Right now you can setup continuous deploy for a Team Project to a hosted service if the Version Control is "Team Foundation Version Control" but If the project is using Git, then you cannot set it up. The reason we want this is because we would like to have all are Team Projects in one type of version control, but we are having hard time getting are developers to be able to work work with TFS on Mac's (git tf works fine, but getting tf to work on mac's is more work then it is worth, having to down grade Java). We do a lot of Azure hosted service work, but a lot of the clients for the service that we make are Mobile Apps. We would like to be able to develop both Azure hosted services and iOS app's using the same Version Control. If publishing to Azure hosted services worked we could then us all git everywhere!

I need the solution for the below problem,I have the solution file which contain aspx page like home, login, main.. After login the home page will be redirected.if suppose am I put the break point in the login page code behind, after complete login action the home page loaded. I have some logic in the home page load event. I would like the load event also automatically execute step by step. because my actuall solution file contain more the 30 pages so I can't able to find which page code behind run at this time. Overall my need is in solution file all the code behinds are executed step by step automatically even am I not put any break point.

In for instance a C# project, if we show the properties of a referenced assembly, the properties list has the element "Version" which currently displays the AssemblyVersion. This element should be renamed "Assembly Version" for clarity. Also, a new element "Assembly File Version" should be added.

The motivation for this is that some third party components keep the assembly version constant (i.e. contract version) between minor releases, to reduce need for recompilation of dependent code. Such minor releases will still increase the assembly file version, so that it's possible to distinguish them. Only showing the assembly version in the VS GUI is confusing to the users (as they think they are somehow not getting the correct version) and might lead to bug reports that target the wrong version).

I order to write software, very dynamic in nature, I find it useful to generate meta-data and meta-knowledge about lower level APIs. I see value and promise in the ability to delay code generation as late as possible based on the premise that one is able to sense and learn (gain knowledge) from an environment. If one looks at low-level API no matter the domain; usage patterns present themselves in fully expressed canonical order and one can stitch them together via composition [DSL]; we have the ability to write LINQ Providers, we just need the ability to dynamically generate code based on parametric properties of what is possible according to the API. Right now I can really express this concept because I’m not able to fully compose dynamically generated blocks to code. I’ve looked at the Roslyn Project but am not sure it’s the right fit for what I need. I need a rich meta-programming framework that allows one to generate blocks of code.

When you already logged-in in TFS using one user password, after some time if you have to work with same server and same project but using different user credential(because some task can only be allowed to be done by some team member in the team, for that special previleged one-two special accounts were created among the team and team lead have given that special user credential to selected team member), then there is no option to log out from TFS and log in using different user credential.

To achieve this thing, you force fully have to close the visual studio and start it again, at starting time visual studio will ask credential and you have to enter the other credential with which you want to work. This is bit frastrating during work.

There's something really annoying in Visual Studio : where I expand or collapse a method or code region, this action is pushed on the undo stack. So if I edit some code in a method, then collapse that method, and then want to undo my change, I have to undo twice : once for the collapse action, and once for the change in code. This can be VERY confusing if you expanded or collapsed several things after editing the code, the editor keeps jumping to different places and you don't know whether your changes have been undone or not... Please, don't record outlining operations, or at least make it optional!