Software


28
Jan 09

Steven Johnson on Research and Writing

Erik Marshall points to an essay by Steven Johnson on his research and writing techniques.  Steven is a brilliant writer; it is interesting and humanizing to see the extent to which he uses technology.  I use BibDesk is a manner similar to how Johnson uses Devonthink, but I think I would benefit from the more unstructured approach in Devonthink.  Any readers use it?

The first stage, which is crucial, is a completely disorganized capture of every little snippet of text that seems vaguely interesting. I grab paragraphs from web pages, from digital books, and transcribe pages from printed text — and each little snippet I just drop into Devonthink with no organization other than a citation of where it came from. This goes on for months and months; I read in a completely unplanned and exploratory way (increasingly online, thanks to Google Books and other sources) and just drag anything that seems at all interesting into Devonthink.

….in the last stage before I actually start writing, I create a little folder in Devonthink for each of the chapters. And then I sit down and read through every single little snippet that I’ve uncovered over the past year or so of research. And as I’m reading them on the screen, I just drag them into the chapter folder where I think they will be most useful. Some snippets get dragged to multiple folders; most don’t make it into any folder. But I read through them all, and in reading through them all, I have a completely new contextual experience of them, because I’m at the end of the research cycle, not at the beginning. They feel like pieces of a puzzle that’s coming together, instead of hints or hunches.

Read the full article: DIY: How to write a book – Boing Boing.


22
Dec 08

Announcing Freedom v0.4

I’m pleased to release Freedom v0.4, a major update to the Freedom software.  This new version is a fresh rewrite, incorporating the ability to access local networks, an extended time period, and many significant bug fixes.  Particularly, Freedom’s authentication mechanism has changed (fixing the password-timeout bug) and Freedom will now appropriately handle suspends and sleeps.

Freedom v0.4 remains Mac only, and it is tested through OS 10.5.6.  This new version was paid for (thank you!) by those who have donated in the past.  I strongly recommend that all users upgrade as soon as possible.

Download Freedom directly (.dmg)

Freedom’s website.

If you run into any problems/have suggestions please leave them in the comments.  I do not plan on developing a Windows version, sorry.

Previous posts about Freedom:

Productive Unit Structures: Introducing Freedom
New Version of Freedom: v0.3
Freedom in the Telegraph


2
Dec 08

Hacking Google Scholar

If you connect to Google Scholar through a proxy (for example, through your library’s proxy), you’ll find that GS is unable to remember your preference settings.  Although Google seems to forget my preferences far too often, in the case of Google Scholar it isn’t their fault.  When you connect through a proxy you appear to Google as a different user every time, and until preferences are tied to your Google account (and not a session/cookie), Google is simply unable to remember them.

To “solve” this problem, I’ve found that you can set a bookmark that will set your preferences each time it is clicked.  While this doesn’t solve the problem of Google forgetting preferences between sessions, it will save you the time and effort of having to reset your preferences each time.  You will need to custom-craft your bookmark.  Here’s mine:

http://scholar.google.com/scholar_setprefs?num=50&scis=yes&scisf=4&submit=Save+Preferences

As you can see, in with I’m passing some options to “scholar_setprefs” – i.e. the mechanism that sets your Google Scholar preferences.  I’m manipulating two options, Number of Results (num=50) and Bibliography manager (scis=yes&scisf=4).  I could also directly manipulate the interface and search language, library links, and if the results opened new windows or not (I don’t because I’m happy with the GS defaults).

The options accept a range of values, which I’ll describe briefly:

Number of results (num), accepts:

  • 10
  • 20
  • 30
  • 50
  • 100

If you’d like 100 results to be displayed, you’d change the url so that num=100.

Bibliography manager (scis=yes&scisf=4).  Google Scholar supports a number of different export formats, and to change their default, you’ll need to change the scisf value.  Here are the corresponding values:

  • 4 (Bibtex)
  • 3 (EndNote)
  • 2 (RefMan)
  • 1 (RefWorks)
  • 5 (WenXianWang)

If you’re a RefWorks user, you’d change the string so that it looked like this scis=yes&scisf=1.

Putting it all together, if you’re a RefWorks user who wants 100 results displayed, you’d set your bookmark as follows:

http://scholar.google.com/scholar_setprefs?num=100&scis=yes&scisf=1&submit=Save+Preferences

Finally, if you’re accessing GS through your library proxy, you’ll need to add the proxy information into the URL. In the case of UNC we place the proxy information directly in the url. Therefore, my proxied bookmark looks like this:

http://scholar.google.com.libproxy.lib.unc.edu/scholar_setprefs?num=50&scis=yes&scisf=4&submit=Save+Preferences

As you can see, I’ve added .libproxy.lib.unc.edu to the beginning of the URL. This will vary by library, so you’ll want to look at other proxied URL’s at your institution to get a feel for where the proxy information goes.  As I noted above, there are a bunch of other options you can change directly. If you’d like to change those, simply view the source of the Google Scholar preferences page, look for the option and value pairs in the form, and tack them into the URL (making sure to add & before the option/value pair).


3
Sep 08

Chrome’s reconfiguration of the web’s geography

I’ve really enjoyed Chris Messina’s two recent posts on Chrome.  His background (Mozilla, Flock) and experience thinking about next generation UI’s and UE’s is on fine display; I particularly enjoy his reconcpetualization of the browser and its experience.

Factory City: Google Chrome and the future of browsers

Factory City: Musings on Chrome, the rebirth of the location bar and privacy in the cloud

In the second, more recent post, Chris discusses the cognitive break inherent in Chrome’s vision of the web.  In removing the URL bar in favor of a single search interface, the web transforms from one spatially and locationally grounded (in URL’s, permalinks and bookmarks) to a fully-mediated, amorphous zone of information.  In this new web, there are no wrong answers or incorrect URL’s, because the algorithm always has information relevant to the intent of our information need.

As Messina notes, the tradeoff is such “that our fundamental notions and expectations of privacy on the web have to change or will be changed for us. Either we do without tools that augment our cognitive faculties or we embrace them, and in so doing, shim open a window on our behaviors and our habits so that computers, computing environments and web service agents can become more predictive and responsive to them, and in so doing, serve us better.

That is, in embracing the mediated web, we trade (to some extent) our agency, any sense of privacy, and most importantly, our extant strategies of finding and reminding for new, less conceptually transparent ones.  To embrace the web in Chrome’s model, we must embrace the algorithm, and essentially invite it into our minds.  This new lens is all-or-nothing, and it casts away our strategies of past, those operationalized in pre-web design patterns.

On one hand, one might be able to argue that the web is so vast that inviting the algorithm home might make sense.  Perhaps it is better to browse with Google on your shoulder, assisting your navigation and selecting your best information choices.  Where I run into difficulty with this model is that Google is placed at a meta-layer above the web; it becomes the lens through which the web is experienced.  This model is troubling at many levels, but I particularly resent the idea the web should be mediated.  Slightly repurposing JPB’s Declaration of the Independence of Cyberspace:

We have no elected government, nor are we likely to have one, so I address you with no greater authority than that with which liberty itself always speaks. I declare the global social space we are building to be naturally independent of the tyrannies you seek to impose on us. You have no moral right to rule us nor do you possess any methods of enforcement we have true reason to fear.

Of course, this isn’t a question of morals; the web is a market, and there will always be a choice to opt-out or not participate in Google or anyone else’s schemes.  The gray area emerges when we consider Google’s place in the market, and the sheer power they exert in the configuration of consumer preferences.  Thinking as an educator – we lament the so-called death of the book.  In five years, will we lament the death of the URL, in an age in which all authority is conferred through the end-product of a citation-based algorithm?

All of this comes with a grain of salt.  Personally, I believe our current spatial metaphors of the web will exist for the imaginable future.  As revolutionary as these ideas seem, we change slowly, and the browsing and searching patterns of billions of web users are already well-established.  Further, this sort of change is essential – we’re constantly reconfiguring the web and our experience of the web – I just question how much we need to do that with Google looking over our shoulder.


2
Sep 08

Google Chrome Privacy Information

Via Vowe.net, the Google Chrome Privacy Policy (scroll down, soon to be located here).  It appears that Chrome will follow a pattern similar to the Google Toolbar – that is, all browsing behavior is sent to Google, but an opt-out is provided.  From the Chrome Privacy Policy:

  • When you type URLs or queries in the address bar, the letters you type are sent to Google so the Suggest feature can automatically recommend terms or URLs you may be looking for.
  • If you navigate to a URL that does not exist, Google Chrome may send the URL to Google so we can help you find the URL you were looking for.
  • Your copy of Google Chrome includes one or more unique application numbers. These numbers and information about your installation of the browser (e.g., version number, language) will be sent to Google when you first install and use it and when Google Chrome automatically checks for updates.  If you choose to send usage statistics and crash reports to Google, the browser will send us this information along with a unique application number as well.

The last bullet is particularly interesting – each Google Chrome browser is fingerprinted so it can be uniquely identified.  It should be noted that Google isn’t the first to fingerprint their browsers – Microsoft tags Internet Explorer with a Globally Unique ID.

As I previously noted, Google is allowing users the ability to opt-out of statistical reporting.  I worry that those who opt-out will not be provided the full browsing experience, compelling users to participate in the statistical reporting.  Furthermore, close attention should be paid to “advanced” features that provide additional reporting, above and beyond the standard statistical reports.  Google Toolbar contains a number of these features that report URL’s, typed information and page content.

The best approach is for Google to be extremely open with Chrome and its privacy practices.  Indeed, open sourcing the code is good – but Google should go a few steps further and meaninfully address the issue in a human-readable format.  Google’s argument about a next-generation browser is solid, and I would be willing to give it a shot.  First, however, Google must win my trust.


29
Aug 08

Firefox 3 Tweaks

I’ve recently moved to Firefox 3, and I’m pretty pleased with the performance.  Firefox 3 feels snappy, seems to handle JS and memory leaks well, and is all-around pretty impressive. Here are my tweaks:

I’m not a fan of the awesome bar – I simply don’t like interfaces (like Google Suggest) that create a lot of activity while I’m typing.  To disable the URL bar, set browser.urlbar.maxRichResults to -1.

Also worth noting is that the malware and phishing protection that come default in Firefox 3 do send your browsing history to Google.  This is not new from Firefox 2, but it is worth mentioning, as you are uniquely identified and correlated in the data.  To turn this off, de-select the two “Tell me…” options under Firefox’s Security settings.  I ran packet traces and verified this does stop Google data collection.


26
Aug 08

Twitter, the enterprise’s third space

When describing Twitter, I use a number of analogies. Most commonly, I think of Twitter as something like a slow-motion chatroom, or even a collection of away messages. I’ve got another one to add to this list: subject-only email.

This new analogy actually comes from my use of the iPhone, where the Twitter interface isn’t all that different from the mail interface. Twitter displays a sender and a brief message, which mirrors the sender and subject elements that you’d see in an email inbox. The main conceptual leap is that with email, there’s often a payload of information, tasks or spam waiting for you. With Twitter, you’ve only got the message (an an occasional URL as payload). This very fact is why I enjoy checking Twitter, and detest my inbox.

Thinking about Twitter this way helped me imagine enterprise integration of a Twitter-like service. I envisioned adding a “Twitter pane” to email clients – a pane for Twitter-like communication aside the inbox. This Twitter pane would act as an ongoing message thread away from the inbox, and its uses would be more conversational, informal and informative.

Imagine the scenario of a guest lecture. Let’s say you’re bringing in a friend to give a talk at your company. You send out the detailed email notice, and maybe a few follow ups. To attend this talk, your coworker must process the email, calendar it, and remember. If the talk is fairly last-minute, that coworker needs to be attentive to email information coming in just-in-time. In these models there’s no space for casual prompting – replying-all to mailing list to say you’re “going to hear Sally’s talk” is generally outside of norms. However, the enterprise Twitter affords a communicative third space – a place for coworkers to discuss, remember and remind one another of the lecture, by virtue of their discussion (and perhaps live-Twittering) of the event.  In this sense, the enterprise Twitter surfaces the collective, prompting observers to action.

The enterprise Twitter gives rise to a new channel of communication that offloads from the inbox, and introduces new forms of communication.  In offloading the inbox, one can imagine common/frequent para-social tasks like casual lunch invites moving to the Twitter channel. In fact, the public nature of Twitter might provide unique opportunities to meet others – it might be a little strange to invite a stranger lunch, but a “who is hungry?” message to the public might allow an ad hoc group to form. In new forms of communication, one can imagine messages that might not pass the listserv test (“Can someone help me with this Perl?”) getting passed to the semi-public of the enterprise Twitter. This presents the opportunity for new connections, more efficient work, etc.

The enterprise Twitter is most interesting for its potential vibrancy. Corporations have adopted internal social networks, and while these networks represent a more robust directory, I doubt many would qualify as particularly vibrant. This may be because corporate social networks don’t really address employee needs,  rather addressing the needs of management in analyzing and diagnosing the “structural holes” of the organizations. An enterprise Twitter does address a very real problem – our ever-overstuffed, mismanaged inboxes – and it introduces a vibrant and relevant communication channel to the enterprise.  The enterprise Twitter might just be the electronic, distributed water cooler of lore.

In implementing an enterprise Twitter, I’d argue that one would certainly want to follow the 140-character limit, allow private, public, and semi-public (conversant) threads. The enterprise Twitter would be inside-the-firewall, and would also follow the limited profiling pattern of Twitter (Name, Bio, Link). Political considerations should be addressed. Perhaps an arbitrary follow limit of 100 would be useful – this would prevent everyone from simply following upper management out of “respect.” Other potential benefits would include the enterprise Twitter as a news or safety channel (posts could go out regarding severe weather and so forth).

Although I’ve used the brand name Twitter thoughout this post, there’s nothing Twitter-specific about the practices I’m discussing. The enterprise Twitter is just a directory, sets of permissions, a messaging protocol and integration into the messaging client. If it sounds like a replication of email, the key differences are only the social affordances and the mental model.

This scheme is not without drawbacks. Primarily, the cost of designing and integrating messaging system is not trivial. I would respond that Twitter has introduced a new type of message, one that our systems and devices should support – therefore this integration is inevitable. Another drawback is distraction. Twitter is a notorious time waster, it is addictive, and it is always on. To this I would drawn an analogy of the modern inbox – it is never off, more corporations expect you to carry mobile devices – so a Twitter management strategy should fit into larger communications-management strategies.

Inside the enterprise, individuals rely on a variety of tools and strategies to stay in communication. None have the unique affordances of the enterprise Twitter, and few offer the common social bridging role of an enterprise Twitter. My thoughts are obviously preliminary, but if anyone is working on such a project or thinking of beginning one, I’d be very interested in your thoughts/feedback.