Eyefood

So this site, as I might have said a few (?!) times, is a closed space. You can;t get in to see things unless you’re a student/member of staff … which is done for specific reasons (that I might also have said). Anyway. Whatever.

The thing is, for it to be a success, there has to be some benefit in participating in the enclosed social network. I mean, what’s the point of whacking your work onto a website, if you can’t (a) do anything with it afterwards, or (b) if nobody is going to look at it.

Obviously, with a site geared around student work, the obvious reason for uploading is as a portfolio, and for peer/staff feedback. But it occurs to me that there are other benefits from sharing - or at least, there should be.

One of the things that I haven’t yet properly implemented, is little easy cut-n-paste code sections, that let you place the work on an external site (you know, like Flickr and YouTube let you). THe old version used to have that feature, and for the life of me the only reason I haven’t written such a simple thing so far (I think) is because I can’t think where to put them on the page, without destroying the so-far-lovely aesthetic.

(Style over content, as I’m perpetually accused of)

But the other thing that I really must get around to writing, is the ability to link your profile on the site to profiles on other sites. For example, images uploaded need to have the ability (even if not taken advantage of) to go to Flickr as well. Videos should be possibly squirted onto YouTube at the same time. Links should plop themselves onto del.icio.us.

And, of course, the other way around. If I upload something to Flickr, then it should find its way into my portfolio. (If I want it to)

It’s not a hard thing — all of those sites have perfectly-well documented APIs that I can take advantage of. All I have to do is write a section that will link with their individual authentication APIs and let me say “Oh hi, this person has an account on your site. Let’s share”

The privacy concerns of the site are about creating a ’safe’ environment for experimentation, but we should (I feel) give students the opportunity to share their work with an outside audience, if they want to.

And the lovely thing is, I’ve done similar code before. I just have to remember how…

Perhaps a misleading title. Oh well.

One of the features that everyone kept asking me for in our ePortfolio tool, was the ability to have private groups. Being able to link together different people and pieces of work, and form discussions around them is fun … but it’s useful sometimes to be able to do that without the glare of the outside world looking in.

(Interestingly, sorry, it’s like a mini-version of the site itself - which is visible only from the inside. Privacy within privacy, he said, digressing)

It wasn’t as hard as I thought it was going to be, when I actually got down to it today (which is usually the case. As long as I can remember how I wrote something in the first place, I can usually change it slightly to add new features). In this case, it was just a question of having a ‘privacy flag’ in the database (which was already there, what with me being a bit more fore-planning this time), and the right checks on group-membership and overall-authentication-levels (and stuff). Wrap it all in an if(), and make sure that all the places where they would normally show up are covered.

Bingo.

Cute huh? The red dots are the private groups, and the orange ones are the ones that are invite-only (to join), but visible to everyone else.

What struck me as a useful thing to do next though, was to be able to specify privacy levels on pieces of work themselves. Things like “public, private, tutors-only, friends-only”. I did wonder a while back about how relevant this sort of thing would be, and erred towards a “Open and Free For All” approach … but I’m starting to think that there are very obvious reasons why thing might want to be private (not least, using the site as a handy place to stick images to reuse on blogs - like I did with the one at the top there)

Anyway, we’ll see how this pans out. Something that did strike me earlier though, was a large sense of ‘community‘ when I was using it a moment ago. I think there’s a significant sense of being part of something because of the Site.

I should write more about that some time.

2. Substructure

(Part one)

I’ve been meaning to write more about my little e-Portfolio tool for some time now, but it’s almost an unnecessary endevour. There are that many similar systems and sites across the whole of the Web 2.0 spectrum these days, that it almost makes no sense to outline what this one does — seeing as its functions are replicated in several other places.

Many people, for example, liken it (alarmingly so) to Flickr … and it’s very true that I’ve been very-much ‘inspred’ by Flickr in the way that a lot of the features of the site work. In a way, Flickr is almost the perfect model for an Art-School-Portfolio site like ours, and I’ve been very aware of the similarities all the way through coding it.

(Having said that, there are some features that are replicated in Flickr, like being able to make Group sets for example)

I have a very pressing need at some point to make an “Introduction to the Site” presentation — not for public consumption, but for the students who will be arriving in September. But in the meantime, I wanted to explain a little about what drives the site, and then follow it up with some screenshots.

…

I’m lucky, in a sense, that I have a very small and carefully-defined target audience for what I’ve made. In essence, it’s never going to extend beyond an active set of users of about five hundred or so. True, ex-students are an increasing part of what goes on on it, but even the most active ones tend to die down a bit in their engagement, and it’s mainly the Undergraduates and Postgraduates that are the focus of activity.

Because of that — and the lesser emphasis of scalability that seem to plague a lot of larger sites these days (*coughtwittercough*) — I’m able to use a very simple file-system for processing the work that gets posted.

Essentially, the main uploading part of the site places whatever was sent to it by a user into a common ‘queue’, to be processed by a series of little scripts — scripts that take advantage of some open source software: ImageMagick and ffmpeg (mainly, but there are others)

It’s like a system of ‘Watch Folders’ if you like — every minute or so, cron scripts check to see what’s in certain places on the server, and fire up if they find things. They move the files to other places, process them into web-ready versions of the original upload, and then move them all over the place (and update the database) when done.

It’s quite an elegant little system, if I do say so myself. There’s even a nod towards being able to expand the file storage if needs be in the future — with the site choosing a specific external server to store the files on.

It also means that it’s relatively simple to add different media-types to the site, just by finding appropriate open source (or, perhaps, not) tools that will squish formats from one to another, and process whatever the students might have on their hard drives into something more web-friendly. I have (for example) an experimental installation of wv — a Word-to-HTML convertor (which isn;t working actually, and I think I need to install AbiWord instead) — that used to take word files and automagically present them as web pages (after being run through tidy to get rid of Word’s hideous formatting)

…

One of the original features that I haven’t yet perfected (and so is still disabled) is the post-processing of moving-image work into Quicktime 7’s h.264 format. What used to happen (when the whole thing ran on my home servers) was that after upload, a quick-and-dirty MPEG4 would be spat out (after about 5-10 minutes), just to acknowledge that something had been uploaded. In the background though, the original file was moved to a second queue, and later that night (always at 3am — when I in theory wouldn’t be working on the machine that did the encoding) it was recompressed into a much better copy.

As I said a few weeks ago though, playing around with Compressor has given me the idea that it might be much easier to set up a render-farm using that, rather than relying on the (rather kludgy) QuickTime scripts that I had before … and it would also give me the options to easily convert the files into many more formats (iPhone anyone?)

…

Anyway, bleh. Rather than talking about each and every feature over the next few weeks, I’m just going to upload some screenshots and describe some of the functions…

People involved with me in the Second Life projects that take advantage of the cagd eportfolio tool that I’ve written, will most likely be familiar with sending snapshots to the site via the postcard --> post@sleeds.org email address. It’s a cunning piece of code that I hacked together, that parses the (somewhat less these days) garbled email postcards that LL send out, filters out the gumpf, leaving you with the image itself, plus some metadata about it — like the coordinates of where it was taken.

(I must publish that code some day — it’s not very complicated and might help a lot of people out)

Anyway, the function didn’t originate as a tool for cagd — initially it was a experimental tool for Flickr. It was shortly after Flickr introduced geotagging — and the ability to see your photos on a map with other people’s dotted around them — and I thought it would be fun if we could see our Second Life snapshots on a map as wel.

The tool that came out of that — sl2flickr (spot the url-change there) — was one of my usual ‘hacky-experiments’. The code works, and the authetication/sign-up part is still there if you want to give it a try. But the map has long since broken, because Linden Lab changed the Map API.

Recently though (although this has nothing to do with the map API really), we had a discussion in the Second Life Flickr Group about using machine tags with a sensible namespace. Now, I must confess, I’m not entirely clued-up about what machine tags are specifically — apart from being tags which are useful to programmes, but not to (and therefore easily hidden from) humans. But what we ended up after a short debate, was a system for geotagging Second Life snapshots. For example, a snapshot taken at the Welcome Area at Ahern could be tagged:

secondlife:region="ahern"
secondlife:x=128
secondlife:y=128
secondlife:z=0

At the time, a few other services (like BlogHUD, I seem to recall — worth checking up on that) agreed that they would start using the same system, but for ages I’ve been wishing that both Snapzilla and SLBuzz.com (that a lot of people, like Torley, use to upload their snapshots from in-world) would do likewise.

Yesterday, through a rather weird little serendipitous email exchange that Thau had with some people from Orange, the two of us found ourselves onstage at Orange Island, talking about the Second Life Flickr Group, alongside Cristiano Midnight (of sluniverse.com) and the Koinup guys — and I couldn’t possibly let an opportunity like that slip without harranging Cris into taking a look through our discussions about tags.

…which he did, and left a post announcing that Snapzilla’s photos will now use the same geotagging namespace. Which is awesome really, because it means that we’ll have significantly more geotagged photos to play with in the future, and that once I get off my backside and re-learn the map API (or convince someone else to), we can have a map just like Yahoo’s and Google’s, peppered with photographs from all over the grid.

An amusing footnote:

I’m almost embarrassed to admit this, but what with there being several names for Snapzilla (well, two — that and ’slunivers.com’), I’d always thought that slbuzz.com was just another url that pointed at it. But, of course, I’ve just discovered that it’s an entirely seperate (and therefore, competing) website. Haha. Ooops.

So when Torley was talking about using SLBuzz.com, I thought she meant Snapzilla. And when she suggested I talked to Mark (the slbuzz owner), I thought she meant Cristiano.

Fortunately though, I didn’t make that mistake out loud. haha.

Semiconductor make moving image works which reveal our physical world in flux; cities in motion, shifting landscapes and systems in chaos. Since 1999 UK artists Ruth Jarman and Joe Gerhardt have worked with digital animation to transcend the constraints of time, scale and natural forces; they explore the world beyond human experience, questioning our very existence.

Warp is a preference pane that allows you to use the mouse to switch between Spaces rather than using the keyboard.

Save spoken text on a Mac to a sound file

FolderOrg is an AppleScript Folder Action that is organizes files and folders by moving them into dated subfolders. This is helpful in keeping files and folders organized by the day they were added, not created or modified.

Burn - open source burning application for Mac OS X