Latest Bugs & Tasks
SO_NOSIGPIPE prevents read or write from sending a SIGPIPE signal if the other side of the socket / pipe isn't connected, and instead will return EPIPE. This can be used for better usage of blocking read/write.
I think this is fixed now.
I do a backup on a schedule using rsync, and this morning, noticed there had been a crash . I don't know any of the circumstances since it happened during an automated backup, but I've attached the debug info. Here's the syntax I was using, which normally seems to work OK:
rsync -v -a -X --exclude=/boot/system/var/swap /boot/* /Backup
Fixed in hrev48611.
Closing this one as it has been 4 years since this bug was tested and a lot of code has changed since Alpha 2. If the issue persists, please open a new bug and reference this one.
Applied in hrev48608.
Should be fixed in hrev48607.
Made the required changes in hrev48606.
Fixed in a different way in hrev48605.
Applied in hrev48604.
Applied in hrev48601.
If I remove an USB key without unmounting them, the kernel crash.
Currently, kOpenSubmenuDelay is set to 225000 (1/4 of a second) which is definitely way too much. This value came from this patch attachment:menu-diagonal-movement3.patch:ticket:284
"open submenus after a 0.2 sec delay to make menu browsing smoother - moving a cursor quickly down a menu doesn't cause submenus to quickly appear and disappear while the cursor flies over an item with submenu)"
However, in R5 there was no delay at all which was one of the reason it felt so fast.
The date of creating a new file in tracker (second button/new/text file) produce a file with the date of the kernel build date instead of the current date.