Forum Posts Feed
myBase Desktop 6.x + Addons Packfor PC on Bits Du Jour today
< Next Topic | Back to topic list | Previous Topic >
Posted by jimspoon
Oct 17, 2013 at 11:37 PM
22111 wrote:
>To finish, I would like to remind a rather important feature most
>outliners don’t have (and which myBase probably doesn’t have either, and
>which every outliner should have to which you switch from another one
>should have as I see it): Search should display the text context for
>your found items (when your search was for text within the content):
>Anybody ever having used desktop search engines like File Locator knows
>how extremely useful such a feature is, in the absence of any
>“relevance” functionality in presenting search results.
This is very important. I have an idea of a method of notetaking in a continuous stream. An “item” might contain no more than one word or number. But the meaning of very brief note would be evident from the context of the items immediately preceding and following in the stream. The availability of this context would make it possible for a note to be very brief, without repetition of anything from previous notes. So - this method of notetaking would be very efficient.
Posted by 22111
Oct 18, 2013 at 12:50 PM
This is both perfect and real-world, meaning I do such things on paper, and later on, I often don’t have the slightest idea what my - too short, obviously - note was meant to trigger from me; I lately put lines on my papers when I begin something “new” there, since without, I too often have asked myself what my note was meant to mean in the previous context, when in fact I had just tried to make economies of paper, and it was a new one…
But you point out to a necessary distinction:
- “Context”: what I meant / am hoping for, is xxxxxxxxx YourSearchTerm yyyyyyyyyyy: even that is scarcely available
- Then, there would be previous line // context line (see above) // next line (and then a line to distinguish this find from the next one, or differently colored background, white/chamois/white/chomois and so on). This would be much better but is not available anywhere in our context, but I have seen such “find tables” in dedicated programs, perhaps in the web only, can’t say where… perhaps in the results for finds in some forum (perhaps it was even finds for the TheBrain forum, but then, it could be wishful thinking on my part).
On the other hand, here and there you can get such “several lines in context” in rare cases, not as context for your search, but as I have said already elsewhere here, MyInfo’s version 5 had an option where for every find the very first 3 lines of content were displayed; this was no help for real context, but if you “organized” your very first lines of your items in a smart way, that function could have been really useful in other “context”. They withdrew it for version 6, promised to reintroduce it later on, but that was about 2 years ago… or is it 3 years now?
- Back to our problem: For one-line entries in your content, probably with a blank line before and after, or in any case where the previous and/or next line is a blank one, the previous thing is not useful, so here “context” (that would not be blank) would mean, “if line before is blank, show the last (???) line of the previous paragraph instead”, and my ??? insinuates that in such cases, the very FIRST line of the previous paragraph would perhaps the better alternative to display instead; for “next line”, similarly, “if next line is blank, display the first non-blank line as third line within the find table”.
Permit me to say that many developers do not want to delve into such “line break problems” even if it’s really useful. As an example, most (or many) file comparers are not even able to consider the following lines as “equal”, by option, and e.g. the developers of (otherwise not bad at all, but then, “moved block” ain’t processed either here, so in the end it’s not that good) “Beyond Compare” acknowledge the problem, but don’t see this as a reason to resolve it:
File 1:
Rather short title of paragraph, for example “Paragraph 1” and then begin of paragph, after just (another) space : Several text lines
of this paragraph
blah blah
File 2 (which for example you want to save in your outliner, in a more neat form):
Rather short title of paragph, for example “Paragraph 1” (but then, you’ll have introduced a line break!:)
Several text lines
of this paragraph
blah blah
Now if you want to compare a new version of this text from the web, for example a law in its latest version, from some legal authority, BC is NOT able to consider your saved version
Title
Then paragraph content
from the new version you’ve downloaded
Title Then paragraph content
And this means that with BC, you simply cannot “edit” your downloads to bring them in this more neat form, since then BC will give you hundreds of false alarms fo co-called “changes”, for a law containing several hundreds of paragraphs: no chance here.
Now compare with what outliners and their search functions can do: No chance…
- But you mentioned another level of “context”: item title of previous item, and item title of next item! As before, for previous/next line, this should indeed be available by option (and technically, it’s not really difficult, neither is the 3-line display described above), so you would have finds table options:
—display line (with context before/afterwards)
—display line with previous and next line (the disadvantage here is you need 3 times as many lines within your finds table, so I wouldn’t like to have this mandatorily, but by option: when really you need it, but only then), and finally
- display line with together in this form (that would be FOUR lines, though!:)
title of previous item
title of item in question (bolded)
line of found term (found term bolded, context regular font)
title of next item
And above, for 3-line display:
previous line (if blank, first line of previous paragraph if there is any)
line of found term (found term only: bolded)
next line (if blank, first line of next paragraph if there is any)
So much for things we will probably never see in our life.
Posted by 22111
Oct 18, 2013 at 01:55 PM
I forgot a detail: Even the 1-line context display isn’t that useful whenever your search term appears right at the beginning of the line (but without it being the beginning of a sentence), or when it appears to the end of the line, especially when a new sentence has begun rather immediately before the term.
Technically, it is possible to have the search term “centered”, by having the program fetching the same number of words/characters before and behind, but such al algorithm doesn’t consider sentence parts enough.
So this brings me to another consideration, since any outliner developer would have to reinvent the wheel, on his own: Perhaps there is a basic problem with outliners, which is the absence of “libraries” for parts of the outliner, except for the tree. You can buy (and even have for free) lots of tree “components”, but it would be a good thing for developers to be able to choose some routine elements from the usual “libraries”, but since there are so few outliners, with so few customers, most of such elements would not exist, or they are not easy enough to find, and then to adapt.
Also, on a bigger scale, software houses buy other software houses / individual developers’ findings, but in the outliner world, without there being the necessary money for such things, every developer minds his business, without ever there being any interaction by which some standards could evolve. And as said, lots of developers’ coding efforts go into html/web rendering/“translation”, and that’s also a reason why they don’t like to spend time with ingenious solutions, and that’s why we are often stuck with a rather primitive state of affairs in outliner details, where ingenious solutions would have perfectly been possible.
Not speaking of missing basics, for example even with “Boolean” search, NEAR is rarely available, let alone things like BEFORE 3 or AFTER 4, in askSam meaning within the 3/4 words before/after the next search term. (As said, AS has got a “command line” for entering search codes, and that’s all quite impressive, even though it’s 25 years old and just the residuals of its quite complete macro language then that, except for the search codes, got lost with its transposition to Windows).
This being said, it would be “advisable” to have a “winner” in this field, a “winner who takes it all”, but under the sole condition that he will then reinvest his gains into real development of his outliner/PIM/WhateverYouCallIt - and we would not see this.
But please permit a frank word here: Judging by the pace of their respective development, it must be assumed that ALL those developers have got a profession for living, and then do some coding in the evening, one day per month, or similar. Even if they applied just all their week-ends to this coding, their new versions should come with tenfold of what they bring now. So this “lack of application” seems to indicate that even for such spare-time developments, those developers want to maximize return on a minimum of hours spent on coding. My remark may seem a little bit harsh, but anybody who knows a bit about programming, cannot help but be astonished by this slow development pace we see almost everywhere in the outliner world.
Posted by 22111
Oct 20, 2013 at 09:23 AM
Quite interesting, the total lack of interest in myBase on bits: Not only it disappeared after just 1 day, but especially, 3 comments on bits in all, which is extremely low, even compared with other, quite “lesser” tools there that often generate about 30 comments or more there, and especially with other outliners on sale there.
This indicates that what I say above on the VISUAL ascpect of myBase (and what the commentator says about it in the cnet link above) both is correct more or less, and is an important factor in the buying decision for such software today, since we convened that technically, myBase is not bad at all, and in fact it is quite one of the “heavyweights”. But its APPEAL is substandard.
This is certainly an important lesson for developers in general, all the more so since we (and they) don’t get any numbers from anyone: Nobody knows even approximate “installation numbers” for any outliner (let alone “active installations” where the user buys updates), except the respective developer for his own one.
Posted by 22111
Oct 20, 2013 at 09:33 AM
“in all” was meant to be “this time”, interest in that program on previous occasions was much more vivid. Btw, it’s the owners of bits that would have, by far, the very best comparative overview of RELATIVE “figures” for the proponents in different software categories (even if they don’t have absolute figures ; since we can assume that the respective bits sales numbers “represent”, more or less, the relevance of those absolute figures, after some years, and some returns of the softwares to bits, all the more so when new buys on bits are cheaper then regular updates from the developer (and the numbers of which are unknown to bits).
By the way, it would be very amusing to compare OUR numbers for those softwares. Do you more or less agree when I say that for MyInfo, I suppose some 900 “installations in use”, for UltraRecall 2,500, for Zoot 650, for TheBrain 1,200, for Rightnote 400, and so on, or do you seriously think we’d need multiply those numbers with 3, 5, 10???