Joined: 26 Mar 2013 Posts: 1856 Location: Hung Hom, Hong Kong
Posted: Sun May 14, 2023 6:33 Post subject:
ho1Aetoo wrote:
https://svn.dd-wrt.com/changeset/52514
now happy?
I am just anxious about this kind of obvious mistake.
"happy"?? Um... what if we could find out when the bug was introduced and what had caused the confusion?
Thank all developers and helpers for this!
BTW, if you thought the bug might affect you:
Aftrer flashing the build with this bugfix, disable Lighttpd, wait a few minutes then enable Lighttpd again. This will toggle the relevant NVRAM value correctly. Then you could also remove any custom firewall rules related to this bug. _________________ Router: Asus RT-N18U (rev. A1)
Drink, Blink, Stretch! Live long and prosper! May the Force and farces be with you!
Joined: 26 Mar 2013 Posts: 1856 Location: Hung Hom, Hong Kong
Posted: Sun May 14, 2023 7:38 Post subject:
ho1Aetoo wrote:
And the bindhost is not a nvram paremeter ... and your enable/disable recommendation is total nonsense
You need to regenerate lighttpd.conf. The easy way is to enable+disable or disable+enable the service, depending the current setup of Web Server Ligttpd.
Quote:
Nobody cares!
Just like your irrelvant spam for days no one care
Anyway... thank you for helping.
I did NOT know that configuration files are re-generated in each boot. What I was merely trying to help. _________________ Router: Asus RT-N18U (rev. A1)
Drink, Blink, Stretch! Live long and prosper! May the Force and farces be with you!
Joined: 26 Mar 2013 Posts: 1856 Location: Hung Hom, Hong Kong
Posted: Sun May 14, 2023 7:42 Post subject:
ho1Aetoo wrote:
The configuration is automatically generated when the service is started at each router restart, so even after the firmware update!
Ok, I did NOT know that. That explained ho1Aetoo's comments. Thanks!!!
But this means the boot-up process might be erratic if the generation of config files finishes AFTER the related process is started. This is possible with parallel running of scripts and processes during startup (race condition). A process might read an incomplete config file that's still being generated.
It also means that the boot-up process might take more time. But yes, it's a simple solution to patching bugs.
I think this deserves another thread. Sorry!
(Side note: Sometimes, unbound and minidlna didn't start up properly in my RT-N18U, especially if the router was contacted by other devices before it finished startup. But that might be related to USB mounting. I am using custom config files in /jffs/etc.) _________________ Router: Asus RT-N18U (rev. A1)
Drink, Blink, Stretch! Live long and prosper! May the Force and farces be with you!