Run Tings Proper

How To Fix Contact Form 7 v5.01 - Reverting Back To reCaptcha v2

2018-12-18T00:36:00.002+00:00

As of the time of writing, v5.01 is broken. There are hundreds of 1-star reviews which have been piling up over the last few days.

It's not clear if the plugin itself is at fault or if the v3 of reCaptcha is just not filtering out the spam as effectively (maybe it's having to retrain its model for each new site, I'm not sure).

Either way, CF7 only supports v3 of the API at the moment.

I'm sure it will be fixed soon but in the meantime, you can solve it be using an extra plugin.

It's called Invisible reCaptcha for WordPress: https://wordpress.org/plugins/invisible-recaptcha/

Step By Step Guide

Go to Admin > Contact Form 7 > Integrations and remove the API keys you have set up:

Then go to Plugins > Add Plugin and search for "Invisible reCaptcha for WordPress" by Mihai Chelaru.

Install it and activate it.

Open a new tab and go to your reCaptcha admin account: https://www.google.com/recaptcha/admin

Generate new keys with the setting reCAPTCHA v2 > Invisible

When you have filled out the form you will get some keys like you have used before:

Copy the Site Key into your clipboard.

Now go back to WordPress and go to Admin > Settings > Invisible reCaptcha

Paste the Site Key into the Site Key field, then go back and copy the Secret Key into the Secret Key.

Bonus: You can set the badge position - I prefer to make it the bottom left, or inline rather than its default right, that is often used up by a "back to top" element of the theme.

Click Save Changes.

IMPORTANT: There is one last step, it won't automatically work with Contact Form 7 forms.

Go to the Contact Forms tab, enabled it for Contact Form 7, and click Save Changes again on that tab:

OK your site is now safely protected again!

Hopefully, this will be fixed in the next few days.

PRO TIP: Don't forget to flush any cache plugins your site might be using.

I’ve turned 10

2016-04-01T13:28:00.001+01:00

Today marks the official day that I have been a professional web developer for 10 years!

That's a lot of buttons pushed, a lot of mice clicked and a lot of clients charmed. That's a lifetime in dog years. Its several lifetimes in internet years.

It got me thinking about how much has changed in the industry over these years. The industry moves forward so fast that I’m always caught up with what new technology I should be learning, not what technologies have faded away.

When I started the technology landscape was very different to the way it is now. Web 2.0 was the new buzzword although nobody could really define what it was. Real geeks of the era got their news via feeds through an RSS feed reader. Facebook was still only available to uni students, MySpace was rocking the social scene. Twitter and Tumblr weren't launched. YouTube was just over a year old.

We were just entering into the second browser wars, Firefox 2.0 was the hot new contender to the Internet Explorer throne. Chrome wasn't going to be a glimmer in Google's eye until two years later in late 2008 and it didn't overtake Firefox until 2011. When I started, I still had to test sites to make sure they worked in Firefox, IE6 and IE5.5 before they could be sent off to clients. There was no such thing as "in private" browsing. Embedded Java applets were still a thing. Splash pages with Flash animations were still clinging on. Ticker tape animations and custom mouse pointer icons were still acceptable. JavaScript was a silly side language for popups and altering your status bar text (although VBScript was long dead thankfully). I can’t even remember how many failed Internet Explorer integration points have come and gone (Active Desktop, Web Slices, Pinned Sites, Tiles… what else have I missed?).

I’ve also seen the rise of web typography, allowing font’s to be embedded into web pages. I remember all the legal kerfuffle when the font studios didn’t move with the times and still had restrictive licenses, and when Google stepped in and made it all nice and simple for us.

Broadband internet had pretty much hit the mainstream by 2006 but this was still a time of dial-up so we had a small file size budget. I’ve been trying to remember what I aimed for and I think it was about 150k for a page, it was considered heavy if it went close to that. This was before the jQuery era of including tons of 3rd party scripts in a page though. I don't think CSS and JavaScript were fast enough to run animated slideshows back then.

The world was just moving away from table based designs over to CSS tableless designs. I actually made my first tableless CSS website a week before my interview at DBS and was quite worried that I wouldn't be able to hold my own in a professional environment. At that time it was all still CSS 2.1 though. I actually just deleted a file the other day with a last-modified dated to 2005 which had some advanced new CSS3 rounded forms features! Of course, this was a Firefox only CSS prefixed feature back then.

I was developing in Macromedia Dreamweaver 8 (which was still to be later bought out by Adobe). I’d already passed through the era of using Notepad and then Microsoft FrontPage before I turned pro.

For years I built all of my pages so that they would validate against the W3C standards compliant validator and I included accessibility features such as high contrast modes for colour blind people and large print mode for people with poor eyesight. I only stopped when the W3C standards lagged so far behind that it was no longer a point of pride to be restricted by it. Over time, web browsers incorporated better built-in tools for browsing with disabilities so my accessibility improvements were also superseded. I tested the page with images turned off (older browsers would give up loading images after a timeout on slow connections, plus some dial up users disabled them for faster browsing) and ensured that all background graphics also had background colours set. I also tested the large print mode to make sure I didn't accidentally set a font size in px (because IE6 wouldn't scale the font size no matter what you did then). Despite their popularity I never used any CSS hacks to smooth over cross browser inconsistencies, as they could be patched in a future version and break the site. I thought if a client paid for a product they shouldn't be handed a ticking time bomb that would need patching. We never once came across a new client in the wild that had an existing website which met all these quality checkpoints. As a result, I was always proud of the work that I produced, even if the clients had no idea of the production quality they had received under the hood.

Geocities was still clinging on, WebRings were waning in popularity. MSN Messenger was still a thing; I got busted several times for chatting to my friends with it on work computers.

Content management systems were still in their youth. WordPress, Mambo, Joomla and Drupal were all slugging it out still in 2006, with no clear winner. Setting them up on a server was a complicated task of paths, permissions and database configurations, there weren't any 1-click installers in those days.

The main free-hosting accounts were basic html servers and if you were lucky you would get SSI (server side includes) so you didn't have to use frames, and a perl-powered cgi-bin folder that you could process your form-to-mail script in. Domain names were still a luxury and it was common to see peoples websites hosted on their internet providers servers with names like demon.net.uk/~someuser331.

Google has always dominated my professional career, but back in 2006 other players were still competing. Bing was still MSN Search, Yahoo was still actually indexing pages. Ask Jeeves still existed but rarely produced valuable results. Search Engine Optimisation still involved buying dashed domains, writing keywords tags and doing keyword stuffing in the footers.

I was still using Winamp and scrobbling my tunes with AudioScrobbler (now Last.fm). I had friends on ICQ and IRC and I still preferred PaintShop Pro over Photoshop.

Responsive design wasn't a thing until quite late in my career. The iPhone came out in 2007, Android in 2008, but the mobile internet revolution didn't really get started until about 2011, and it wasn't until 2013 that the first "year of responsive design" was declared.

My work desktop has gone from Windows XP up to Windows 8.1, moving from a single 17" crt monitor up to the three 19" bad boys I have today. I've taught my desk some new tricks, it can now "stand" as well as "sit" but it's forgotten how to have drawers in the process.

My mouse would still have stayed the same as it always was but my beloved Microsoft Intellimouse was viciously killed off when Microsoft discontinued it sometime around 2012. Being left handed meant that I've always had a limited choice of mouse to choose from so I was always faithful to this brand. Also because of the amazing build quality I didn't go to purchase any new ones and realise they had been discontinued until about a year later - way after the last of the stock had been sold. Eventually I took a chance on a Steel Series Kana v2 which after a short period of adjusting to the new side buttons has become a welcome tool in my hand.

I actually still have the same Microsoft Digital Media Pro keyboard that I've had for well over 15 years now I think. It's never failed me and despite my casual searches I've never seen a mechanical keyboard that has quite the right layout for me (I want the UK layout, with double height enter button, and some multimedia buttons like volume and a calc button - and I don't want it to look like some techno-future-bot).

So... if you've read this far then you must be a fellow geek. What was the latest technologies that you remember from when you first became a pro-developer?

Skype won’t send any pictures I try to send

2015-10-10T09:30:00.000+01:00

If you can’t get Skype to send pictures and keep seeing an error saying “this photo is unavailable” then hopefully this quick tip will get you back up and running with exchanging media.

For me I kept trying to send the pictures, and even other files like PDFs but I just kept getting the same error:

I even tried sending from my iPhone, restarted Skype several times, force closed and then finally I even rebooted the PC.

For me the solution was found when it occurred to me that I had recently changed my Microsoft Account password to make the password more secure.

It seems that Skype will carry on working if you change your Microsoft Account password and even survive a reboot with a valid auth key. Whatever it does under the hood with your images however it cannot pass them over until you have re-authorised.

Solution

I logged out of my Skype account and then logged back in again using my new Microsoft Account credentials.

After this the images started working for me again!

Orchard CMS Quick Tip–No constructors found by AutofacUtil ConstructorFinderWrapper

2015-10-07T09:30:00.000+01:00

While writing an Orchard CMS driver class for a tutorial recently I found myself confronted with this error message. If this has happened to you then your gut instinct might be that you are going to have to learn something about Autofac and it might get complicated. Read on for a quick solution.

The exact error message that I was presented with was as follows:

No constructors on type 
'Orchard.LearnOrchard.FeaturedProduct.Drivers.FeaturedProductDriver' 
can be found with the constructor finder 
'Orchard.Environment.AutofacUtil.DynamicProxy2.ConstructorFinderWrapper'.

My first reaction as I’ve said was that this wasn’t going to be solved until I had started digging into the Autofac documentation.

After taking a second to regroup and realised this was a pretty straightforward issue.

Solution

I’d forgotten to make the constructor public. If you don’t specify it then it defaults to private so Autofac couldn’t find it.

So if, for example, your code looks something like this:

FeaturedProductDriver(IFeaturedProductService featuredProductService) {
  _featuredProductService = featuredProductService;
}

Then you just need to add a simple public modifier to the start:

public FeaturedProductDriver(IFeaturedProductService featuredProductService) {
  _featuredProductService = featuredProductService;
}

And Autofac will be happy.

Orchard CMS Quick Tip–If you cast it they will come, or Accessing your field of dreams

2015-10-06T09:30:00.000+01:00

I see a lot of people coming up against issues trying to get at the fields they have added on their content types. If you can’t figure out how to access the data then read on to discover the important points you need to understand which will make accessing fields easy.

The first important thing to understand is: Fields are always in a ContentPart.

The second thing to understand is: If you’re in a class then you will need to cast the object to dynamic to get around IntelliSense.

The third thing to understand is: The field name is a ContentField type not a string.

Fields are always in a content part

If it looks like its just in the Fields section of a content type then you need to know what that type is. This is the type not an individual so you are looking for something like Page or BlogPost not “About Us”.

This means to access a ProductId field which has been attached to the Page content type you would use:

var productId = CurrentContent.ContentItem.Product.ProductId.Value;

Cast to dynamic

The second thing is that if you’re in a class file you will get stuck trying to access anything after the ContentItem because IntelliSense can’t see the dynamic properties of the class at development time.

It’s simple to access this information, just cast the ContentItem to a dynamic type:

var dynamicContentItem = (dynamic)CurrentContent.ContentItem;
var itemProductId = dynamicContentItem.Product.ProductId.Value;

The field is a content field type

The last bit is that you don’t just access the value of the field by its name ProductId because that is a class. If, for example the field type is a “Text Field” then this is actually a class called Orchard.Fields.Fields.InputField. This class exposes its actual data under the .Value property.

Other fields types use different methods to access their internal values depending on the type of data they are exposing. Make sure you look at the implementation to figure this out as IntelliSense won’t be available to you so otherwise you won’t notice you have made a mistake until an exception is thrown at run-time.

Sebastien Ros has created an Orchard Cheatsheet which covers common properties that you might want to access on each of the built-in Orchard content fields.

Follow these three simple lessons and you should get the information you need without too much hassle.

Orchard CMS Quick Tip–Zoned out, or Why changing Placement.info still doesn’t move your module

2015-10-05T09:30:00.000+01:00

This is a common problem that can catch out Orchard users and developers but after you learn this easy to remember tip you should be able to solve it quickly whenever you experience in the future.

You might be trying to move the default placement.info in your module, or overriding it in a theme. No matter what you try the module stays firmly in its original place. You’ve checked and double checked the zone name, the casing, the weights and

Frustrating right? Well let me turn that frustration from annoyance with Orchard into kicking yourself:

You missed the forward slash off the front of the zone.

Yep it was just one character that was stopping your module from moving to its desired location.

You see, Orchard has two types of zones, the local zone and the global zone.

By default the zone your <place> tag refers to is a "local zone" which means that it only sees the zones in that modules shape file (the .cshtml). This is useful to move it around in the immediate template to places such as header, content, footer, meta, etc.

However, if your zone is a "global zone", as in one that is defined in the theme.txt and is used somewhere in your global layout.cshtml file, then you need a forward slash at the start. This is how you move it around the template to zones like AsideFirst, AfterMain and FooterQuadThird.

Solution

Add a slash at the start eg from:

<Place Parts_Accreditations="AfterMain:0"/>

To:

<Place Parts_Accreditations="/AfterMain:0"/>

Orchard CMS Quick Tip–No Persister for: SomePartRecord

2015-10-02T16:31:00.001+01:00

If you’re making a module with a record class and you start seeing errors saying that there is no persister for the part record class you’re working on then you have probably made this simple mistake.

Once recently I started writing a quick module and thought to myself, I don’t need all of these extra namespaces, everything can just go in the general module namespace.

I set about coding and everything seemed to be going well but then I see this error:

No Persister for: SomePartRecord

You might also see this error if you accidentally made your record class in the root folder of the module and then dragged it over to the .\model\ folder.

The problem is that for Autofac to work it requires two criteria for it to be able to find your model:

Ensure that the models (ContentItemPart and ContentItemPartRecord) are in the .\Models\ folder
Ensure that the namespace wrapped around those classes ends in .Models

If you fix these two items then you will also fix the No Persister for: SomePartRecord error.

Orchard CMS Quick Tip–A duplicate value cannot be inserted into a unique index

2015-10-01T09:30:00.000+01:00

Work on an Orchard CMS widget had been going pretty well. It display OK and didn’t seem like it had any problems but each time I tried to save anything I’d get an exception thrown.

The exception started off like this:

A duplicate value cannot be inserted into a unique index. [ Table name = Common_IdentityPartRecord...

I had added an IdentityPart to the widget because I thought I had read somewhere that adding the IdentityPart was a best-practice for supporting the import / export module.

It turned out it was not.

After a quick search I found a post by Piotr Szmyd which explained the reasoning behind this:

You can't use Unique constraint when dealing with content part records. The reason is that those are created early, when no data has been passed to it yet (with all properties being empty/null). This happens inside ContentManager. Hence, all properties on those kinds of records need to be nullable and non-unique.
But you can use the unique/not-null constraints on classes that are mapped to records you create and use via IRepository directly.

If you have added the IdentityPart to a content part by accident to a content part then you can remove it with a data migration.

Open up your Migrations class (often called “Migrations.cs” or in the “.\Migrations\” folder).

Assuming that your content part is called “” the code would look like this:

public int UpdateFrom1() {
    ContentDefinitionManager.AlterTypeDefinition(
        "FeaturedProductWidget", cfg => cfg
            .RemovePart(typeof(IdentityPart).Name));
    return 2;
}

Obviously you should also update the UpdateFrom1() to whatever update version your migrations class is currently on.

Orchard CMS Quick Tip–Curling up with Orchard, or brace line breaks without ReSharper

2015-09-30T09:30:00.000+01:00

Orchard CMS requires that all commits to the core use the JavaScript style curly brace positioning, where the opening brace is on the same line as the method signature. If you use ReSharper then it comes with an automatic .resharper file that applies this out-of-the-box but non-users can still get this functionality with plain vanilla Visual Studio.

This isn’t a post against ReSharper at all, I think its great, but its not-free so not every developer has it. If you are wanting to contribute to the Orchard core, write a blog post about Orchard with code or just have your custom modules look like they are part of the rest of the code this quick tip will let you get the brace positions the same.

What’s the difference?

Lets take a simple example. A normal install of Visual Studio would format a piece of code like this:

namespace Orchard.LearnOrchard.FeaturedProduct.Models
{
    public class FeaturedProductPartRecord : ContentPartRecord
    {
        public virtual bool IsOnSale { get; set; }
    }
}

The opening curly braces each start on a fresh line. Orchard favours the more terse JavaScript style approach with the curly braces:

namespace Orchard.LearnOrchard.FeaturedProduct.Models {
    public class FeaturedProductPartRecord : ContentPartRecord {
        public virtual bool IsOnSale { get; set; }
    }
}

Update – Orchard also supports the free Rebracer extension

It was pointed out to me (#1, #2) that Orchard actually also supports a free add-on which will manage the settings described below. If you’re using a version of Visual Studio that supports extensions then you should probably install this instead:

Rebracer Visual Studio extension by SLaks

It’s a free extension that adds in support for automatically managing the settings for brace line-breaks based on the configuration values located in a rebracer.xml file.

Orchard comes with one of these in the root of the solution so you just install it and your Orchard solution will automatically pick it up, no further configuration required. The benefit of this is twofold:

Zero configuration
Your non-Orchard solutions can have their own (normal) brace line-break settings.

So unless there is some restriction with your environment it makes sense to use this.

How do I do it with pure Visual Studio settings?

It’s just a few checkboxes within the options panel:

Click Tools | Options…
Scroll down to the Text Editor node
Expand the C# node
Expand the Formatting node
Click on the New Lines node
You will see a list of options like in the image below which give you full control over when Visual Studio should put your open brace on a new line. Remove the check’s from all of the checkboxes in the “New line options for braces” section.
Click OK

When you change this setting it will change it for all of your projects, so if you switch between Orchard and non-Orchard projects you will have to come back and re-enable these settings.

Older versions of Visual Studio

You might not see these options. Some of the older versions of Visual Studio showed simplified settings by default. It seems like (in Visual Studio Community 2015 at least) that it defaults to showing all the settings now.

If you can’t follow the steps above then look for a Show all settings checkbox in the bottom corner of the Options dialog.

Orchard CMS Quick Tip–Getting at the Orchard.Layouts JSON for your template defaults

2015-09-29T09:00:00.000+01:00

If you have started using the Orchard.Layouts module which has seen some pretty significant updates in the last few releases you might be wondering how you get at the JSON data that you’ve been finding in content part default layout data setting. There are two ways but both are a bit tricky.

UPDATE

It’s been pointed out to me by several people that this technique doesn’t work any more. I don’t know an alternative way to solve this, sorry.

What do I mean exactly?

When you have a LayoutPart attached to one of your content items you will see a section like this in its Content Definition page:

This data pre-populates the page when you create a new content item.

If you want to customise your own content types with LayoutParts that come with defaults its possible but their currently there isn’t anyway to access this through the admin panel directly. Instead you need to take one of the two courses listed below.

XML export

The first option is to export the data using the import/export module that comes with Orchard CMS. You can so it this way but then you will get the JSON data in xml encoded format so it will need extra processing.

Access the clipboard data

(Special thanks to Daniel Stolt, one of the core Orchard.Layouts developers, for this idea)

In Chrome you can select an element, hit Ctrl+C and then that JSON data is actually placed on the system clipboard, along with the plain-text of the element. It's how copy-paste works.

The trouble is that you probably don’t have an application installed that will paste clipboard data with the type “Chromium Web Custom MIME Data Format” so if you try to paste it back again all you will get is the plain text representation.

To get at this data you can use LINQPad to extract this hidden JSON data:

Just follow these simple steps:

Download LINQPad 4

Paste this query into the query window:

var s = (System.IO.MemoryStream)System.Windows.Forms.Clipboard.GetData("Chromium Web Custom MIME Data Format");
System.Text.Encoding.Unicode.GetString(s.ToArray()).Dump();

Set the Language to C# Statement(s) for multiline support. Note the (s) on the end.
Query > Query Properties > Additional References > Add > Search Forms > Add System.Windows.Forms.dll reference > OK
Run the query
Bonus tip: File > Save to save the snippet for next time. You can load it back from the My Queries tab in the bottom corner.

Troubleshooting - NullReferenceException: Object reference not set to an instance of an object.

You copied something else into the clipboard while setting this up, copy the original back from Orchard.

Orchard CMS Quick Tip–Stop “Unable to launch the IIS Express Web server” SSL clashes

2015-09-28T11:27:00.000+01:00

When I’m working with Orchard I often want to have several copies of Visual Studio open, running several copies of Orchard for testing and comparison reasons. This is how I stopped the “Unable to launch the IIS Express" Web server” errors when I tried to launch two at the same time.

Each time I would try to spin up a new dev environment server it would come up with an error like this:

The second project was fighting the first project for the SSL port. I wasn’t using any SSL in my project but it was loading it anyway.

I’d gone into the project property pages by right clicking on the Orchard.Web project, selecting properties and configuring the Server in the build tab which had solved the initial clash but there was seemingly no option for the SSL settings.

It turns out that this is actually in the other project properties screen. If you select the Orchard.Web project then press F4 on your keyboard it will bring up the properties dialog:

You can then just double click the SSL Enabled = True option to disable the SSL port and stop the clash.

Orchard CMS Quick Tip–Orchard suddenly starts showing a 404 error for every page including the root

2015-09-25T11:18:00.000+01:00

As your working on your Orchard CMS website or maybe right after you have deployed it, you might suddenly find that nothing is working. If you’re getting a simple 404 page no matter what you try then this is the first thing you need to try.

Check the logs.

Orchard keeps logs in some text files inside the ~/Orchard.Web/App_Data/Logs/ folder.

The last time this happened for me it turned out I had created a circular dependency:

2015-09-01 17:50:21,718 [6] Orchard.Environment.DefaultOrchardHost - (null) - A tenant could not be started: Default
 (null)
Autofac.Core.DependencyResolutionException: Circular component dependency detected: Orchard.Environment.DefaultOrchardShell -> Orchard.Environment.DefaultOrchardShell -> Orchard.Mvc.Routes.IRouteProvider[] -> Orchard.Mvc.Routes.IRouteProvider[] -> rtpDesign.ExampleSite.Routing.Routes -> rtpDesign.ExampleSite.Routing.ExampleSiteSlugConstraint -> Orchard.Alias.Implementation.DefaultAliasService -> Orchard.Mvc.Routes.IRouteProvider[].
   at Autofac.Core.Resolving.CircularDependencyDetector.CheckForCircularDependency(IComponentRegistration registration, Stack`1 activationStack, Int32 callDepth)
   at Autofac.Core.Resolving.ResolveOperation.GetOrCreateInstance(ISharingLifetimeScope currentOperationScope, IComponentRegistration registration, IEnumerable`1 parameters)
   at Autofac.Core.Resolving.InstanceLookup.ResolveComponent(IComponentRegistration registration, IEnumerable`1 parameters)
   at Autofac.Core.Activators.Reflection.AutowiringParameter.<>c__DisplayClass2.<CanSupplyValue>b__0()
   at Autofac.Core.Activators.Reflection.ConstructorParameterBinding.Instantiate()
   at Autofac.Core.Activators.Reflection.ReflectionActivator.ActivateInstance(IComponentContext context, IEnumerable`1 parameters)
   at Autofac.Core.Resolving.InstanceLookup.Activate(IEnumerable`1 parameters)
   at Autofac.Core.Resolving.InstanceLookup.<Execute>b__0()
   at Autofac.Core.Lifetime.LifetimeScope.GetOrCreateAndShare(Guid id, Func`1 creator)
   at Autofac.Core.Res2015-09-01 17:50:21,718 [6] Orchard.Environment.DefaultOrchardHost - (null) - A tenant could not be started: Default
 (null)
Autofac.Core.DependencyResolutionException: Circular component dependency detected: Orchard.Environment.DefaultOrchardShell -> Orchard.Environment.DefaultOrchardShell -> Orchard.Mvc.Routes.IRouteProvider[] -> Orchard.Mvc.Routes.IRouteProvider[] -> rtpDesign.ExampleSite.Routing.Routes -> rtpDesign.ExampleSite.Routing.ExampleSiteSlugConstraint -> Orchard.Alias.Implementation.DefaultAliasService -> Orchard.Mvc.Routes.IRouteProvider[].
   at Autofac.Core.Resolving.CircularDependencyDetector.CheckForCircularDependency(IComponentRegistration registration, Stack`1 activationStack, Int32 callDepth)
   at Autofac.Core.Resolving.ResolveOperation.GetOrCreateInstance(ISharingLifetimeScope currentOperationScope, IComponentRegistration registration, IEnumerable`1 parameters)
   at Autofac.Core.Resolving.InstanceLookup.ResolveComponent(IComponentRegistration registration, IEnumerable`1 parameters)
   at Autofac.Core.Activators.Reflection.AutowiringParameter.<>c__DisplayClass2.<CanSupplyValue>b__0()
   at Autofac.Core.Activators.Reflection.ConstructorParameterBinding.Instantiate()
   at Autofac.Core.Activators.Reflection.ReflectionActivator.ActivateInstance(IComponentContext context, IEnumerable`1 parameters)
   at Autofac.Core.Resolving.InstanceLookup.Activate(IEnumerable`1 parameters)
   at Autofac.Core.Resolving.InstanceLookup.<Execute>b__0()
   at Autofac.Core.Lifetime.LifetimeScope.GetOrCreateAndShare(Guid id, Func`1 creator)
   at Autofac.Core.Resolving.InstanceLookup.Execute()olving.InstanceLookup.Execute()

Disabling the offending code fixed it straight away and I was back up and running in minutes.

SOLVED: Audio buzzing and crackling whenever the GPU is active or the mouse is moved

2015-09-24T15:00:00.001+01:00

This problem seems to be somewhat common out there and mostly the solution is unique to your particular setup. Recently I was struck with a bad case of audio feedback from my computer whenever anything was happening on the screen. It was so bad I couldn’t even play computer games any more as the noise was too distracting. After trying a ton of different combinations I finally figured it out.

I had recently purchased a new computer desk and in anticipation of it I’d tidied up the room, hoovering behind it and sorting out my cables.

However at the same time I had been trying out some old hard drives in my computer and also randomly my PC power button had stopped working. Plenty of variables to work with then.

I’ll cut to the chase. I have two active speakers, a six socket extension cable and another 4 socket extension. It turned out that actual issue was that during my tidying up I had put one speaker on the left hand extension cable and another on the right hand extension cable.

Having two different wall sockets involved with the speakers and the PC was creating a feedback loop.

If you’re having audio problems double check that your PC and your speakers / amp are on the same plug.

Orchard CMS Quick Tip–Always Start Without Debugging

2015-09-16T19:26:00.001+01:00

When you’re working on Orchard CMS in Visual Studio you will probably be tempted to press F5 as you normally do to start the solution in a web browser. Don’t do it!

You’ve probably already noticed that it takes a long time for it to spin up.

A quicker alternative is to press Ctrl-F5 (Start without debugging). A lot of the time you probably won’t need the debugger anyway. When you do need it you can just attach it without having to restart.

As a bonus you won’t see “Edit and Continue” issues every time you try to modify the source code when its running without debugging:

Starting the debugger manually

Click Attach to Process… in the Debug menu:

(Once you get used to this you will remember it’s Ctrl-Alt-P)

Then select your webserver, most likely iisexpress.exe:

And click Attach.

It will quickly attach debugger. Switch back to your web browser and carry on using it as normal. Do whatever you need to do to make it trigger the breakpoint or throw the exception you wanted to debug.

SOLVED: Web Deploy fails with The "CollectFilesinFolder" task failed unexpectedly.

2015-08-30T23:42:00.001+01:00

If you’re seeing this error when you’re trying to build a package with Web Deploy then you have some deeply nested folders in your project which is causing the problem. A common problem web developers of 2015 have to wrestle with is fighting the Node Package Manager on a Windows developer box. Read on to get this issue quickly solved so you can deploy your site.

Problem

When attempting to build a Web Deploy package Visual Studio fails with an error message similar to this:

Error Message
Severity    Code    Description    Project    File    Line
 Error        The "CollectFilesinFolder" task failed unexpectedly.
 System.IO.PathTooLongException: The specified path, file name, or both are too long. The fully qualified file name must be less than 260 characters, and the directory name must be less than 248 characters.
    at System.IO.PathHelper.GetFullPathName()
    at System.IO.Path.NormalizePath(String path, Boolean fullCheck, Int32 maxPathLength, Boolean expandShortPaths)
    at System.IO.Path.GetFullPathInternal(String path)
    at System.Security.Util.StringExpressionSet.CanonicalizePath(String path, Boolean needFullPath)
    at System.Security.Util.StringExpressionSet.CreateListFromExpressions(String[] str, Boolean needFullPath)
    at System.Security.Permissions.FileIOPermission.AddPathList(FileIOPermissionAccess access, AccessControlActions control, String[] pathListOrig, Boolean checkForDuplicates, Boolean needFullPath, Boolean copyPathList)
    at System.Security.Permissions.FileIOPermission..ctor(FileIOPermissionAccess access, String path)
    at System.IO.FileSystemInfo.get_FullName()
    at Microsoft.Web.Publishing.Tasks.CollectFilesinFolder.CollectAllFiles(String rootPath, DirectoryInfo dirinfo, Boolean recursive, List`1 result, Boolean defaultInclude, Regex[] fullPathMatchs, Regex[] relativePathMatchs, Regex[] excludeFullPathMatchs, Regex[] excludeRelativePathMatchs)
    at Microsoft.Build.BackEnd.TaskExecutionHost.Microsoft.Build.BackEnd.ITaskExecutionHost.Execute()
    at Microsoft.Build.BackEnd.TaskBuilder.<ExecuteInstantiatedTask>d__26.MoveNext()    Orchard.Web        0

You have some deeply nested folders, most probably a node_modules folder.

Solution

At this point the only solution I know of is for you to delete the node_modules folder.

It seems that even if it’s excluded from your solution, Web Deploy will still try to scoop up the files.

The node_modules folder is a special folder used by Node Package Manager, or NPM. It’s gained a lot of popularity recently as one of the main way for developers to distribute JavaScript based code and keep it up to date.

Don't worry about deleting the folder, its a special folder and it can be automatically repopulated when you reopen the project in Visual Studio 2015 or use the NPM command line.

It's not required for your site to run on the webserver. It is just used by whatever pipeline you have set up for development; like compressing CSS, minifying JS, compiling pre-processors etc.

Basically if you don't know what it is then you aren't likely to need it unless you're following a tutorial explaining how to use it. That said, if you don’t know what it is then let this be a wake up call for you. Its an intrinsic part of many modern development techniques and it means you are probably getting pretty behind in your web development skills.

Anyway, because the path is so long Windows Explorer might not be able to delete it. The best way I have found is to use 7-Zip. If you just open 7-Zip up via the start menu you can navigate around your hard drives using it as a file system. The quickest way is to copy / paste the path to the folder into the path bar at the top. When you delete it, you must press SHIFT-DELETE otherwise it will still fail, even in 7-Zip, because the path is too long to put it into the recycle bin.

Luckily, after fierce debate in the NPM GitHub issues, it seems that NPM 3 is adopting a new flat file hierarchy which should stop most of these problems. Not long now…

SOLVED: Error CS0246 The type or namespace name 'Orchard' could not be found while upgrading Orchard CMS

2015-08-21T16:51:00.000+01:00

Orchard CMS is a great asp.net MVC based CMS which is really flexible but as it stands its upgrade process is a long and complex process far from the simplicity of WordPress and other competitors. After you have followed their many-step upgrade process you will need to bring any custom themes or modules back into your main solution. At this point I started getting hit with compile time errors. With a huge solution like Orchard it can be be daunting task to figure out the solution but after some digging around I figured out what was causing this particular issue.

The issue first began when I took one of the more popular Orchard modules, Vandelay Industries. When I dropped the source in and went to “add existing project” the references for Orchard all had alerts on them.

My gut reaction was that something had changed in Orchard since the last version and new versions of the modules would be needed. However, when I checked the website it seems there hadn’t been any new releases since Orchard 1.9 had come out.

The next stop was thinking that the references were somehow pointing to the wrong folders because I had copied the projects over from the old install. Deleting and recreating them didn’t help.

Changing the target framework

At this point I widened my search and dropped the mention of Orchard CMS, instead just looking for a stripped down version of the error message. That’s when I had the breakthrough.

This obscure message was actually complaining because the old module projects were compiling against .NET Framework 4.5 but now Orchard uses .NET Framework 4.5.1. The modules were getting confused because they were trying to pull in Orchard assemblies that had been compiled against newer versions of the .NET Framework!

Fixing the issue

The solution is to set the .NET Framework to the same version as the rest of the solution.

Here is a step by step process:

Right click offending project in the Orchard solution
Click Application tab
Set Target framework drop down to .NET Framework 4.5.1 (or if you’re reading this in the future, whatever framework version of .NET Orchard is currently compiling against)

GOTCHA – Don’t use your keyboard

There is a bug which is in definitely in Microsoft Visual Studio Community 2015 but probably in other versions as well. When you are changing the target framework you must click the option with your mouse not using an arrow keys / enter combo.

There is an extra event which is triggered when you properly change the selection. When you click on the new framework version you should see a dialog box offering to update your web.config to target the newly selected version. If you don’t see something like this then you need to try again:

As soon as you press Yes the project will update and close.

Repeat this process for each of the troublesome projects and you will soon have a compiling version of your Orchard CMS based website!

Getting control of the WordPress Galleries

2015-08-12T14:25:00.000+01:00

For a few versions now WordPress has had a gallery shortcode feature that you can add in via the Add Media button. Getting it to play nice with your theme however can take a bit than is really needed. Here are a couple of tips I picked up when I had to customise it for a recent client project.

It inserts inline styles by default

By default when you put a gallery into your post it will inject some inline styles. A bit naughty yeah. I only noticed because it was forcing a 2px border around my thumbnails but after a quick search I found this post on CSS Tricks explaining how you can remove this “feature”.

The simple solution to this is to add a filter into your functions.php that will disable this output:

add_filter( 'use_default_gallery_style', '__return_false' );

Updating your theme to use HTML5 Support

This worked ok but then I lost all my styles so I started digging a little deeper because all I wanted to remove was that 2px image border.

It turned out that since WordPress 3.9 they have implemented an alternative version which uses HTML5 tags and skips the default styles by default. This seemed like a better solution and its also simple. Remove the last filter if you added it above and replace it with:

add_theme_support( 'html5', array( 'gallery', 'caption' ) );

Now we are at least using modern techniques and our 2px border has gone but so has all the other styles.

Getting some basic styles back in

After you switch over to HTML5 you need to provide all the formatting yourself. It occurred to me that this problem has likely already been solved in the built in themes that WordPress come packaged with. I opened the /wp-content/themes/twentyfifteen/styles.css file and a quick search showed me what I needed:

/**
 * 14.2 Galleries
 */

.gallery {
    margin-bottom: 1.6em;
}

.gallery-item {
    display: inline-block;
    padding: 1.79104477%;
    text-align: center;
    vertical-align: top;
    width: 100%;
}

.gallery-columns-2 .gallery-item {
    max-width: 50%;
}

.gallery-columns-3 .gallery-item {
    max-width: 33.33%;
}

.gallery-columns-4 .gallery-item {
    max-width: 25%;
}

.gallery-columns-5 .gallery-item {
    max-width: 20%;
}

.gallery-columns-6 .gallery-item {
    max-width: 16.66%;
}

.gallery-columns-7 .gallery-item {
    max-width: 14.28%;
}

.gallery-columns-8 .gallery-item {
    max-width: 12.5%;
}

.gallery-columns-9 .gallery-item {
    max-width: 11.11%;
}

.gallery-icon img {
    margin: 0 auto;
}

.gallery-caption {
    color: #707070;
    color: rgba(51, 51, 51, 0.7);
    display: block;
    font-family: "Noto Sans", sans-serif;
    font-size: 12px;
    font-size: 1.2rem;
    line-height: 1.5;
    padding: 0.5em 0;
}

.gallery-columns-6 .gallery-caption,
.gallery-columns-7 .gallery-caption,
.gallery-columns-8 .gallery-caption,
.gallery-columns-9 .gallery-caption {
    display: none;
}

This almost worked for me except for two things. First I removed the padding in that .gallery-item class:

.gallery-item {
    display: inline-block;
    /* padding: 1.79104477%; remove this line */
    text-align: center;
    vertical-align: top;
    width: 100%;
}

Also you need to set the figure tag to 0 margin if your boilerplate hasn’t already done so:

figure {
    margin: 0;
}

After that it displays nicely.

Finishing Touches – The lightbox

Instead of just clicking through to the raw image or an attachment page I wanted the images to load up in a lightbox. I’ve head JetPack includes a nice feature which does this for you but you have to sign up an account for the client and getting the verification code through their email address has just seemed like an unnecessary waste of time.

Instead I used this popular plugin, Lightbox Plus Colorbox, which uses the incredibly popular jQuery colorbox script as its underlying javascript. The only catch I found was that you need to make sure you set the gallery to point at the media file when you are inserting it. By default it will point at the post attachment page:

Conclusion

In this article we have learned a couple of simple ways to tap into the deep power of customising WordPress. By simply changing a couple of internal settings we can enhance the functionality and meet clients needs more closely. We’ve also seen that sometimes it can be simpler to grab a quick plugin to fill in gaps where WordPress or our theme let us down.

Make a mental note: Its time to update your chkdsk skills

2015-08-11T13:30:00.000+01:00

You might not need this right now but if you’re using a Windows 8 or newer computer then you should have this information tucked away for when that rainy day inevitably comes pouring down on your precious data. In this article I explain the updates to chkdsk and why you don’t need to chkdsk /f any more.

Traditionally when your hard disk has a melt down you will reach for chkdsk /f to save your bacon. The /f flag is for telling it to attempt to fix the errors and not just list the ones it finds.

Thankfully this experimental update to my skills is not being provided at the expense of my own precious data. The hard drive I’m working with at the moment is a family members from their laptop. The drive is terminally broken but the laptop is really just used for online services.

This means there isn’t any real data to be lost but we are right slap in the middle of Microsoft releasing their new free Windows 10 update. Obviously we have no install media so I’m hoping to prop this computer up for one last boot just to get the free upgrade and then install that on a new replacement drive. Depending on the licensing laws this may or may not work but its worth a shot to get a free, legal version of the system.

Ok, hopefully that will be enough back story for this article to stick in your memory when your fateful day next arrives.

There is a great article here from before Windows 8 was released which explains all this in much more geek detail:

http://blogs.msdn.com/b/b8/archive/2012/05/09/redesigning-chkdsk-and-the-new-ntfs-health-model.aspx

However, today I just want to give you a quick introduction to the new system.

Instead of chkdsk /f you should now use chkdsk /scan.

This will do an online scan of your drive (previously you would need to take the drive offline while it was scanned) and it will attempt to fix the issue there and then.

It seems that the old chkdsk /f does the scan first and then attempts to fix them all at the end. For me this resulted in about 5 hours of scanning and then it failed to cement its findings with a “Insufficient disk space to recover lost data” notice and then it aborted its mission.

But this time around with the /scan option it is attempting to fix each of the issues as it comes across them:

Which is a lot better! According to the documentation, if it comes across anything it can’t fix while online then it will add it to a fix list. You can then perform a targeted surgical strike on your drive by running chkdsk /spotfix to take the drive briefly offline, apply the listed errors and seamlessly bring the drive back up.

Instead of the drive being offline for potentially hours you are now looking at it being offline for mere seconds even with 300 million files. The graph below is taken from the MSDN blog article I linked to above which demonstrates this:

So, in conclusion: Use chkdsk /scan on your drives if its a Windows 8 or higher, NTFS based file system. If there are some things it can’t fix online then run chkdsk /spotfix to fix the remaining issues.

SOLVED: Fixing an “Unidentified Network” error with our TL-WA850RE Wireless Repeater

2015-08-07T16:00:00.000+01:00

We have a wireless repeater in our house which boosts the signal to a full five bars at even the farthest reaches of the house. It’s worked pretty well except needing to be rebooted here and there but recently it completely stopped working. If you’re connecting to your repeater but then getting “Unidentified Network” and no internet connectivity then read on to see how I solved this.

Its pretty simple, you just update the firmware.

I only wanted to write this article in the hopes that somebody else in my position will see that its safe to do so. I spent a few hours trying various combinations, rebooting, reconnecting, restarting the main Wi-Fi, the list goes on.

I was reluctant to update the firmware because my edition as printed on the back is a TL-WA850RE(UK).

The UK bit isn’t mentioned elsewhere. On the website it lists two firmware’s with nothing after the main model and then latest firmware on the site lists (EU). I figured, why bother putting it on there if it doesn’t make any difference so I wanted to solve it without potentially bricking the device.

After exhausting all other possibilities I took the plunge and installed the firmware.

The one I downloaded was from this page:

http://www.tp-link.com/en/download/TL-WA850RE_V1.html#Firmware

It was called “TL-WA850RE(EU)_V1.20_140826” released on 26/08/14.

I made a wired connection to the repeater, logged in to its admin panel at tplinkextender.net and updated the firmware. After a brief reboot it came back online successfully.

The only thing I would note is that I then kept getting “you are not authorised to view this router” errors when trying to visit tplinkextender.net in my browser.

I think that this was just a caching issue with the browser however for a quick solution to this, the new firmware adds tplinkrepeater.net (repeater instead of extender) as an alternative admin panel access point.

Logging in via this domain allowed me to quickly set up the router. I chose to give it a different name as the main network so that I could be sure I was connecting to it.

Then I moved it back to its location, waited for it to boot back up and the house finally had 5 bars again.

Starting a local instance of IIS Express with PHP support

2015-06-04T00:50:00.001+01:00

If you're a developer that works in Microsoft Visual Studio you’re probably quite used to the handy local web server that appears when you summon it with an F5. There are many alternative ways to spin up a local web server. Recently I’ve used virtual machines, Node, even PHP comes with its own built in dev server. They all have their strengths but this article shows you how to keep it in the family with a quickly deployable instance of IIS Express that supports both .Net and PHP.

After you have got used to the instant feedback of a browser refreshing immediately after you save your changes it can be difficult to go back to the old save / upload / refresh cycle of yore. Maybe you have a pre-processor cycle to build your CSS files or lint your JavaScript. Whatever your reason you can’t help but appreciate the advantages of running a web server on your dev computer.

Today I’m going to show you how quickly and easily it is to get the best of all worlds; a server that can run your asp.net sites, serve up the static html sites that you might have built a while back using Dreamweaver templates or some such and it wont even flinch when it comes across your PHP powered contact form.

The server that can do all of this is IIS Express. It will already be installed on your computer most likely. It comes with Web Matrix, one of Microsoft’s hobby development tools. It is now also the server of choice on modern versions of Microsoft Visual Studio.

The next few sections will guide you though the one-off installations and then the final section will show you the one-liner you can use in the future to start your PHP & Asp.net enabled server.

Get the Web Platform Installer

All of the required tools can be found in Microsoft Web Platform Installer. Search your start menu for “Web Platform”. If you don’t have that you can normally get it from its homepage here:

https://www.microsoft.com/web/downloads/platform.aspx

At the time of writing though, that page seems to have been broken for the last few days. In case you can’t access it still then you can get a direct download here:

http://www.iis.net/learn/install/web-platform-installer/web-platform-installer-direct-downloads

Get IIS Express

You probably have this already installed. Check by looking for this folder:

C:\Program Files (x86)\IIS Express\

If that doesn’t exist then load up Web Platform Installer and search using the box in the top right. Just type IIS Express and press enter:

For me it wasn’t listed as installed because my copy came with either Web Matrix or Visual Studio so its worth checking that folder first.

Get PHP For IIS Express

After you have IIS Express installed you need to get the PHP addon for it. Steps are as follows:

Load Web Platform Installer
Search "php express” in the search box in the top right
Click the Released column to sort the results by release date. This is useful so you can make sure you install the latest PHP version. At the time of writing the latest version for me is PHP 5.6.0 For IIS Express.
Your installer should have a button marked Add where the screenshot above says Installed. Click it.
Click Install at the bottom of the Web PI Window.
For me, it also wanted to install a second package named Microsoft Drivers 3.2 for PHP v5.6 for SQL Server in IIS Express. Let it do this and confirm the licence agreements when it asks you.

That’s it! From now on you can start up your server with one simple command.

After I show you how in the next section, please make sure to read the two gotcha’s that I experienced while setting this up so that you don’t get tripped up.

Launching a web server from your project folder

The crucial command you need to type is a simple script file called runphp.cmd. Follow these steps:

Open a cmd window using your favourite terminal software. If you aren’t already thinking of your preferred command line window then you can do it using built in tools by pressing CTRL on your keyboard and right clicking on the folder. Pressing down CTRL gives you two extra options in the context menu which are normally hidden.

One is Open command window here which would be very handy as it would put us right in the place we wanted to be but unfortunately it doesn’t open it with Administrator privileges.

Instead select Copy as path. This will copy the path which we can paste in later. It should be just near Send to (so quite far down if you have a lot of utilities installed).
Open your start menu, type Command and then when the search result comes up that says “Command Prompt” you need to right click and select Run as Administrator.

Confirm that you will allow it to make changes to your computer when asked.
Once you’re in that command prompt get to the correct folder by typing cd /d and then right click and choose paste to paste your path in. The should be a space between the /d and the path you paste. Press Enter.
Now that you’re in the correct project directory just enter runphp.cmd and press Enter.
The screen will fill with lots of scrolling text that will make you briefly feel like the ultimate hacker and you will hopefully see this at the end:
Set the window aside, you will want to leave it open for the duration of your development session.

That’s it! It was a long journey to get here but just remember that next time all you really need to do is open an command window in administrator in your project directory and runphp.cmd.

Viewing your website

In the system tray near your clock you should see a little icon with stacked silver boxes.

If you right click it and then choose View All Applications you should see a simple window like this:

Click on the localhost link and it will show your site in the browser. For me, I just added a simple index.php to the folder with a phpinfo() call to be sure my PHP was working:

Gotcha to look out for

WebMatrix will auto install an older PHP v5.3 package the first time it sees a PHP website. If you already installed a PHP for IIS Express which targeted a newer version of PHP then it will break things and all your sites will now run on PHP v5.3.

The solution is to just re-order your PATH environment variable so that the version of PHP you want to run is the first one in the list (or the only one, up to you).

Edit your PATH env variable by just typing environment into your Start Menu or into the Control Panel search box. You should be presented with an option something like “Edit the system environment variables”. Click it and then click the Environment Variables button in the bottom right of the screen that opens.

PHP for IIS Express installs to C:\Program Files (x86)\IIS Express\PHP\ by default.

For me I had v5.3 and v5.6 paths set in my PATH and ended up with v5.3 first in the path so runphp.cmd always ran the older php v5.3 each time rather than the v5.6 like I wanted.

Conclusion and lessons learned

So that’s it. You can now quickly and easily open an administrator command prompt in your project folders and runphp.cmd to spin up an IIS Express server. You can run Asp.net, PHP or plain old html sites.

There is actually a Node for IIS Express package out there as well but that’s a tale for another day...

How to reset a WordPress password on the database

2015-05-25T13:03:00.000+01:00

You’ve gone and forgotten your password? You didn’t set up your recovery email address? No, it wasn’t your fault, the client has forgotten? You inherited the site and the records aren’t right? Well whatever the reason, don’t panic. This guide will show you how to recover that account without having to do anything too drastic.

This guide isn’t for total noobs. I’m assuming that you can fend for yourself to a certain extent. If not you need to show this guide to your web developer and ask them to help you out. It also assumes that you have access to a copy of phpMyAdmin which is installed on virtually every hosting platform out there.

NOTE: If you know your way around phpMyAdmin then jump to the exciting conclusion for the actual password reset technique.

So first you need to locate your phpMyAdmin install. If you have Plesk on your server, log in to that, drill down to the problem site and then into the database settings for that account. If you have cPanel then you will be able to get to phpMyAdmin through that too. If not then you probably have access to it via your main hosting admin / billing panel. Try searching your hosting companies support section if you can’t find it.

You might have more than one WordPress installed in the account and if you installed it with a script installer then it could have a cryptic name like wp_wordpress5 or something. If you’re not sure about the database name that the site is using then check wp-config.php which is in your root folder. Normally the first line in this will be something like this:

define('DB_NAME', 'somesite_wp');

Thats your database name so make a mental note to look for that when we get into phpMyAdmin.

Ok follow the procedure you discovered earlier to get phpMyAdmin loaded up. Over the years phpMyAdmin has had some facelifts and changed slightly. You might see some slight visual differences in if your host hasn’t kept up to date but push forward and everything should work ok.

The initial screen will likely already have your database selected and you will see a list of tables down the left hand side which looks like this:

If not, click the Databases tab along the top. Then in the list of available databases just click on the database name you made a note of earlier.

The account details are all stored in the wp_users table. Click on that either in the left hand side menu or if you have the structure tab open you click also click the name or the browse button in the main page.

This will take you to a page which has the Browse tab selected at the top:

Look down the table and find the account you need to reset the password for:

If you have a lot of users you might need to use the pagination buttons at the bottom of the page:

Or if its a huge site you can search by clicking the “search” tab along the toolbar at the top:

This will take you to a page that searches the current table. You will probably want to search on the user_login line and make sure “LIKE” is selected in the Operator column (this will match partial usernames too):

When you find it you need to click the edit button for that row:

Now we’re in business.

The Password Reset Technique

You will see a row called user_pass. In it there will be a line of gibberish text. To update your password just type your new password into the Value field:

But wait! There’s more. The text was originally in gibberish because the password is encoded using an MD5 hashing technique. Open up the empty select box in the screenshot above (the Function column) and select MD5. This will apply the MD5 hash function to your text and you too will be the proud owner of a line of gibberish.

Just save it out to complete this process. At the bottom of the page, make sure Save is selected and click Go:

You will now be able to log in to your WordPress site with the username and the new password you just encoded.

ASP.NET 5 Beta 4 “Couldn't determine an appropriate version of KRE to run”

2015-05-23T13:57:00.000+01:00

I’ve started getting to grips with the new ASP.NET 5 project system and during the course of this process I’ve encountered a few glitches. This time a project I downloaded based on earlier beta3 code wouldn’t run for me in the browser. I found the solution to this issue in this article.

NOTE: This article is a vNext article. It is based on a pre-release version of ASP.NET 5, specifically beta 4. You might find some differences as you follow along. If you find any please add them in the comments and I’ll incorporate them into this article!

The project (eventually) compiled fine but when it tried to load it up in IIS Express I got this error:

Couldn't determine an appropriate version of KRE to run

I didn’t dig deep enough to figure out what underlying setting this affected but I solved it by simply doing the following:

In the Solution Explorer, right click the project in question
Select Properties (at the bottom of the list)
Tick Use specific DNX version
Tell it exactly which one you want to use, for me that was 1.0.0-beta4, .NET Core, x64
Close the properties window and it will automatically save the new preference

The project then ran in the browser for me.

ASP.NET 5 Beta 4 "The Debugger cannot continue running the process. Unable to start debugger"

2015-05-22T13:00:00.000+01:00

I’ve just started getting to grips with the new ASP.NET 5 project system and its friends. As everything is still in development there are still some kinks being worked out. Here is a quick solution if you see an error saying "The Debugger cannot continue running the process. Unable to start debugger".

Like I’ve just said this is based on pre-release software so I don’t expect this tip to have a very long shelf life but hopefully it will get you back on your way quickly if you experience this issue:

An error message displays showing "The Debugger cannot continue running the process. Unable to start debugger".
The debugger / start button in the toolbar just says “Start” instead of listing its currently selected option like “IIS Express”
The debugger button drop down which normally allows you to select other debug commands is empty.

If this you right now then just do this:

Restart Visual Studio

Yep, its just got confused and closing down the app and restarting it got it back on track.

ASP.NET 5 Beta 4 “The 'prepack' script failed with status code 99”

2015-05-21T12:05:00.001+01:00

I’ve started getting to grips with the new ASP.NET 5 project system and during the course of this process I came across a compilation error “The 'prepack' script failed with status code 99”. It was a pretty easy fix which I’ve detailed below.

The entire error I was handed back by the compilation system was as follows:

1>------ Build started: Project: ExampleProject, Configuration: Debug Any CPU ------
1>  C:\Users\exampleUser\.dnx\runtimes\dnx-clr-win-x86.1.0.0-beta4\bin\dnx.exe --appbase "C:\Users\exampleUser\Documents\GitHub\ExampleProject\src\ExampleProject" "C:\Users\exampleUser\.dnx\runtimes\dnx-clr-win-x86.1.0.0-beta4\bin\lib\Microsoft.Framework.PackageManager\Microsoft.Framework.PackageManager.dll" pack "C:\Users\exampleUser\Documents\GitHub\ExampleProject\src\ExampleProject" --configuration Debug --out "..\..\artifacts\bin\ExampleProject"
1>  grunt-cli: The grunt command line interface. (v0.1.13)
1>  The 'prepack' script failed with status code 99.
========== Build: 0 succeeded or up-to-date, 1 failed, 0 skipped ==========

The fix is pretty simple. It just means that the grunt-cli package is not installed in your project. You just need to initiate a package restore to get it back on its feet.

If you’re a command line person then you can type this in your project folder:

npm install

Most of the time though its a lot simpler to stay in the safe nest of Visual Studio and you can initiate a package restore by following these steps:

Go to the Solution Explorer window
Look in the src folder then in the name of your project then in Dependencies then find NPM
Right click on the NPM node and choose Restore Packages
Wait patiently while it sorts out its packages.

If you’re bored while waiting or just curious to see what its up to you can watch the progress by taking these steps:

Select Output window (either docked somewhere at the bottom of your screen or in View > Output or pressing CTRL-ALT-O
In the drop down at the top labelled Show output from: select Package Manager Log

When its finished it will notify you by updating the status bar at the bottom with “Installing packages complete”. You should now be one-step closer to your working project!

SOLVED: Why my Joomla contact form was refusing to send its email

2015-04-13T02:10:00.000+01:00

While working on a Joomla website recently I had the annoying problem of a contact form failing to send. It took longer than I would have like to spend looking for the solution so hopefully you will find this article before spending the time I did trying to get to the bottom of it. As usual, in hindsight its actually quite simple.

Looking in my mail servers logs I could see why the email was failing to send. For some reason the from address of the email was set to @adminemail@. Naturally the receiving mail server was rejecting it based on the fact that it couldn’t do an SPF check on this non-address.

Why oh why was I being cursed with this problem? Well I never did totally get to the bottom of it but I did figure out the broad strokes. My web server is a Windows based one running Plesk 10. Originally this version of Plesk didn't have PHP 5.3 support so they released an addon to install it.

The other part of this mystery is if you recognise @adminemail@from somewhere you have probably leafed through php.ini recently and seen this:

; For Win32 only.
; http://php.net/sendmail-from
;sendmail_from = me@example.com
sendmail_from = @adminemail@

It is the default admin email address that mail() uses. You’re suppose to set it to your admin from address. Trouble is, the server doesn’t just host one site though so I can’t set it manually in the php.ini.

Now what I assume was supposed to happen was that Plesk should insert a sane value into this so that every client doesn’t have this problem. For whatever reason, it doesn’t.

Normally this isn’t a problem, if you’ve written form processing code before you’ve probably done something like this to cut down on the likelihood that your precious cargo is going to be flagged as spam:

ini_set('sendmail_from',$fromaddress);
$mail_sent = mail($to, $subject, $msg, $headers, "-f" . $fromaddress);
ini_restore('sendmail_from');

However when you’re detailing with a CMS like WordPress or Joomla or some such you find yourself restricted by what would be simple issues but are obfuscated by the layers of the system that build this up. Plus if you go editing the core you will have to remember to re-apply these modifications every time you update.

We are agreed then that this isn't the best solution all round. Is there a better option? Well you don’t think I would have dragged you through these last 8 paragraphs if I didn't have something up my sleeve do you?

The solution is to turn back to Plesk and explicitly tell it to set the from address, like so:

Log in to Plesk
In the top right hand corner click in the search box and type the first few characters of your domain. This is such a handy feature. I’ve been using Plesk for almost a decade now and I don’t know why but it was only in the last 6 months that I realised how handy this is. So fast! Anyway select the offending domain.
Click the Websites & Domains tab
Click the name of the domain in the table at the bottom of the page
Click the PHP Settings tab
Scroll to the bottom of the page and find your Additional configuration directives textarea:
Enter this snippet in to it:
```
sendmail_from = emailaddress@example.com
```
This is a programming website, I don’t have to say replace emailaddress@example.com with your desired email address – do I?
Press OK to save this information.

Hopefully you will now test your form again and see your inbox light up with a delivered form email!