Good evening everyone, as the title states, with bb 1.37 i was hit with the sad events unfold by the attached script (tailored for this report). "Do i consider this behaviour as a bug or a feature?" is what i'd like to ask. If the latter then i'm led to wrap lots of my code in unsavory safeguards to painfully obtain, somewhat, stuff working in a barely satisfactory way.
So what happens in there? See the sample output : [----] ----------- while loop --------------------1 so1... a 29986: [1]<29 117>xxxxx 29986: oooo9 117>x 5 so1... 29986: <a> ----------- read alone --------------------1 so2... b 29986: []<29 118>xxxxxx 29986: ooooo ----------- short sleep -------------------1 so3... c 29986: []<29 119>xxxxxxx 29986: <OK> 2 $ bc [----] The read alone was introduced to confirm that the "while" instruction did not interfere withe test. So, as the script runs you have to play with the geometry of the window of the terminal. As we see, all the output is made from a single process (28815 in this case). As far as the read command is concerned, an interruption by SIGWINCH(28) ends it with $?=1. Whereas, unwavering, 'sleep' keep... sleeping the whole time it was asked (we never get to have winChg() called). so bug or feature to be expected? Thanks for yor attention and blissful night (at least in my TZ) JClu 202510164
wc+sl+rd.sh
Description: application/shellscript
_______________________________________________ busybox mailing list [email protected] https://lists.busybox.net/mailman/listinfo/busybox
