jostber
My new software favorite: Personal Brain: James Fallows article:

http://jamesfallows.theatlantic.com/archives/2009/08/my_new_favorite_software_perso.php

TheBrain 8.0.2.0 Slackware 14.1 KDE 4.10.3 Java 1.7.0_25 / (Windows 7)

Quote
dyslucksia
A very interesting article indeed. The only thing to beware of is trying to run too many similar applications (Mind Manager, PB, etc.) in an attempt to use the features of one to compensate for their omission in another. I'd love to get hold of these guys after two years and see how many PIMs they still use regularly. Bet they can be counted on the hand of one finger.
PB 5.5.2.1 on Windows XP, J-1.6.0_17
Quote
dyslucksia
jostber, are you aware that James Fallows is a member of our forum and has posted here:
http://thebrain.websitetoolbox.com/post?id=3633968

James strikes me as a perceptive man, a thinker looking for somewhere to park his thoughts, as David Allen put it. He appears to be having a honeymoon with PB, calling it "My new software favorite". I remember thinking that - once. His article tries to distil the essence of what PB does, and I think he expresses it best by quoting Karel van Wolferen in his article:

Quote: it (PB) may serve as the application that ties it all together and cuts straight through the labyrinths of things we cannot afford to leave out.

James admits he has to keep his magazine afloat, so he is obviously under some pressure to produce articles, of which this is one. That is not to denigrate the article, but it becomes apparent to me that this is not really an in-depth analysis of PB's strengths and weaknesses as much as a rather superficial review, padded with other people's examples and views.

Something does not ring true for me in the way he proposes that PB will solve his problems, and that is the unstructured way in which he believes that adding information to a Brain will help him retrieve information later on. Throwing stuff in there is fine, but as they say, garbage in, garbage out. Been there, done that. I don't think the excuse that PB quickly allows us to find what we need using Instant (or even Advanced) Search holds water any more.

Here's the acid test. After six months of heaving stuff into a Brain, are you really sure you know where to look for it? Are you sure you included enough keywords, aliases and cross-references to help you out? Sure, PB has a great habit of dragging out associations between things you never imagined, one of its great strengths, but that's really icing on the cake. I want a bulletproof way of retrieving stuff I know is in there, and, just as importantly, I want to be able to be sure that if I can't find something, it's not in there. Can PB guarantee that?

No, it wasn't built to do that. What I find challenging about it is the way its imperfections force me to reconsider what an ideal personal database should be. Maybe, having wrestled with PB for the better part of a year (and with previous versions before that), I've gone a little further down the road than James. For me, the honeymoon is well and truly over. In fact, this wrinkled old harridan and I have discussed divorce on more than one occasion, but right now we're still together. What's the secret?

To answer this, I like to recall Ronald Reagan's favorite joke*:

Quote: Worried that their son was too optimistic, the parents of a little boy took him to a psychiatrist. Trying to dampen the boy’s spirits, the psychiatrist showed him into a room piled high with nothing but horse manure. Yet instead of displaying distaste, the little boy clambered to the top of the pile, dropped to all fours, and began digging.

“What do you think you’re doing?” the psychiatrist asked.

“With all this manure,” the little boy replied, beaming, “there must be a pony in here somewhere.”

Maybe it's not the kind of pony that zenrain hopes will accompany the next release of PB, but the concept has become a kind of holy grail for me; that is, that searching PB should be like searching a closed universe. When we pile stuff into a Brain, we are relying on ourselves to think up and enter all the necessary associations we will need to retrieve this information in six months. This is usually 70-95% good enough, but the better it is, the more effort we have to put into crossfiling it, which requires mega man-hours. Hey, I thought computers were supposed to make life easier for us, not more labor-intensive.

We are on the threshold of the next big breakthrough; an ontological lingua franca that can be bundled with any application such as PB. Requiring a fair amount of AI to deploy, it's not here yet, hence we toil in individual misery. PB's design has several ergonomic tricks up its sleeve, as James observed, but are they enough to manage infoglut?

The way I see it, we need three tools:
  • A suitable database structure. PB, mind mappers and concept maps come pretty close to this, though they approach it from different directions.
  • A deep understanding about how to design a knowledge database in whatever application one chooses. There is so much I am just starting to realize about how to design a proper knowledgebase in PB to make it do what it should really be capable of doing. Regrettably, there is very little published about this. Don't expect to find much about this in the User Guide.
  • Readily available, internally consistent classification frameworks as I alluded to earlier, starting with Roget's Thesaurus and including applications such as WordWeb, but ideally expert systems built into programs such as PB.
Then, and only then, will we be certain enough to know whether or not the pony really exists.

*For a longer version of this joke, see here, also here.

PB 5.5.2.1 on Windows XP, J-1.6.0_17
Quote
zenrain
Dyslucksia, nice post, it raises some interesting points.

Several comments:
Quote: Something does not ring true for me in the way he proposes that PB will solve his problems, and that is the unstructured way in which he believes that adding information to a Brain will help him retrieve information later on. Throwing stuff in there is fine, but as they say, garbage in, garbage out. Been there, done that. I don't think the excuse that PB quickly allows us to find what we need using Instant (or even Advanced) Search holds water any more.

I've tried quite a lot of PIMs. Evernote, DevonThink, OneNote, MindManager, Together, Journler, VoodooPad, the standard OS file structure, Leap. Ummm, that's all I can think of right now.
In my experience, I'm more likely to find information in PB than all of the programs listed above. There are several reasons for this.
  • Scaleability. I can stuff more in PB than I can in the other options, and have it remain useable, and maintainable. Creating folders and subfolders only works so well for so long. Tagging is the same thing. Sooner or later I've ended up with a large list of tags, or extensive network of subfolders, or both. PB keeps what I need, but it's out of sight unless I need it.
  • Searchable. Instant search covers 85% of what I need. Spotlight (no, not windows live search or whatever the hell it's called now) is the best equivalent to PB's advanced search, except spotlight is faster and covers the complete OS files. DevonThink's search beats everything I've ever tested or come across. But (and it's a big one), the next bullet point makes up for PBs search functions being below these two options.
  • Related structure. PBs ability to show related information is to me invaluable. At a glance I can see not only the main topic I've searched for, but all the information I've deemed as related to it. Yes, it's user powered, but my opinion to what I need to see when viewing a particular subject beats out any automated option I've found. Even DevonThink.
  • Visual Presentation. This is because I'm a visual thinker. Because of this, I retain information much longer, because I can remember how I filed stuff and how it's related to other stuff (at least important stuff) much easier than a flat file. For me, this helps it stick.
I've not found a bulletproof way of dragging out stuff I've filed, especially when we are talking about a lot of filed stuff (and I mean a lot). But PB for me comes the closest.

Quote:
  • A deep understanding about how to design a knowledge database in whatever application one chooses. There is so much I am just starting to realize about how to design a proper knowledgebase in PB to make it do what it should really be capable of doing. Regrettably, there is very little published about this. Don't expect to find much about this in the User Guide.
  • Readily available, internally consistent classification frameworks as I alluded to earlier, starting with Roget's Thesaurus and including applications such as WordWeb, but ideally expert systems built into programs such as PB.


I completely agree that this is the holy grail. I also understand everyone thinks and classifies things differently. I ran across this post a long time ago where greenwood mentioned using WikiPedia's classification system. I've actually adopted quite a bit of the system where it works for me, and it has improved my organizational structure considerably. I can't say this will work for everyone, but it's a good post to revisit for ideas.
Ultimately, what works for one person may not work for another. With the glut of information available, it's up to us to find the solution that works best for us, and organize it in a manner that makes the most sense. This has been a work in progress for me, and from posts on this forum, it seems like many others as well.
In my opinion, there's no perfect solution, but we all grow as we apply and refine what we have now.

As Richard would say, For Pony!

Windows 7
J-1.6.0_22
--
OSX 10.6.3
Java SE 6
Quote
dyslucksia
Thanks zenrain. I particularly liked your reference to greenwood's suggestion about using Wikipedia's categories as a starting point. One could do lots worse. They've already worked out a good ontology that is in tune with today's world. Why should we reinvent the wheel? I might give that a try too, taking it only as far as needed in whatever Brain I happen to be using (follow the breadcrumb trail at the bottom of any Wikipedia page, where it says Categories).

I was interested in whether Wikipedia's category system is home-grown, or "borrowed" from a professional ontologist, and if so, whose? No answer to that yet, but their Categorization page is also interesting and I see that they make use of AT&T's Graphviz for constructing relationship diagrams in the form of directed graphs.

Since Graphviz, like Apache Derby, is open-source software, I wonder whether Harlan might consider adding it to PB as a plug-in? After all, WikidPad already does that in Windows. And Linguine Maps, mentioned on Graphviz's page (maybe it uses Graphviz technology) is an open-source Java library that does just that, producing printable directed graphs that don't wobble, shake or snap. Who needs Expanded View?

BTW, I enjoyed Ryan & Lar's Looking for Group comic strip that you linked to in your post. For pony! The same guy also said:

Quote: No idea what he is talking about, but I respect his attitude and willingness to kill.

PB 5.5.2.1 on Windows XP, J-1.6.0_17
Quote
zenrain
Glad you enjoyed it. And excellent quote.
Windows 7
J-1.6.0_22
--
OSX 10.6.3
Java SE 6
Quote
Steeph
dyslucksia wrote: Here's the acid test. After six months of heaving stuff into a Brain, are you really sure you know where to look for it? Are you sure you included enough keywords, aliases and cross-references to help you out? Sure, PB has a great habit of dragging out associations between things you never imagined, one of its great strengths, but that's really icing on the cake. I want a bulletproof way of retrieving stuff I know is in there, and, just as importantly, I want to be able to be sure that if I can't find something, it's not in there. Can PB guarantee that?

No, it wasn't built to do that. What I find challenging about it is the way its imperfections force me to reconsider what an ideal personal database should be. Maybe, having wrestled with PB for the better part of a year (and with previous versions before that), I've gone a little further down the road than James. For me, the honeymoon is well and truly over. In fact, this wrinkled old harridan and I have discussed divorce on more than one occasion, but right now we're still together. What's the secret?


zenrain wrote:
I completely agree that this is the holy grail. I also understand everyone thinks and classifies things differently. I ran across this post a long time ago where greenwood mentioned using WikiPedia's classification system. I've actually adopted quite a bit of the system where it works for me, and it has improved my organizational structure considerably. I can't say this will work for everyone, but it's a good post to revisit for ideas.
Ultimately, what works for one person may not work for another. With the glut of information available, it's up to us to find the solution that works best for us, and organize it in a manner that makes the most sense. This has been a work in progress for me, and from posts on this forum, it seems like many others as well.
In my opinion, there's no perfect solution, but we all grow as we apply and refine what we have now.


@Dys. I think that PB is build especially to do that. However as zenrain puts it, it's holy grail of Information Technology. If you can't find something it's actually your weakness (or rather of Homo Sapiens wet brain) than its PB's weakness. You are the one in charge who decides how to categorise stuff. Sure you think of tons of tools to help you out, but in the end it is you who decides.

Be aware that you are dealing with a problem here that I think is fundamentally unsolvable. It's tackled by great minds at Universities, MS, Google, philosophers, and who not. In centuries of Information Shuffling it hasn't been solved yet.
It might be circumvented somewhat through sheer computing power eventually, as Google does for example, but I don't think it can be solved. Because information changes. society changes, languages changes, interpretations changes, history changes, culture changes and constantly new things are discovered.

I think you should give PB some slack here for not being able to solve such things. Sure things can be improved, but the problem is hideously complex and ever changing.
PB user since 1998

Mind over matter?
I don't mind and it doesn't matter.
TB 8.0.2.1 Pro on Win10.1 Pro 64bit JVM 1.8.0-112
Quote
mbaas
Steeph wrote: I think you should give PB some slack here for not being able to solve such things. Sure things can be improved, but the problem is hideously complex and ever changing.

Thanks guys, lots of good thoughts here and I wish I had the time to get into this discussion - but I have not. However, Steeph has put it nicely, I agree
And BTW, Ontology is overrated - this article has helped to cure me from that (just read how some of the "professionals" developed their categorization-schemes...)
Quote
jostber
mbaas wrote: Steeph wrote: I think you should give PB some slack here for not being able to solve such things. Sure things can be improved, but the problem is hideously complex and ever changing.

Thanks guys, lots of good thoughts here and I wish I had the time to get into this discussion - but I have not. However, Steeph has put it nicely, I agree
And BTW, Ontology is overrated - this article has helped to cure me from that (just read how some of the "professionals" developed their categorization-schemes...)


The Ontology article is a very interesting one, thanks!


TheBrain 8.0.2.0 Slackware 14.1 KDE 4.10.3 Java 1.7.0_25 / (Windows 7)

Quote
Steeph
jostber wrote: mbaas wrote: And BTW, Ontology is overrated - this article has helped to cure me from that (just read how some of the "professionals" developed their categorization-schemes...)

The Ontology article is a very interesting one, thanks!


Indeed it is, another thanks!
PB user since 1998

Mind over matter?
I don't mind and it doesn't matter.
TB 8.0.2.1 Pro on Win10.1 Pro 64bit JVM 1.8.0-112
Quote
dyslucksia
The design of an ontology is beyond PB's specifications. I'm not suggesting it should try to enforce one on us (though who knows what they might tack on after a spell checker...). The important thing is that it should allow us to design our own without hindrance, and this it does fairly well. As I said, I like the idea of borrowing Wikipedia's one, though any of the good directories out there (Yahoo!, Google) would do. And for a specialized Brain, the classification used by an existing specialized Web portal should do quite nicely, thank you.
PB 5.5.2.1 on Windows XP, J-1.6.0_17
Quote
Darkstar

Humm--- I've used Personal Brain as my primary digital information organizer since version 1, and what I've found that works best for me is: keep every brain focused on what it does.

Now, you might ask why--- and the answer is the "holy grail" of information keepers---- SO I CAN FIND THINGS IN THEM MANY YEARS LATER!

That's it.

With a "brain with a purpose", it only contains items that are useful to it. I have a general work brain, that contains a listing of all the common tasks I do for work. It lists my major work projects--- but those projects major details? That's all contained in their own brains. I do have a few duplicates because of this (for example, my "shortcut" thought to my SQL+ tool occurs many times, as it is in my current work brain, as well as each project brain that I needed to use it), but that is acceptable to keep my brains focused. When I am searching for the emails "Kristy Smith" sent me related to the requirements for the Constellation Action Tracker project, I go into my Constellation Action Tracker project brain and do a big search for her name. I'll get only the "signficant" emails she's sent me about it, (as well as any work order and trackers with her mentioned). If I had did that in a "One Brain for everthing", I'd get ALL her "somewhat importish seeming" emails and mentions--- and at that point, I might as well let my email app do the searching--- as they are going to be very similar results, only the email app will ONLY have the emails in it--- and not all the various OTHER documents where she is mentioned or originated.

When I have to go back into a brain that I haven't used in NINE YEARS--- it is CRITICALLY important that it only have the pertanent information contained in it. I don't need the intervening nine years of general information detrius that accumulates over the course of a career (or life, for that matter).

Frankly, I am not missing the gaining of "insights"--- as there is rarely any insights to be gleaned from that sort of information flotsam. I can gain insights from reviews of old brains and comparing them--- particularly side by side--- with later brains on similar subjects. But you can't actually do that in PB (as of version 5.x) if you put it all in one brain--- as you can no longer have multiple instances of PB open to the same brain. So just to see your major network groups in a big brain, you have to export it or copy it to a new brain and then you can finally do side-by-side comparisons (although--- if you spend the time in Expanded View, you can set them up in it--- with a lot of arranging and pinning--- an extreme lot of bother if they are not very interlinked clusters).

I'm not claiming my way is the best--- other than it works well for me. But the reason it does is that by keeping the information isolated means that it is easy to search, and easy to know where to look for the info. And frankly--- depending on the information DENSITY, it might not even be in my PB. I use other tools for certain kinds of information storage. So if the information would be better SEARCHABLE in another tool, then it is probably in that other tool. (for example, I use Zoot to maintain an "information collective" of all the space related web articles I've found interesting for years--- so if I need to refresh my mind on a the later trends in cosmology or how a particular commercial launcher has spun its releases for a particular period, I go into it, and use its superior search capabilities to find it. I don't have to PRE-CATAGORIZE or otherwise PRE-MARK an item in Zoot--- I can build smart searchers on the fly to do that for me.)

Information organizers/Information management really have 2 competing goals:

  1. To help put the information you often need right at your fingertips. This first goal means you want all that info floating around, and zooming down to the user the instant they need it. So you want easy controls for the user to "recall" that info. Bookmarks, pins, recent used lists--- that's all mechanisms to put the information you use often, or you find so important you want to easily jump to it, right at your fingertips. 
  2. to be able to find any item of information stored in the archive, and make it easy for the user to find that information with a minimum of knowledge remembered about it.
PB meets the first goal easily and well with Instant Search, Pins, and Recent Thoughts.

PB isn't so good with 2. It is trivial to look through a brain with just 30 thoughts, and see if any of those thoughts are what you were looking for. But if you have 30,000 thoughts, you have just scaled out beyond what you can easily or trivially find. If that item isn't marked in some way to make it easy to come up with a small subset of items that I can review to find it, it is good and lost. At this point, my information system has failed me when I really needed it. I don't need it for "quick launching" apps or web pages--- as there is literally thousands of apps that will do that for me. What I really need is a long term information retrieval system that is superior to my own mind. So to meet number 2, you need to mark items in as many ways as possible.

So how is 1 and 2 competing goals? To meet #2, you need to really mark up any new items--- and even revise older items with the new classifactions that you evolve as you add in new items and new markings you can (hopefully) easily find the information years in the future. This means the more items you have, the more markings you have (and you may have several different systems at work at once). But "the information you often need right at your fingertips" emphasises how often you want to access that bit of information, so you end up with just 1 or 2 ways in--- ie, thoughts linked to pins (or a thought that has been set as home). And that encourages you NOT to heavily mark items--- as you are just using a few hubs (or trunks) to find all the "important" stuff. This leads to piling up the "non-important" stuff so you can get on with doing what you need to get done (work, journaling, shopping, whatever).

In a purpose focused brain, the information in it has already gained context--- by just being in that brain. Furthermore, it will be organized in some organic fashion--- whether you set out with one particular kind of organization at the start or just had one form over the creation and use of that purpose focused brain. Also, it will have less items in it then an "everything" brain, so you will have less items to look through on your searches. And that is one of the main value of these tools.

It's something to think about--- in the future, once you've left that brain (or that area of the brain, if its a big brain), how will you FIND that information again in 9 years, when you've changed how you think and how you view the world? After you've had a while to forgotten frankly everything--- even the "significant" associations between thoughts in that brain? Or do you not care, because it (probably) won't be important to you then? Is it more important doing things quickly? Or do you need a balance between quick and being able to find things? Think about it--- as this will determine what will be the best way to arrange your brain.

Have fun, and enjoy your thoughts!
-Darkstar
Quote
dyslucksia
A nice, thoughtful post. Newbies to PB in particular should read this one.
PB 5.5.2.1 on Windows XP, J-1.6.0_17
Quote

Add a Website Forum to your website.

Newsletter Signup  Newsletter Signup        Visit TheBrain Blog   Visit TheBrain Blog       Follow us on Twitter   Follow Us       Like Us on Facebook   Like Us         Circle Us on Google+  Circle Us         Watch Us on Youtube  Watch Us       

TheBrain Mind Map & Mindmapping Software     Download TheBrain Mind Mapping Software