Pragmatism is my methodology

Skyrim. What ever you do, don't hurt the chicken

2011-12-22T10:56:00.001+13:00

System.ServiceModel.FaultException: The security context token is expired or is not valid

2011-12-07T15:32:00.001+13:00

A colleague and I encountered this error when trying to connect to CRM 2011 via the Organization.svc from a custom web page. It takes approximately ten minutes for the process to run and partway through we would get the error

“System.ServiceModel.Fault.Exception: The security context is expired or is not valid”.

Some basic googling point us in the direction of the security tokens on the OrganizationServiceProxy but when we dug deeper it turned out that there was no token assigned; very confusing. Eventually we found this blog comment which talked about when using Kerbros authentication if you use negotiateServiceCredential="false". When calling the service it must be running under the same machine name SVN. So we changed the calling custom web page to run under the same credentials as the service running CRM 2011 and lo and behold it started magically working.

I hope this saves someone else the 4 hours we spent bashing our heads against the computer.

Quote of the Day

2011-11-29T09:39:00.002+13:00

Quote of the Day:

Isaac Asimov:

There is a cult of ignorance in the United States, and there has
always been. The strain of anti-intellectualism has been a
constant thread winding its way through our political and cultural
life, nurtured by the false notion that democracy means that “my
ignorance is just as good as your knowledge”.

I found this link on Daring Fireball which i thought was worth republishing

Awesome Solar Activity

2011-11-16T15:02:00.001+13:00

Vodafone Warriors 2012 Draw

2011-11-10T13:18:00.001+13:00

The NRL draw for 2012 is out. The Warriors draw is below, you can click on it for a bigger image. 2012 is going to be our year. I’ve already got my season tickets sorted.

CRM 2011– Publish Customizations Failed

2011-11-01T14:21:00.001+13:00

My colleague Greg Olsen and I have been trying to publish an (admittedly massive) 6.5MB CRM solution for the past few days. Unfortunately we have been getting a very unhelpful “Generic SQL error”. We knew it was related to timeouts but after doing the usual timeout fixes (tweaking the reg file, changing the web.config etc) we were still getting the same error.

Fortunately we are not the only team to have encountered this issue and I found this blog by Darren Liu. While all the steps outlined in his blog are useful (and highly recommended if you are having timeout issue), the fix for us was editing the MSCRM_CONFIG database. To change the SqlCommandTimeout value for deployments. The default value is 30 (seconds) which is fine for small/medium size CRM solutions but for one like ours we needed a bit more of a buffer, so we upped it.

USE MSCRM_CONFIG

GO

UPDATE DeploymentProperties SET IntColumn=9000 WHERE ColumnName='SqlCommandTimeout'

Problem solved. Thanks Darren

What we SHOULD have been taught in our senior year of high school

2011-11-01T12:36:00.000+13:00

Typical gold from the Oatmeal

What we SHOULD have been taught in our senior year of high school:

View

SQL Server – Error 40

2011-10-29T21:24:00.000+13:00

This is a very useful blog post by Penal Dave. In it he nicely covers off the common troubleshooting steps if you happen to get this most meaningful of errors.

How to peel a garlic in less than 10 seconds

2011-10-29T21:12:00.001+13:00

Our World Champion All Blacks

2011-10-25T15:46:00.001+13:00

This video wonderfully captures the night from an All Black fan perspective.

Our World Champion All Blacks! from Jared Brandon Productions on Vimeo.

Adding a button to the ribbon – CRM 2011

2011-10-19T13:27:00.001+13:00

I got quite stuck trying this out recently, it turns out it’s not that intuitive. It requires a bit of xml editing of the Customization.xml file. This blog nicely covers the steps required.

Testing Web Services

2011-10-19T13:19:00.001+13:00

A client put me into this tool called soapUI which is hands down the best tool i’ve seen for testing web services (even if it does run on java).

It can test any web service and any WCF service that uses HttpBinding as well. It’s has a very nice UI and is simple and easy to use.

How badly wrong can a newspaper be?

2011-10-04T14:21:00.001+13:00

Turns out very. As the UK Daily Mail leads the way to the gutter in the Amanda Knox verdict

Earth in 60 seconds

2011-09-19T11:04:00.001+12:00

Awesome video of the earth from space

NZ Warriors make the last 4 of the NRL 2011 Finals

2011-09-19T08:52:00.001+12:00

So it was a great weekend for NZ sport, in particular if you’re a NZ Warriors fan where they managed to beat the crowd, the Tigers and also the ref to grab one of their most memorable victories of all all time

Wahooooo

2011-09-14T14:11:00.001+12:00

So after a decent hiatus i’m back blogging. A combination of domain name expiring and awe inspiring laziness.

I’m back baby!!

Words of the day – Sesquipedalian Loquaciousness

2010-10-14T16:30:00.001+13:00

“A predilection by the intelligentsia to engage in the manifestation of prolix exposition through a buzzword disposition form of communication notwithstanding the availability of more comprehensible diminutive alternatives.”

A colleague pointed this phrase out to me at the office today. So far I’ve spent the last ten minutes practicing trying to pronounce it.

[ses-kwi-pi-dey-lee-uh n] [loh-kwey-shuh s ness]

And it’s available for you right now for only ten measly syllables…

Who messed with my config file?

2010-10-05T16:06:00.001+13:00

I hit this doozy of an error in CRM when deploying code to a production environment today. Specifically when i was publishing changes to a workflow.

“AN ERROR OCCURRED WHEN THE WORKFLOW WAS BEING CREATED. TRY TO SAVE THE WORKFLOW AGAIN”.

WTF? So apparently when you have made unsupported changes to the CRM config file, when subsequent rollups are applied they don’t update the config file, As it detects that it has changed and does not apply the latest updates to prevent the unsupported changes from being overwritten.

I’m looking forward to finding out who the tool was thought it was a good idea to makes changes to the config file on the production environment and then not tell anyone. There’s a straight shooter with upper management written all over them.

Thank god for Pablo Peralta and this blog he wrote last month, describing the cause and the fix. He probably just saved someone's life.

TDD Kata

2010-07-07T10:20:00.001+12:00

Courtesy of Roy Osherove. A nice TDD Kata that helps to introduce the applying of the fundamentals of TDD to any language. Red, Green, Refactor!

Enabling a wireless adapter on Hyper-V

2010-05-19T15:50:00.001+12:00

I was working from home last night and kick started my Hyper-V VM to VPN to the client development environment (in order to remote to the development machine.. don’t get ne started on the 3 or 4 computer hop) when i found that my image wouldn’t connect via my wireless connection. I did some digging and it turns out that Hyper V does not support connecting via a wireless adapter by default. (There are apparently good reasons for this.)

It requires a small amount of configuration kung-fu. I found this great article by Ben Armstrong a virtualisation manager at Microsoft outlining how to resolve this issue

To paraphrase

Create a new internal network on the virtual network manager – i called mine wireless
Open network and sharing and find the wireless network adapter on your machine (make sure you have the actual adapter and not a wireless network you can connect to).

Right-click and find properties and select the sharing tab
Check the Allow other network users to connect through this computer's Internet connection checkbox

You’re done. Assign the internal network adapter to your hyper-v image and enjoy.

Test framework comparison

2010-05-06T09:37:00.001+12:00

A great comparison of the various .NET testing frameworks available

SIM to MicroSIM

2010-05-06T08:59:00.001+12:00

Great “how to” article by John Benson. The meat cleaver is nice touch too,

Google to launch digital books

2010-05-05T09:43:00.001+12:00

Currently called “Google Editions” this might hopefully start Amazon and Apple to start unshackling their respective Kindle and iBook stores.

“The company is hoping to distinguish it from offerings from incumbents like Amazon by allowing users to access books from a broad range of websites using a broad array of devices”

Speaking as a consumer in New Zealand who can not access e-book content on either platform (without some serious geek kung-fu), because we suffer form some archaic form of regionalised protectionism (we’re a global economy, media companies… come join us in the 21st century) i would just like to say…

It’s about time!

Some useful CRM links

2010-05-03T16:00:00.001+12:00

What’s with all the CRM?

2010-04-20T13:11:00.001+12:00

In case you’re curious I have shifted teams at work and now I am working with Microsoft’s Dynamics. Specifically CRM at the moment. So for the last two weeks and for the next month and a half I am working with Microsoft CRM. And quite enjoying it too.

Change is good!

Writing to a CRM date control from a child form

2010-04-20T13:00:00.002+12:00

Despite the fairly innocuous title this issue had been bothering (driving me crazy) for most of the last day.
Problem:
The clients CRM Contact form has a button that uses JavaScript to launch an external web page (we’ll call it the child form). This form would perform a search and return a list of matching users. Each user would have a button and when you clicked a user’s button it would use JavaScript to populate the parent form’s (in this case the CRM contact form) fields such as address, gender, ethnicity and date of birth.

e.g. opener.document.crmForm.all.address1_line1.DataValue = [address text];
Now this worked fine for the text fields, and pick lists but when I tried to write to the date control it did not play nicely. Now to populate a date control in CRM using JavaScript you set the DataValue to a JavaScript date
e.g. opener.document.crmForm.all.birthdate.DataValue = new Date();
However when I called this code from the child form it would populate the CRM Contact Form’s date of birth control successfully but that’s as far as it got. It would then hang if I tried to close or save the form. It would say a CRM error occurred and it that it had to close the form.
However when I tried putting the exact same piece of code without the parent reference in the Contact form’s CRM load event
e.g. crmForm.all.birthdate.DataValue = new Date();
it would work fine. It would save or close the form with no problem. Awesome!! Nothing like inconsistent behaviour.

Solution:
My work colleague Steven Foster wins the chocolate fish for the “magic” workaround (I thought magic sounded better than dirty) that saved the day :)
The solution was to have a hidden text field on the CRM Contact form that the child form would write the date to as text e.g. “21/11/2009”. Fortunately the child form would always write the date in a consistent manner so I could rely on it being in the format “dd/MM/yyyy”. Then I called the hidden fields FireOnChange() event to trigger the CRM OnChanged event. I then added a CRM OnChanged event to the new hidden text field that would parse the string and convert it into a JavaScript date and use that date to populate the date of birth control.

The JavaScript code on the child form
opener.document.crmForm.all.[hidden text field].DataValue = [date of birth text];
opener.document.crmForm.all.[hidden text field].FireOnChange();

The JavaScript on the CRM Contact from hidden text field OnChange event
var textDate = crmForm.all.[hidden text field].DataValue;
var dateParts = textDate.split(“/”);

var birthdate = new Date();
birthdate.setDate(dateParts[0]);
birthdate.setMonth(dateParts[1]-1);
birthdate.setFullYear(dateParts[2]);

crmForm.all.birthdate.DataValue = birthdate;

So there you go problem solved!

How to migrate iTunes libraries

2010-04-19T08:26:00.001+12:00

Surely this process should be easier…

Do not develop CRM on a machine that you are also deploying into the GAC

2010-04-15T16:22:00.002+12:00

I was getting a System.TypeLoadException as it couldn’t find the file i had just added in my project. Every time i ran my test against the project i kept getting the same result. Enter plenty of colourful language and head bashing.
Fortunately in my googling i came across this helpful explanation on StackOverflow. It turns out that if a version of the file exists in the GAC then it will use the GAC version (assuming the version number is the same) and not the version you think you are debugging. Thank god for Google but how did i get here?
I am currently on a Microsoft CRM project and i had to build a plugin (dll) and deploy it. The development environment is hosted by the client and is also the test environment. Anyway it turns out when you want to deploy two dlls as part of a plugin (for example the plugin dll and some domain or business logic layer dll) then you have to deploy the secondary dll (not the plugin) to the GAC so that it can be referenced… nuff said there really.

So the short version is try and avoid debugging projects that are in the GAC as well. I think Chandler from friends sums it up quite nicely “oh god, can open… worms everywhere”

MP3 store Comparison

2010-04-14T12:35:00.001+12:00

It is very interesting to see who is putting in personal tracking details in the music they sell. As a bit of an Apple fan i am pretty disappointed

Base 10 file

2010-03-30T08:38:00.001+13:00

A friend of mine on twitter @benmcdowell (i recommend his Harmonica iPhone app by the way) recently twittered about the file size 1 kB being 1024 versus 1000. The reason being that Ubuntu has just made the shift to base 10. With Apple having made the shift in Snow Leopard it leaves Microsoft still clinging to a slightly outdated version of measuring file size.

It’s an interesting topic as it seems logical to shift to base 10 but it is obviously a lot easier to execute the ostrich manoeuvre and just stick your head in the ground and hope it goes away

Kind of reminds me of the classic metric versus imperial debate

The NY Times on the tea party movement

2010-03-29T09:02:00.001+13:00

The week before the health care vote, The Times reported that births to Asian, black and Hispanic women accounted for 48 percent of all births in America in the 12 months ending in July 2008. By 2012, the next presidential election year, non-Hispanic white births will be in the minority. The Tea Party movement is virtually all white. The Republicans haven’t had a single African-American in the Senate or the House since 2003 and have had only three in total since 1935. Their anxieties about a rapidly changing America are well-grounded.

Mercurial

2010-02-25T23:01:00.000+13:00

Today i was planning to write about my decent into the world of Distributed Version Control Systems (DVCS. Specifically i was going to talk about Mercurial but it turns out that most of the heavy lifting has been done for me.
I felt very motivated to use a DVCS in my next personal project after reading Martin Fowlers bliki and then writing about them last week. Based purely on Martin Fowlers recommendation, some light googling and my magic 8-ball I decided to go with Mercurial. Also the fact it has strong windows and mac support (unlike GIT) and is a fairly mature product might have had something to do with it :)

Coming from a Subversion background i had a bit of unlearning to do. My first mistake was forgetting that as a distributed version control system every machine needs to have its own repositories. Whoops.... that little assumption took me about half an hour of fluffing around trying to understand why my mac client (murky) kept tanking on me.

Anyway it turns out that there are some great articles/tutorials out there of those who have done the hard yards for you. So why reinvent the wheel?

Joel Spolsky has put together a fantastic series of six tutorials at http://hginit.com outlining the key concepts. My favourite is the first tutorial where you are forced to unlearn your bad subversion habits
StackOverflow has a great wiki on all you would ever need to know on the who/what/where/why/how of Mercurial
A great Mercurial client for the mac by Jens Alfke called Murky
And for windows subversion users there is of course TortoiseHg to ease the pain of transition

So enough blogging. Time to cut some code

What happens when you type in a url

2010-02-25T22:29:00.000+13:00

This is a great blog post by Igor Ostrovsky at Microsoft about what happens when a url is entered.

This should be compulsary reading for any web developer or anyone at all interested in the actual workings of the internet. This is where the magic happens :)

Source control

2010-02-18T22:02:00.000+13:00

Martin Fowler has just posted a really interesting blog on source control... my favourite part was this comment about Visual Source Safe

"I've heard too many tales of repository corruption to trust it with anything more valuable than foo.txt."

Brillant! I think anyone who has touched VSS would agree with these sentiments.

The blog actually got me thinking about source control and continuous integration which is a natural level of progression from it.
I was never introduced to to any type of source control or even an overview of what source control was, while i was studying at university. Although that shouldn't really surprise me i guess... i hear they still teach the waterfall software development methodology anyway.
The point is, is that these are major tools in any developers arsenal, from one man start ups to large developer teams. Yet graduates are coming out of their studies and we have to teach them from scratch, the why as well as the how. Being language and platform agnostic there is no reason for universities not to teach these fundamental skills.

Anyway I have been keen to take a look at a distributed version control system (DVCS) for some time now so perhaps this is the push i need. The only question is now GIT or Mercurial?

What is your preference and why?

From C# to Objective-C

2010-02-11T23:16:00.001+13:00

The book The Pragmatic Programmer recommends that as part of a programmers ongoing learning that we should learn a new language every year. So this year i thought I would actually follow through on this and learn Objective-C.

The reasoning was pretty simple. I love my iPhone, so messing round with writing apps for it in my own time kind of appeals. Also i own a still shiny late 2009 MacBook Pro, so i can just download the SDK and start having a poke around. Finally i have always wanted to learn C in some way shape or form, (in fact Joel Spolsky has quite a strong opinion on this) and as Objective-C is classed as being a "Strict superset of C" i thought this would be an ideal way to knock off two goals of mine.

Now as a Java/C# developer by study and then professional life I'm not quite used to the different syntax or degree of flexibility that a lower level language live Objective-C offers. For example an Interface is not what i understand an Interface to be, for that you need a Protocol (or are they more abstract classes?) Good times!!

Fortunately in my potting around i've found some good links/resources which i've included below. I'll continue to update this list so please feel free to drop me any of your tips or hints.

This is a great post by Scott C Reynolds who also gave me the inspiration for this post as well.
Very useful article on going from C# to Objective-C and some of the pitfalls encountered. Also a nice reminder of some very helpful design patterns.
Great blog by Chris Small with heaps of detailed code examples comparing C# to Objective-C
A useful question raised by someone in a similar situation on StackOverflow

Let me know what works for you

Google as an ISP?

2010-02-11T13:39:00.004+13:00

Anytime that Google wants to come to New Zealand and try this out is fine by me

http://googleblog.blogspot.com/2010/02/think-big-with-gig-our-experimental.html

1 gigabit per second over fibre straight to your door! I don't think the day is far away when houses will start needing fibre in the house, not just leading to it.

Xcode - Can't find outlet or action server on the inspector panel

2010-02-09T21:38:00.004+13:00

Useful link

Turns out it's on the library window on the classes tab. As they refer to the class not the instance.

So it turns out that vaccines do not cause autism

2010-02-08T13:16:00.002+13:00

What i don’t get is how come it took the lancet so long to print the retraction

Great comment on flash… or the lack thereof

2010-02-05T13:23:00.002+13:00

I really enjoyed this little article by Barry Dorrans talking about how Apple not having flash (or the little blue brick) on the forthcoming iPad is not such a bad thing
My favourite line “It’s not a little blue brick people, it’s a little blue condom …”

Lamb on a spit

2010-01-31T21:03:00.002+13:00

Just a friendly warning, if you are squeamish then you might not want to scroll down. There are photos of a lamb being prepared for cooking.

So today was the day of the trial run. Despite the grey weather and one of the two burners playing up i think i can safely mark it down as a successfully day.

We were guided through the process by our resident spit expert Anthony. So without further ado.. lets get amongst it

The lamb

Step 1 (see above). Insert the rod at the business end. Insert skewer with the wing nut facing up, if you hit bone use a hammer to punch the skewer through it. Ensure you have room at the other end of the rod (move the lamb down the rod if necessary). Tighten the wing nut.

Step 2 (see above). Insert the skewer at the neck end with the wing nut facing down this time. Again hammer the skewer if it gets stuck on bone.

Step3 (see above). Adjust the position of the lamb by hammering the prongs at the neck end (making sure you remember to loosen the wing nuts) until feet are just short of the grove for the spit. Pull the hind feet up and hold against the rod. Thread the wire through the tendon and around the rod and then loop around again this time going all the way under and the feet. Then get pliers lift up to get tension on the wire and twist until the wire is tight. It is important to note that you do not want to over tighten the wire as this will weaken it and cause it to snap.

Step 4 (see above). Take the neck and hold against the rod. Thread the wire through the hole in the rod and wrap around the lamb several time and then tighten by lifting up the wire and twisting (as in step 3)

Step 5 (see above). Take about 1 foot of wire and shoe horn it so it looks like a very skinny version of the letter “U”, then about two thirds of the way down the lamb punch it through the body either side of the spine trying to aim between the ribs. This might take several attempts. Then pull the wire through with pliers and lift, tighten and twist (as in step 3). Then about one third of the way down the lamb repeat the process.

Step 6 (see above). Take a carving knife and punch 20-40 holes (depending on the size of the lamb) all over the lamp. Then insert a clove (or half clove) of garlic into each hole and add a small one inch length of rosemary. Then rub a generous amount of salt through the inside and outside of the lamb.

Step 7 (see above). Take two bits of wire both about a foot (or less) long. Close up the chest cavity at the head end and take the two bits of wire and tie up the cavity. One at just below the highest point (picture on the right) and the other about half way or a bit further down.

Step 8 (see above). Remove the front legs with a boning knife. This is quite tricky and you need to make sure you cut through all the tendons while leaving the knuckle on the lamb. Then on each leg hold the meat side up and half way down cut all the way to the bone (as if you were trying to cut it in half) then put the two front legs aside.

Step 9 (see above). About 1/3 of the way between the top of the chest and the bottom of the chest make two 1 inch cuts running parallel to each other on either side of the lamb. Then pull the skin though like a button hook and feed a front leg through with the cut facing up so that the ‘button hook’ fits in the cut. Then repeat again for the other foot about 2/3 of the way down.

Step 10 (see above). Take a whole onion and any remaining garlic and put into the cavity of the lamb. Then close up any remaining gaps with more wire. Finally rub some more salt on the outside.

Step 11. (see above). Throw onto the spit and cook for 3-4 hours (depending on weight). We cooked this 15.3 kilogram lamb for 3 hours.You probably need to allow 4+ hours for 17-18 kilograms.

Step 12 (see above). Remove when cooked and cut all the wire. Then cut up the lamb and enjoy!

Steven Fry on the iPad

2010-01-29T08:44:00.003+13:00

He writes an insightful article about the iPad and why you should try it before you judge it.
My favourite line
“Slightly annoying that the iPhone autocorrects iPad into upas – which is a kind of poison mulberry I believe…”

CI versus automated build tools

2010-01-21T08:35:00.002+13:00

Great blog post by Roy Osherove
http://weblogs.asp.net/rosherove/archive/2010/01/14/the-difference-between-continuous-integration-tools-and-automated-build-tools.aspx
Looks like the ultimate combo is Team City for CI and FinalBuilder for build automation

Interesting article on how many people have ever lived

2010-01-21T08:31:00.002+13:00

Just over 100 Billion give or take apparently.
http://www.prb.org/Articles/2002/HowManyPeopleHaveEverLivedonEarth.aspx

Run as different user – Windows Server 2008 R2

2009-12-04T14:05:00.002+13:00

I didn’t realise but as part of the upgrade from Windows Server 2003 to 2008 Microsoft removed the ability to run as different user as a context action. You can still do it from the command line using the command “runas” but on the whole that’s not very useful.
Fortunately Microsoft Sysinternals have a very useful too called ShellRunas.
It can be found here

AX reaches for the cloud

2009-12-01T08:41:00.002+13:00

Interesting blog by Brandon George about AX ROI and also AX starting to offer some services over the AX Cloud Platform

Payment Services for Microsoft Dynamics ERP
Connect for Microsoft Dynamics ERP

Kindle goes bad

2009-07-18T08:57:00.003+12:00

Are you kidding? It appears that owning a book on a kindle means that you own it at the publishers discretion and not your own.

I have been following the Kindles progress with great interest and have been patiently waiting for Amazon to introduce it down under to New Zealand and Australia.

However behaviour like this is a deal breaker.

Windows 7

2009-07-05T22:08:00.007+12:00

Part of my justification for getting a MacBook Pro was the fact that with bootcamp i can install windows. As i am a .NET developer this is kind of dealmaker or dealbreaker :) The process was easy.

1. Create partition in bootcamp
2. Boot to windows and format partition
3. Install windows
4. Drink beer

MacBook Pro

2009-07-05T21:09:00.006+12:00

So after much debating and googling i finally bought myself a new laptop. Its been on the cards for quite a while and i did a lot of looking around.
I was quite drawn to the macbook pros when i first saw them but the cost kept put me off. I had a look at the dell XPS range, the IBM thinkpads and a quick nosy at the sony vios but i just kept coming back to the macs. The design, the build quality and the fact i could maybe play at some iPhone development were all just too tempting.

So i ordered one about two weeks ago and it arrived last week. Aside from the fact that the numpty TNT courier dropped it about 10 feet in front of me when he was trying to put in on the receptionists desk -it survived the fall- i have been impressed.

The unboxing was treat. I thought that when i open my iPhone that that was an awesome experience, however it pales into comparison when opening a MacBook Pro, the whole experience puts you in a delightful frame of mind and then you simply open the aluminium lid and enjoy.

I have been using the laptop for about a week now and so far it has been a dream. Easily the most satisfactory piece of hardware i have ever owned

JQuery Ribbon Manager

2009-05-28T10:38:00.002+12:00

Saw this on a Scott Hanselman tweet. Some nice JQuery candy
http://jqueryribbon.codeplex.com/

Booting from a VHD with Windows Server 2008 R2

2009-05-14T23:33:00.006+12:00

*Updated 11 June 2010* - fixed some incorrect text when attaching the VHD and added a gotcha when building the VHD in a drive other than your primary c: drive

To skip straight to the technical setup click here
I have been developing on virtual machines for quite some time now. First with Virtual PC/Server and then a little more recently with HyperV and although there are many well documented major benefits throughout the web to this approach to development, there is a downside

You have to share processor power and RAM between two Operating Systems
You get no decent native graphics hardware support so no luck with duel monitors (i do realise you could use /span in remote desktop but it is a pain to configure the setup and if anything crashes or you need to reboot…. you get to start all over again)
No native graphics support also rules out developing for WPF or XNA in an meaningful way

With the release in the last few weeks of the RC versions of Windows 7 and Windows Server 2008 R2 came the latest virtualisation offering from Microsoft ‘booting from a VHD’. A couple of quick points to note. First, this is not meant as a replacement for running VMs on HyperV, VMWare, et cetera, it is just another arrow in the developer quiver. Secondly, if you are not a fan of duel booting your machine then this approach will probably not have much appeal to you either, as essentially this is the same behaviour but with some extra niceness. You can boot from the VHD and get all the *RAM/cores/processor speed your dev machine has to offer plus all the native hardware support (duel monitors baby!!!) and then like any VM when are done with the project you can park it.
*There is a small amount (~120MB) of RAM that is set aside to run the VHD that the VHD cannot see.
For me the decision to go down this path was a no brainer. My current machine is no slouch -duel core 2.83GHz, 4GB RAM and running windows server 2008 with no themes installed and desktop experience turned off . However when I was using HyperV to run my development VM, the host OS was only running Outlook and my browsers and I was having to set aside 1.5GB for the base OS when i was creating my VMs. Think about that for a second… 1.5GB to run Outlook, IE and Chrome/Firefox! While this in of itself is not necessarily a bad thing, when you combine this with my development VM setup (2.5GB RAM ) which is running SQL Server 2008, SQL Reporting Services, Navision 2009 and Visual Studio 2008 (with ReSharper obviously) running a WinForms solution of some 125915 lines of code (not including comments, region tags or curly braces) then you can probably start to see my pain point. So short of applying some dilithium crystals to my VM i needed a way to realise all the processing power on my machine.

The technical stuff

To boot from a VHD requires either Server2008 R2 (RC) or Windows 7 (RC)
There are two ways to set this up you can upgrade from a previous version or run through a clean install which also happens to be the MS recommended approach.
I choose the upgrade path and went from Server 2008 to Server 2008 RC. To do this upgrade requires 15GB of HD space and if you are running HyperV you have to disable the role as the upgrade will not proceed with it enabled.
I ran through the upgrade process and it took about 80 minutes. Much longer than the 20 minute install but worth it to keep my system intact.
Once the upgrade/install is complete reboot from the Server2008R2/Windows7 disk and when you get to your first splash screen (for me it was the select keyboard configuration) press SHIFT+F10 to launch a command window

*Important* Be aware that if you are going to build your file at a location other than the c: drive then the letters might not match up. I assume this has something to do with reserved letters not being allocated at this point and the partitions just get allocated when they appear. Hence what might be f:\ for you might be e:\ at this point.

type diskpart at the promt. Then
type create vdisk file c:\Server2008R2vhd.vhd type=fixed maximum=30000
You can call your file whatever you like I called mine Server2008R2vhd.vhd. For the size of the VHD, I choose 30GB (30000) which is huge for a VM but really whatever suits you.
type select vdisk file=c:\Server2008R2vhd.vhd (selects the VHD)
type attach vdisk (this used to be called 'surface'). Essentially it mounts the VHD as a visible partition
we're done. Type exit and then exit again
Continue on with the installation configuration i.e. pick an OS version etc. until you get to the point where you choose your partition to install to.
Select the newly added partition which should be the same size as you added (fingers crossed) and hit go.
Sit back, configure as required and enjoy.

Things to watch for

You cannot use an existing VHD i.e. a backup, HyperV, Virtual Server or PC versions. It must be a brand spanking new clean version. There probably is a nice workaround for modifying an existing VHD for this purpose. If anyone knows what to do then please let me know :)
The VHD was added as boot menu option with the exact same name as my host OS (they both said Windows Server 2008 R2.) The top option is the newly added VHD. Which is useful as the machine needs to reboot several times as part of the install process. You can rename this options using BCDEdit (see below)
I was able to rename and remove my first few unsuccessful attempts through BCDEdit
typing bcdedit gives a list of the all profiles for booting
bcdedit /set {ntldr} description "My Awesome boot VHD" with {ntldr} being the identifier of the profile or {current} or {default}
you can delete by typing bcdedit /delete {GUID} /cleanup

Good luck and happy booting

Staying home when sick

2009-05-11T20:42:00.001+12:00

I saw this very interesting article on www.stuff.co.nz some months back and have been meaning to post it as it it struck a chord.

Bottom line. If you are feeling sick.. stay home!! End of story

Decimal versus Double

2009-04-03T14:04:00.005+13:00

I was trying to remember the difference between a decimal and a double and came upon this handy little reference on MSDN

Essentially decimals have a far greater level of precision but doubles have a much larger range. However in most case when dealing with UI why would you ever need the range of a double?

My ten cents anyway

Mind you, when you save it to the database it shrinks it down to whatever your DBA thought was appropriate anyway :)

Sql Express 2008 - saving changes

2009-04-02T22:29:00.005+13:00

Recently I installed SQL Express 2008 on my home machine and on my first outing with it I received this error

Saving changes is not permitted. The changes you have made require the following tables to be dropped and re-created. You have either made changes to a table that can't be re-created or enabled the option Prevent saving changes that require the table to be re-created.

It turns out that the ability to edit a table that will result in the table being recreated in SQL Server Express 2008 is turned off by default. Awesome!!!

To enable this you need to go the the menu
Tools --> Options --> Designers
and then Table and Database Designers. Uncheck the checkbox marked "Prevent saving changes that require table re-creation"

Problem solved!

What's in a name?

2008-10-23T21:01:00.010+13:00

I have been musing over a decent name for my blog for some time now. It has been my favourite go to excuse for not following through with writing any blog topics that have come up.
The conversations in my head usually run along the lines of "i should really blog about my 3G iPhone (they rock by the way, everyone should have one) but really i need a cool name for my blog first". Lame excuse i know but the only person i had to convince was me so it was not really too hard.
Anyway i was in a meeting today at work and my manager Joe made a comment that really resonated with me (bullsh*t bingo phrase i know but the term is still relevant.) About how in our current project life cycle we've gone from XP to Agile to some form of Waterfall back to Agile and then to some hybrid form of a combination of all of these or as it was referred to.... being pragmatic. The phrase seemed to mesh nicely with how I view the world of development so i thought i would reuse it. For example whilst it would be great to use DDD, TDD, CI, pair programming, MVC/MVP on every project that comes up, the reality is that sometimes it just does not make sense to go down that path and use those particular disciplines and part of being a good developer is knowing when that is.

Anyway for me one of the key tenants of being a successful developer is recognising when someone has done something good and stealing it and claiming it as your own.
So cheers Joe I owe you a beer

Autocomplete textboxes

2008-10-23T13:56:00.000+13:00

So we had this interesting issue on this WinForms project I have been working on. We have been making quite extensive use of autocomplete textboxes in our application and one of our testers raised a continuity issue that when they selected an auto-complete option from the list with their mouse it behaved differently from when a user hit the tab button. After some quick investigation it became apparent that there were no events to subscribe to on the auto-complete part of the control to modify the behaviour. I ran this issue past a number of my peers and we were quite stumped.

However thanks to some great work by a colleague who took it on as a personal challenge (thanks Faranak and your uber googling skills ) we managed to resolve the issue we were having. It appears that although Microsoft have effectively buried all the events and monitoring of the auto-complete part of the text box way down in the windows process there is a relatively simple fix.
When you click on one of the auto-complete text options or when you scroll down and press enter then the same thing happens. A KeyDown event is thrown with the ‘Enter’ key marked as being pressed regardless of whether it was a mouse click or the enter button pressed. Thus all we needed to do was monitor for the KeyDown event and the enter key, check that the text matches the list of codes (the auto-complete text source) for the textbox and implement our required outcome. (Cue much rejoicing)

Bens first blog

2007-06-06T09:07:00.000+12:00

Welcome to my blog. Well i have been meaning to start blogging for a while so i thought there is no time like the present. I have quite a wide range of thoughts and opinions so this blog will be quite wide ranging. Let the blogging begin.