Forum Posts Feed
1 clipping, 6 items (re tagging "vs." outlining)
< Next Topic | Back to topic list | Previous Topic >
Posted by 22111
Jan 31, 2015 at 01:49 PM
spin-off of http://www.outlinersoftware.com/topics/viewt/5721/0/organizing-lots-of-thought-snippets since I realize in that context, it IS ot.
“I agree that a single outline hierarchy can become unwieldy. [...] tagging [...]”
As readers interested in theory and better practice (the latter asking for the former as a prerequisite, and yes, there are even outliner bloggers who ain’t interested, but that’s their choice and their business, the web is full of such things which ain’t thought out, or where even the attempt to think a little bit about the subject you’re writing on, has never been made, and will never be made either - I find it very amusing that 99 p.c. of those sites are done in the usual blog tools, i.e. those thinking-avoidance-inclined authors also seem to refrain from learning some bits of html et al., by a presumed-by-me majority), you will remember that I wrote quite a lot about “outlining vs. tagging vs. how to combine it perhaps to best effect”.
You will also remember that I said, multiply your outlines (even within some big db, but then you need better fractionizing functionality than is currently implemented in UR e.g.), then recombine them again (which, within the big-db concept, is very well realized in UR, but not so in MI e.g. - no details here since I gave them before, it’s just a reminder).
But in my previous writings (where I explained the superiority of the outlining concept for many, but not all, application cases, I left out one other very important aspect which I would like to mention here, where it’s far from ot:
In fact, the op asks for the best ways of organizing thought snippets; I acknowledge that most of the time, these would be rather short and rather not connected, so order / outlining / ordering within outliner parent items is far from mandatory here - ok, there is always the aspect of how and when to do sub-groups whenever such lists-of-equals get too long.
But as for web clippings, e.g., not only do I preach (as you know) plain-text (plus pics/tables in case), but without all the surrounding, unwanted things, and then do your own formatting, and I call this “clipping hygiene” - and it’s a relief for your further work whenever you really DO something with such clippings later on, instead of just collecting them.
Also, I regularly do cut off longer web articles and such into a main outlininer item, together with some more child items (which are automatically called “ditto” if I don’t use the dialog to give them more specific names); for one because I hate vertical scrolling, at least when it’s more than just ONE dn / up again, and also, to better separate distinct ideas in the original text, better than the original author technically did this (perhaps because he didn’t want to multiply sub-pages: bear in mind that in an outliner, navigation between siblings is very easy by kb, whilst most web pages are too lazy to make available kb navigation, and thus multiplication of short pages is usually very cumbersome). (alternative a)
The same applies to technical pages, i.e. programming help pages, where there is one question, followed by several answers: Once it will have become just a selection-and-then-1-key thing, it’s a pleasure to gain, by doing so, much better access to your information than you had before (alternative b)
Then, I often bold some these sub-items since (for me), most of the time, some of them are more relevant than others (but which ain’t irrelevant by that: those last I would leave out altogether). Btw, it’s also this my splitting up info that has risen my items’ number to a six-figure number, and thus this is relevant info to put that last number into perspective (cf. several thousand margin numbers in a 600-page legal textbook: it’s perfectly adequate, but you must know what these numbers mean in order to realize they make perfectly sense).
Now let’s reconsider alternatives a and b. Only in SOME cases of b the original order is meaningless: Even in most cases of several answers to one question, at least some or one of the answers will refer to an answer “higher up” in the list, so even then it makes sense to not mix the original order up, and all the more so almost any case of alternative a, i.e. an author developing some ideas or presenting some facts, i.e. in the order in which he decided to present these facts, and even if you leave out paragraphs (irrelevant, or irrelevant for you(r means)), original order is valuable information you would not want to sacrify…
but that’s exactly what you would do if ever you had the (in that context, very bad) idea to put any clipping into several items in your IMS; in other words, this extremely helpful means of even further optimization of IM
(i.e. by speeding up visual access, and also, by more precise cross-referencing (which in most outliners where it’s available to begin with, is possible only on the item level, not on the paragraph level (where MI excels, but those cross-references don’t survive any export, hence their reduced usefulness if you’re not willing to spice them up with some more code of your own ( cf. http://www.outlinersoftware.com/topics/viewt/5083/47 ))))
will simply not be available in a tag-based environment, whilst it’s one of the aspects in which outlining really shines; combine this with specific cloning (i.e. not of the respective parent item, but of one or several of those sub-items you will have created, and cloning into different targets contexts of any one of those sub-items), and you will come quite near a really optimized IMS in what currently available solutions offer (cf. my threads here re cutting up info and then recombining it again in multiple ways (individually optimized for your purpose in question) and the fact that no current solution to resolve this dilemma can be called “near-ideal” (cf. 1-, 2-, 3-pane outlining, let alone alternative realizations).
(Since “slatibartfast/ianb (here)” alleged, over at DC, that my English is so poor he doesn’t understand what I say (or then, my English is so poor I did not even understand what HE said), please feel free to ask me whenever some of my ideas may appear obscure to you. (And people who are mentally into witch hunts, burning and stoning, should seriously consider physically joining some (recognized or self-declared) state in a region of this earth where they could freely act out.))
Posted by 22111
Jan 31, 2015 at 04:15 PM
It just occurred to me that you can simplify this further:
a) the current / most recent (i.e. last-done) web clipping can be stored, with its url name)
b) it can be checked if you just have selected something (e.g. some text paragraphs)
c) it can be checked if you press a (= any) key within a certain time period after selection has become complete (e.g. 1.5 sec.)
d) the current url can be checked (ditto if it’s a pdf or other file format like .docx et al); it may or may not be identical to the previous url / file name (a)
Hence:
- IF b) is identical with a) AND there is no key pressing after selection within the given time period (b, c)
- THEN the system will automatically create a new sub-item (named “ditto”) under the item a) and put the selection into it
alternative 1: You press a key (e.g. a mouse key) within that time frame, by which instead of the sub-item to be created to be named by default, you will name it individually, by a dialog window
alternative 2: some key again, but by this, the new sub-item will be named by the first line within the selection (this is advisable only in texts the original author of which will have given meaningful (and short) sub-titles (which is the reason this should not be the default, by “no key”)
Of course, there is always the (necessary) alternative of another mouse key again which puts the current selection after the content of the CURRENT item / sub-item (obviously, this is needed for “selective selection”, i.e. copying in several installments, in order to avoid unnecessary web page elements e.g.: whoever ever has tried to manually copy the main body of a web page will have seen such a non-discriminatory, general selection very often will get you many unwanted elements, too).
The general rule being: Any often-used step within IM must be available with a minimum of effort (in time, in action, and in clarity (mnemonics, easy to remember), in order to be used on a regular basis: Whatever is cumbersome in real use, will be applied less often, by that reducing the overall quality of your IMS’ content. (In particular: Splitting up original texts into siblings (each with the original url/path-file and other wanted metadata, of course) is highly advisable (see above), but you will not do it if it takes too long, and even some additional seconds for any additional refining step will appear quite long and thus will have to be avoided.)
At this time, I only have two variants:
- store
Posted by 22111
Jan 31, 2015 at 04:18 PM
(Previous para = residual of shortness of text input field and ugly scrolling (no preview).)
Posted by 22111
Feb 1, 2015 at 12:34 AM
Similar for pdf’s. I fully understand the need for archiving the originals (and de-secured copies, from which to copy, too), but it’s evident that for most cases of then “working on them”, it’s preferable to have excerpts, with just the paragraphs that apply, instead of reopening page 23 of a 60-page document again and again, then perhaps even scrolling on that page.
On the other hand, from many such pdf’s (or other sources) you will not make just one excerpt, but several, and then there’s room for an intermediate step which seems unnecessary at first sight, but which is, imo, immensely helpful for upholding clarity in your work process: Have an intermediate copy file (e.g. .xdoc, and of course automatically named identically to the pdf in question), or, better, an intermediate parent item in your outliner (see above), and then do clones (or copy from your Word file) of these siblings there, which you insert into their respective work contexts, i.e. avoid to do this directly, since by the latter you will miss a (very handy) compilation of what you will have copied from the source file; I suppose in the old days, for every book, you will have had some A4 / letter-format list sheets on which you listed your photocopies made, in the form of 23:3-5, 28:1, 4 (meaning paras 3-5 and 1 and 4 on pages 23 and 28, respectively), in order to track these (the above description implies work on library books, of course; were they your own, you would also do some marks in the books themselves I suppose (and we all know why library books often have some glue on their pages, from all those Post-It’s that later on replaced those sheets)).
This being said, it’s evident that you can do without such an intermediate parent item, for all excerpts from a given source, if you have a good search function for immediate recollection of all this material spread “all over the place”, and some good means to enter the common source reference. Also, bear in mind that the order in the hit table will very probably be by order of the respective target locations, whilst you will need this list ordered by order of original locations, in the source, hence the need to semi-automate the insertion of a leading page-and-paragraph-number indication by which the hits will then be sorted; after the common title (indicating also the source) there could be other comments which would not affect the sorting. If your IM tool allows for very fast “getting to” such an ad-hoc list (i.e. it should be a one-key type only when the source is selected in a source list), you could avoid that intermediate copy group from which then to clone, so this latter concept seems to be far more promising. (You see here again that I’m not touting use and abuse of outlining or worse, unnecessary complications of your work flow; it’s just (and I repeat myself here) that in order to optimize workflow and bring optimized results with minimized efforts of the user, the sw has to become quite sophisticated, internally.)
Posted by 22111
Feb 1, 2015 at 01:00 AM
My fault: Page and paragraph numbers don’t need to be leading:
You will have a standardized, common title for any excerpt from a given source.
So this standardization must be assured, similar to tagging, where tags should not be entered manually, but clicked from a list/tree. Thus, similarly, you should have a list with current sources (this also implies several concurrent “workspaces”, for several projects you’re working on), and any source will there be named (or even renamed, with automatic propagation of the rename everywhere!) just ONE time, and any further clipping from that source should then be automatically (!) named equally, except for the 22:3-5 AFTER that common denomination, and then, any differing comment is allowed of course: At the condition that the search function is able to sort 22 BEFORE 100 (“intelligent sort”), this system will work admirably (and if it doesn’t, you will need page numbers like 022 indeed).
As for the automated naming of the excerpt items / documents, from a given source document, that’s very easy, technically: Your “clip” routine will check the title of the document (which is on your hdd then, which means you could also have changed and/or shortened the original title (which would be preserved within the source metadata replicated in every such excerpt item/file, and retrieved from that “source list” entry, and then put this title into any such excerpt’s title, completing it with the page (and perhaps para) info.
The next step would be to program the creation of these list entries to be in some recognized bibliography format; then, a bibliography-creating routine would create the correct bibliography for publishing, from the citations/references that you will create from your excerpts, also in different possible formats, but always by using the “centralized” bibliography info from the list entry, and combining it with the page and paragraph info derived from the excerpt item/file.
As you see here, even academic work could greatly profit from enhanced outlining, and btw., the scenario described above is nothing more than very smart tagging applied to outlining, i.e. you use tagging whenever tagging is superior to outlining, WITHIN outlining.