A frequently asked question in the MSDN forums is “how you can get access to the user who is interacting with your workflow?”. For example, the user modifying a task. The workflow is likely running in a different context and/or session to your browser session so there’s not an obvious tie-up.

However, in this scenario, you can get the login name of the person who modified the SharePoint task, via the Executor property of the OnTaskChanged event. Simply bind the Executor property to a string (e.g., “taskLastModifiedBy”) and whenever the task changes, SharePoint will copy the user ID to this property in the format of DOMAIN\LoginName. You can then get an SPUser object for that login name with, e.g.

SPUser user = workflowProperties.Web.AllUsers[taskLastModifiedBy];

Note: On a related subject, and the thing that prompted this post - if you’re trying to update a Person or Group field on your workflowProperties.Item, then you must pass it an SPUser object! This is bizarre, because other types of list (e.g, the task list) you can pass it a string and SharePoint will do the rest. I spent ages and all kinds of different things and always getting the “Invalid data has been used to update the list item. The field you are trying to update may be read only.” error. Annoying.

One method for persisting data in a SharePoint environment is the use of Infopath forms. Infopath stores data in your form in XML format. I have a form which has a “comments” box where people add new comments as they progressively update the form and a “consolidate comments” box which shows all previous comments*. However, Infopath forms don’t natively support appending new data to existing data - and new comments added may blow away any previous comments.

Microsoft markets that Sharepoint Designer workflows are directly exportable to Visual Studio workflows. And in many ways, this is absolutely true. It’s trivial to create a .wsp from an SPD workflow and import it to VS. However, if you then inspect the workflow that has been created for you, a little concern would not be underestimated. If your SPD workflow was what you wanted, then you could be reasonably confident that its VS version would do the same job. But presumably your reason for importing to VS is because you need more control over the workflow in the longer term, in which case, you’re likely to be left underwhelmed by what the import process creates for you. Imagine creating a webpage in Microsoft Word and you’ll get the drift.

So when faced with a similar proposition, recreating the workflow in Visual Studio from scratch immediately became apparent as the best solution. After all, the intent and purpose of the workflow was clear; recreating it in VS ought to have been a doddle. However, to complicate matters, the SPD workflow used Infopath edit forms. This turned out to be easy in SPD, but non-trivial in Visual Studio. Getting an Infopath form to register with a VS workflow project requires various items, that VS may not do automatically for you.

Quick hint about debugging SharePoint projects in Visual Studio 2010. You may find a bizarre error where deploying projects/solutions is fine, but trying to debug doesn’t work. You see an error alert with the text:

“Could not load the Web.config configuration file. Check the file for any malformed XML elements, and try again. The following error occurred: The given key was not present in the dictionary.”

Checking the Web.config shows no malformed XML at all. One thing to check into is the configuration of alternate access mappings in Central Administration. For debugging to work, the Site URL property of your SharePoint project must match the URL of the Default zone for your web application.

It seems to me that every SharePoint developer, at some point in their career, will encounter this particular error. It’s one of those infuriatingly obtuse errors that could be caused by a multitude of different issues - either individually or cumulatively. In my case, I was in the middle of developing a SharePoint timer job and I had a particular piece of code that I wasn’t too sure about, and so wanted to test very quickly, over and over again - without the rigmarole of deploying/retracting solutions, hunting through Central Admin to run the job, etc. I just wanted to run this particular section of code against the SharePoint object model and see what happened. As is a common technique for this, I started up a fresh Console Application project, pasted in my code and figured that would be enough.

I finally got around to solving a longstanding irritation I’ve had with Finder. Well, when I say that, what I mean is, I posted the irritation to a forum, and some helpful folks there gave me the inspiration for the solution.

The irritation:

Quite often I copy a bunch of images off my phone on to my laptop. In Finder I open them up and use the quickview (using spacebar) to quickly view and delete items [using the keyboard only]

Read more

When SQL Server 2000 became SQL Server 2005, a huge number of things changed. DTS became SQL Server Integration Services, and as part of the package we saw the likes of Analysis Services and Reporting Services. All these additions to the SQL Server family necesitated a change in the suite of management tools - and that meant the death of Enterprise Manager and SQL Query Analyzer, tools loved by everyone, everywhere. They were combined and rebranded as a single tool: SQL Server Management Studio. DBAs and SQL developers cried out in horror - the new tool was, well, new - it was very different and required a change in working practice. What’s more, there was no choice - if you wanted SQL Server you had to use SSMS - there’s no using Enterprise Manager (although you can use SSMS to manage SQL Server 2000.)

Consider the following. You have a set of data. Within the data are multiple subsets that share an identifier. WIthin your main set, you need to number the subsets for each item within it, e.g.,

Data

ID SubsetID Name Lots more columns…
———- ———-
1 ABC Dave
2 DEF Trev
3 DEF Bob
4 DEF Steve
5 HIJ Jim
6 HIJ Bilbo

and you want something like

ID SubsetID SubsetPosition Name Lots more columns…
———- ———-
1 ABC 1 Dave
2 DEF 1 Trev
3 DEF 2 Bob
4 DEF 3 Steve
5 HIJ 1 Jim
6 HIJ 2 Bilbo

Requirement: multiple site collections on port 80 on unique URLs in SharePoint 2007, e.g,. http://democlient1, http://someotherdemo
The scenario: you have one development machine which you use for multiple clients. Because of this, you want to host multiple site collections and allow a unique URL for each, without having to specify the port number in the URL. For example, if your machine name is SPDEV then by default your primary web application will respond on http://spdev. But you also want to have http://democlient1, http://someotherdemo, etc. There are a three ways of doing this. The two common ways are 1) assign unique IP addresses to each web application each responding on port 80, but this can be more complicated to setup and manage at the machine level, or 2) Use a non-standard port number for each web application, but this creates uglier URLs, e.g., http://democlient1:18765, http://someotherdemo:18734 etc.

Solution: It’s possible to achieve the requirement.

After our fairly epic 12 hour ascent of the Weismies, we took a rest day. Not surprising really - all four of us were in varying stages of pieces. It meant we had to abandon an ascent of the Lagginhorn - that looked like a full on day. In reality, I felt I might have made it up, but must admit I wasn’t too upset to find myself on a cable car going down the valley from the Weismies Hut. We were back in Saas Almagell by mid morning.