Friday, June 22, 2018

Reading versus Writing

I noticed this claim today in Christian Tietze's blog:
Some software nudges you, sometimes even pushes you, towards system design decisions. Take Wikis as an example. Most of them have two different modes:
  • The reading mode.
  • The editing mode.
The reading mode is the default. But most of the time you should create, edit and re-edit the content. This default, this separation of reading and editing, is a small but significant barrier on producing content. You will behave differently. This is one reason I don’t like wikis for knowledge work. They are clumsy and work better for different purposes.
This is, unless I am very much mistaken, a subjective reaction not an objective observation. It isn't my experience, in any case.
  1. "The reading mode is the default." There may be some wikis that make reading mode the default. The personal wiki I chose, ConnectedText, let's you decide whether you want to always view topics in view mode. I did not turn on that option. But even if I had, hitting Alt-E for edit (or getting out of edit) is automatic. There is no barrier in my experience.
  2. {Wikis] "are clumsy." Not my experience either. On the contrary, I find the separation between topic names and topic identifiers that is a basic feature of The Archive much motre clumsy. It may be true that "to create links between notes" in an application like nvALT, "you need to define how to target a note first. One response is to use file names. If you want clickable links the full path to the note could be used. However this is a fragile solution. It breaks when file names change or the location of the file changes. A better answer is to use an unique ID for each file. The Archive uses a timestamp ID. These timestamp IDs are by definition unique" (from a review). In a real wiki, the program keeps track of name changes for you.

I do not want to knock The Archive: to each his or her own. I would find the opaque first layer a much larger problem.

Sunday, April 8, 2018

CintaNotes, One Last Time?

I recently downloaded a free version of CintaNotes, just because it appeals to me on so many levels. When I activated Simplenote synchronization, almost 11,000 notes appeared in it. And I can testify that it has absolutely no problem dealing with that many notes.

The notes must have come from previous nvAlt adventures with my ConnectedText exports. I thought I had deleted them, but I must not have. So they did show up in CintaNotes. However, the joy did not last long, as I managed to hose the database by trying to do massive changes to entries, using a keyboard macro program.

In any case, I would suggest that the author of the program look more deeply into SimpleNote synchronization for getting stuff into CintaNotes. The only reason someone like me cannot use it as a serious alternative is that I have many electronic notes that I cannot import into the program's database. (Nor will it be possioble to export them in a format usable in another program.)[1]

1. Export also may be a major problem with TiddlyWiki, but I was still not able to test it.

Saturday, April 7, 2018

A Review of InfoQube

Paul J. Miller has written a real review of Infoqube. It's the first one I have read, and it is well worth reading.

Did it make me want to go out and try the program? No, and that even though I liked Ecco (which it tried to emulate and expand on). The reason is that I have come to believe that outlines are not the best way to store long-term information, and that I therefore like better a hypertextual approach (say wiki).

But I admit that not everyone will feel that way and that there are some that will brave the steep learning curve of InfoQube. I am too old for it.

Friday, April 6, 2018

Notebooks in The Guardian

There is today on interesting post in The Guardian on the notebooks of Henry James, Paul Theroux, Susie Boyt and Amit Chaudhuri. Two quotes stand out for me, one by Henry James, the other by Paul Theroux.

James found "the only balm and the only refuge, the real solution of the pressing question of life, are in this frequent, fruitful, intimate battle with the particular idea, with the subject, the possibility, the place.” And thus found his notebook almost a necessity of life.

Theroux claims: "Note-taking is not just a method for remembering. It is a way a writer tells himself, or herself, a story – and this becomes a process of life, a mode of being. Writers are constantly talking to themselves."

I find myself in my own practice closer to to Henry James. My notes do not tell a story to me—at least not directly—but then I am not a "writer", even though I write daily.

The Archive

I have more or less given up on the Mac. The Mac Mini will probably not be updated again, and running ConnectedText on Parallels became more and more difficult. (More precisely Parallels became more difficult. Every time my Internet connection broke, Parallels would not start. I eventually got it to work, but I never understood what it was that made it work. My best guess, it had to do with the keychain function of OSX). But whatever may have been the reason, I now use a NUC7i5BNK with 8 GB of memory and Windows 10 (64 bits) and have no problems.

I can still run the Mac mini I have (It's hooked up to the same monitor as the NUC), but I rarely do. Therefore, I have not tried The Archive and I cannot review it myself. But there is a very informative review on Welcome to Iherwood. As far as I can see, the most important feature of The Archive is
The Archive will automatically set a unique number to each note you create in the format of yearmonthdayhourminute that the note was first made. Call that the note ID. You can append a note title to provide a clue as to the content of the note. Together those will make up the file name of the note — each note is saved as a separate plain text file in the designated folder.
As people who regularly read this blog know, I think the unique number is superfluous in modern database systems. I use direct links to entries (in ConnectedText). Luhmann who used unique numbers in his paper-based system, thought the numbers were in themselves beneficial because they allowed for branching and continuation of notes.[1] He may have been right, but the simple unique numbering of the Archive does not seem to allow for this—unless there is something I miss.

It is, by the way, fairly easy to furnish a program like ConnectedText with this capability through AtoHotkey. I think that some users have adopted "yearmonthdayhourminutesecond for this purpose (to allow for notes that are only seconds apart).

As I said, I don't find I need it, but I understand that different people have different needs.

I recommend the review.[2] And, on the base of it, the program (as long as you understand what you get into.[3]

1. See also Different Kinds of Links on this blog.
2. One nitpick: Luhmann did not "invent" or "develop" the slipbox. He only developed a numbering system for the notes (that is most certainly intriguing.
3. But see also the newest post (on the three layer structure of notes in: The Archive. "My archive became opaque like the sea: You can see a couple inches into the deep but you know there is much more that you can’t access. You can dive deep, but still you just see a couple of inches at any time. Therefore, I thought of it in terms of unexplored territory for which I need mapping methods and such." It would be my claim that direct links to topic names would be less opaque, even though it also benefits from structural notes.

Thursday, March 22, 2018

From ConnectedText to Zim?

It is always good to look for alternatives—or so it seems to me. I have been exclusively using ConnectedText for my notes since 2005. Recently, there have been rumors that ConnectedText won't be updated in the future, and some of us were considering alternatives to ConnectedText. It seems that these rumors are not correct. Connectedtext may be in "maintainance mode, but it will be updated as needed.

Still, I wanted to see what other options there are. So, I tried to determine whether and how compatible ConnectedText's files are with Zim. I had considered it before, but I decided to give it another chance. (I will never give Wikidpad another chance, as I found it highly unreliable, even if it was quite easy to change its markup to something closer to that of ConnectedText [see here for a start]).

I was able to transfer more than 11,000 ConnectedText topics into Zim. I used ConnectedText's backup files (*.cbk) and transformed them with the first AhK script I made (i.e. I gave the entries more friendly titles).[1] I did not use the second script (to remove the first line with "#CT#" as that produced files Zim had difficulties recognizing. (May have to do with UTF-8, but I don't know. I then cleaned up the text files with fnr.exe ( (Find & Replace Tool). But be careful, it's easy to make mistakes that spoil the files.

The titles of Zim's text files do not allow spaces. I used Advanced Renamer to replace the spaces by underscores (see I probably could have done all this with AhK, but was to lazy. You should also note that Zim is very fussy about what character it recognizes in titles. It does not recognize question mark, for instance. And a bracket at the beginning of a file name does not get recognized (so you have to remove brackets in topics like (Kuehn 2002) which I extensively use in ConnectedText for bibliographical entries.

The resulting files work, as Zim uses text files, not a database. But there are some problems:
  1. the load time is about 3 minutes on my system (off a SSD with 500MB per second transfer time). After that, it is reasonably fast in loading pages, etc.
  2. every time you change the underlying text files, it needs to re-index which can take up to 20 minutes
  3. many of ConnectedText's more advanced functions are not available
  4. I did not succeed in transferring pictures automatically
  5. it's impossible to print directly from Zim. You can only print to your browser
  6. I don't like that I have to use underscores in naming pages

Zim is clearly is not a good replacement for ConnectedText, but it is a replacement, if such is needed. It would do. in a pinch, and some of the plugins might provide more functionality that I can use. But for now and the foreseeable future, I will stick with ConnectedText. (See also my Reassurance.)

The transfer took about half a day to accomplish (and is still a work in progress, as I discover new things that need to be changed, like properties I had forgotten).

The other thing I recommend as a limited replacement for ConnectedText are Notetab outlines. They also work and are much faster. But they have even fewer features. I have written about them before as well (on this blog; just search for notetab). I use them as another "backup" for ConnectedText, that is, I create yearly files, like "Notes 2016.otl", "Notes 2017.otl", etc.

1. For the scripts, see here.
2. I first published this on the ConnectedText Forum, but I thought even people who don't frequent it, might have a passing interest.

Monday, March 5, 2018

Umberto Eco's Library

Here a video of Umberto Eco taking a walk through his private library. I can only wish ...