stéfan sinclair :: scribblings & musings

We’ve made available a preview release of Digital Texts 2.0, an attempt to experiment with social networking practises in the context of interacting with electronic texts. Although we have a fairly detailed scholarly agenda for this project, one of the things I’m most curious about is whether or not students would be interested in using a Facebook application to interact with texts, whether it be for pleasure or for course work. Similarly, can instructors find innovative ways to incorporate such tools into the classroom?

Some key features currently available:

• Add and browse Texts (via Amazon lookup, or manually)
• Organize your texts into Collections
• Join Groups of like-minded Readers
• Comment on and add Tags to Authors, Collections, Texts, and Groups
• Share your findings with Friends

Some upcoming features (probably by the end of the summer):

• Import/Export feature set
• Citation Generation tools
• Hybrid Searches combining Authors and Readers of Texts
• Text Recommendation system

Are you planning on using Digital Texts 2.0? Please let me know!

Thanks to the Digital Texts 2.0 team and especially to the heroic efforts of Johnny Rodgers, the programmer and designer, and Shawn Day, who has provided outstanding feedback.

Geoffrey Rockwell and I have just completed our first full experiment in pair text analysis: Now Analyze That. Using text analysis, we wanted to try to say something interesting about a corpus of texts, and I think we succeeded; the point wasn’t to do a thorough and theoretically meticulous reading of the corpus (as both of us are trained to do), the point was to use analytic tools to relatively quickly identify some potentially interesting directions for further study. Actually, we had more objectives than that:

• experiment with doing pair text analysis (one person at the keyboard, one person focusing on the bigger picture, as per principles of pair programming)
• experiment with a form of rhetoric that harmoniously combines prose and analytic evidence (or even embedding tools directly into the analytic text)
• begin a series of mini analytic exercises that could serve as methodological examples (for teaching purposes, etc.)
• work with our analytic tools (Taporware and HyperPo) to identify their strengths and weaknesses for this type of analysis

Geoffrey and I will be doing more experiments with the rhetoric of text analysis and tools methodology…

The Text Analysis Developers’ Alliance is a very loose kludge of coders (to coin a collective noun) – actually, it’s not just coders, but designers and users of text analysis tools as well. The group, founded in 2005, has as a mission to encourage collaboration among researchers and to provide useful resources to developers and users. TADA has slowly but surely been expanding its Wiki, which now includes a wide array of information of potential interest to those involved in text analysis.

Shawn Day, who wrote most of the very useful TAPoR Recipes, has just completed a nice redesign of the front page of the wiki – thanks very much to him for that.

TADA is currently planning a tools contest – watch this space for more information about that very soon.

I ordered my MacBook Air soon after they were announced in January 2008 – I needed to upgrade, and I was intrigued by the ultra-thin and light form factor, as well as the idea of having an SSD, flash-based drive (similar to memory cards in digital cameras), which I’m convinced will be the norm in a couple of years. I’ve been thoroughly delighted by my choice: this machine is precisely what I wanted. I don’t miss the integrated optical drive (I have the external one), the machine is plenty fast (faster than my previous Intel-based MacBook Pro on most things), and – a bit to my surprise – I don’t even miss the larger screen (13.4” instead of my previous 17”). I can even live with the minimalist ports (a mere three: audio out, USB, external display), though life requires a bit more patience without Firewire 800.

Unfortunately, my drive died. Some drives are simply lemons, and it certainly doesn’t help that this is a newer technology as applied to internal drives for computers, but the irony is that Apple trumpets the reliability of SSD, which has no moving parts for enhanced durability.

Fortunately, I’ve learned my lesson before, and I’m pretty diligent about making backups, especially now that it can be fairly automatic with Time Machine and my Time Capsule (wireless base station with integrated drive for backups). Automated backup has been fairly trivial for desktop machines for quite a while (with scheduled software and an external drive), but the magic of the wireless solution is that it works so well for laptops as well, since it backs up whenever it can and tolerates interruptions.

Which isn’t to say that there weren’t any problems (which is actually the main motivation for this post). One challenge I was aware of is that Time Machine backs files up, but it doesn’t create a bootable drive, which can make accessing the backed up content a bit difficult (there are solutions to having having Time Machine co-exist with bootable disks, though I’m not sure any of them are as tolerant to the mobile nature of laptops and transient connections). Another challenge is that Time Machine backups made through a wireless connection aren’t directly browsable – the sparse bundle disk image can’t be navigated as easily as when Time Machine works directly with an external drive, because it’s actually using network storage. For whatever reason, I couldn’t get “Browse Other Time Machine Drives” to work properly either, perhaps because it was a network backup (that option is available when you Ctrl-Click on the Time Machine icon in your Dock – it’s not available from the menu bar or from the preference pane). The best work-around I could find was the following:

• start the migration assistant application (in your Utilities folder)
• select migration from Time Machine backup
• select the correct drive and connect (don’t migrate)
  • open Terminal (also in your Utilities folder)
  • type: ls /Volumes
  • note the likely name of the Backup image
  • type: open /Volumes/[backup image name above] (type in the first couple of letters of the drive and press tab to auto-complete the rest)
• browse into “Backups.backupdb”, the name of your computer, the backup you want…

Not the easiest procedure, but it will allow you to retrieve files in a pinch, if – like me – you can’t navigate to them otherwise.

Anyway, I did finally get my MacBook Air back with a new drive and I performed a full restore from the Time Machine backup, which worked fairly smoothly. (Again, there were a couple of hickups: you can only restore when booting from the installation DVDs, but the DVD drive takes up the one USB port on my MacBook Air that I’d prefer to keep free for my Ethernet dongle in order to restore faster from the network drive – I should have tried my USB hub but didn’t. When I booted with the DVD in another machine, which is now possible for MacBook Airs, I was unable to mount the network drive properly (permission issues that weren’t a problem when booting locally). So I had to boot from the external DVD drive and complete the backup over wireless, which was slower, but that’s ok since I was sleeping through most of it.)

As reported elsewhere, Apache didn’t work properly until I manually created the needed logs directory sudo mkdir /var/log/apache2 (the logs directory deliberately isn’t backed up by Time Machine). Also as noted elsewhere, Mail had to re-index all my mailboxes, but failed to do so properly, which would have left me without a huge number of my messages (some messages were simply corrupted, but for some reason all the messages added since my previous migration to the MacBook Air weren’t imported properly). The best solution I found was to open the relevant directory in Time Machine (~/Library/Mail/Mailboxes/Archives) and to restore the directory as it was before the disk crash, and then to manually rebuild the index (Click on the Mailbox file menu and choose rebuild).

Finally, if I were starting over, I’d probably try these instructions to avoid Time Machine starting a backup from scratch after the restore instead of just continuing from where it was.

Sorry for the longish post – hopefully someone will find some tip of use if they run into any of the same problems that I did. And now I’m back from the dead.

So I asked my trusty (design, etc.) collaborator Johnny Rodgers to mock up some business cards. As usual, he did some great work, but I mentioned that maybe the logo could be a bit bigger. He promptly responded with a link to Make My Logo Bigger Cream. Don’t mess with the designer. ;-)

Anyone who does web development on Mac has a major thorn in his or her USB port: how to test content for the way 4 out of 5 people see it in Windows Internet Explorer (may Mac Internet Explorer rest permanently in peace). Since Mac moved to Intel chips, there are excellent virtualization solutions (like Fusion and Parallels) that allow you to run OS X and Windows simultaneously, but those solutions can be resource intensive (not to mention costly, once you factor in a legal license for Windows).

ies4osx is a great solution that allows you to run different versions of IE (5, 5.5, 6, 7 – which you can’t do in a Windows installation) without running Windows, thanks to Darwine, a port of the Wine libraries for Mac OS X. Turns out you can have your cake and eat it too.

Stephen Ramsay, Digital Humanist extraordinaire, posted a blog entry about wanting to learn Groovy, and made a comment about Java being the predominant language in Digital Humanities. I posted a comment indicating that scripting languages like PHP and Perl seem to me to be more common, for noteworthy cultural reasons, like the availability of student research assistants and the desire to get something up and running more quickly. Neither of us really cares that much which language rules, but both of us care very much about the pedagogical implications.

In a follow-up post entitled Language Games, Stephen mentions that his students are generally happy to learn whatever language he chooses to teach. He must command much more respect from his students, because mine don’t always seem as convinced by my choices ;-). As I’ve mentioned before, I spend a lot of time thinking about what languages to teach students and to use in my projects. I certainly don’t begrudge that intellectual effort – that seems like an essential part of humanities computing: not only doing the technical stuff, but thinking about why we’re doing it the way we are.

Anyway, at the end of his post Stephen asks readers (especially digital humanists) to chime in about the languages they use. Given that I don’t have to moderate the discussion on his blog, I’d like to cast the net even wider: if you’re a student in the humanities, what language(s) would you want to learn and why? If you have a moment, please drop by Stephen’s blog and add your pennies. This may make very interesting anectodal evidence to add to the results that Susan Schreibman will hopefully present this summer based on her survey.

The Adobe CS3 Master Collection is a treasure chest of well-respected applications for graphic design (Photoshop and Illustrator), video editing (Premiere and After Effects), web authorship (Dreamweaver and Fireworks), and various other useful tools (like Acrobat Pro). All these goodies don’t come cheaply, of course: the regular version is $2,500 USD, though the academic institutional pricing is very attractive at about $500.

I’m actually not a regular user of any of these applications, but I was still fairly excited to install the latest versions on my Mac, especially to get the speed boost of the universal binaries. Imagine my shock when I fired up the installation disk and received this error:

This software cannot be installed because the file system of the OS volume is not supported.

No thanks to the informative value of this vague error message, a bit of searching revealed that the issue was that my disk was formatted to be case sensitive. What? You advertise the added creativity, productivity, and efficiency of your new software and then you expect people to reformat their drives and re-install their systems just because you don’t support case sensitive systems? Now, granted, a case sensitive disk isn’t the default for Macs, but I always reformat my drive to be case sensitive for several reasons, including the benefits of having the case sensitivity of my local development environment match that of most deployment servers. Having a case sensitive disk isn’t exactly an obscure practice. Come on, Adobe, let’s be sensitive here.

So, yes, I did bite the bullet and obediently waste most of a day reformatting my drive and re-installing everything – open software isn’t returnable, ya know. Fortunately it was time to do a clean install anyway. The more difficult question might be why I purchased the CS3 in the first place, given that I prefer web development in Eclipse and that there are so many great open source and low-cost alternatives for graphic design, such as Gimp, Inkspace, and Pixelmator. Adobe products are installed in our Multimedia labs at McMaster, so I’m going to blame technical support for our students, ok?

Given the recent irregularity of my blog posts, I don’t imagine there are a lot of regular visitors to my site, but those who do visit occasionally might have noticed that I’ve finally removed my “temporary message” claiming that I’d be filling in additional contents within the next few days (I had added the temporary message a couple of months ago).

I often feel overwhelmed by all of the interesting things happening, and there’s certainly no shortage of material to blog about. And yet, there are spells where I have difficulty overcoming the inertia of silence (as I’ve mentioned before). In any case, I’m going to now embark on a set of posts related to some of my research projects. Historically I’ve been reluctant to talk too much about my own work, perhaps in part because I view this blog as more outward looking (noting things of interest happening elsewhere), but maybe these posts will be of interest to some of you as well. And even if not, it’s a good way to discipline myself to start filling in my research page (don’t let the “coming soon” fool you ;-)).

I try not to think of myself as a Mac nut, but sometimes it’s hard to ignore the evidence, like the irrepressible urge to upgrade immediately when Leopard was released. I had no specific reasons to upgrade, aside from a vague sense that there would be a few enhancements, and the desire to be “current” (at least in terms of operating systems). The novelty wore off within about half an hour, and I was more or less back to business as usual. Time Machine does relatively little for me (literally, since I disabled it after getting a decidedly unhelpful error message), Stacks are poorly implemented IMHO, iChat effects are only funny the first time, etc. The one thing I do find useful is the RSS reader in Mail – much more useful than in Safari. In any case, I’d accepted the idea that I’d upgraded by an uncontrollable compulsion, but that it would be difficult to rationally justify. Until today. I found a mention of a mosaic screensaver function in Leopard that is absolutely stunning. I feel like getting a dedicated machine and a huge wall-mounted display just to show the mosaic feature all day, every day (but I won’t ;-). Essentially one picture pans out and you see a mosaic of other pictures until that mosaic becomes another picture, and so on. It’s hard to do it justice, there’s a video that gives you the gist. So now my upgrade to Leopard is a phenomenal screensaver (the irony is that I normally don’t even use a screen saver, I just power off the screen), and a bunch of other secondary features thrown in as a bonus…