I had a conversation yesterday that reminded me how confusing I found WS Make Connection the first time I saw it in action. It’s a really simple specification designed to do something really simple, but the names used in the messages can cause no end of confusion.

Though Make Connection can be used standalone, it’s currently primarily used to provide retransmission capability for WS-Reliable Messaging 1.1, so I’ll focus on that scenario.

When providing reliability in an environment where the service provider cannot open a connection to the client (e.g. because it’s behind a NAT device), the service provider still needs a way to send messages to that client.

An example would be retransmitting a response message to a client because the original transmission of it failed.

In this case the client needs to connect to the service provider to provide a connection for the service provider to use to send the response. That’s what Make Connection is. It’s a specification which defines a single one-way message which a client can send to a server, which allows a server to send a message to the client using the HTTP response channel.

i.e MakeConnection message=’Hi, I’m client xxx, and here’s an HTTP connection you can use to send me a message if you want to’.

The confusion: People used to session oriented technologes (including WS-RM of all things) see a ‘MakeConnection’ message and assume it’s setting up a long running connection that will be used later on. This is not the case. A MakeConnection message only has meaning for the length of the HTTP request/response interaction it is sent in. For subsequent messages from the server, additional MakeConnection messages must be sent by the client.

Last week’s announcement letter means that my current project at work is coming to an end.

I’ve been working for that last 6 months on the WebSphere ESB 6.2 release with a particular focus on web services and policy. So the bullet points most relevant to me are:

• New policy-driven ESB mediation, including integration with WebSphere Service Registry and Repository
• Support for the WebSphere Application Server Web Services Feature Pack and SOAP 2.1 (sic)

The policy driven mediation is one answer to Katy’s recent question “What’s so great about WS-Policy?”. WS-Policy provides a standard format to describe requirements, and with the policy driven mediation we exploit this standard to describe requirements on a mediation flow (or more precisely the primitivies within one). Using WS-Policy means we can use the existing support for WS-Policy in WSRR to author an attach policies, and allows users to use their existing WS-Policy expertise (or more likely for the moment - develop expertise which they can use later in other areas).

The policy driven mediation support is also an example of componentisation in action. We’ve spent a chunk of the last few months extracting the WS-Policy engine from WAS 7 to use it in WESB 6.2. WESB uses it in a very different manner than WAS, so we’ve made some changes on the way, but I think what we’ve ended up with will be useful in both products (and more) for a long time.

I’ve been less involved in the Web Services Feature Pack support. I’ve mainly been helping colleagues understand all the great new technologies, how to use them, and how we can best exploit them in WESB and WPS. It’s always eye-opening talking about something you’re so familiar with because so much that’s just background knowledge to you is advanced and complex to people with a focus elsewhere. I’m looking forward to seeing this second (or is it third because of WAS 7?) phase of customers working with JAX-WS, WS-RM and more.

With development for this release winding down I’ll be spending a bit of time out of the office talking to customers about what they can do with the current and coming releases - I look forward to hearing a bit about what they’re up to.

Not trips that are bought and paid for by work, but trips I quite fancy anyway, and will cost lots…

ApacheCon US - To see what all the fuss is about, and meet lots of interesting Apache people

New York for Hogmanay - Because it’s supposed to be right up there with Edinburgh

South by Southwest - It looked great from the edges this year, so doing it properly would be fun.

So now all that’s left is to work out what I can afford, and to find people to do go with…

I have an iPhone 3G which I’m using to write this post using the wordpress application from the AppStore. A detailed review is probably best left for a real keyboard, but the interim verdict is… 5 stars

After completing the WS-Policy Round 1 implementation a few weeks ago and getting pulled in to more WS-Policy stuff at work I was looking for another spare time project. About then, Rich told me about the CurrentCost device he had set up and it seemed like a fun thing to work on.

We had a tweetup at work where it seemed like lots of the problems were already solved…. so, being that I was looking for a fun challenge, I’ve ignored all that for a few of weeks, bought a device independently, built a cable, and this bank holiday weekend connected it to a slug, written a really suspect python script to read the data from the device and publish it to an illsley.org XMPP id. I’ve also written an XMPP client which is running on a box at work retrieving the data and putting it into a .csv file. One of the good things about using an illsley.org XMPP id is that the google backend records all the messages so I could not have the client and extract the data from google later.

Even that’s not ideal, so the other options I’m planning to explore are Amazon SQS and Amazon SimpleDB, before probably caving and recognising that the MQTT solution is best and asking the guys at work nicely if I can properly participate.

There are lots of interesting things to do with the data, which I’ll start to play with once I have a meaningful volume of it. One possibility which surfaced this week when I saw an internal demo is snaffling an early build of WebSphere Business Events and seeing what interesting things it can do with power usage events.

A topic I’ve considered a fair amount over the last 18 months is where ‘the cloud’ ends and the personal computing device begins.

It was first triggered in my mind by frustration at the amount of time needed to download large podcasts. It’s a simple scheduled task which can’t happen when my laptop is off, but when I get up in the morning I want to quickly get any new podcasts to sync to my iPod, not wait for a 100MB download. There is, however, a wireless router with a processor capable of http downloading sitting connected to the internet 24/7, and if it downloaded the podcast, it would be very quick to copy the file across when the laptop reappears on the network.

I considered scratching that itch with a Slug, but that approach would have missed the point - I want to continue using iTunes in the normal way. I want iTunes to automatically hand off that work when my laptop isn’t on.

More recently, I’ve been considering power consumption and the kinds of tasks that really don’t warrant the 60 watts the MacBook can pull, or even the lower power levels it pulls, simply for having the screen on. I want these tasks to happen somewhere, when and where is appropriate.

After 3 failing hard disks in the last 2 years, backups are a topic close to my heart. Ever since the Time Capsule was announced I’ve wanted one, and last Saturday was the day I caved and spent the money. I’m happy with it so far - though it does run pretty hot and I’ve not taken that extra step of trying a restore from backup.

The list of things it can do is pretty impressive, and made me reconsider these cloud thoughts. If there’s a company that can pull off linking home devices in a seamless, usable manner, and if there’s a company looking to leverage it’s software to sell consumer hardware, it’s Apple. They are also a company which has shied away from building a ‘home server’ - a product category that I can’t imagine resonates with more than 0.5% of consumers. A cloud of routers, home entertainment devices, capsules, and more all of which can communicate, and interface well with software is far more up their street and something I’d be willing to shell yet more money into.

Of course, I’d also like such a home cloud to extend to the ‘big cloud’ for appropriate tasks, but I’m guessing that’s yet further years away.

[I did start this post last week before the Forrester report]

One of the perks of working at Hursley is the occasional trip to the pub at lunchtime. Today was one of those days, and in-keeping with previous trips with that particular group, I got a bit of a ribbing about participating in a relatively small number of social networking type activities. First I think it was blogging (the, um, rabble rouser of the group now has a blog), then it was having a SecondLife avatar, and facebook, and today it was twitter.

Every time I start by defending the technology because I’m under (gentle) attack and feel I should, then sip on a pint and the discussion disappears so we never really address the issues. This is a brief attempt to do just that, because I think it’s an interesting discussion, not because of the comments.

Blogs - I think the potential value of blogs is clear. My blog is generally of interest to me, and occasionally to others. There are other blogs, written by authors, politicians, technologists, and CEOs which get these people closer to those who are interested and have opinions than before. A good thing.

SecondLife - shrug It’s an interesting experience, and probably represents a style of interface we’ll use regularly in the future. I’m not at all hooked and may never log in to that particular metaverse again.

Facebook - An excellent way to keep in touch with friends I don’t see every day, or indeed very regularly at all. Useage has dropped off a bit recently, but that’s at least in part because I’ve been experimenting with twitter. It’s a closed data platform which really isn’t something I’m comfortable with. I want to own my social graph, so facebook having proved it’s value to me, I’ll probably move as much out of their environment as possible.

Twitter - It’s been nearly 2 months, and this experiment is ongoing. I’d liken this to working in open plan writ large. I know what lots of people are doing. Lots of it is unexciting, and easily ignored. Some of it’s interesting, and stimulates ideas or laughter, and even seems to foster a sense of team… see the #currentcost activities in hursley spilling onto twitter as an example. Twitter deserves a full post at some point, but the short version is that while it’s quite compelling, at least over the short term, I don’t really know where it goes from here.

[UPDATE] The other interesting thing about twitter from a work perspective is that it’s large scale consumer pub/sub and event notification.

Dopplr - Every so often I get a notification that someone else is sharing their trips with me. I don’t travel enough for this to be interesting to me, though I know some people who do, so I should probably speak to them to work out if it’s as useful as the theory sounds.

So some successes, some not, and some undecided. Overall, I’ve got what I wanted from them. These technologies/services are all a means to a particular end for me. The end? Fun experimentation and keeping an eye out for interesting technology trends. I don’t need these technologies to be great for the experiment to be worthwhile.

I’d been playing around writing a python WS-Policy framework a couple of months ago and put it down because of other things going on. Google then announced AppEngine, and, being the only python code I had lying around thought I’d put it up. The result is a very experimental WS-Policy implementation deployed on AppEngine, available at wspolicy.illsley.org

I’ll follow up with my thoughts on AppEngine in another post, once I’ve experimented a little more.

Triggered by a tweet, I re-read roo’s blog on backchannels today. The only backchannel experience I’ve had was on the W3C WS-Addressing Working Group. W3C Groups share information and take minutes straight into an IRC chat. ‘/me’ messages are considered to be out-of band comments, and are not included in the minutes. The IRC chats are used in both telecons and face-2-face meetings, with slightly less going on in them during the face-to-faces, though the IRC may be on a projector.

Initially I didn’t really understand the way it is used, particularly some of Paul’s comments.

In both scenarios, it’s really useful to type while others are talking if you’re looking for a clarification, or have a useful comment to share, and don’t want to interrupt the flow or someone talking… which is particularly difficult to do on a call anyway where you can’t see how people are reacting. It is also used on occasion to undermine people and arguments before they are fully formed which can be useful or really frustrating.

The third use, that Paul is particularly good at is lightening the mood. Often with truly awful jokes. Nevertheless, even a bad joke can jolt bored people into life again, and reinforce the idea that we’re humans working together, rather than faceless representatives of our organisations. With the ‘/me’ notation, these jokes are very easily ignored if you’re not interested, so they are very low cost.

I miss all of these uses when on telecons without a backchannel, which is most that I’m involved in at the moment. In those that do have backchannels, the same social norms about the channel aren’t the same, particularly there’s no recognised way to indicate that your comment can easily be ignored.

All that I can take from that experience to the idea of a conference backchannel, which is a slightly different environment is a belief that it’s probably a useful idea, that social norms can prevent abuse, and that including a mechanism to allow ignorable comments to be easily identified would probably be important.

I guess IBM like announcing things with similar numbers at the same time… also April 1 seems to be a favourite.

The WebSphere Application Server v7 Open Beta which contains the fruits of my, and many other peoples labour (including multiple open source communities) has been announced for May.

Additionally the WebSphere MQ v7 release has been announced for June. There’s been a lot of pizza consumed getting that release ready, so congratulations to everyone on the second floor for getting it done.