https://issues.apache.org/SpamAssassin/show_bug.cgi?id=6939
--- Comment #80 from Karsten Bräckelmann <[email protected]> --- > No, you're mistaken in the general sense. That depends entirely on what a > program is intended to do with it. But regardless, even in the context of > spamc, it's still wrong as I've already demonstrated. So you say, the calling program should never have passed an empty string as a value to an argument that requires a file? Agreed. Evolution should have checked that. > > Wrong. STDIN needs to be accessed, to decide whether to contact spamd at > > all. See comment 38 for the max-size option you outright dismissed before. > > However, this is entirely irrelevant. > > Wrong again. spamc calls read_args()'s before setting up transport. ^^^^ Read. Not the same as verify arguments' contents. And do FS check to test the argument exists, is a socket, and the apparent listener seems to be a spamd instance. Neither for TCP. So what? > > * Your recent arguing is about apparently hanging spamc processes, which > > can only happen with stalled STDIN. > > No, we tried it even with input (see above), and it still hanged. False. If you strongly believe I am wrong, please tell us at least the comment number rather than "above" only. > > Kip, can you coherently and in a single statement (unlike the squirming in > > the previous 70+ comments) explain the issue again, please? > > It wouldn't matter what I said, besides the fact that I've already met this > request, You didn't. > you're going to continue to troll I won't. > and not listen anyways. The problem has been patched now. It has not. (Unless you refer to your private system only.) > If you don't want to use it to improve your > code base, that's fine. We can patch it downstream. ^^ By "we" you mean... Please add a comment to this bug report, if/when "we"/you ever patch it downstream. -- You are receiving this mail because: You are the assignee for the bug.
