The lost outpost

In the past week I’ve had two separate conversations with people who wanted to know a way of posting large(ish) files on the web for temporary purposes, i.e. just to let a couple of people download something without going via email.

I don’t have a definitive answer, of course. The traditional way would be something like an FTP server. There’s Amazon S3 too.

The service I’m increasingly using is drop.io - a really simple way of temporarily sharing files up to 100Mb in size. There’s no sign-up or account required. You simply specify a drop name (so I could create a name of “andyptemp” or similar, and it would end up having a URL of http://drop.io/andytemp) and then specify a time limit of between 1 day and 1 year after which the drop will be deleted. Then you can add as many files as you like up to 100Mb for the drop. You can add a password for access if you like. You can specify whether other people can just download / view the files, or add their own. And that’s it.

If you decide to use the service, one hint I’d give is to set the “optional” admin password for your drop when prompted, as it means that you can go back in later and see how many people have downloaded files, as well as adjusting the “self-destruct” time of the drop.

There are some other really cool features like the ability to have an RSS feed of the drop, get email alerts, post MP3 files via a phone number, fax documents into it… a bunch of things I’ve just not needed to play with yet… but it’s a nice service, and appears to work well.

(NB that drop http://drop.io/andytemp is live for the moment, and it is set to read/write, but in time it will delete itself. Have a play if you like…)

(update: actually on reflection I’ve made it read-only as I should have realised that this means anyone can upload anything and I can’t vouch for whatever is uploaded, which was a bit short-sighted - ordinarily of course you’d only share the URL with folks you know)

Here’s something I’ve wanted to get my hands on for a while now… an Eye-Fi SD card. If you don’t know about these things, essentially they are standard 2Gb SD cards that fit into any camera that will take the format (or others, with e.g. a Compact Flash/SD adapter). The good part is that they make the camera wireless-capable….

So I picked up my Eye-Fi card and the first thing that I noticed was the cool packaging… pull on the tab on the right-hand side of the box, and the box slides out to the left, revealing a USB dongle and the card already inserted. You need the dongle, because you need to use the computer to configure the card.

Once I plugged the dongle into the machine, an Eye-Fi item appeared on the desktop… it was pretty simple to just install the Mac software. Once I’d done that, I hit a small snag… I got a message about the Eye-Fi Manager software being unable to initalise the card. I tried running the Eye-Fi Manager a few times, but the same thing happened… until I took the dongle out of the USB port on the right-hand side of my machine, and plugged it back in on the left. That time, I got a set of dialogs enabling me to register an account. Not sure what happened there!

Actually this seems to be an issue on my MacBook Pro… for some reason the Eye-Fi Manager software will never “initialize” the card when the dongle is plugged in on the right of the machine (although it still shows up as a mass storage device, and Lightroom sees it and offers to import images from it). Worked fine over on the left, but then the dongle is a bit too wide to enable the Magsafe power plug to be connected at the same time. Actually it seems a little random, unfortunately. I raised a problem with Eye-Fi support and they basically talked me through steps for checking that nothing else is using the port, plugging and replugging - nothing specifically useful. YMMV.

Right, so here’s how this thing works. You start the Eye-Fi Manager software, which opens a web page to configure the card. Here, you can add wireless network details (it supports a whole range of network settings including WEP and WPA keys), rename the card if you want, and configure a huge variety of online services. I have configured mine for Flickr… but the software supports Facebook, SmugMug, WebShots, SnapFish, Picasa, Photobucket… and a gazillion others that I’ve not heard of before (oddly, Movable Type, Vox and Live Spaces, but not WordPress - hmm!). Once you’ve done that, you put the card in the camera, and it will automatically connect to the network and start uploading shots any time you take them.

What appears to happen, is this: the camera uploads to Eye-Fi’s site, which then transfers to your chosen / configured photo service. The next time the Eye-Fi Manager sees the Eye-Fi site, it then mirrors the photos to the local disk (you can specify a location in the Eye-Fi Manager). I’m not 100% certain that this is how it works, but that’s what I’ve observed.

So now what about the downsides to this? Well for starters, the only supported file format is JPG. That’s OK, but of course Flickr now supports video too, for instance. Oh, and by the way, this is going to upload all your photos, anytime you take any, so I’ve set the default privacy option to private for Flickr uploads so I can review and tag etc. before publishing. The photos are obviously not titled or anything when the Eye-Fi uploads them, and they get a simple tag “Eye-Fi” set, but that’s all. So you will want to go and change title, tags, description, potentially rotate and so on once the image has been uploaded. Now that Flickr has Picnik integration, you can of course do some simple editing later as well. This does all bypass my “standard” photo workflow of Lightroom import, catalog, edit, and then upload, though.

One thing that the Eye-Fi does not support is wireless networks with certificates. Other than that, Open, WEP, WPA/WPA2 are all OK. It’s only going to work with networks it knows about, too (although you can configure more than one) - there’s no UI on the camera for configuring the card, you have to use the Eye-Fi Manager software while the dongle is plugged in to the camera.

Also, because there’s no UI on the camera side, there’s no visual indication as to what is happening… the Eye-Fi will silently upload your shots, and there’s actually no way of knowing that it is doing it, or when it has finished doing it. Of course it would be amazingly difficult for this to integrate with every camera if the makers had tried to build the Eye-Fi into the camera’s user interface, so I understand why this is the case - it’s just a little bit disconcerting! One nice feature is that there appears to be support for “interrupted” uploads, I see there’s a “Receive interrupted” comment in the Eye-Fi manager UI, so I think it will support partial upload and then resume.

Overall, it’s a neat idea, and certainly pretty cool for quick shoot-and-upload scenarios. Of course I often want to catalog my shots and touch-up on the computer first, but I can see cases where this could be really cool. Very handy for conferences etc. (oh, and that USB dongle can act as a reader for any SD card, too - handy). A qualified thumbs-up!

Here’s a link to a nice review, and here’s some news about the new models coming soon.

(post updated 14th May 2008 - a couple of additional details about workflow, the card initialization error, and the screenshot of the local machine import was added)

Until recently I’d hankered after one of the higher-end Intuous graphics tablets from Wacom, but a few friends have bought Bamboo models lately so I began to think that this might be a good option for me[1]. So far, it seems that way. What’s this all about? Well I’ve wanted a tablet for a while to help with photo / graphics work, and to try out sketchcasting (see below).

The product

The Bamboo Fun is a bundle which includes the tablet and stylus, plus a mouse (slightly redundant given I have a Mighty Mouse already) and Photoshop Elements - only version 4 for the Mac, annoyingly, but I guess I could always upgrade. To be honest, I’ve not even installed it yet.

The Bamboo comes in very stylish packaging reminiscent of something Apple would make… the box unfolds neatly, each item is wrapped in that thin foam bag packaging, and the driver CD is in a square box exactly like the ones that Apple uses for OS X CDs! So, first impressions are good. After that, basically it’s just plug-and-play… there’s a driver to install which provides some System Preferences to customise the tablet sensitivity and behaviour of the shortcut buttons, but that’s it.

Negative marks go to Wacom for having their registration page (and most of the website, it seems) “temporarily unavailable” for over a week. Not cool, and they are ignoring my emails too.

Usage - OS X and a tablet

In use, it’s been something of a mixed experience so far. The tablet itself is great, but it takes a lot of getting used to over a mouse (which is something I fully expected). The issues have been around the software support, and specifically in my case Lightroom. Two major annoyances - one that the zoom wheel at the top of the tablet doesn’t work in Lightroom, and secondly that although a single tap/click will zoom in, it is then really hard to get Lightroom to zoom back out with the stylus (should just be a simple tap/click again, that’s how it works with the mouse anyway).

On the plus side, it works beautifully well in Pixelmator.

Although OS X has handwriting recognition built in (the “Ink” system), actually I’ve not found this entirely reliable so far. [For éampe ltd to intSome word Son tbeTABLET] For example, I tried to write some words on the tablet just then, and you can see what happened. There doesn’t seem to be a way for the system to learn handwriting styles either. There is some software called inkBook that looks promising and somewhat more functional than the in-built software in OS X, but I don’t think I need it just yet (here’s a review). Handwriting was never going to be the primary focus of this purchase.

Oh, and it turns out that there are some issues with Ink and 64-bit apps in 10.5.x … I found that iScrobbler started to crash, and it looks like Ink is responsible. Weird.

Sketchcasting / sketchblogging

One of my main interests with the Bamboo was trying out Sketchcasting. Dave Briggs blogged about this a couple of months ago, and I’ve been fascinated since then. My first effort, though, was not the best… it’s way over-long (mainly as I fumbled around to work out how to get the thing to work) and not as well-planned as it might have been. Actually I think the Sketchcast site is somewhat limited… there’s no way of making things private, no friending, few social features at all. So instead, I picked up ArtRage and will try using that and Screenflow to create sketchcasts, and probably share them via my Viddler account, which will at least enable things to be embedded on WordPress.com.

More recently I’ve also noticed that Sacha Chua has been sketchblogging… using her Nintendo DS. Sacha is very creative and this seems like a really cool way of using a DS, although it looks like you need a bunch of homebrew software to make it work. Worth a look if you are interested, though.

[1] and, evidently, the only way I’m going to get a custom header image is to MAKE ONE MYSELF. This comment is aimed at no-one in particular. That is all.

It’s a three-part miniseries on gadgets now that I’m back from Las Vegas. If you’ve followed my Flickr then you’ll know what I’m about to talk about. Techie geekery starts here. Yes, I probably shouldn’t get quite so worked up about technology, but here I am.

One thing that I’ve not found that easy to get hold of is an external enclosure for SATA disks. IDE/USB caddies have been pretty widely-available for a while now, but most disks now use the SATA interface so I needed something that would let me backup my Thinkpad onto a spare 200Gb 2.5″ disk I’d been given.

I had a look through a bunch of enclosures in Frys and I have to say I just pretty much went with what seemed simplest (a NexStar SX from Vantec), without really spending too much time on the features… the decision was mainly based on cost ($20) and size. So when I finally (!) got around to opening the box today, I was completely astounded at the quality of the product, given the cost.

The box itself was solid cardboard (not the flimsy sort of stuff I’d expected). Inside was a soft carry case, and inside that was the metal enclosure itself. The product also ships with a USB cable with a “pass through” socket which seems to enable more devices to be attached, a driver disk, some screws and a screwdriver - very complete.

Once I took the enclosure out of the carry case I had another pleasant surprise. It’s metal and very compact. There are two small squeeze clips at the back… press them in and the top and bottom are released to slide off. The other USB enclosure I’ve previously purchased was really poor compared to this, with ugly screw lugs on the top and a very loose lid… this thing is really elegant.

The only small issue I had was that the Hitachi drive I’d been supplied for my Thinkpad initially would not fit into the actual case, because it is so compact. The side screws made the drive just a little too wide to fit. I had to remove the top cover of the drive (see photo), but since Vantec provided a screwdriver this was no big deal.

That’s really all I have to say, but I just wanted to point out that this seems to be a really good quality bit of kit at a decent price. I probably ought to look for a 3.5″ version and then I can whip out the drives from my old Linux tower before decommissioning it!

Tara Hunt has posted a great piece recently on Twitter for companies (by the way, if you don’t follow Tara’s blog, you really should take a look). Here’s a a small illustrative anecdote from WSTC last week.

I gave a presentation at our internal WebSphere conference that talked about Twitter, among other things (ordinarily I’d Slideshare the presentation, but it did contain some stuff I can’t post externally, sorry about that… maybe I’ll write an external version sometime).

One slight technical hitch. The conference organisers were set up to record presentations from a Windows laptop and Powerpoint using Camtasia. I was presenting using Keynote on the Mac, so I chose to use ScreenFlow to record the session (I noticed that Jason McGee did the same thing for his WebSphere sMash pitch earlier in the week). I tested beforehand and everything seemed good, but just at the end I switched virtual desktops to do an unscripted demo of something, and found a stack dump from ScreenFlow on the screen. I’d spoken for an hour, taken 15 minutes of questions… and lost the lot.

I twittered my frustration and headed out into the corridor, thinking about re-recording.
Within minutes, I checked my email and found an email offering support from one of the guys at Vara Software, who produce ScreenFlow.

Well there’s yet another illustration of the power of Twitter. Let’s be clear, they haven’t in fact been able to recover the data for me, but that doesn’t matter… this level of responsiveness and support just makes the relationship I have to that particular piece of software much more positive and sticky. I’m pretty impressed. I assume that they were using Tweetscan or Summize (Twitter mashups / “search engines”) to watch for references to their software and decide whether to respond. This was just great.

Oh, and this worked brilliantly to illustrate my point about why folks should take a look (or second look) at Twitter in the talk itself. This is increasingly becoming an important channel of communication.

IBM has had some guidelines for employees who want to blog for a while now. They’ve been very helpful for those of us wanting to go out and start blogging externally, and the principles also apply to our behaviour internally, and across many of the social media tools that we use. I always refer to them as “lightweight”, by which I mean they take a relatively light touch, don’t attempt to impose control, encourage us to be respectful and sensible, and generally fit in with our existing standards of behaviour. They’re largely common sense.

Over the past couple of months we’ve been working collaboratively to update the guidelines. We did this in the same way that we developed the original document - via a wiki, and some lively discussion on our internal blogging platform.

The new IBM Social Computing guidelines can now be found on the IBM website. They’ve been extended to talk a little about new technologies beyond blogging (social networks, media sharing sites, and so on), and just generally tweaked to ensure that we’re taking account of changes in online behaviours over the past couple of years. One of the things I like is that we’ve actually made very few changes to the document, and certainly no major revisions… I think that again demonstrates the “light touch” and flexibility of the original guidelines and the level of trust that IBM is showing in its employees. For me, this is a good place to be.

I haven’t had a chance to write up many of my thoughts from the WebSphere conference I’m attending in Las Vegas yet - the agenda is, as usual, absolutely jam-packed with content so I’ve not had much time to form notes into blog entries. A bunch of people are twittering from the conference.

If you took one of my cards at the poster session last night, you may be interested to find out more about me and links to the tools we were discussing via my internal blog.

I’ve made the now-annual pilgrimage to Las Vegas for the internal WebSphere Services Technical Conference. The direct Virgin flight from Gatwick actually seemed to pass relatively quickly despite the lack of working entertainment system in my seat.

So this year I guess the social software coverage is likely to be a little wider. We’re using the tag wstc2008 for blog posts; I’ve created a Twitter hashtag of the same (I know of a few others who will be here); and photos may appear on Flickr or other sites with the same tag (although not from me - I don’t have a camera with me this time around).

The conference agenda looks very good, and very varied. I’ll be presenting a poster session on Social Software on Tuesday evening, and a full talk on the subject on Thursday morning. In the meantime, I’ll share what I can.

Unix?

Originally uploaded by andyp uk

Spotted this building in Male’ when we were in the Maldives last month.

OK, geek joke, but it was a photo that was begging to be taken.

I was just editing a couple of posts and spotted that WordPress is now inserting “possibly related” links:

On further investigation it looks like this was a feature they rolled out at the end of last week, which can be disabled via the Design->Extras Dashboard panel. It seems interesting and I’ll keep an eye on it… with the HackDay post shown above, it would have been really cool if it had linked across to Kelly’s blog, but never mind, one of my old posts is fine…

Update: it only shows up on the main post page (not on the front page) so you have to follow the URL to the page to see it… so far I’m not that impressed, on one post it has created two links to the same old article of mine (which isn’t at all related), but we’ll see how it gets on and I’ll disable it if it becomes too random.

The buzz
There’s a bit of a buzz going on at work at the moment - a bunch of us from “the Hursley crowd” have started playing with Current Cost meters. These devices are intended to enable consumers to see exactly what their energy usage is and, hopefully, modify behaviour to save electricity accordingly. The idea, simply, is that it provides real-time information about energy consumption.

Rich, James, Nick and Ian have all written about their Current Cost meters already, amid much twittering and the support of @andysc.

The product
The device itself comes in two parts. The unit that goes inside the house is a wireless LCD display which shows the current usage in watts, the current estimated cost per day assuming that usage is maintained, a bar chart with yesterday’s usage, overall KWH in the past day and month, and the time and temperature. The other half of the device is a somewhat larger and heavier transmitter (shown at the top of the picture, the top of the two black boxes inside our cupboard) which sits next to the electricity meter, with a clip that gently attaches around the cable (you can see that hanging off the cable at the bottom of the picture). The product is completely non-invasive and it’s incredibly easy for anyone to install: there’s no rewiring, just a clip. I was extremely impressed. It “just worked”.

Update: I should point out, given some comments, that we’re using an early batch of the meters and I’m not certain when they will be generally available.

Update: Roo points out that Eco Gadget Shop have them for sale to consumers, minus data cable.

The impact

One of the other features of the device is that it can be plugged in to a computer, and the data can then be captured and analysed over time. We are using some homebrew software to do this, pulling the data from the serial port (most of the meters use 9600 baud, it turns out that mine is set to 2400 for some reason).

It’s kind of scary to see some of the spikes in the graph, and just watching this has certainly made me adjust my behaviour in terms of switching things off and unplugging chargers and so on when they are not in use. We’ve all got our meters hooked up via a Microbroker, and this has been my first opportunity to really play around with MQTT technology… I’ve obviously been aware of it for a very long time, but it’s nice to have something tangible to hack around with. It has also led me into a bunch of interesting discussions about home automation, tweetjects and low-power servers. Fascinating stuff.

The ideas

I have a bunch of thoughts about this. I have it hooked up to an old Linux box, but I’ve also successfully attached it to my Macbook Pro and a Windows Thinkpad. Currently the software is sending the MQTT data to a Microbroker and a Java app is drawing the graph shown above, but it would be fairly straightforward, for example, to squirrel the data locally and do some interesting analytics using Project Zero (aka WebSphere sMash) and some AJAX-y Google Chart goodness. I can also capture ambient temperature over time. It’s all just a matter of finding the hacking opportunity!

I just uploaded a small set of images from the local Hackday event at IBM Hursley to Flickr. Feel free to add tags or notes if you took part.