<div dir="ltr"><div dir="ltr">Le mar. 9 juil. 2019 à 13:24, hex <<a href="mailto:hex7c3@gmail.com">hex7c3@gmail.com</a>> a écrit :<br></div><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<div bgcolor="#FFFFFF">
<div class="gmail-m_-3506881588657602860moz-cite-prefix">Now that I look at it it seems very obvious. I still have doubts,
though:<br></div>
<p>I guess I am seeing the contents of a build directory as somewhat
volatile. For most files I want that, to either clean the project
or override the files. <br>
</p>
<p>But what about files I want to archive, such as a tarball or
zipped documentation: does it make sense to place them into the
build directory? The files belong to the project, though are not
source controlled but aren't install targets either. </p></div></blockquote><div>How would CMake know that these tarballs are in a somewhat final version that you want to archive as opposed to a state where they haven't passed the most basic tests yet? I would implement this outside CMake: wherever is the information that guides the decision to archive. When that decision is made then some script would simply "exfiltrate" the tarballs and whatever else from the build/ directory before it gets cleaned. Sounds like something CI does every day.</div><div><br></div><div>This being said, CPack seems to have a ton of documented and undocumented (read: Stackoverflow) options, hopefully some of them do what you want.</div><div><br></div><div>Marc</div><div><br></div></div></div>