BSW unique abilities

[PureMoxie]

Splitting this off from the previous thread...

> For example, if I wanted to note

> something under a particular project, I wouldn't need to navigate

> anywhere. As long as I were located in the right Brainstorm model

> (and one single model would be more than enough to use as a

> universal scratch file) all I would do is:

>

> - Type the standard title for

> my project; I use naming conventions such as [CUSTOMER-PROJECT].

> Immediately, a namesake would be created which would include any

> subsidiary entries already there - Promote the title to heading

> (press Home)

> - Press Ctrl+PgDn to navigate to the end of the subsidiary info -

> Type what I had in mind - Demote the heading (press End)

> - Delete the namesake; the info I entered is safely stored in the

> original location

That's a great idea - it never occurred to me to use a temporary namesake like that, but that allows one to "teleport" to some remote section of the model without leaving your place.

> - Carry

> on with my work; again, I haven't changed location, so the

> interruption hasn't really been longer than finding my actionables

> file and noting my info there. What's more, I can use this approach

> for anything, not just actionables. I could also place a word like

> <action> or <note> just before the snippet, which would help in

> later retrieval.

I do the same thing in terms of having a simple "tag" like <action> before the item. In addition to aiding in searching, you can actually get a separate list (under a different heading) fairly easily (and this works for any search term) by writing to a file with one level only and containing the tag match (e.g., <action>); then merge this file back into a new heading, making sure "Bypass intelligent merge" is not checked. The new list has all of your <action> items, and each item is a namesake with its original context.

In fact, I use namesakes quite heavily and in several different ways, although the usage tends to be incidental in most cases rather than planned out in advance. But it's what makes BSW more network/wiki-like (and interesting) rather than strictly hierarchical. I use it a lot for having multiple sort orders on the same items. For example, I use four different sorts on active clients depending on what I am trying to maximize at the time:

1. top 20% clients based on spreadsheet analysis of prior 12 months revenue

2. immediate cash flow - clients sorted by proximity of next bill date

3. overall revenue - clients sorted by hourly rate, descending (not all of my clients have the same rate)

4. client aging - how long it's been since I've touched their project(s)

While in some ways, a true database would make this sorting easier, I am only dealing with 10-15 clients at a time, and I can quickly sort them under each list heading while all of their project details are still right there as children of the client entry. I use this approach a lot to get different views and temporary priorities on the same underlying data.

Namesakes work really well for very specifically named items. When using shorter and more general terms as entries, you will often have to disambiguate items that are really representing two different things (unless you want the general term to be available in multiple contexts, which you certainly can do in tandem). But this is the same as having to rename wiki pages if you found you made a page name too generic. I find sliding around laterally between namesakes as well as the process of deciding whether they are really the same thing or not actually helps me see connections between things that otherwise wouldn't be immediately obvious.

[Joseph Buchignani]

You've obviously taken the namesaking feature of BSW quite far. 

In Cyborganize, I try to avoid namesaking because it requires a mental load to keep track of naming conventions, name specificity, and general outline structure. 

For example, handling actionables in a separate BSW file obviates the need for the <action> tag and sorting procedure. 

It's an interesting feature for certain specific applications like the client sorting trick you mentioned. I just prefer to avoid that level of complexity when I'm working. However for individual-specific info needs like the client sorting scenario people should create their own files to manage this. Cyborganize can just reduce the workload on those specific files by handling the main bulk of the info processing. 

Hope that makes sense. What I just wrote could be easily misread any number of ways, so don't take it as dogma if you disagree. 

--
Ignore the following. It is a nonsense sentence that disables Google ads from displaying next to my emails by triggering sensitive keywords.

I enjoy the massacre of ads. This sentence will slaughter ads without a messy bloodbath.

[Chris Murtland]

It does make sense. But here we're starting to get into stuff that has no answer - this is about how meaning is encoded into the data and where that meaning lives. Here, you're using the file container to establish the boundary of "I can take action on this," whereas the <action> tag establishes the boundary at the item level. There is no right answer, because each method of encoding the boundary has its own pros and cons. And the establishment of the boundary (or even what to use as a boundary) ultimately becomes arbitrary (e.g., why not a separate file for each type of action, or just an Actions heading in a file that also has notes and quotes, or the extreme example of a separate text file for every thought?)., All categorization tends to become something like Borges' encyclopedia (http://www.multicians.org/thvv/borges-animals.html)

There are directories, files, tags, database fields, outline or record items in a single proprietary file, etc. The metadata, if you decide to even use metadata, about an item can live in plain text within the item itself (simple text strings), or it could live in a specific database field or even a tag. Databases have the advantage of making it easy to create ad hoc queries on any set of criteria the database knows about; plain text has the advantage of being fast and flexible with no foreknowledge of the ultimate structure needed.

So, the scenario and the user's willingness or ability to deal with certain trade-offs will determine to a large degree which approach to take. I think you are right to establish firm conceptual boundaries and containers for a personal productivity system, because consistency of approach that can be applied in a generic sense certainly makes one more productive. And there is a big difference between the theoretical analysis of what can happen technically and what actually happens at the pragmatic level of doing things every day. I just think the conclusions about trade-offs are ultimately going to be subjective preferences.

[PureMoxie]

Other than cases where I start out intending to deal with namesakes (client sorting example), at other times when they appear, it is a sort of primitive machine intelligence that is prompting me that either "maybe these things are related" or "are you sure that's what you mean?" In cases where there is a relationship, on rare occasions it provides a worthwhile new insight. In the "are you sure" case - unless I am sure - it prompts me to be more specific about one or both of the items so that the namesake doesn't occur; this feels to me like the model is evolving to be more accurate and precise, but I don't know why I think that is important.

The other benefit is a form of exploration - sliding laterally through the model via namesakes is sometimes just a way to revisit sections of the model even if I have no purpose in mind for doing so. Not something I do a lot, but sometimes it can trigger thinking about something that has been dormant for a while. I'm sure that could be accomplished in other less oblique ways.

It's true that namesakes add complexity and a degree of cognitive load. I can't say that I'd recommend them to anyone who is just trying to get a lot of things done as quickly as possible.

[Joseph Buchignani]

the kind of connective exploration you're describing is something that PersonalBrain does even better than BSW. However I rejected that approach because I found a more efficient and powerful method of gaining the same benefit - the longform loop, which is better adapted to rich interconnected thought and ages better than web models because the nuances of connections are made explicit. It is also better adapted to spaced repetition and creates a more shareable final deliverable.