real work" for this "secondary activity".

I know there is "search", but the more you write with an extended vocabulary, the less you'll be able to find relevant things, since "manually", you simply don't think of all possible variants of terminology, and today's search tool will not be of big help to find them either.

So this means, Luhmann probable too much relied on his memory since "calling up" his carded knowledge was simply too much effort for him and/or his secretary, and what we get from today's outliners and from today's file system (in Windows), is a little the same phenomenon, that we often have to make do with what we immediately get, since digging further is too much effort and thus not justified - better software could be of big help here to include more relevant things, as for Luhmann would have some categorization instead of just using an automatted pagination stamp.

I beg your pardon, citation's from "Part I", blog entries being sorted the commercial sorting way, not the administrative sorting way. Sorry.

I beg your pardon also for all those typos above, "remaining 45" not seen, out of the 85 lot, etc.

I would like to add that Luhmann's pagination stamp wasn't something "idiotic", as it might appear today at first sight, since all this referencing would have to be redone, manually, if he tried to constitute groups of cards, meaning if you put card number 2490 out of its context (if there is a context), into a new context 843, all those references to this card "2490" have to be changed to "843", and you will probably not even know where they are, or you must put the references, "backwards", onto the card itself, too: "this card has been referenced in the following contexts" - but then, you must do a lot of rewriting at numerous places whenever you re-assign some "group"-card to any other location... and there is the "xerox" problem quickly coming in, because it offers easy solutions to "shifting 1 card around, but having to update 322 index numbers in consequence".

So my point is not that Luhmann did it wrong, but my point is, this paradigm shift to computers/software (where updating thousands of index entries is done in the fraction of a second) did not offer us enough from what Luhmann et al. could do with physical files: Technology would offer it, but software doesn't follow.

Btw, the very first version of askSam was promoted as an electronic Zettelkasten version, as was HyperCard (and that's why HyperCard was abandoned, and AS took many years to get its tree-on-the-fly: Before, it was all about remembering key words (whether in fields, or in body text, for AS), without a real "key word management" offered by the software: just tagging, and this made many users had left AS even before their intro of the tree-on-the-fly, which thus came too late.