| 2 | | > I can't build test it right now, but from cursory grepping I see no possibility how it could have worked in 4.8.16, because the code in the tarball doesn't contain any relevant references to `SUBSHELL_OPTIONAL`. I don't believe in magic subshell disabling fairies, you know. If you can bisect it to show which commit exactly between 4.8.16 and 4.8.17 broke it, this will be much appreciated. |
| 3 | | > |
| 4 | | > > And why throw-it away? |
| 5 | | > |
| 6 | | > If I am right that it was lost 5 years ago, and nobody has complained so far, then it doesn't sound like something that users can't live without. And if there is no test for it, it can vanish anytime without anyone noticing until it's too late. |
| | 2 | > I can't build test it right now, but from cursory grepping I see no possibility how it could have worked in 4.8.16, because the code in the tarball doesn't contain any relevant references to `SUBSHELL_OPTIONAL`. I don't believe in magic subshell disabling fairies, you know. |
| | 3 | OK, try-it. Your gonna be amazed. Fairies or not. |
| | 4 | For me, every version worked fine on AIX and Linux (cento 6 and 7, fedora, openwrt). |
| | 5 | |
| | 6 | > If you can bisect it to show which commit exactly between 4.8.16 and 4.8.17 broke it, this will be much appreciated. |
| | 7 | Ok. I will try on the next weekend. |
| | 8 | |
| | 9 | > If I am right that it was lost 5 years ago, and nobody has complained so far, then it doesn't sound like something that users can't live without. |
| | 10 | No one complained because it just worked. |
| | 11 | |
| | 12 | > And if there is no test for it, it can vanish anytime without anyone noticing until it's too late. |
| | 13 | Agreed. |
| | 14 | |
