tfs | Microsoft TFS API Python client

Preamble: the dependency resolution in the C# program was not picking up v12 where v9/6 was requested automagically; it was only doing so because the config file of the compiled program was telling it so, and that only happened once and because Newtonsoft v12 was being referenced in the project. Thanks to @n0rd for pointing that out. Resolving strongly named dependent assemblies to a higher major version is not a default behavior in .NET 4.5-8.

Modifying the config of Powershell to achieve the same might be possible, but I didn't go there. The original piece that needed this logic will be eventually running on servers that I don't control, so the less administrative overhead, the better. Now, for the working answer.

You can provide a resolve handler in Powershell 5 after all, telling .NET to use the loaded version of Newtonsoft in lieu of any other one. It goes like this:

I think the issue can be related to the ingress classname. May be you need to explicitely provide it in your .tf:

Try using either:

I made a few changes to your script, it's working on my environment when downloading the sources from the same Azure DevOps organization. The changes are as follows:

• Instead of relying on manually setting the sparse config, I use the sparse-checkout set command, it will automatically set other required flags and configs.
• Instead of relying on (insecurely) setting the security token in the clone URL, instead I pass the variable in as an extraheader.
• Instead of relying on a PAT (which is fine locally), I instead pick up the access token from the build/release job.

This results in the following task:

I miss having the ability to encode a query and have the same question. Nice work with finding out it's in base64.

I think reverse engineering is the only way to find out how things are encoded. For example, the stuff after the underlines is most likely binary-encoded.

See the below for economy:

Git LFS uses a different HTTP and TLS library than the one in Git. Git uses libcurl, and Git LFS uses the Go HTTP library. As a result, the supported authentication logic is different, although both programs will use the Git credential helpers and other credential lookup logic.

Since you mention Azure DevOps, my guess is that you're using NTLM. In 3.0, Git LFS removed NTLM because it had known bugs and nobody was interested in fixing them, and because it uses cryptography known to be insecure since 1995. Azure DevOps is the only major site known to use NTLM, and the Git LFS maintainers asked if they'd like to be involved in helping maintain it, and they declined.

NTLM can be handled in one of two ways: via the NTLM authentication scheme or the Negotiate scheme. The latter is also used by Kerberos, which both Git and Git LFS do support, and which is secure. Currently, if you have NTLM set up to use Negotiate, Git LFS simply won't work, since it prioritizes Negotiate over Basic. In the upcoming 3.1, expected out this month or next, Git LFS will fall back to Basic if Negotiate fails, so you'll be able to work even if you have NTLM enabled on your instance.

I strongly encourage everyone to get rid of NTLM because it's so insecure. There's really no defensible reason to use it anymore: even Microsoft tells you to turn it off. If you turn off NTLM on your instance, or switch to Kerberos, things should just work. Otherwise, you'll need to wait for Git LFS 3.1 or explicitly set the authentication method to basic in the configuration.

No, git-tfs doesn't have this option to use an existing workspace because it would have been difficult to check if the existing workspace have been created from the good TFVC path.

And also, if you want to clone with branch support, you need to create multiple workspaces at different places so git-tfs need to handle them itself.

If you are in charge of the migration, your best chance is to ask your tfs administrators to grant you this right at least until you finish your migration.

I had to give search criteria in below format and it worked:

Ok, so after my code finished, I worked the calculations backwards,

Case1, is 9c1 x 19 = 171

Case2, is 9c2 x 19 x 19 = 12,996

Case3, is 9c3 x 19 x 19 x 19 = 576,156

That's a total of 589,323 combinations.

Here is the code for all 3 cases, you can run them sequentially.

You also requested to join the array into a single string, I have updated my code to reflect that.