Whilst trying to configure ls' colors on Cygwin (specifically the distro babun), I found that the custom extension colors that were in the LS_COLORS environment variable weren't being used. Here's an example: di=34:ln=35:so=32:pi=33:ex=31:bd=34;46:cd=34;43:su=30;41:sg=30;46:tw=30;42:ow=30;43:*.tar=1;31 Normally, the ls command should output white text for every file EXCEPT .tar files, which should be red. However, what I'm getting is white text for everything INCLUDING .tar files. I wonder if it's a problem with MSYS ls, or if it's a problem with my shell (zsh). Basically, the custom file extension colors that I try and define fall back to the default file color. Here's the corresponding part of my .zshrc where I've defined the ls alias and run the output of dircolors. eval "dircolors /etc/DIR_COLORS" alias ls="/bin/ls --color=auto"

On Msys2 env.: $ cd /c/msys64/mingw64/lib/ $ ln -s libz.dll.a libzlib.dll.a But confusingly why ln command did copy instead of linking it symbolically ? $ ls -l libz* -rw-r--r-- 1 abdu None 108700 Jan 23 2024 libz.a -rw-r--r-- 1 abdu None 65914 Jan 23 2024 libz.dll.a -rw-r--r-- 1 abdu None 65914 Jan 23 2024 libzlib.dll.a -rw-r--r-- 1 abdu None 2298 Dec 13 03:15 libzoneoc.a Please one help out of confusion to solution

I have found a similar question on https://askubuntu.com/questions/144547/how-to-package-a-single-text-file-into-a-deb - unfortunately there is no answer I can use. Basically, I want to manage a text file, say myapp.txt.config, in git; and then, I'd want to create a .deb package that copies it on Raspberry Pi in /usr/share/myapp/myapp.txt.config. However, most of the work I do in this repository is in MSYS2/MINGW64 on Windows 10, so I would like to avoid the round-trip to a Raspberry Pi, just to build the .deb package. Since this is just a text file, I'd like a single .deb to be valid for all distributions (stretch, bookworm for Raspbian). How would I go about implementing this? I can see that a .deb file is an [ar archive with tar archives inside](https://en.wikipedia.org/wiki/Deb_(file_format)) - and there is working ar, tar, install, bash, python3, perl, make on MSYS2/MINGW64, however there are no Debian tools in the MSYS2/MINGW64 repos, as MSYS2 uses pacman instead. I have seen in https://askubuntu.com/questions/90764/how-do-i-create-a-deb-package-for-a-single-python-script that the basic source package layout would be:

myapp-config/
    -- myapp.txt.config
    -- debian/
        -- changelog
        -- copyright
        -- compat
        -- rules
        -- control
        -- install

... but I'd have to run dch --create to "create a properly formatted debian/changelog entry", and debuild --no-tgz-check to build; though dch is a Perl script ([devscripts/scripts/debchange.pl](https://github.com/Debian/devscripts/blob/master/scripts/debchange.pl)) , and so is debuild ([devscripts/scripts/debuild.pl](https://github.com/Debian/devscripts/blob/master/scripts/debuild.pl)) ; but I cannot tell to what extent they will work on MSYS2/MINGW64. I thought also of creating a Makefile with a [single install entry](https://stackoverflow.com/questions/39892692/how-to-implement-make-install-in-a-makefile) :

install:
	install -D -m 644 myapp.txt.config /usr/share/myapp/

... and then calling [checkinstall](https://github.com/ruxkor/checkinstall/blob/master/checkinstall) on it to create a deb package; checkinstall seems to be bash, but I cannot tell if it will run on MSYS2/MINGW64, as it probably needs sudo - also that github version is from 13 years ago, and it might even be that checkinstall is [no longer being maintained](https://askubuntu.com/questions/1138384/why-is-checkinstall-no-longer-being-maintained) . So, what would be the right approach to build a .deb package in my case? If there is no way to build a .deb package on MSYS2/MINGW64, what would be the right way to organize the project, so I build this single-text-file .deb on Raspberry Pi with least amount of effort?

I'm running git bash 2.45.2.1 on a windows computer (for work) and I've noticed that running even basic processes can be extremely slow. For example,

$ time /usr/bin/echo 1 > /dev/null
real    0m0.102s
user    0m0.015s
sys     0m0.015s

$ time /usr/bin/echo 1 | wc > /dev/null
real    0m0.151s
user    0m0.031s
sys     0m0.030s

Why would it take this version of bash a whole 0.1 s to spawn a process? Is this a Windows problem? Can this be fixed? This can get **very** slow if I pipe even a hundred lines to xargs. I don't think the slowness can be the hardware's fault. This computer's processor is a 13th Gen Intel Core i9-13950HX.

I work on a legacy system originally written in C for 3b2 SVR3, later in C++ for Sun/Solaris, now mainly for Linux. We have a few users who require our apps to run on Windows Servers, so we have accommodated that for the last 18 years using MSys / MinGW, cross compiled from our *nix-based development machines. We're now trying to upgrade to MSys 2 and have run into what appears to be an intractable problem that no amount of searching has turned up an answer to. Our application works across multiple time zones simultaneously. We can't rely on a static system time zone setting to operate properly. We get our time zone definitions from the IANA database, installed as zoneinfo files in the traditional locations, depending upon environment. We update zoneinfo for our users as the need arises. For as long as I can remember (I've worked on this for nearly 30 years), it was adequate to putenv("TZ="); tzset(); to get methods like mktime() and localtime() to produce the information we needed. That has stopped working now that we've moved to MSys 2. Additionally, most of our code doesn't care what the initial time zone is, but one method insists that the TZ variable have some value (any value, it doesn't care). It really doesn't matter why TZ must be set - that code exposes another symptom that we might not otherwise have noticed so quickly: upon entry to any of our programs, the TZ environment variable has been removed from the environment. The final curiosity is that MSys 2 comes with a **date** program that seems to work just fine with the TZ environment variable, as do **printenv** and **ls**, seeming to indicate that what we need to do *can* be done. We have a short test program we've been using to demonstrate the problem and with any luck, prove that we've found a solution: #include #include #include int main(int argc, const char *const *argv) { const char *tz = getenv("TZ"); const char *ptz = tz ? tz : "TZ not set"; printf("at entry: %s\n", ptz); time_t now = time(NULL); putenv("TZ=US/Pacific"); tzset(); printf("in %s it is now %s", getenv("TZ"), ctime(&now)); putenv("TZ=US/Eastern"); tzset(); printf("in %s it is now %s", getenv("TZ"), ctime(&now)); return 0; } When I run the above on our Linux dev machine, I get the following output: $ printenv TZ US/Central $ date "+%F %T %Z" 2024-08-15 10:20:55 CDT $ ./msystest at entry: US/Central in US/Pacific it is now Thu Aug 15 08:20:55 2024 in US/Eastern it is now Thu Aug 15 11:20:55 2024 Under MSys 2, just seconds later, I get the following: $ printenv TZ US/Central $ date "+%F %T %Z" 2024-08-15 10:20:57 CDT $ ./msystest at entry: TZ not set in US/Pacific it is now Thu Aug 15 16:20:57 2024 in US/Eastern it is now Thu Aug 15 16:20:57 2024 Using **ldd** on MSys 2, we've noticed that there are different DLLs referenced by **date** and **msystest**: $ ldd /usr/bin/date ntdll.dll => /c/Windows/SYSTEM32/ntdll.dll (0x7ff962d20000) KERNEL32.DLL => /c/Windows/System32/KERNEL32.DLL (0x7ff961b30000) KERNELBASE.dll => /c/Windows/System32/KERNELBASE.dll (0x7ff960340000) msys-intl-8.dll => /usr/bin/msys-intl-8.dll (0x430b30000) msys-2.0.dll => /usr/bin/msys-2.0.dll (0x180040000) msys-iconv-2.dll => /usr/bin/msys-iconv-2.dll (0x5603f0000) $ ldd msystest ntdll.dll => /c/Windows/SYSTEM32/ntdll.dll (0x7ff962d20000) KERNEL32.DLL => /c/Windows/System32/KERNEL32.DLL (0x7ff961b30000) KERNELBASE.dll => /c/Windows/System32/KERNELBASE.dll (0x7ff960340000) msvcrt.dll => /c/Windows/System32/msvcrt.dll (0x7ff962790000) So our question is - is our problem the result of the differing DLLs, and if so, how do we get the msys DLLs built into our programs rather than the msvcrt DLL? We've done a lot of searching online to try to resolve this to little avail. We usually get 1 of 2 answers: 1. Windows doesn't work that way so you can't do what you want with MSys 2. This is how you build a DLL with MinGW Needless to say, the answers are either (1) wrong, or (2) answering the wrong question. Any help anyone can provide/point us to would be invaluable.

I recently started using tmux (in addition to screen), both in a Linux environment and in msys2 (which is also new to me and for which I cannot install screen). Very similar to screen, starting tmux disables the shift-PgUp scrolling (and the scroll bar in the terminal emulator). For both terminal multiplexers the Internet does provide an easy solution. In the case of tmux it is adding this setting to the configuration: set -g terminal-overrides 'xterm*:smcup@:rmcup@' This works fine in the Linux environment but not in msys2. $TERM is set to xterm. Unfortunately I am not familiar (yet) with the internal details of terminals so I neither understand why this setting is necessary nor why it could fail. After all before starting tmux the msys2 terminal emulator does have this capability. The Internet contains a lot of questions about tmux scrolling but I could not find any for this problem with msys2. **Meta** I was thinking about whether this is the most suitable site for this question (and it may well not be). So if this does not result in a sufficient answer but you have an idea where it might make more sense to ask this then feel free to let me know.

Let's say I have the following trivial script, tmp.sh:

echo "testing"
stat .
echo "testing again"

Trivial as it is, it has \r\n (that is, CRLF, that is carriage return+line feed) as line endings. Since the webpage will not preserve the line endings, here is a hexdump:

$ hexdump -C tmp.sh 
00000000  65 63 68 6f 20 22 74 65  73 74 69 6e 67 22 0d 0a  |echo "testing"..|
00000010  73 74 61 74 20 2e 0d 0a  65 63 68 6f 20 22 74 65  |stat ...echo "te|
00000020  73 74 69 6e 67 20 61 67  61 69 6e 22 0d 0a        |sting again"..|
0000002e

Now, it has CRLF line endings, because the script was started and developed on Windows, under MSYS2. So, when I run it on Windows 10 in MSYS2, I get the expected:

$ bash tmp.sh
testing
  File: .
  Size: 0               Blocks: 40         IO Block: 65536  directory
Device: 8e8b98b6h/2391513270d   Inode: 281474976761067  Links: 1
Access: (0755/drwxr-xr-x)  Uid: (197609/      USER)   Gid: (197121/    None)
Access: 2020-04-03 10:42:53.210292000 +0200
Modify: 2020-04-03 10:42:53.210292000 +0200
Change: 2020-04-03 10:42:53.210292000 +0200
 Birth: 2019-02-07 13:22:11.496069300 +0100
testing again

However, if I copy this script to an Ubuntu 18.04 machine, and run it there, I get something else:

$ bash tmp.sh
testing
stat: cannot stat '.'$'\r': No such file or directory
testing again

In other scripts with the same line endings, I have also gotten this error in Ubuntu bash: line 6: $'\r': command not found ... likely from an empty line. So, clearly, something in Ubuntu chokes on the carriage returns. I have seen https://unix.stackexchange.com/questions/355559/bash-and-carriage-return-behavior : > it doesn’t have anything to do with Bash: \r and \n are interpreted by the terminal, not by Bash ... however, I guess that is only for stuff typed verbatim on the command line; here the \r and \n are already typed in the script itself, so it must be that Bash interprets the \r here. Here is the version of Bash in Ubuntu:

$ bash --version
GNU bash, version 4.4.20(1)-release (x86_64-pc-linux-gnu)

... and here the version of Bash in MSYS2:

$ bash --version
GNU bash, version 4.4.23(2)-release (x86_64-pc-msys)

(they don't seem all that much apart ...) Anyways, my question is - is there a way to persuade Bash on Ubuntu/Linux to ignore the \r, rather than trying to interpret it as a (so to speak) "printable character" (in this case, meaning a character that could be a part of a valid command, which bash interprets as such)? EDIT: *without* having to convert the script itself (so it remains the same, with CRLF line endings, if it is checked in that way, say, in git) EDIT2: I would prefer it this way, because other people I work with might reopen the script in Windows text editor, potentially reintroduce \r\n again into the script and commit it; and then we might end up with an endless stream of commits which might be nothing else than conversions of \r\n to \n polluting the repository. EDIT2: @Kusalananda in comments mentioned dos2unix (sudo apt install dos2unix); note that just writing this:

$ dos2unix tmp.sh 
dos2unix: converting file tmp.sh to Unix format...

... will convert the file in-place; to have it output to stdout, one must setup stdin redirection:

$ dos2unix 

... and then, in principle, one could run this on Ubuntu, which seems to work in this case:

$ dos2unix 

However, - aside from the slightly messy command to remember - this also changes bash semantics, as stdin is no longer a terminal; this may have worked with this trivial example, but see e.g. https://stackoverflow.com/questions/23257247/pipe-a-script-into-bash for example of bigger problems.