-
Notifications
You must be signed in to change notification settings - Fork 654
Problem with 1.2.0 release of GitVersion #246
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Comments
Actually, this is the same issue as I was seeing here in #226 which I thought was down to a modified version of GitVersion that I was using. Seeing this even with the official release. I will close the other issue, sorry for duplicate. |
When I originally raised this issue, @SimonCropp called out to @nulltoken for some help. Can either of you help with this? Is there something wrong with my repo? Is there any additional information I can provide to help solve this issue? For now, I am going to fall back to using the older version of GitVersion, but I would like to move onto the new released version as soon as possible as I would like to take advantage of some of the newer features. Thanks! |
Looks like someone else, unrelated to GitVersion is also seeing this issue: |
To confirm, I have just rolled back to version 1.0.0 of GitVersion, and it generates the following:
No other changes in the Build Configuration, or Repository, were made. Any thoughts? |
Sorry for that. It looks like I completely overlooked this call for help. 😿
That may hint at a regression at the libgit2 level. I'm afraid that without a repro case that will be very difficult to troubleshoot. Would the repo you're working against be public or have you experienced similar problems with a public one which url you could share with us? |
Not a problem at all, thanks for getting in touch! Unfortunately no, this repo is not a public one, so not in a position to share it :-( I have not seen this happen in any other public repo, but my coverage of using GitVersion is limited to my own internal repository. What I can say is that this is relatively new repository, which was created in Stash. I have just tested with Version 1.1.0 and 1.1.1 of GitVersion, and both of them result in the same error message, only version 1.0.0 seems to work correctly. I don't know if there was an upgrade of libgit2 between 1.0.0 and 1.1.0. Any thoughts on how to proceed? Thanks! |
Looks like LibGit2Sharp jumped from version 0.17.0.0 to 0.18.1.0 between Version 1.0.0 and 1.1.0 for GitVersion. https://github.com/Particular/GitVersion/blob/1.0.0/GitVersionExe/packages.config |
And it looks like GitVersion relies on new functionality within the latest version of LibGit2Sharp, so it is not even as though I can re-compile with a reference to the earlier version. |
Adding a comment to the SO question might help to get a potential public repository to work with.
Which reduces the set to only 328 commits... Yeah! (@carlosmn any idea?) |
Ha ha, somewhere to start anyway? I have just tried to re-compile GitVersion with both 0.17.0.0 and 0.18.0.0, to try to narrow down further, but GitVersion relies on functionality in 0.18.1.0, so without modifying GitVersion, I can't reproduce. Is there some small sample application that I could create to try to reproduce the error, taking GitVersion out of the equation? That way I could test on different version of LibGit2Sharp. Should this issue really be moved to LibGit2Sharp? I will go an post on StackOverflow just now. |
I think it should be mainly handled in the LibGit2Sharp repo with libgit2/libgit2sharp#794. Once it's fixed there, we should eventually be able to close this one. |
I suspect the workaround for Windows remote filesystem oddities, which introduced memory-mapped writing into the packfiles. I've commented with more detail in the libgit2sharp issue. @gep13 from the stacktrace, any clone or fetch under the same conditions should be enough to trigger this issue. |
@nulltoken should we roll back libgit in the short term? |
@SimonCropp From a GitVersion standpoint, I don't know the impact of moving GitVersion back to LibGit2Sharp v0.17. With libgit2/libgit2sharp#794, we've been able to identify a repro case. It's possible one of the libgit2 team member may work on try and fixing it in the following days. Proposal:
|
Sounds good
|
Looks like this is going to be covered in the new release of LibGit2Sharp: This will require pulling the new version into GitVersion when it is released. I have tested this on my Sample Application, using Stash as the target to Clone from, and I am happy to report that it is all working. |
I have just upgraded my build server (TeamCity) to use the latest release of GitVersion.Portable from Chocolatey.org, and I am now seeing this failure in the TeamCity Build Log:
I should mention that I am using the MetaRunner for GitVersion that was created by @robdmoore, but I don't think that this is the issue.
Based on the error message, I thought it could be an issue with a modification to the AssemblyInfo files causing a problem, so I removed that as a test, but I get the same error for both.
Any thoughts on what could be going on here?
Thanks!
The text was updated successfully, but these errors were encountered: