Outlining and concept maps (Scapple, TheBrain, ConnectedText...)

< Next Topic | Back to topic list | Previous Topic >

Posted by 22111
Nov 11, 2013 at 10:21 AM

Scapple

“The export files don’t seem to preserve any kind of hierarchy. I realize that may be by design”

No import from your outliner, and no export to your outliner but in a flat list if I understand this “design” well - this is unacceptable.

Let’s remember this is brand-new software, but if they insist on not developing the according functionality soon, forget about this program.

General Considerations

Of course, for such a concept map, a real tree structure is not possible, but that is no reason to not try to preserve any structure to begin with.

Export: By frequency of interlinks, the program could create groups, called Group 1, Group 2, etc., with the items as children (siblings, from their point of view). Then, by option, it would be possible to automatically add several types of comments like “(also in Group 2)” and “(cf. x)”, “(but y)”, etc., these by checking the types of connections (alternatives, objections, difficulties, possible solutions, details…).

Then, it should be possible to NAME those groups, even before export, in order to facilitate their clarity, especially for such automatic comments.

From the screen, the groups could be retrieved beginning from top left over top right, then bottom left and bottom right, if that is sensible, or, automatically, in three ranges, middle left, middle right in-between those, or just “middle”, or top left, top middle, top right; middle, bottom left, bottom right, i.e. the program should check for groups, by frequency of interlinks, and then decide - some decisions might be erroneous, but most will be helpful.

Also, directed links should be checked, anything is good that will break up all-flat lists, and multiple “clones” creation, with the correct automatic comments, should be another target, which means, where, by multiple vectors, within the concept map cloning is NOT necessary by implicit, one item being a “factor”/“element” in multiple contexts, the export should create multiple “subtrees” instead, with lush “cloning” - if you want this or not, in any case, could be decided in common by yourself and the program:

“Export” would present another “view” first where the program would present “things it will do” by coloring/bolding and other visual indicators, and in which view you could also color/bold or de-color/de-bold the selections the program will have made for you: All this will greatly reduce manual touching up of the export result in your outliner then.

Import should of course create sub-webs, children/siblings/dotted lists, within the space of that virtual sheet of paper (= the screen deployment of all imported elements), in “groups”, from top left to bottom right again, and here again, the program should decide for itself if it’s reasonable to create those groups in 2 or 3 ranges, or even 4 (of course, it’s always sensible to not create monster maps, and from there at least, splitting up the too big a map into several reasonably-sized ones should be possible).

On import, too, there should be options, in order for your deciding upon directed vectors, bidirectional links, just grouping, etc.

Also, instead of fetching all the text, or none, there should be the possibility to just fetch its first paragraph.

I think that for planning/“thinking” about things, a concept map is invaluable, but then, for “execution”, with all the “material” that such “putting into practice” involves, an outliner structure is better, so the export function of a concept mapper should be quite elaborate, whilst the import function could be quite rudimentary, but then, it’s an iterative process, so even import should not be too primitive, and of course it would be ideal to have some real “synching”, i.e. program functionality within the concept mapper, on import _to an existing map_ for checking, what’s new, what has been changed within the outliner - to some degree at least. This means you would not have to “rearrange it all”, for every new “import”, but you would have a secondary “import” function, which tries to preserve your existing map, just updating it a little bit - it goes without saying that this would be tremendously helpful.

ConnectedText?!

All this makes me wonder if the in-built mapping functionality (which avoids all these extreme export/import/synchronization problems between mapping and outlining in separate programs) does “replace” such a “concept mapper” to some degree, and if yes, to what extent: Known problems, known weaknesses, but also known strenghts?

It’s interesting there has been written so much about CT, from Kühn et al., but this mapping functionality of that program is not that much mentioned, let alone detailed, so some insight here from real users would be really helpful: It could be that from this point of view, CT might be the ideal program.

TheBrain

Also, it would be of interest to know what TB “does” with interconnections and such when it builds up its “outlined” structure. As far as I “know them” (but I’m just guessing here), they just retrieve the main tree and cut off all these additional linking elements? Or do they do something really smart there?

Posted by Hugh
Nov 11, 2013 at 06:11 PM

And you’ll want Scapple to do all that, and still pay $14.99?

Scapple is an electronic form of “napkin concept-mapping”. As that implies, it really isn’t intended to be competitive with the “big” mind-mappers, concept-mappers or outliners (as the developer keeps saying), all of which cost at least double and in some cases 10, 20 or more times as much. I wouldn’t ever use Scapple to outline a book - except in the most generalised “outline” form - but I would and do use it to outline a scene, a memo or a business letter, for which I find it appropriately useful. I certainly wouldn’t put it in the same category as The Brain or Connected Text.

There’s a tendency to aggrandise and hyperbolise software that is continuously adding functionality and tries to do many things. And then a little later we tend to knock it down again for being bloated, too big and difficult to learn. So it’s interesting when a developer says: “This is what it is: it’s designed to be simple and it isn’t going to indulge in function-creep”. I can’t think of many such applications in the realm of outliners and mind-maps. Opal may be one, Tree may be another, Taskpaper and its Markdown-influenced, outliner relations Folding Text and Oak may be others.

Posted by Stephen Zeoli
Nov 11, 2013 at 09:37 PM

I don’t think Scapple is intended to deal with hierarchy. It’s about relationships. I suspect it sells well to Scrivener users, who can drag the notes directly into the binder. It’s a handy little app, but it isn’t ever going to be a substitute for an outliner or mind-mapper.

Steve Z.

Posted by Franz Grieser
Nov 11, 2013 at 11:37 PM

>It’s a handy little app, but it
>isn’t ever going to be a substitute for an outliner or mind-mapper.

It has never been intended to be. It’s just a small tool, indeed. I use it for clustering and for producing material I feed into a mind mapping tool. Before Scappl, I did the clustering on paper or index cards and had to type the words again when using Xmind for building a structure. Now I can easily use the text file I export from Scapple.
That alone is more than worth the $15 for me.

And: Yes. I did try to create simple illustrations using Scapple - and gave up. It’s simply not intended for producing pretty graphics.

Franz

Posted by 22111
Nov 12, 2013 at 11:57 PM

“And you’ll want Scapple to do all that, and still pay $14.99?”

Hugh, I take this for a rhetoric question, addressed to the general public; in such cases, I put the “you” into quotes to indicate that - as you know, I’m willing to pay for good software, in accordance to its value (and so, a price increase for Scapple would not be a problem if the features were there).

I’m quite amused you all throw yourself on the Scapple thing - I just mentioned it because it was there (since new in Windows, and mentioned here a month ago), and thus triggered my thinking of the need for perviousness between outliner and concept map, both ways.

And yes, deep hierarchy there would not be on my list (anymore) - I accept that “thinking process enhancement tools” like concept maps (but outliners are not such tools, that is also generally seen I think, and Steve said so some years ago here, and since he’s right, we need interoperability for our data between those two tools categories!) are useful for RESTRICTED sets of data.

But even with these limitations accepted, we’re speaking of several dozen of items, AND ARRANGED IN A CERTAIN WAY in that map, and then we want to do some “real work” upon those ideas, hence our import need to the outliner, and then, we want to REFINE our thinking process, but in the meanwhile, our work on that set of data, within our outliner, will have de-synched the map, and thus, a function that “memorizes” the previous state of the map when you RE-import that “same” = a similar set of data, from your outline again, would be of extreme usefulness: If you have to manually check print-outs of your map again and again, from possible re-arrangements, add-ons and deletions within your outliner, you will have a lot of (painful and unnecessary) work to do!

So, again, Scapples was just my starting point (but then, their writing software is quite elaborate and very friendly-priced, so perhaps a really good Scapple would only be 80$ instead of 400 MindJet would charge for it…).

But then, I mentioned CT, etc., which are not concept maps, but CT graphics especially would be of real interest here, since, in case, “you” ( ;- for the quotes) could use it for thinking, putting the meat to it, re-thinking… iteratively?

And TB as well? I never found a real-life use for it, but perhaps here’s a chance to make it really useful.

Of course, TB is quite hierarchical in the sense that independent items are quite new there and that their development has not gone very far, up to now, so I doubt it would be possible to have several strong, alternative “central points” (and their respective sub-webs) interacting with each other.

Similar for TC, but TC is said to be a wiki, not hierachical by design, so perhaps there are chances TC could do it better.

IF they can do it in some way, both programs would present those “visuals” “from within” - whilst on the other hand, I don’t know any concept mapper with an inbuilt “database”.

There is certainly “room”, open space to be filled by, and for such an integrative approach.

I know it’s very difficult to integrate concept maps and outlining, and so people endlessly try with mind mappers - but we all know that for thinking, mind mapper are NOT the thing, concept maps are. So some data translation should be made possible, half-automated, with SOME user interaction in the process.

As said, the mapper could identify new items by graphical identifiers like bolding, coloring or by other means, and it also could identify items that are NOT in the map anymore, judging from the “new” outliner/hierarchical data import, e.g. by totally thinning them, and/or by greying them out, black becoming light grey, etc.

Similar for renames: If the new element is in place of a deleted one, or even if it’s in the same sub-structure, the mapper could “ask” you if it is bound to replace the deleted one, and anyway: Every “NEW” element (really new, changed, displaced (judging from the tree), etc.) could be placed in some “waiting state” by the mapper, i.e. it would be “thinned”, so that you could easily and fast rearrange, by mouse, the element to that specific place within the map you want it be be.

Which means, the map is not a faithful graphical representation of the tree - this would be impossible -, but the map, on each iteration, would “remember” the previous state and do just the absolute minimum on it, by its own own means, but present you any changes in a way to enable you to quickly do the necessary changes yourself… which the mapper would then remember again for subsequent iterations…

Similar for the re-imports into the outliner: No overwriting of existant items (and their respective contents), but just careful indications - technically, that would be another problem, of the outliner in fact, which should be able to indicate (by bolding, underlining / coloring vs. thinning out / greying out / italicising): “This item is not present anymore in the new synch from mapper”; “That item has been added within the mapper” - and yes, in some cases, both represent one single, renamed item, for which you then will have some manual work to do.

All these - big - problems (and no, I don’t have the slightest idea about programming of graphics) strongly indicate the usefulness of a one-program solution that does both, serious data storage AND graphical, not-hierachical data representation.

I suppose many people will try to use such a “mapper” as sort of a “by-stander” to their data repository, but in the end, that’s a lot of “double work”, in the sense of manual “search”: Which element in the outliner is where in the map, and vice versa?

Which brings me to the idea that in the end, perhaps it’s “best” (= in the absence of an integrative solution, by it one-program, be it two programs working entrenched) to have a concept map (= NO mind-map) that (hopefully: easily!!!) permits systematic referencing single items in your outliner database (cf. the threads about cross-referencing outliner items), and then to not leave one element in your map without its link to the outliner - but that’s hard work, notwithstanding its being one-way linking only.

Btw, you’ve got a similar problem with Warnier designs, no software being able to synch between your design, and between your code repository; that’s one of the reasons why Warnier’s almost forgotten, when in fact, “together” with Yourdon, the honor to have buried spaghetti code for good, belongs to him. But that’s another story.

Back to topic list