# CMakeError.log confuses new users
**URL:** https://discourse.cmake.org/t/cmakeerror-log-confuses-new-users/5322
**Category:** Development
**Created:** [March 29, 2022, 8:16pm UTC](https://discourse.cmake.org/t/cmakeerror-log-confuses-new-users/5322 "2022-03-29T20:16:08Z")
**Posts on this page:** 7
**Page:** 1
### Author:  [@buildSystemPerson](https://discourse.cmake.org/u/buildSystemPerson)
#### Post date: [March 29, 2022, 8:16pm UTC](https://discourse.cmake.org/t/cmakeerror-log-confuses-new-users/5322/1 "2022-03-29T20:16:08Z")
At work when a user has problems with CMake the very first thing they do when contacting me is send me a copy of their `CMakeError.log`
The problem is that the `CMakeError.log` almost never has the problem they are experiencing. `CMakeError.log` is really only useful when the `project` call fails.
I then provide them instructions on what I need to figure out their issue.
This is mainly a workflow issue with new CMake users. More experienced developers don’t generally have this problem.
Thoughts?
---
### Author:  [@ben.boeckel](https://discourse.cmake.org/u/ben.boeckel)
#### Post date: [March 29, 2022, 8:48pm UTC](https://discourse.cmake.org/t/cmakeerror-log-confuses-new-users/5322/2 "2022-03-29T20:48:43Z")
Huh. I usually need to point people to the `CMakeError.log` file at all (it is also used in `try_compile` situations I think). I’m not sure what CMake itself can do here.
FWIW, my bad reports usually involve red herrings in the form of harmless warnings or missing context such as lacking the backtrace or it being some domino error that doesn’t matter because things fell apart in the scrollback and that got missed.
---
### Author:  [@buildSystemPerson](https://discourse.cmake.org/u/buildSystemPerson)
#### Post date: [March 29, 2022, 8:52pm UTC](https://discourse.cmake.org/t/cmakeerror-log-confuses-new-users/5322/3 "2022-03-29T20:52:37Z")
> [@ben.boeckel](#):
>
> I’m not sure what CMake itself can do here.
Here is usually what I ask for in these situations:
```auto
# Redirect stdout/stderr into 1 log file with DEBUG enabled
cmake -B build/vs --log-level=DEBUG *> cmake_log.txt
```
---
### Author:  [@brad.king](https://discourse.cmake.org/u/brad.king)
#### Post date: [March 30, 2022, 1:32pm UTC](https://discourse.cmake.org/t/cmakeerror-log-confuses-new-users/5322/4 "2022-03-30T13:32:22Z")
There is related discussion in [CMake Issue 22131](https://gitlab.kitware.com/cmake/cmake/-/issues/22131).
---
### Author:  [@buildSystemPerson](https://discourse.cmake.org/u/buildSystemPerson)
#### Post date: [February 3, 2023, 10:31pm UTC](https://discourse.cmake.org/t/cmakeerror-log-confuses-new-users/5322/5 "2023-02-03T22:31:03Z")
Addressed as of CMake 3.26 👍
```auto
The “CMakeFiles/CMakeOutput.log” and “CMakeFiles/CMakeError.log”
files are no longer populated by CMake’s builtin modules, and
“cmake(1)” no longer suggests looking at them after a “CMake Error”
occurs. Information previously logged to those files is instead
logged to “CMakeFiles/CMakeConfigureLog.yaml”, the “cmake-configure-
log(7)”.
```
---
### Author:  [@VictorEijkhout](https://discourse.cmake.org/u/VictorEijkhout)
#### Post date: [August 11, 2025, 7:59pm UTC](https://discourse.cmake.org/t/cmakeerror-log-confuses-new-users/5322/6 "2025-08-11T19:59:44Z")
What is the function of the phrase
> [@buildSystemPerson](#):
>
> the “cmake-configure-log(7)”.
in your answer?
---
### Author:  [@craig.scott](https://discourse.cmake.org/u/craig.scott)
#### Post date: [August 11, 2025, 9:48pm UTC](https://discourse.cmake.org/t/cmakeerror-log-confuses-new-users/5322/7 "2025-08-11T21:48:14Z")
It refers to the following page of the CMake docs: [https://cmake.org/cmake/help/latest/manual/cmake-configure-log.7.html](https://cmake.org/cmake/help/latest/manual/cmake-configure-log.7.html)