Microsoft Visual Studio cannot shut down...



  • The only modal dialog that is open is this one telling me Visual Studio cannot shut down because there's a modal dialog somewhere! And it pops up every time I try to exit Visual Studio.


  • kills Dumbledore

    Something something Linux hardware.

    Are you sure there isn't a modal off screen somewhere?



  • Strange. If there was a real modal dialog, you shouldn't be able to even click on the close.


  • Grade A Premium Asshole

    I have a laptop that absolutely refuses to uninstall VS2013. It will start the uninstaller, run for a long time, say it needs to reboot, and nothing has been removed.



  • @cartman82 said:

    If there was a real modal dialog, you shouldn't be able to even click on the close.

    Exactly, I was able to do stuff with Visual Studio while a modal dialog would block everything off until it's dealt with.

    CTRL + ALT + DEL time.



  • @Jaloopa said:

    Something something Linux hardware.

    The Linux Hardware Errors must be leaking out of the unplugged Ubuntu hard drive I have sitting on the table next to my tower.


  • Discourse touched me in a no-no place

    @Polygeekery said:

    I have a laptop that absolutely refuses to uninstall VS2013. It will start the uninstaller, run for a long time, say it needs to reboot, and nothing has been removed.

    The words "I'm sorry, I can't do that, Dave" don't appear anywhere, do they?



  • @Polygeekery said:

    I have a laptop that absolutely refuses to uninstall VS2013. It will start the uninstaller, run for a long time, say it needs to reboot, and nothing has been removed.

    (You sure?) Of course, uninstalling "Microsoft Visual Studio Professional 2013 with Update 4" leaves the machine with about 2 dozen other packages that have to be manually uninstalled. And some of them complain because they depend on other packages, so you have to guess at the proper uninstall sequence.

    Last time, I just nuked my VM. It was easier.


  • Grade A Premium Asshole

    @dcon said:

    (You sure?)

    Yep. It acts like it is uninstalling for a long time. Then it reboots and nothing has been changed. VS still works fine. No real errors or otherwise.

    I CBA to just delete everything and clean out the registry, etc. I will probably just re-install Win7.



  • @mott555 said:

    The Linux Hardware Errors must be leaking out of the unplugged Ubuntu hard drive I have sitting on the table next to my tower.

    Fuckin open source hardware can't even get basic shit like "not doing anything when unplugged" right



  • The Linux is coming from inside the house!

    Trigger/Time warning: TVTropes.



  • I was running the uninstaller (or the updater?) for VS once and it got stuck in a loop or something at some point. An indefinite progress bar just kept cycling for like an hour. No sign of what to do.

    Hit cancel: The dialog switched to "Cancelling..." and stayed in an infinite loop.

    Attempted to kill it in the task manager: The dialog remained and the task would reappear every time I killed it.

    Rebooted: When the machine came back up, the VS dialog reappeared and resumed its infinite loop.

    Cried: Successfully relieved some of the stress.

    I forget what I had to do to resolve the issue. They went to a lot of trouble to make sure that process would complete it's job no matter what. But all I wanted it to do was stop.



  • @Onyx: how do phone systems (and PBXen) in your part of the world behave when you self-dial?

    My work phone system dumps you into your own voicemail inbox (consistent, but not particularly useful), while my cellular provider (Sprint) has it set up so that you actually can access your voicemail that way -- which is a rather nifty way of doing it, IMO...


  • BINNED

    My mobile phone gives me a busy signal. I don't have the voicemail set up there though.

    Analogue stuff... not sure, should check, I expect busy.

    Now, for the local PBX systems... I only dealt with multi-channel SIP lines, so it will let you dial back in, not sure if I ever did it while using the same caller ID as the number I dialled. The voicemail... well, in our standard setup, you'll actually get there, yes. But not because it's a special check or anything, it will simply see your line is busy and dump you to voicemail because of that. That is if no queues or callgroups are set up, of course, I'm talking 1:1 extension to caller ID mapping here.

    Other modern SIP stuff will work the same no matter where you are I guess, it only depends if your telco lets you dial back in. For older stuff like Phillips or Siemens, I skipped those, went straight to SIP. My boss would know, though.



  • @Onyx said:

    The voicemail... well, in our standard setup, you'll actually get there, yes. But not because it's a special check or anything, it will simply see your line is busy and dump you to voicemail because of that. That is if no queues or callgroups are set up, of course, I'm talking 1:1 extension to caller ID mapping here.

    Yeah -- my work phones are an IP-phone setup (likely SIP, the current hardware's Avaya though), and act as you describe re: self-dialing and voicemail -- as I said, it's a consistent behavior (if busy, then send to voicemail), just not particularly useful (how often do you want to leave yourself a voicemail? :P ).

    What my cellphone provider does, though, is actually connect you to the "listen to your voicemail inbox" function -- far more useful than simply being asked to leave yourself a message!


  • BINNED

    @tarunik said:

    What my cellphone provider does, though, is actually connect you to the "listen to your voicemail inbox" function -- far more useful than simply being asked to leave yourself a message!

    That's cool, but nothing I'd even attempt doing that on a local PBX.

    1. You might have a limited number of voice channels available. Dialling yourself takes up two already and clogs up the works. I don't want to encourage such behaviour.
    2. You might have more extensions than registered numbers. That means you can do neither DID nor DOD, you have to remap multiple extensions to same caller ID. Meaning, when you dial "yourself" it might actually be mapped to multiple phones (and ring all, or first available, or whatever). You also don't have an unique caller ID. So there's no reliable way of detecting if it was really you dialling yourself. It would be doable either in systems that do DID, or by using custom SIP headers, but you have to trust your provider to leave them intact. To specific and / or fiddly of a setup to seriously consider.

    What we do is have a generic voicemail code and it routes you depending on your local extension (with optional password, of course). Also, accessing other voicemails is possible by using <code><extension_number> (with obligatory password prompt in that case). I know, nothing revolutionary, just mentioning it for completeness sake.



  • @Onyx said:

    It would be doable either in systems that do DID, or by using custom SIP headers, but you have to trust your provider to leave them intact. To specific and / or fiddly of a setup to seriously consider.

    I'd only implement it for a POTS-equivalent common carrier provider, or a PBX with full DID -- without DID, such a trick is pointless IMO anyway :)


  • BINNED

    It still uses up the available channels, even with DID. Maybe it's just the fact that I only worked with small systems so far but that has always been a concern on stuff I worked with.

    Most of the phones today can be set up to call whatever you want when you hit the voicemail button anyway.



  • @cartman82 said:

    Strange. If there was a real modal dialog, you shouldn't be able to even click on the close.

    Seriously, WTF?

    Why would this nothing no-effort fly-by post get 14 likes!?

    Low-effort crowd feedback is strange and unfair.



  • @cartman82 said:

    Low-effort crowd feedback is strange and unfair.

    You get what you pay for.


  • Discourse touched me in a no-no place

    @cartman82 said:

    Why would this nothing no-effort fly-by post get 14 likes!?

    15 now. I'll help by not adding one, since it seems to bother you.


  • FoxDev

    @FrostCat said:

    15 now. I'll help by not adding one, since it seems to bother you.

    16 now for some reason. -looks innocent-


  • ♿ (Parody)

    @cartman82 said:

    Why would this nothing no-effort fly-by post get 14 likes!?

    Presumably surrogate action for people who would have said something similar. Or quoted you with QFT or This or some other equivalent thing.


  • :belt_onion:

    That's exactly what you're supposed to use likes for! Come on guys, we're better than this!

    Filed Under: Civilized discourse ?= wtdwtf



  • I do hope they still ask you to enter your code when doing this. I remember that the cellphone providers here in the Netherlands didn't and that made it trivial to listen to every voicemailbox by just spoofing the caller id.



  • @martijntje said:

    I do hope they still ask you to enter your code when doing this.

    They do



  • I just ran over a nice problem: I use the provided Team Foundation Server for my pet project to keep a backup somewhere else, for versioning and for relatively easy synchronization between my laptop and my desktop.

    Recently I forgot to get the latest version on my laptop, resulting in some (minor) conflicts upon checking in. Resolved them, downloaded the latest version on both desktop and laptop - only to discover that Visual Studio now can't load one of the projects in the solution.

    C:\Users\xyz\Documents\Visual Studio 2013\Projects\abc\abcUniversal\abcUniversal.Shared\abcUniversal.Shared.shproj : error  : The composition produced a single composition error. The root cause is provided below. Review the CompositionException.Errors property for more detailed information.
    
    1) No exports were found that match the constraint: 
    	ContractName	Microsoft.VisualStudio.ProjectSystem.References.IBuildDependencyProjectReferencesService
    	RequiredTypeIdentity	Microsoft.VisualStudio.ProjectSystem.References.IBuildDependencyProjectReferencesService
    
    Resulting in: Cannot set import 'Microsoft.VisualStudio.Azure.Deployment.ProjectReferenceMaintenanceService.ProjectReferencesService (ContractName="Microsoft.VisualStudio.ProjectSystem.References.IBuildDependencyProjectReferencesService")' on part 'Microsoft.VisualStudio.Azure.Deployment.ProjectReferenceMaintenanceService'.
    Element: Microsoft.VisualStudio.Azure.Deployment.ProjectReferenceMaintenanceService.ProjectReferencesService (ContractName="Microsoft.VisualStudio.ProjectSystem.References.IBuildDependencyProjectReferencesService") -->  Microsoft.VisualStudio.Azure.Deployment.ProjectReferenceMaintenanceService
    
    Resulting in: Cannot get export 'Microsoft.VisualStudio.Azure.Deployment.ProjectReferenceMaintenanceService.WireUp (ContractName="Microsoft.VisualStudio.ProjectSystem.ConfiguredProject.AutoLoad")' from part 'Microsoft.VisualStudio.Azure.Deployment.ProjectReferenceMaintenanceService'.
    Element: Microsoft.VisualStudio.Azure.Deployment.ProjectReferenceMaintenanceService.WireUp (ContractName="Microsoft.VisualStudio.ProjectSystem.ConfiguredProject.AutoLoad") -->  Microsoft.VisualStudio.Azure.Deployment.ProjectReferenceMaintenanceService
    

    The fun part: When I try to revert to an earlier checkin, Visual Studio simply crashes.

    Then again, it might also be connected to an upgrade to the Azure SDK I did. I'm a bit stumped, however, how I'm supposed to fix this.



  • @cartman82 said:

    Seriously, WTF?

    Why would this nothing no-effort fly-by post get 14 likes!?

    Low-effort crowd feedback is strange and unfair.

    Imagine how guilty I felt about this one...


  • FoxDev

    why didn't i like that before?


  • FoxDev

    Because you're a daft fox ;)


  • FoxDev

    that seems plausible. :-P


  • ♿ (Parody)

    @Keith said:

    Imagine how guilty I felt about this one...

    I imagine not at all. Especially since you're now trolling for more likes.



  • Since @Rhywden is busy turning this into the VS WTFs thread, I'll toss one of my own into the ring:

    So, it's release-day today, and I have to do a bit of editing to bump the version number for the release, both in a header and in the app's resource file. One of the fixes that was to go into this release touched an unrelated part of said resource file in order to add a GUI element; it had already been merged in as it had passed its tests, though.

    So, I edit, save, commit, and go off to build the release. At the last moment though, I decide "wait, I'm going to do a quick check to make sure something isn't catastrophically broken." So I do that, and the form with the GUI element I added won't load, throwing an "Attempted an invalid operation." error. Googling is fruitless, so I go off to debug it; however, my first debugging attempt causes Visual Studio to crash with some sort of unhandled .NET exception that paralyzes anything that comes in contact with it, even the postmorterm debugger. I manage to get past that by reloading VS, and it turns out the problem is because the dialog data exchange routines are trying to operate on an uninitialized control. :wtf:

    So, I look around, trying to figure out "why, oh why, could this control not be getting initialized?" It turns out that even though the ID for the control is in the header, the control somehow was deleted from the resource file! After trying to blame SVN for the problem and failing miserably (as it was in the branch I had merged the fix from, and SVN steadfastly refused to re-merge the change), I look at a diff of what happened to the resource file with the version bump.

    :headdesk:

    VS' resource editor hadn't realized that the resource file had changed when the branch was merged, and as a result had rudely stomped on the merged-in changes when I saved the edits to the resource file as part of bumping the version. So, I open the resource file in Notepad and copy-pasta the clobbered change in from the diff, then save and commit, which puts me back on track, although I have to blow away and remake the release tag because I had tagged it before I noticed that it was broken.

    Sadly, I don't think I'll ever know why VS blew its top the way it did during that first debugging attempt...



  • That's why I use a header containing a set of #defines for application version, and reference those in the RC as well as anywhere that needs the version.

    So I don't (directly) touch any other files at all when updating the version number, it's very clear that there is simply one place.


  • Discourse touched me in a no-no place

    @lightsoff said:

    So I don't (directly) touch any other files at all when updating the version number, it's very clear that there is simply one place.

    Being able to inject the version number into the code during the build process is a great feature. Reduces the fuck-up likelihood when releasing quite a lot.



  • @Keith said:

    Imagine how guilty I felt about this one...

    39 LIKES!? WTF is wrong with people?!


  • BINNED

    We like to spread the love.


  • :belt_onion:

    Yeah, it's 40 now...

    I did not like, for the record...


  • 🚽 Regular

    Between yesterday and today a deleted post of mine got a like (my history is public):



  • @Zecc said:

    Between yesterday and today a deleted post of mine got a like (my history is public)

    Even deletion can't stop your admirers.


  • BINNED

    Maybe someone wanted you to reconsider but didn't want to flag? I think I did that a few times. Probably got deleted anyway. Meh. I tried.


  • FoxDev

    @cartman82 said:

    @Zecc said:
    Between yesterday and today a deleted post of mine got a like (my history is public)

    Even deletion can't stop your admirers.

    Doesn't stop whoosh :badger:s either 😒


  • Discourse touched me in a no-no place

    I think that was me because I read the edit history and thought it was funny.


  • Discourse touched me in a no-no place


Log in to reply