run duration has been increased after upgrade to the latest version

[carpet92]

I've been using the v18.27.1 version of oh-my-posh for a long time. Today, after upgrading to the latest v19.29.1, I noticed that the run time has been increased to 50 ms, whereas v18.27.1 took about 10 ms.

After that I downgraded back to the 18.27.1 using:

$ winget uninstall JanDeDobbeleer.OhMyPosh
$ winget install --version 18.27.1 JanDeDobbeleer.OhMyPosh -s winget

A couple of screenshots:

---

---

---

• PowerShell 7.4.2
• Windows Terminal Preview Version: 1.21.1272.0
• Added all exclusions for Windows Defender (for pwsh.exe, windowsterminal.exe, wt.exe, oh-my-posh, for all related processes etc.)

[JanDeDobbeleer]

@carpet92 I'm going to need the output of those calls to see what's taking more time. You can pipe to clip using oh-my-posh debug --plain | clip and paste here.

[carpet92]

@JanDeDobbeleer when I run the oh-my-posh debug --plain | clip command, it produces no output for me. So, I just uploaded the results of oh-my-posh debug > debug_18.27.1.txt and oh-my-posh debug > debug_19.29.1.txt to gist:

• debug_18.27.1.txt https://gist.githubusercontent.com/carpet92/c479d74773019b439e3df82729d39c72/raw/4eaf525b6632f1f433fa693aadf18523360a6bf6/debug_18.27.1.txt
• debug_19.29.1.txt https://gist.githubusercontent.com/carpet92/c479d74773019b439e3df82729d39c72/raw/4eaf525b6632f1f433fa693aadf18523360a6bf6/debug_19.29.1.txt

[JanDeDobbeleer]

@carpet92 log wise, yes, experience wise that shouldn't make a difference.

[carpet92]

@JanDeDobbeleer isn't this problem reproduced in your setup? If not, then the problem may be on my side. Perhaps silly question, but ... can't this be caused by an old version of go installed locally? I have go 1.17.8 and I didn't use it since this version so it's such an old.

[JanDeDobbeleer]

@carpet92 it should be compiled so no. But I need to update my local VM to reproduce. I'm on my mac ATM.

[carpet92]

@JanDeDobbeleer just a heads-up. I tried to iteratively install oh-my-posh version by version and check run duration, and found that the discussion topic started to appear since v19.16.0. Log: debug_19.16.0.txt

[JanDeDobbeleer]

@carpet92 so it is the upgrade to go1.22. I'm going to file a bug on their end for this one.