TYPO3 Core

Could you have a look at this, and verify my findings? Thanks!

Thanks Karsten, yes you are right. Please correct it as soon as HEAD is open again (btw. can you please send a reminder to Kasper as well?).

Maybe you should also change the variable which is called $forceReturnPath the same way.

Important: This must be fixed in the 3.7, 3.8 and in the HEAD branch!

Ok, I'll fix this in all branches as soon as you give your okay to HEAD changes. I'll change the variable name as well then.

Thanks for the quick feedback.

I agree, this is an error (Sorry for that). Please correct it (when CVS is open again).

On a side note, I think that the variable should be forceReturnPath, and not enable return path (Thus in my opinion t3lib/class.t3lib_htmlmail.php should be corrected, and not t3lib/config_default.php), since the returnpath really is forced by setting the -f flag to sendmail, which forces the return-path to something different than what is stated in the header of the mail (The header of the mail is then rewritten to include the correct return-path by senmail).

Thanks for explaining the variable naming. This way it makes sense, I had looked at it from a different perspective (where nothing was forced, if there wasn't any value to set the return path to).

Then I'll fix it in the htmlmail class, and probably change wording of the settings' description a bit.

Perfect. You are very welcome to change the wording. The only change needed (I think) in htmlmail is to change

$this->forceReturnPath = $GLOBALS['TYPO3_CONF_VARS']['SYS']['enableReturnPath'];

to

$this->forceReturnPath = $GLOBALS['TYPO3_CONF_VARS']['SYS']['forceReturnPath'];

I changed t3lib_htmlmail so it checks for forceReturnPath instead of enableReturnPath.

Sorry for not catching this error ealier.

Thanks, Jan-Erik. I have just fixed the bug in the 3.8 branch as well.

Thanks to you two, sorry for not acting as promised. Has the explanation been touched as well?

That was not needed since "enableReturnPath" is not mentioned anywhere else.

Michael, that's true. But it doesn't force the return-path - there isn't any return path set through this option. It rather forces the usage of a certain method of specifying it, if it should be set. No?