Archive for July 2005
Today we had a seminar from Dave Lemphers, who is a Developer Evangelist from Microsoft in Melbourne.
He talked about upcoming stuff coming out of Microsoft, but the most memorable line of the talk from my perspective was when he was talking about Visual Studio 2005. He said "For those that don’t know, ASP.NET is our PHP-style technology". How the world has changed.
- .NET framework 2.0 has generics, like C++ or Java
- User interfaces for Longhorn will be done with a thing called Avalon, which is like Flash.
- They have a new source control system as part of their Team System, which will suck much less than Source Safe, and has a nice concept called "shelving" which is saving code to the server without checking it in.
- SQL Server 2005 is tuned for access from .NET and will be slower from ODBC, ADO etc.
I am having an argument with a colleague at present. He says:
"No offence meant to anyone, but as a software engineer I do not agree with
adoption of any home-grown solutions, which is not subject to robust testing,
when there is packaged one available (where we reasonably expect there to have
been robust testing). "
We are in the process of trying to find a good way to manage student relationships - basically we need CRM. I have to look around and check out some Open Source CRM packages to see if there is anything suitable. We have a homebrew system which either needs expanding (functionality-wise) or replacing. As you can see, he is not in favour of doing it ourselves.
After the very public software problems the university has had, I would never assume that packaged solutions are better than homebrew.
If you know of any good student oriented CRM tools, drop me an email or comment. We are also talking about an internal wiki which might be sufficient (and is definitely easy).
I got the call for papers for WWW2006 today. One of the tracks for next year is Developing Regions, and I’m hoping that our School will do a paper on our involvement with the African Virtual University.
This is an amazing project where we are helping a number of African Universities to deliver Computer Science Diploma and Bachelor programs by providing courseware, training, and support. The first students of these programs graduated this year. Eventually (after a four year partnership) the African unis will run the programs independently.
At least it looks that way. Explosions reported on the Underground and on buses, people have died. The Underground is closed. I don’t know what to say.
Now here’s some good news. The EU Parliament voted overwhelmingly (648-14) to reject the directive on software patents. Everybody seems to agree that this won’t be the last we hear about this unfortunately.
I should have mentioned: I gave my paper ("Mining User Preferences from Web Access Logs and User Public Information") this morning, and I was really happy with how it all went. The AusWeb site has both the full paper and the slides (just read the slides if you need an overview, short and sweet) which put it in a context of what I was thinking when I did this work (a little while back) and explains a bit more about the motivation behind PUP.
The main conference is now over but I still have the Unisites SIG to go tomorrow and then head home tomorrow night.
AusWeb has been an interesting conference as it has a blend of hard technical and some more educationalist papers, and this produces an interesting mix of people to talk to.
I’m sitting in a talk by Liddie Nevile at AusWeb’05 on User-centred accessibility. This is more on the topic that I was discussing here a week or two ago - accessibility turning into adaptivity. Rather than trying to built websites that are universally accessible, the goal is to adapt websites to users’ individual capabilities. You can read the paper here.
It dovetails well with ideas that I have about user personalization on the web: users should control the way they interact with web applications, the format of the data, and the way it is presented. This is the basis of the PUP (Personalization with User Privacy) protocol that I’m developing.
This afternoon it’s Brad Kasell from IBM. I start off agreeing with him.
He says small is the new big. In other words, lots of things that are suddenly cool are little things…blogging, simple idea, del.icio.us, so simple and yet so good. Greasemonkey, which lets you write client side scripts to make web sites work the way you want it to - for example, surf Amazon but with links to book availability in your local library. Disposable applications - love these - the things you write to make your life easier, you write the code, use it, throw it away. (PHP is perfect for many of these by the way.)
He then spends 10 minutes talking about Web Services and SOA…guess what? More on the same theme…RESTian web services and why they are gaining in popularity.
So all of this so far goes towards the theme I see everywhere: simple, clever, elegant technologies. Just to go on my web services rant again, in Brad’s words, technologies most of the time only have to be "good enough". We don’t need wads of standards, platforms, environments, etc for 80% of applications.
Anyway, then he went on to talk about Open Source and I began to disagree, although I still really enjoyed the talk. He had a nice quote from Business Week…Open Source is "give a little, take a lot". (Or if you wanted to start a completely different argument you could say from each according to his abilities, to each according to his needs…) He then went through some of the same old arguments I have heard - and heard refuted - before.
Question: Corporations want to buy support and what will they do if the project/company dies? Answer: In many Open Source cases support is available from multiple vendors so you have the freedom to choose. If none suit, you can always support it yourself. (In many cases, if you think that having a vendor to support your software means you’re safe, you can think again. If a vendor goes belly up with proprietary code - or decides to stop supporting it - or won’t make the changes you need to support your business - what would you do?)
He talked about Open Source business models being unsustainable but I felt his argument was pretty flawed. He said only 3% of users of JBoss (for example) buy support, so that was unsustainable. As long as that 3% is enough to support the business to support their users, why isn’t that ok? Not every company has to be a big company. JBoss has been around for a while - as has Red Hat - as has MySQL (10 years this year) - so it must be fairly sustainable. More sustainable than a lot of dotcoms anyway. Further, if only 3% of customers buy support, isn’t the argument that "Companies need support" a myth? Hard to have it both ways.
Two further Good Things Brad mentioned were :
- The Guide to Open Source Software for Australian Government Agencies
- communitycode.org, ala "Hack for the Dole". This is remarkably cool. From their website:
CommunityCode is a group of Australian open source / free software
people who want to give Newstart recipients a way to count any Free Software
work they do towards their mutual obligation requirement, and to involve
people who might not otherwise have a ‘way in’ to Free Software."
Overall another really good talk.
Another day, another keynote, this one by Terence Huwe, a librarian from UCB. This was a great talk, full of inspiring ideas. He spent some time talking about Open Access research - a great idea, which is to academic research what Open Source is to software. (The University of California has a summary of their perspective. I’m really excited about this idea and I would love to hear more about this in a specialist venue like OSCON.
Terence also had some interesting things to say about the role of blogging in digital libraries. Quote: "Sovereignty over the story is shifting" (comparing blogging to journalism).
He also cited some numbers showing that in this electronic age, physical libraries are becoming ever more popular, with library members at an all time high in many parts of the US. For example, in San Francisco, 70% of registered voters hold a library card. He described the library renaissance going on at UCB - the library is open 24/7, has WiFi, an espresso bar, and places to lounge around and discuss and blog.
There were so many interesting things mentioned in this talk, from the idea of fugitive resources in libraries, to spray on and fog based projectors, that my head is spinning with ideas. The slides will apparently be online after the conference so I’ll provide a link when they become available.