Expect spawn patch is a workaround not a solution

[Matthew Burgess]

According to the above bug report, our Expect patch that fixes spurious failures in the GCC testsuite isn't the correct approach. I would be grateful (as I'm sure TCL and Expect upstream would be) if someone could take the time to write a simple Expect script and/or a simple TCL script that tickles the same bug! Good luck :)

[alexander@…]

Greg Schafer pointed me to this when I asked the same question:

​https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=43310

[Matthew Burgess]

Well, I've tried that test case. I've increased it to 150 lines and it still works as I'd expect (no pun intended!) it to. This is without patching expect at all, and compiling it up against tcl-8.4.11.

[archaic@…]

What is the status of this bug? Has anyone done more testing?

[archaic@…]

Dropping priority and moving to unspecified milestone. I'd like to close it as WONTFIX, but Matt may have other ideas.

[Matthew Burgess]

I'd prefer if we keep it open. It'll take a more committed person than me to try and get the TCL and Expect devs to cooperate and recognize there is a problem. Last time I tried, folks couldn't even reproduce the problem (IIRC one person tried, but they were running the test case on an already patched version of Expect that was in Ubuntu!).

[Matthew Burgess]

We're not going to get this addressed by 6.3. Shifting to 'future'.

[gerard@…]

Closing this for now. Not very hopeful in seeing it resolved any time soon. Reopen at a later date if somebody decides to tackle this one. Seems majority is content to leave it status quo.