Flickr as image backup

A good friend is a tenured scholar at a small school that does not routinely provide secure network storage space to its faculty. She described art and architecture students keeping their digital work on USB drives and revealed that her second book manuscript resides on her laptop and, occasionally, gets backed up onto a CD. Her digital images live exclusively on her hard drive.

I have gotten so used to debating the merits of tiffs vs jpegs and worrying about the archivability of word-processing and CAD files that I was astounded to learn that there was a degree-granting institution in North America that doesn’t provide even minimal network storage space to its faculty. Am I hopelessly out of touch? Is this, in fact, common? I do know that there are plenty of unaffiliated scholars (and tiny museums) who have to make do without IT support but I had assumed that network space went along with office space as the minimum provision for university-level teachers in 2008.

What do people recommend for those scholars whose work increasingly depends on digital files but who lack the institutional support to maintain them? Here I would include independent scholars, adjuncts, professors at small schools, and graduate students. This problem is especially acute, I suspect, for those of us who work on visual materials and have personal catalogs of digital photographs numbering in the many thousands.

What I suggested to my friend was, first of all, that she email her manuscript to her husband, so a copy gets onto her and his email servers, where it has a fighting chance. Second, I suggested that she start migrating her digital images to Flickr, which offers unlimited storage of image files under 20MB for $25 per year with its pro accounts. I also suggested she might invest a couple hundred dollars in a RAID backup, though this doesn’t get the files off site.

Is anyone else, besides my friend, losing sleep over this? What other solutions are out there?

And yes, I know about hard-disk recovery solutions, like this one, which saved my bacon once to the tune of $2400. But relying on this as a backup strategy is, well, short-sighted.

THATCamp

edwired has a series of posts on the recent THATCamp at CHNM at George Mason. It sounds like it was an exciting, productive weekend. The website and blog associated with it are rich resources, overflowing with useful links for people interested in technology and the humanities. There are a lot of very interesting projects underway out there, so click around…

Teraflops for the Rest of Us

I missed this announcement while I was on the road but back in April, the Chronicle of Higher Ed reported that the NEH Office of Digital Humanities is sponsoring a grant program that will offer 100,000-hour bits of supercomputing time at the Lawrence Berkeley National laboratory. The plan is to get humanities researchers joining chemists, physicists and mathematicians developing research projects that require high-performance computing.

My first reaction to this was “whoa, cool.” My second was, “we’re so not there yet.” At least if the “we” is scholars of the material world. Unlike text-driven scholars, our stuff simply isn’t digitized in large quanitities yet. On the other hand, some of the relevant documentary materials ARE digital, like historic census and probate records, city directories, etc.

Brett Robley, director of the ODH, points out that scientists, too, had to learn about supercomputers before devising projects that could take advantage of them. The official announcement is here.

The mind reels.

Cataloguing Buildings in Zotero

For anyone who writes with footnotes, Zotero is a fabulous tool. With a click of a mouse, it imports catalog records from a library database, or JSTOR, or even Amazon, allowing a scholar to create a personal reference datbase on his desktop. Better still, it permits extensive annotations, keyword tagging, and hyperlinks both to other items in the database and to external materials. Some users know that it can catalog images, too, pulling metadata from Flickr, or HABS. If you already run Zotero and need to work with images, try it. The possibilities are mind-bending for those of us who work with visual resources.

Its handling of images in a structured way makes me wonder: could Zotero be used to build a massive database of buildings? In other words, if the metadata reveals that multiple images represent a single building, could Zotero be made to recognize the fact? Since it’s open-source, I assume the answer is yes, and that the only real question is how difficult it would be.

Imagine a web-based resource of architectural photographs, such as the one that the SAH is building with ArtStor, called the AVRN. Now imagine that the metadata for each of those images was carefully controlled, so that City and State fields were populated consistently, as well as Building Name, Date, etc. Zotero could grab the thumbnail, link to the image file (optionally downloaded), and make a record of the relevant metadata (including latitude and longitude) in the database. A handful of fields would need to be added to Zotero’s structure for this to work well for architecture.

In time, as hundreds of images were ingested, Zotero would beocme an index to buildings, sortable, say, by region, or date. The image records could be linked, furthermore, to secondary resources, like JSTOR articles, or texts referenced in Google Books.

Now imagine a user who kept his own image library in tip-top shape, with metadata carefully applied in a custom XMP panel for architecture. If Zotero could pull XMP/IPTC metadata from image files, it could incorporate, as well, locally held material. And if images posted on Flickr had carefully structured metadata in XMP fields, these, too, could be brought in (at the moment, I think Zotero is just pulling metadata from Flickr-specific fields, not XMP).

Using Zotero to manage image references would be a marvelous way to maintain research collections on architecture or, i presume, any visually oriented subject. The problem is that, unlike books or journal articles, there is no single standard for cataloguing buildings. The AVRN might develop a standard for its own materials but this would only be as exhaustive as its own collection.

A Union List of Buildings would be a fabulous resource for those of us who study architecture in any capacity: find a structure on a map, or in an index, and retrieve the basics–location, date, primary construction material, and architect or builder, if known. With a union list and a MARC-like record for each building, Zotero or a similar tool could, with a click, import correct, fielded metadata for an image, and be a central tool for bringing together all sorts of references, from images and texts to police blotter entries and property tax lists. Best of all, with good metadata, there would be no need for any of the original resources to be held locally: the Zotero-like database could be, like Zotero itself, a lightweight annotated list of references and links.

Of course all of this depends upon the existence of a Union List of Buildings, which doesn’t exist. Who will build it? And how?

Computers and Fieldwork

After being diverted and delayed for 24 hours due to bad weather in Chicago, I returned late last night from the 2008 annual meeting of the Vernacular Architecture Forum in Fresno, CA, where, among other things, I was elected to the board of directors, to thunderous applause. At least that’s how I remember it.

At recent VAF conferences, there have been anxious discussions about the status of fieldwork and the utility of digital technology. These have not usually been brought together until this year, when it happened in three separate sessions, including both of the ones in which I was a panelist (incidentally, does anyone else feel that it’s time for us to extend the number of time slots for sessions from 3 to 4?).

In her discussion of the Gaspe field school, Tania Martin observed, quite pragmatically, that all the new digital gizmos we have at our disposal are only tools, with particular strengths and weaknesses, and that it’s up to us to understand how and when they are useful, and when the old ways are better. Immediately after Tania’s sensible and seemingly uncontroversial remark, two of the other panelists  were both compelled to point out that they didn’t think too much of new technologies in the context of fieldwork, pointing out that they require electricity, for example, and that the visual qualities of hand drawing are superior to CAD drawing.

I suppose this is true as far as it goes, but why do VAF participants still feel they need to advertise their skepticism (or worse, their ignorance) about digital technologies in general? Film cameras require a complex chemical process to make a photograph but that doesn’t seem to bother anyone. We need to call a moratorium on analog posturing at these meetings, like bragging that we don’t have a cell phone. It’s irritating, unproductive, and takes time away from more sensible, thoughtful conversations about our work that are sincerely open to new possibilities.

This conversation might be easier if we re-focused the discussion away from particular technologies, or tools, and toward research activities. The question need not be: is CAD better or worse than ink-on-mylar? It should be: when does CAD help me do my work?

The two techniques that came up in this context on Saturday were CAD and laser scanning. Laser scanning is a form of “remote sensing,” or measuring a thing without touching it. The technology makes use of a spinning laser and a measuring device that is attached to a computer. A technician sticks the device on a tripod, sets it to run, and walks away. For a few minutes, the laser whips around the room, recording distances at regular intervals and passing the results along to the computer. With enough measurements, the system can create a good representation of the space via a “point cloud,” or a series of measured points that can be translated into a 3-D model in CAD software. With the best devices, the point cloud can be dense enough that it captures quite refined details, like molding profiles and door knobs.

It’s a cool technology, especially for developing models of very complex spaces or ornate details, both of which would be extremely laborious to measure by hand. It’s a fantastic way to record the spatial characteristics of a building or a sculpture, for example, and probably the best way to capture, say, the ornate carving on a capitol. The process is time consuming, however, requiring a skilled technician to translate the points into planes and curves. It is, for the moment, expensive. It also fails to record color, and lacks the resolution to pick up cracks and other subtle changes in wall surface that can be important evidence in old buildings. It is not, in short, a replacement for looking carefully. It is a mathematical representation of a building’s formal characteristics that does not make analytical sense of it. If we define fieldwork as a process of understanding a building through recording it, laser scanning is not a great tool for fieldwork.

Clearly, however, it has enormous potential as part of a larger enterprise that includes close looking and careful annotation. In a complex building with many changes, a knowledgeable fieldworker will be required to make sense of how it developed over time and relate that information to the model. The newer, and simpler, the building, the more complete a representation the laser scan becomes. It is surely not a way to record the Peyton Randolph House in all its complexity but it might well capture most of what we want to know about, say, a modern warehouse, or a Target store.

It would be marvelous to have a scan of a building in hand before doing the analytical fieldwork that we still see as fundamental to our scholarship. If a technician can be given the task of making a scaled model, from which one can generate careful scaled plans, sections, and elevations, a field scholar can concentrate on annotating the building with descriptions, notes, etc. I might also add other kinds of documents to the model, such as photographs, allowing it to become the digital framework upon which I hang my analysis.

Modeling, in general, would only ever be one piece of the documentation puzzle. It is a means of representation, not a means of recording. Laser scanning makes modeling simpler, and perhaps makes fieldwork easier, but is not a substitute for fieldwork any more than board drafting is a substitute for field measuring.

And so we return to drawing. Drafting, whether by hand on a board or in a CAD program, is a technical skill. There is no romance to it, nor has there been since the professionalization of architecture in the nineteenth century defined the production of construction drawings as manual wage labor. CAD programs certainly reflect the technical nature of this process and perhaps we are right to resist market-driven definitions of our work.

On the other hand, CAD does bring the production of clear, clean, competent drawings within reach of even the clumsiest of us. We all love to look at a beautiful, hand-drawn illustration, such as the ones by Cary Carson and Chinh Hoang in “Impermanent Architecture” and Tom Hubka’s in Big House, Little House. But the ability to produce such images is a very rare skill. We can’t all be David Macaulay and we delude ourselves if we think otherwise. For most of the drawings that most of us need to produce, a CAD plan is likely to look better than a hand-drawn one, and can be completed much more quickly.

Some of us, to be sure, will continue to derive enjoyment from drawing with ink on a board but for most of the VAF, and for most of our students, it is foolish to imagine that ink-on-mylar has any future beyond a romantic hobby. Those who can draw beautifully will continue to do so, and my hat is off to them. The rest of us will be better served by learning to use the computer.

Computer History Museum

Last night I arrived, for the first time, in California. I’m staying in the Bay Area for a few days before the 2008 meeting of the Vernacular Architecture Forum in Fresno. One of my first stops was at the Computer History Museum in Mountain View, a stone’s throw or two from the Google headquarters (one of tomorrow’s stops). This is a very interesting place and well worth a visit for anyone who has the slightest curiosity about the subject. I haven’t fully processed the experience, so this post represents a first crack at what I saw today.

First off, I should say that, for such a new institution, the place shows itself well. Exhibits are well lit and well presented, the building is spacious and new, and best of all, admission is FREE. There were about 20-25 other visitors in the hour-and-a-half or so while I was there, which made for a pleasantly quiet visit.

I’ve wanted to see the Computer History Museum for about a year, as I am fascinated by the idea of a museum devoted to objects whose significance can’t be grasped (or even guessed at) by just looking at them, at least for most of us. They need, badly, to be interpreted. It’s hard enough presenting history in a museum environment, which tends to require either acres of wall text or small armies of interpreters (the CWF approach).

On the main level, there are currently three exhibits: one devoted to “Silicon Valley Pioneers,” which consists of short biographies and photographs of 20th-century movers and shakers in the industry, accompanied by a couple of objects each and mounted in display cases. This is familiar stuff in a familiar format. Substitute governors, or abolitionists, or baseball players, and you can imagine a similar display in your favorite history museum.

A small exhibit devoted to the history of chess-playing machines was, likewise, familiar in format, with large banners, lengthy wall text, and illustrations telling the story from the Mechanical Turk to Deep Blue. Because this is a modern museum, there are video displays, too.

The main event in the museum, however, and certainly the main reason to go, is its collection of historic computing devices, from abaci and slide rules to calculators to a couple of supercomputers and rows upon rows of personal computers. These are all arranged in a large room, organized by type and then, seemingly, by chronology. The name of this exhibit is “Visible Storage.” I only had about an hour in it but I must say that i was very sorry not to have had more.

There is very little interpretive text in this exhibit. For the most part, labels describe the particular innovation that one device or another represents: a leap forward in memory storage, for example, or one of the first Cray supercomputers. But if you’re looking for basic information, such as “what does that button do?” this is utterly lacking. And of course it is. For the more ancient devices, such as a 1959 DEC PDP-1, or a Cold War missile guidance system, knowing what the “On” switch is wouldn’t help much. The ONLY way, really, for a non-technologist to make any sense of this is with some context.

Context, then, came in the form of an enthusiastic docent, who was leading a tour around for much of the time that I was there and generally seeming to chat with anyone who would listen. I’m not sure how else you could explain this stuff to anyone.

The exception to the need for extensive interpretation, of course, are the computers that a visitor has had some direct experience with. From listening a little to some of the other visitors, it was clear that I was the tyro in the room (“do you know why they wired these things this way? It’s because they needed to…” “Ooh, boy, I threw away two of these just a couple of years ago.”) Still, it was strangely affecting, and somehow validating, to see the first computer I ever used, a Radio Shack TRS-80, on display. And then there was the Commodore 64, and a couple of Apples, and a box for Zork I, and the fancy HP calculator I had in high school. And then, whoops, on a table a few feet away, here’s the on-board guidance computer for a Minuteman I.

Despite the destructive power at the business end of a few of these devices, and despite the scrubbed appearance of some of the corporate computing machines, there’s something very ad hoc about most of the displays that make it clear how much of this history was made up along the way. There are those little raised-letter, stuck-on labels (original!) everywhere, and tiny notes stuck onto panels (“Patch the input first and the output second, or, turn off the plate power.”) Though I have no idea what most of these early machines did, it’s not hard to sense some of the delight their original users had in just making them function, to say nothing of getting them to solve particular problems.

The weirdest object in the room, and the one that surely warrants a more extended discussion than I have the time or patience for, is a sleek “Kitchen Computer,” offered by Neiman Marcus for $10,600 in 1969. Its interface consists of a series of obscurely labeled switches and LED lights. Long and low and made of lovely red and white plastic, it was meant to store and retrieve recipes. The panel describing it notes that “there is no evidence that any Kitchen Computer was ever sold.”

A Blog? Seriously?

How can anyone with any claim to scholarship spend time on something as seemingly frivolous and self-absorbed as a blog? To be sure, there is no shortage of dreck on the web but a weblog is no more and no less than a medium for written communication and therefore a useful tool. I have taken up this tool for two reasons, both centered around the utililty of computing for architectural history.

First, there is little understanding evident among my peers about the potential of digital technologies for architectural research. Our friends in history, archaeology, architectural design, literary studies, the social sciences, and even art history have begun to explore digital possibilities for their work with much more enthusiasm. So, I hope, this blog can help shine a little light on what architectual historians have to gain from computers.

Second, I want to explore, through writing, the possibilities of working digitally and the larger implications of rendering buildings in digital form. What is gained or lost in representing architecture in a digital photograph, or a digital model, or a database? What happens to our work as its publication moves from print to digital, and from the library to the web? To help understand these questions, I am beginning to explore the history of architectural representation in writing, drawings, and models to help me place digitization within our particular disciplinary context.

My hope is that short posts on particular digital topics–imaging, CAD drafting, database construction, for example–will, over time, develop as useful web-based reference materials for architectural research. I have posted several notes to email discussion lists on some of these subjects over the last few years, and given advice to friends privately, but this material is not widely available even though experience suggests that it should be.

Finally, I hope that the discipline of writing short public pieces on a regular basis will help me clarify my thinking on architectural representations and motivate me to push my research along. Because I work in a museum, I do not have a regular audience of students for testing ideas but a blog seems a fine venue for semi-baked notions that are not ready for a conference paper or publication. The very qualities that make blogs seem to lack seriousness (they are unvetted, spontaneous, and ephemeral) are the same ones that make them an ideal testing ground for new material.