VB.Net Using Delegates for Form Updates from Worker Threads
by Tim on September 21, 2011
At Geekup Preston this week a conversation with @surlydev led to an explanation of how to use delegates to update a UI from a worker thread. Talk is cheap, and a code sample goes along way to help explain.
Imports System.Threading
Public Class DelegateDemo
' imagine a form with Button1 + Label1
' the thread
Private _worker As Thread
' declare a kind of delegate with the signature you want
Private Delegate Sub UpdateDelegate(ByVal s As String)
' declare an implmentation with matching signature
Private Sub UpdateStatus(ByVal s As String)
Me.Label1.Text = s
End Sub
' kick-off the worker thread
Private Sub Button1_Click(ByVal sender As System.Object, ByVal e As System.EventArgs) Handles Button1.Click
' start a worker
If _worker Is Nothing Then
_worker = New Thread(AddressOf WorkerMain)
_worker.Start()
End If
End Sub
' Worker thread
Private Sub WorkerMain()
Dim x As Integer = 0
Do
' increment
x += 1
' update UI + sleep
' this executes delegate + params on thread that owns the form
Dim msg As String = String.Format("Hello {0}", x)
Dim del As UpdateDelegate = AddressOf UpdateStatus
Me.Invoke(del, msg)
' half a second snooze
Thread.Sleep(500)
Loop While x < 30
' thread's dead baby (will terminate when it exits this sub)
_worker = Nothing
End Sub
End Class
BarCamp Nottingham 2011 #bcnott
by Tim on July 28, 2011
At weekend I attended BarCamp Nottingham hosted at the NottingHack Hacker Space and what a great event it was.
As well as lots of interesting people and talks, the hacker space itself was an awesome place to be in and a great facility that Nottingham should be proud of. I felt very priviledged to be able to look around and share this cool space with so many interesting and inspiring people.
The talks I went to were:
Are we asking to much? @mheap‘s look into bad recruiting experiences which echo my own. The ratio of good candidates to bad candidates is massive and finding the right people is really tough. An emphasis on investing in people and “aptitude to learn” vs. “got all the skills today” seemed to be a consensus.
Less (for CSS) by @46bit was an overview of Less and the various ways it can be used to streamline CSS maintenance. Apparently, once you start using it, returning to projects without it is painful.
Turning Data into People by @101ofawolf was a look into the quest of name research and piecing together anecdotes of scoundrels from historical sources of information other than the conventional census and birth, marriages and deaths register.
Public Speaking Tips from @proactivepaul with the helpful suggestion of turning umms and errs into dramatic pauses and other protips.
High-speed Flash Photography by @cazm was another great talk showcasing stunning photography taken using home-made gear. The slides can be enjoyed here.
Build Your Own Museum by @katemonkey was a look at the publishing tools available and showcased some great reasons why you should share your passion. There was also a collective mourning for the loss of Geocities which for many of was an cyber-childhood neighborhood.
Round Robin Databases by @jbjon (the man who got me hooked on trololo) asked whether “we really need all that log data?” and showcased a great way to store time-series data using predictable fixed-size buffers and some popular tools for doing this.
Inventing Money with @RichardSmedley was a look into the origins of currency and a look at complementary currency schemes for exchanging wealth and value. Richard introduced the Ripple Project an idea of peer-to-peer based money lending system based on trust, where your credit rating is the sum of how much your friends would be willing to lend you based on their judgement of your ability to repay. A great idea that resonates with how most of us mentally assess lending our mates a tenner – in the interest of research I tested this econmic theory out on @caius and @martinrue who both must believe I’m good for a jar of pickles.
Other realisations Richard passed on were that role of speculative capital in markets is to supply liquidity. And that the amount of speculative capital available on the planet now far outweighs all primary and secondary value production – IMHO this is directly responsible all the speculative bubbles we have been witnessing recently.
Introduction to Coffee Script by @thetombell was a walk-through the various uses for Coffee Script and its abilities to simplify some of the messier aspects of Javascript.
Build Your Own Museum part II looked at the actual The Textile Manufactures of India online museum built by @GarethDoodles to showcase the fabric sample collection curated by John Forbes Watson in 1866. The goals of the project were to provide access to the collection and bring it and its meta data to life. This Joomla-based site totally nailed that. Great job!
3D Printers – not really a talk per se – but the Hacker Space was where I met my first 3D printer – a clever device capable of printing most of the components required to build a replica of itself. Awesome!
BarCampers Guide to Bitcoin by @wisemonkeyash delved into the cryptographically complex world of BitCoin and explored the tools, technology and ecosystem that is emerging around this new form of alternative currency. See the slides here.
The economic theme-de-jour of the current excessive supply of money was revisited with the friendly and helpful advice given: invest for the future in gold, tins of beans and shotguns.
Much to @ruby_gem’s fascination, this talk provided fresh fuel to earlier discussions about the digital currency, its characteristics, limitations and applications for money laundering.
Tattoos 101 by @msemtd gave an all you needed to know talk about tattoos, how they work, a look at the tools and the trade and the qualities of the tattoo artist. In particular the people skills required by a *good* tattoo artist who might advise you against a tramp-stamp on your face. And we tattoo’d an orange!!
Lock Picking: The Key Fundamentals by @ShannonTony was a high baud presentation into the workings and attack vectors of door locks. Really interesting and alarming to understand how trivial it is to bypass locks.
As you can see from this epic list, BarCamp Nottingham was great. On top of these sessions, there were countless fun, chats and catch-ups with new and old friends. Thank you @ruby_gem for organising and running the day – I look forward to BarCamp Blackpool!
TagWalk’s round-up of #bcnott
Social Media Fragmentation
by Tim on July 19, 2011
The arrival of Google+ has prompted me to rethink my use of social networks like Facebook and Twitter for publishing information and status updates. Already, some friends have moved away from Twitter and it has made me realise that the networks are gatekeepers to my friendships.
Something I read about social networks a while ago stuck in my mind… that ourselves and our relationships are being commoditised – we are the product.
Without question, social networks make it easier than ever for us to stay connected and make new connections – but I fear our voices are getting diluted across the ever growing social media landscape.
It occurs to me that as the number of social networks I participate on increases that my presence on each diminishes and my friends become more distributed. Furthermore, the energy required to maintain a cohesive mental model of who belongs where is becoming too great.
For many years I maintained a blog, but moved into Twitter and Facebook as the barrier to sharing tidbits of news and occasional photos was substantially lower than publishing a new post on my blog. As my friends were all users in the same spaces it made little difference – but now my friends are fragmented all over the place.
This post symbolises a change of strategy. I don’t think the solution is some kind of aggregator, but is to relocate.
I think I’m going to hangout here for a while.
And to see how easy it is to publish with modern tools, here’s a photo of my dog Millie.
Namaste.
Consuming DataSift Streams with Node.js
by Tim on February 17, 2011
This Node.js code sample consumes a DataSift stream and invokes a callback with each message object as it arrives.
The DataSift streams contain one JSON object per line, and may receive occasional empty lines to act as a “you are still connected” signal. The sample callback echoes each message out to the console.
/*
* Consuming a DataSift stream using node.js
* To use this sample, you will need your DataSift username, API key and a stream identifier.
* Copyright 2011 Tim Hastings.
* Released under the MIT License with no warranties given.
* Have fun!
*
**/
var http = require('http');
function consumeDataSiftStream(username, apikey, stream_id, callback) {
var buffer = '';
// parse the JSON of a single line and invoke the callback
function processLine(line) {
try {
var obj = JSON.parse(line);
callback(obj);
} catch (err) {
// parse failure or error from callback
console.log(JSON.stringify(err, null, '\t'));
}
}
// handle the HTTP response and attach handlers
var response_handler = function(resp) {
// data handler
resp.addListener('data', function(chunk) {
// append to buffer
buffer = buffer + chunk;
// consume any newlines
var pos = buffer.indexOf("\n");
while (pos >= 0) {
// take the line from the buffer
var line = buffer.substring(0, pos);
// remove from buffer
buffer = buffer.substring(pos + 1);
// pass to line handler if not empty
if (!line.match(/^\s*$/)) {
processLine(line);
}
// any more lines in the buffer?
pos = buffer.indexOf("\n");
}
});
// close handler
resp.addListener('end', function() {
// any remaining buffer will not have a newline in, so invoke
if (buffer.length > 0) {
processLine(buffer);
}
});
};
// create the connection and attach response handler
var path = "/"+stream_id+"?api_key="+escape(apikey)+"&username="+escape(username);
var client = http.createClient(80, "stream.datasift.net");
var headers = {host: "stream.datasift.net"};
var req = client.request('GET', path, headers);
req.addListener('response', response_handler);
req.write('');
req.end();
}
// test it
var username = 'USERNAME';
var apikey = 'APIKEY';
var stream_id = 'STREAM_ID';
consumeDataSiftStream(username, apikey, stream_id, function(obj) {
// is this an error or idle msg?
if (obj.status) {
console.log(JSON.stringify(obj, null, '\t'));
} else if (obj.tick) {
console.log();
} else if (obj.interaction && obj.interaction.author) {
console.log(obj.interaction.author.username+"\n "+obj.interaction.content);
} else {
console.log(JSON.stringify(obj));
}
});
Welcome 2011
by Tim on January 21, 2011
A new year and a new start, so here is the obligatory new blog. This year sees the final and necessary step in my long, drawn out plan to step into the world of freelancing and start-ups. Actually doing it.
In the year ahead I am looking forward to working with many of the friends I have met online and in person at Geekup, many BarCamps, PodCamps, hackathons and conferences. From a technology perspective, I would like to make full use of the tools and APIs I have been using in my projects rather than the more conventional enterprise technology stacks.
As well as different people and tools I am also looking forward to a change in lifestyle with a reduced commute to the office and a better work/life/wife balance. As I type, Ryan is sat next to me doing some virtual painting :-)
This blog will, from time to time, be an outlet for unashamedly geeky content, but for today, it looks like I have successfully managed to avoid a long list of geeky tech-speak.
Anyhoo, I must dash and crack on.
You can follow me on Twitter: @timhastings
