The opinions expressed in these materials are my own and are not necessarily those of Microsoft.
Copyright © Microsoft Corporation. All rights reserved. Unless otherwise indicated, all source code provided is licensed under the Microsoft Public License (Ms-PL).
Keyboard: CTRL + SHIFT + F Menu: Edit -> Find and Replace -> Find in FilesCommand: Edit.FindinFilesVersions: 2008,2010Published: 1/3/2010Code: vstipFind0002
You can customize your Find in Files results to show what you want to see and how you want to see it.
Example: You don't want to view the entire file path shown in the Find Results tool window.
So instead of this:
You want this:
No problem! :)
DANGER: This involves hacking the Registry so use this tip at your own risk!
Here are valid values you can use in the string:
$v drive/unc share
$x end col if on first line, else end of first line
$L span end line
$C span end col
$0 matched text
$t text of first line
$s summary of hit
$T text of spanned lines
Keyboard: CTRL + ALT + K (anywhere); CTRL + K,T (cursor must be on a member name) Menu: View -> Call Hierarchy Command: View.CallHierarchy; EditorContextMenus.CodeWindow.ViewCallHierarchyVersions: 2010Published: 1/19/2010Code: vstipTool0005
For the C# folks this is one heck of a great feature: The Call Hierarchy. It allows you to visualize all calls to and from a selected method, property, or constructor. To see how it works just right-click on any method, property, or constructor in the Editor and select View Call Hierarchy:
You should get a window similar to this one (amount of information will vary):
Notice there are "Calls To" and "Calls From" areas related to your selection. You can expand them out:
When you click on a node in the tree, the Call Sites window updates so you can visit the call if you want to:
You can continue expanding the hierarchy to see more "Calls To" and "Calls From" information:
The best part is you can right-click on a symbol and get all kinds of options:
Here are some options you may come across:
Context Menu Item
Add As New Root
Adds the selected node to the tree view pane as a new root node.
Removes the selected root node from the tree view pane. This option is available only from a root node.
You can also use the Remove Root toolbar button to remove the selected root node.
Go To Definition
Runs the Go To Definition command on the selected node. This navigates to the original definition for a method call or variable definition.
You can also press F12 to run the Go To Definition command on the selected node.
Find All References
Runs the Find All References command on the selected node. This finds all the lines of code in your project that reference a class or member.
You can also use SHIFT+F12 to run the Find All References command on the selected node.
Copies the contents of the selected node (but not its subnodes).
Collapses the selected node so that re-expanding it displays current information.
Versions: 2010Published: 1/29/2010Code: vstipDebug0004
For those not familiar with WPF Trees, check out this article: http://msdn.microsoft.com/en-us/library/ms753391(VS.100).aspx
Now. For those that ARE familiar with them have we got something for you: WPF Tree Visualizer. Yes, it started out as a CodePlex project and ended up in the product itself as a visualizer. But how do you use it?
Just an update on the book. It appears that it may be over 300 pages as originally planned so I am seeing if that is an issue with the publisher. Would you have a problem if the book was larger?
Also, I've gotten another draft of the book as we move along and, just as I have always done, I would like to get your thoughts. I've attached the PDF for your review. Let me know your thoughts. Bear in mind is it a rough draft and some things will change. For example, "Free Your Document Windows" will become something like "Multi-Monitor Support" as I review the titles.
Also, I will most likely be trimming out the more trivial tips. The tricky part is that one person's "trivial" is another person's "cool" so I'm erring on the side of too much information rather than not enough.
This post is a bit overdue but better late than never. So let’s talk about Visual Studio Update 1…
Soma originally introduced Update 1 in his November 26th post here:
Since some of you may want to install this update immediately there are a few options for you.
From inside the IDE, go to Tools | Extensions and Updates:
Then look under the updates section and install from there:
If you want a more direct route you can click this link: http://go.microsoft.com/?linkid=9821199
You will get this dialog and can just click Run to execute the small .EXE and begin the online install:
Although many believe you can only install via online resources this is actually not true. A quick look at vsupdate_KB2707250.exe /? shows us the /layout switch:
So when I run vsupdate_KB2707250.exe /layout c:\ziptemp\vsup1 on my system this is the dialog I get:
This will download the source files to the designated location so you can make them available on a network share, thumb drive, or other media for stand-alone installation.
The most obvious question is, “Why should I care?” Despite the bug fixes and new features, this particular update represents a major shift in the timing for delivery. We have now committed to a shorter update cycle so we can deliver critical fixes and incorporate new features more quickly. What this means for you in real terms is that blocking issues may be resolved much more quickly and/or you will get a new feature that will help you get work done better/faster/stronger.
So what exactly did we deliver? Fundamentally there were two things we provided: Bug Fixes and New Features. You can find all the details here:
I’m going to repeat the bug fixes here and then address the new features in another article right after this one.
After an extended period of time looking for the list of bug fixes included in the package I was initially unable to find a comprehensive list anywhere. It looks like the team recognized there was a gap and updated the KB article on 1/13 to include all the relevant information. I’ll just repeat the information here verbatim for convenience.
Quick Housekeeping Note: I’ve deprecated the old Visual Studio Tips extension and have tried to get it removed from the gallery so it isn’t available anymore but have had little success. I am currently not planning to replace the tool with a new version but am open to suggestions as to a viable alternative. For now I suggest you use your favorite RSS feeder or visit the blog directly. Now on with the post…
With the BUILD conference (http://www.buildwindows.com/) just around the corner and lots of announcements coming I thought now would be a good time to pause the usual posts and talk about the next version of Visual Studio (vNext) and how you can prepare for some of the goodness to come. This is a high-level look at what is coming and I will be blogging in much more detail on the features throughout the vNext cycle. With that said, I’ve intentionally left out any images of the IDE for this post in favor of brief feature descriptions. Additionally, I will be blogging and tweeting directly from the BUILD conference while I am there to keep you up to date on the latest revelations that are coming out. To get a sneak preview of some of the features make sure to see the following video:
Visual Studio 2010 introduced a great new extensibility model that made custom additions very easy to make. You can find these extensions at the Visual Studio Gallery (http://visualstudiogallery.msdn.microsoft.com/) and, if you haven’t explored some of these great additions, I suggest you browse through the gallery to see the items there. When asked which extension I suggest everyone download the number one answer I give is always been the Productivity Power Tools (http://visualstudiogallery.msdn.microsoft.com/d0d33361-18e2-46c0-8ff2-4adea1e34fef?SRC=Home). Download it now. Seriously. Now.
We have (and will going forward) use extensions to field test new features that will be included in future versions of Visual Studio and it is a safe bet that some of the Power Tools will find their way into the vNext. If you watched the video mentioned above you will hear Weston Hutchins mention this as well. What better way to see what people want than to let them vote by actually downloading and using a feature? Granted it’s a little harder to focus on individual features when a lot are packed together but, even then, we get plenty of feedback on what people like and don’t like.
Once we get beyond VS2010 then we can start looking at the new features that have been publicly revealed so far. Here is a list of things you can expect in the next version:
One of the bigger pains in prior versions of Visual Studio is moving a project between versions. It was an all-or-nothing proposition typically due to schema changes to the proj files. If you opened a VS2008 project in VS2010 and converted it to the new project format then that project could no longer be opened in the prior version. For those that have experienced this you know this wasn’t entirely true and you could do a little proj file hacking to go back but it wasn’t fun or intuitive to do so. Visual Studio vNext will solve that issue. You will now have the ability to open a project in the new version of Visual Studio and still go back to prior versions. The net effect is you can have different teams on different versions of Visual Studio but still able to share projects with each other!
With the dizzying array of menu items, toolbar buttons, options, files, etc… it is often difficult to remember where a particular item exists. VS vNext adds the ability to quickly search the IDE for these things. A good analogy would be to compare it to the ability to search programs and files when you click on the start button in Windows. I don’t know about you but I don’t look for files anymore for the most part, I just hit my Windows key then type in what I am looking for then select what I want. Imagine that level of functionality inside the Visual Studio IDE and you have an idea of what this new feature does.
vNext supports creating new instances of Solution Explorer so you can have focused areas (one project in one instance and another project in a second instance) for working on multiple monitors. Additionally these instances can be nested inside windows that have been pulled outside the IDE for use on other monitors.
In the Productivity Power Tools we introduced the Solution Navigator with some cool new features that have found their way into the new IDE. The ability to drill down into the classes then members from Solution Explorer now exists. You can right-click, say, a method in Solution Explorer and see some new terminology: Calls, Is Called By, and Is Used By. Calls represents the call hierarchy for the method; Is Called By represents a list of methods that call the current method; and Is Used By represents all the specific file, line, and column positions where the method is mentioned.
Much like the new search for the IDE, we now have a new search for just the items in Solution Explorer. You can quickly type in a file, class, reference, etc… and see a filtered list in Solution Explorer of the items you are looking for.
A long-awaited feature that will collapse everything in Solution Explorer so you can expand items “fresh” as you navigate though the various hierarchies.
Another feature introduced in the Productivity Power Tools, this feature will let you keep interesting tabs surfaced on the left of the tab well while non-pinned tabs get pushed out of the visible area based on usage.
You can now rip off documents from the IDE and have multiple tabs grouped together in one or more floating tab wells. This is great for multi-monitor situations where you want groups of related files on separate monitors.
There is now a dedicated search box for the Toolbox window so you can quickly filter items to find the controls you want.
The most notable change to the Add Reference dialog is speed. It is way, waaaaay faster to add a reference with the new dialog than ever before. Also, a search box has been added to allow quick filtering of assemblies.
The Quick Find / Replace has been cleaned up and streamlined to make using the experience easier overall.
As with many other tool windows, the Error List window now sports a search box you can use to filter the list of errors, warnings, and information messages.
The Error List window also now supports a scope filter that lets you show only those items in Open Documents, the Current Project, or the Current Document.
Probably one of the coolest new changes is the Preview Tab. Essentially it allows you to look at a preview of a file and decide if you want that file to be opened up in the tab well. I’ll show much more on this feature later but wanted to mention it here to round out the list of items I will be covering in the short-term.
There are a lot of very cool changes coming in Visual Studio vNext. This post is meant to make you aware of the coming changes at a high-level to prepare you for a deeper discussion of each of these features in future posts.
Continuing on our odyssey of exploring the features in vNext, I thought we would look at some of the performance enhancements that have been publicly announced. Naturally, at this time, I can’t go very deep into the inner-workings of the improvements but will endeavor to list them out and explain somewhat based on information currently available. Also, be aware that this is a very early look at these features so there will be many improvements made as the product gets closer to RTM. I’m actually running two versions of vNext currently; the one that will be made available to the public as a pre-release at some point and the one we build internally so I have a pretty good handle on what you will see and what is coming out. I can say that there are some great things that are being done with the product you will definitely enjoy. With that said let’s take a look at some of the things we have revealed so far…
First off, if you haven’t seen it yet, most everything I mention comes from the great TechEd session done by Weston Hutchins that can be found here:
For vNext we are focused on four areas of investment to improve the product:
As I review the areas of improvement expect a great deal of overlap as many of the improvements fall into multiple categories. With that said, this area is pretty self-explanatory: We want vNext to enable users to quickly go about their work without waiting on the IDE. These improvements manifest themselves in a variety of ways but all have the singular goal of making the product more responsive. Below are some of the areas that address this goal that we have made public so far.
One of the IDE enhancements revealed so far is Add Reference dialog box. First and foremost it has been supercharged to load almost instantaneously. How? We index the assemblies beforehand to have a ready-made list on demand. We haven’t gone into detail yet on what exactly is being done but expect that revelation in due course. Also, as with most of the IDE, we have added the ability to search/filter the list so you don’t spend a billion years looking for the assembly you want.
We have moved some activities out-of-process to give control back to the IDE quickly. See Memory Utilization.
vNext will take advantage of multi-core systems much better than VS2010 which will result in more overall responsiveness. See Long Running Operations.
Building and debugging are examples of a long-running operations that we are addressing. For example, Visual Studio 2010, while a great improvement over previous versions, really doesn’t take full advantage of multi-processor machines. Our goal is to speed things up so you can get your work done faster with less time spent waiting.
Build operations have now been moved to background threads so you can actually do other things in Visual Studio while a build is happening.
Another great new feature is the ability to build projects in parallel. vNext comes with an option to have parallel project builds based on the number of cores your system has. In fact, the default value is automatically set to the max cores on your system. You can reduce this value if you want.
Responsiveness also includes the overall efficiency of Visual Studio. We are working on reducing the memory utilization overall based on usage.
vNext uses much less virtual memory internally. Much of this is due to running builds out-of-process (see below) as well as other enhancements we have made that haven’t been released yet.
C++ has had this feature for a while and we are now brining it to C# and, in a future release sometime after vNext, VB as well. We now perform builds outside the Visual Studio process. Instead of loading numerous referenced assemblies inside the VS process they are now loaded into an external process then discarded when we are done. This results in significantly less memory usage inside Visual Studio which results in a more responsive, stable product.
This one is really simple. How do we get the best information possible when there are issues with the product so we can continuously improve it for you? Solving for this question is never easy but we have devised ways of obtaining performance data from your actual use of the product to help strengthen our knowledge of areas that need to be addressed.
The number one tool you can download TODAY to give us feedback on VS2010 and vNext is PerfWatson. You can get it from the Visual Studio Gallery here:
This tool allows us to get up-to-date, immediate information on the issues that are impacting performance. This tool is currently built into the pre-release version of vNext and, I suspect, we will include it in the RTM version as well with an option for you to turn it off if you don’t want it.
If you haven’t gone to http://connect.microsoft.com/VisualStudio/ yet then you need to. This is one of the main ways we have been getting directed feedback from you for some time now. It’s a great way to get to see what others are saying and to communicate directly with the VS team.
NOTE: This is an older feature I’ve updated the information for VS2013
Ever been writing some code and you want to leave a reminder to yourself to do something on a particular line or area? Did you know about the "to do" comment feature or custom tokens? They seriously rock if you have never used them and, because they go in source code, everyone can have access to the information when you check-in the code!
So here's how TODO comments work:
In VB you just put any comment in that begins with the letters "todo" (case doesn't matter):
In C# and C++, it's pretty much the same thing (again, case doesn't matter):
In C++ you have to explicitly turn this feature on. Go to Tools | Options | Text Editor | C/C++ | View and change Enumerate Comment Tasks to True:
Regardless which language you use the result is an entry in your Task List:
Note: You can Double Click any entry to automatically have the editor go to the respective line in your source.
If you don’t see the entries, make sure you have set the task list to see comments:
We actually have several tokens you can use in addition to TODO. To see all the existing tokens go to Tools | Options | Environment | Task List:
You are welcome to use HACK and UNDONE as well. The MSDN documentation is horrifically bad when it comes to describing these tokens and I don’t agree with the description of use necessarily. Here is what it says:
“Items labeled TODO, HACK, and UNDONE in the Task List window indicate code problems that do not keep the project from compiling, but that do cause run-time errors. You should correct these errors before running the project.”
Also, be aware that the number of entries that appear in the Task List changes depending on the type of project you are working on. With VB and C# projects, the Task List displays all of the comments in the project whether the file is open or not. With C++ projects, the Task List displays only the comments that are found in the files currently opened for edit.
Feel free to create your own tokens for your use. Creating your own tokens is very simple, just pick a name for your token and type it in the Name textbox:
Note: UnresolvedMergeConflict looks like an error but isn’t it is an actual token that will make a high priority item in the Task List.
Next choose a priority level:
Then click the Add button to make it an active token:
You will see it in your list:
Now you can use the token in your comments:
Be aware that any tokens you create ARE NOT SHARED with other developers so you may want to come up with a standard set of tokens for everyone to use if you intend to leverage this feature company-wide then export them (http://blogs.msdn.com/b/zainnab/archive/2010/07/14/exporting-your-environment-settings-vstipenv0021.aspx) and have folks import them (http://blogs.msdn.com/b/zainnab/archive/2010/07/15/importing-or-changing-your-environment-settings-vstipenv0022.aspx).
Tokens are a pretty nice feature to keep track of places in your code you need to revisit. I don’t suggest them over, say, tasks in Team Foundation Server but they are a great short-term reminder for things that need to get attention. If you decide you would like to create Task List items programmatically you can do that as well. Here is a link to some guidance to get you started:
We just had a meeting the other day and are getting closer to the final draft for the book. The piece I am most concerned about is how best to convey keyboard shortcut information so it is easy for you to use. Here is the table structure we were originally thinking:
Notice the actions are on the left and keyboard settings are along the top. During the discussion I thought it would be more logical to put the keyboard settings along the left and the actions across the top. My thinking is that when you use the table you will start with your setting and then want to move right along actions you can take. What do YOU think? Do you like it the way it is or with the settings and actions switched?
Keyboard: CTRL + ALT + I Menu: Debug -> Windows -> Immediate Command: Debug.Immediate Versions: 2008,2010 Published: 9/28/2010 Code: vstipTool0095
When using with the Immediate Window, you can work with class and object members directly:
You can use any method or property as long as it is in context. So, for example, when you are in debug mode, you can call any method that is in scope:
A less-known feature is you can work with properties and methods while in design mode. If you have static methods on a class, for example, you can just execute them right away:
For object members, obviously, you need to create an instance of the object before working with the members:
WARNING: When working with members at design-time a build will occur. This could have unintended consequences so make sure you have experimented with this feature a bit before you use it.