Trying to figure out why the httpd server sometimes required to be restarted, after boot, I find finally the solution: this is not an thttpd malfunction, unfortunately (or fortunately?)! In releases < 10.x, all custom setup was in runlevel 5; in 10.x, trying to move any muLinux-specific script in the runlevel 4, all daemons are started (but httpd is the master) at the runlevel four. Now, the "init" daemon (the bad guy) is used to _kill_ any process, before to change runlevel, so (the bad guy) kill without delay also the httpd. As workaround, you have to restart the daemon by hand, when prompt is reached (it happens at runlevel 5): # setup httpd restart Better solution is, for installed systems, to move the row ". /etc/rc/setup" from the bottom of /etc/rc/4 to the top of the /etc/rc/5 boot script. I'm sorry for the stupid error, but muLinux is being very close to a complete Linux system, and a single man can't stay ahead to every changes. For the sake of the humanity, I will release soon a 10r5 version. I also added, between other, the automatic detection of available RAM in setup/tmp; so, if RAM is enough, it is maximized to 4096 (4M). Why? I noticed the upx compressed binaries crash silenty when /tmp has not enough space for decompression. This is not the case in current published release of muLinux (they use no upx), but it happens for me. Michele -- I'd like to conclude with a positive statement, but I can't remember any. Would two negative ones do? -- Woody Allen