About A WorkLife FrameWork | Purchase

Notebook History comment...

Posted on: Friday 6/15, 2007;  10:33 AM

I just posted another note to MathGroup that I think is pertinent to this blog. So I am posting it here...  It is in response to this question:

Szabolcs  Jun 13, 7 : 47 am
In Mathematica 6 there is an entry in the Cell menu : Notebook History. Does it do anything apart from tracking the modification times of cells? Is anyone here using this feature?  If yes, how do you use it/why do you find it useful?
I am just curious about this new feature.

Steve Luttrell un 15, 5 : 09 am
I' m using it experimentally to replicate (part of) the effect of using a time - stamped "laboratory notebook".

The time - honoured way to time stamp material (e.g.for patent priority purposes) is to keep a countersigned paper record, but this doesn' t work if you do everything in electronic form (e.g.Mathematica notebooks) unless you use some additional procedures to generate verifiable time stamps.The Notebook History feature looks to me like a first attempt to generate time stamps for Mathematica notebooks.It doesn' t go all the way because you can' t determine the exact state of a notebook at a given point in the past, but at least it shows where you have been going back to "fiddle" with earlier results, and thus gives a last - updated time for each cell.As for whether this sort of time stamp would be recognised in law, that presumably would depend on whether the time stamps were difficult to fake.I think that time - stamped copies of your notebooks are the only convincing approach here.

My "laboratory notebook" has been a web of Mathematica notebooks since version 3.0, and up until now I have carefully added date labels to every group of cells that I create and/or edit, AND I periodically make a time - stamped copy of active notebooks.I don' t yet know the precise way in which I will end up using the new Notebook History feature; at the moment I regard it as an experimental feature.

Steve Luttrell

There are a variety of uses for this feature.  One simply is that it showcases a number of FrontEnd technologies at work since it is built entirely with top level Mathematica code.

Internal to the notebook, cells have the option CellChangeTimes (as long as the notebook has the option Track CellChangeTimes→True which is the default).  And it is these data that are used to track the notebook's evolving state.

When a notebook gets very large though, the Notebook History notebook has a hard time keeping up with the current state of the notebook. For this reason, it might not be as useful without the addition of a means to turn off its Dynamic updating and only display the notebook's state when specifically asked by the user.  I think that this would generally be a better default state (or something like this).

I agree that use of this as a means of verifying Intellectual property creation might be useful.  But I don't believe that the US Patent Office yet recognizes any electronic means of date stamping material.  However, I suspect that it could still be used in some fashion along with traditional countersigned with print \outs of the material.

In A WorkLife FrameWork Diaries are tagged automatically with when they were modified (even in Mathematica version 5.x) and one of the intentions of this is to be able to use them as laboratory notebooks.  But a broader intention is that it allows me to locate cells by their tags (I use CellTags rather than CellChangeTimes because it allows me to associate a variety of dates with the cell and its content).

Also, in A WorkLife FrameWork, there has been a facility similar to the V6 Notebook History menu item in A WorkLife FrameWork's Analytics Palette since the first version:  

http://scientificarts.com/worklife/documentation/analyticspalette.html

In fact, this was the possible origin of the V6 Notebook History menu item.

An example of my use of this is to see when I have interacted with a particular WorkLife FrameWork Diary or Notebook.  If the Diary is one that I have created to keep items associated with a particular client, for example, then this allows me to see at a glance when I worked on things having to do with that client.

Also, once when I was looking at the history of the notebook that I was writing the code for A WorkLife FrameWork in (and autosave package), I noticed that there was a big gap in when I added new material to the notebook.  I wondered about this and then realized that that was the time interval during which I was writing the documentation.

So, things like this are one sort of tool for mining and visualizing one's personal data.

It is important to remember and realize that Mathematica is not just a technical computation tool in the traditional sense, it is a System.  And within the system one can make tools that extend considerably beyond the conventional.  I wanted to catalyze new directions in this by creating A WorkLife FrameWork...

Tagging of all sorts is a very powerful way to work within a notebook and to selectively locate and extract content.  

see

http://scientificarts.com/worklife/documentation/taggingpalette.html

and

http://scientificarts.com/worklife/screencasts/tagging.mov

I hope that some of this helps...

--David

Permalink