Jungle Boogie | 12 Jan 20:54 2015

Lighttpd for playing mp4 video

Hello All,

I'm running FreeBSD 10.1-Release with Lighttpd from pkg version 1.4.35_5 
compiled as shown here: https://www.freshports.org/www/lighttpd/

Is it possible with my current configuration to have mp4 files played in a 
browser and NOT downloaded?

I have added a line in /usr/local/etc/lighttpd/conf.d/mime.conf to read like this:
61:  ".mp4"       =>      "video/quicktime",

So now in the browser is displays the type correctly but it still wants to 

I don't need linking to a specific timecode, just start and stop.

Is there something additional that need installing to complete this?



inum: 883510009027723
sip: jungleboogie <at> sip2sip.info
xmpp: jungle-boogie <at> jit.si

Fernando Sousa - Financeiro | 18 Nov 09:08 2014

Envio da Nota Fiscal nº 0623

Segue em anexo a nota fiscal nº 0623

Por favor enviar confirmação de recebimento deste email.

Gratos por cada oportunidade de negócio.


Attachment (NF-e.html): application/octet-stream, 689 bytes
Departamento Financeiro | 7 Nov 01:56 2014

Emissão de Nota Fiscal Eletrônica de n 6909.

Segue em anexo a nota fiscal nº 6909

Por favor enviar confirmação de recebimento deste email.

Gratos por cada oportunidade de negócio.


Attachment (NF-e.html): application/octet-stream, 665 bytes
Edgar Fuß | 14 Feb 19:08 2014

phsysical.basedir vs. physical.doc_root

I have a Lua script that modifies lighty.env["physical.doc-root"], yet, in the FastCGI handler, the
value of the DOCUMENT_ROOT environment variable doesn't reflect that change.
I found out that, in mod_fastcgi (and mod_cgi and mod_scgi), DOCUMENT_ROOT is set from physical.basedir,
while in mod_ssi, it's set from physical.coc_root.

So, what's the difference supposed to be between physical.doc_root and physical.basedir? The choice of
which of both to pick looks erratic to me, but that's probably due to my lack of understanding.

In response.c, both are set to the same value, but from there on, their use disgresses in a way not obvious to
me. mod_alias modifies basedir in a way I don't understand (it looks erroneous to me); mod_simple_vhost,
mod_indexfile and others operate on doc_root as I would expect.

Can someone please enlighten me?

Renata Santos | 30 Aug 20:24 2013

Agora voce pode ter o que as estrelas tambem tem!!

Conheca agora o Vibrado Rabbit. 
Ele possui uma combinacao perfeita de movimentos para lhe proporcinar prazeres nunca antes sentidos!!!

Prazeres Simultaneos.

A vibracao e intensidade do coelhinho estimulador clitoriano ampliam em dezenas de vezes o prazer.
Você vai querer sempre mais!

Promocao Especial de Lancamento: Compre agora o seu Vibrador e  receba 2 Presentes maravilhosos..

Somente enquanto durar o estoque.

Saiba mais sobre o Rabbit e compre com total discricao acessando o site


Renata Santos | 30 Aug 05:33 2013

Ja ouviu falar em Perfume com Feromonios??

Ja ouviu falar em feromonios? e sobre seducao pelo cheiro?

O que voce faria se pudesse seduzir alguem somente pelo seu cheiro, sem precisar falar nada?

Com o avanco da ciencia de hoje, isso agora e possivel. Veja como:

Cientistas ha muito tempo descobriram que certos hormonios, denominados feromonios, provocam
instantaneamente um grande apetite sexual nos animais.
"Essa e a razao que, por exemplo, os caes ficam excitados na presenca de uma cadela no cio".

Onde queremos chegar? O corpo humano tambem produz esse hormonio que influencia em nivel subconsciente, o
mesmo e recebido pelo sexo oposto atraves de um pequeno orgao no nariz chamado VNO. Infeslimente devido a
evolucao do homem, nosso corpo passou a nao produzir a quantidade de feromonios necessaria para resultar
algum efeito.

A novidade: Conseguiram desenvolver um perfume a base de feromonios sinteticos com o mesmo poder
afrodiziaco dos feromonios encontrados em nosso corpo.
Milhares de pessoas satisfeitas ja compraram mais de uma vez, os resultados sao otimos e comprovados!

Vale a pena conferir, a solucao da seducao e conquista foi reduzida a um unico perfume. Comprove!

Acesse nosso site:

David Narvaez | 10 Jan 18:08 2013

Code Coverage Mode


Is there any way to make Lighttpd run in "Code Coverage Mode" for PHP?
What I'm trying to achieve is a way to run, e.g., XDebug's Code
Coverage functionality in a non-intrusive way (without modifying the
PHP code to include code coverage  statements) once the server starts
and have a log of all lines and counts once the server shuts down (and
while the server is running, with a refresh interval).

I've read around and it looks like WebSphere + Rational Application
Developer provides such functionality, is there anything similar for

Thanks in advance.

David E. Narváez

Anthony L. Awtrey | 3 Aug 19:39 2012

Lighttpd plugins and debian packaging


Short version:

Can someone more-or-less officially provide guidance to distro
maintainers on how to build lighttpd plugins outside of the source tree?

Longer version:

I am migrating from Apache to Lighttpd and want to port some of the
custom plugins I've written. I found the official way to develop
plugins for lighttpd is by working in the source tree, copying the
mod_skeleton.c and hacking up the build to compile it together with
the lighttpd daemon.


This has some obvious drawbacks from my point of view. I run Debian
and if I maintain a custom lighttpd package version with my plugins,
then I have to update my lighttpd package manually every time a new
lighttpd release comes out that fixes bugs or provides security fixes.

The Debian maintainers don't want to provide a lighttpd-dev package
because they are convinced it deviates too much from upstream. When I
asked about it, I was told that:

"well, upstream would need to provide a set of headers, libraries and
tools which are explicitly meant to develop. This isn't the case for
Lighttpd. We only could pick some random headers and maybe the skeleton
example and package it as -dev packages. These may, or may not be stable
APIs and headers - we can't know."

Full thread here:


Based upon this other, 6 year-old lighttpd thread it appears to be
relatively straight-forward to build and maintain plugins outside the


Another point is that Ubuntu seems to deviate from Debian's lighttpd
package and *does* provide a lighttpd-dev package for this purpose, so
it appears to be at least possible.

Can someone more-or-less officially provide guidance to distro
maintainers on how to build plugins outside of the source tree?

Thanks in advance! Looking forward to using lighttpd more often in
production once I can get issue this resolved.


lighttpd + php_cgi = 100% cpu usage


I just installed a lighttpd server with fast cgi running media wiki on an AWS EC2 Instance.
The wiki is running fine while only one person is connected, if 3 simultaneous access occurs it's
impossible to use the Wiki.
2 instances of php_cgi consuming a total of 100% of cpu.
Using alone the application, each page access raises the cpu usage of the server to 35%.

I would like to know what could I do to have the server working properly.

Follow some information about my configuration:



php-cgi -v
PHP 5.3.10 (cgi-fcgi) (built: Feb  2 2012 23:45:39)
Copyright (c) 1997-2012 The PHP Group
Zend Engine v2.3.0, Copyright (c) 1998-2012 Zend Technologies


ighttpd -v
lighttpd/1.4.29 (ssl) - a light and fast webserver
Build-Date: Jul 16 2011 04:55:02


$SERVER["socket"] == ":80" {
        $HTTP["host"] =~ "example\.com\.br$" {
        url.redirect = ( "^/(.*)" => "https://example.com.br/$1" )
    server.name                 = "example.com.br"

$SERVER["socket"] == "example.com.br:443" {
ssl.engine = "enable"
ssl.pemfile = "/etc/lighttpd/ssl/example.com.br/example.com.br.pem"
ssl.ca-file = "/etc/lighttpd/ssl/example.com.br/root.crt"
server.name = "example.com.br"

server.document-root = "/var/www/lighttpd/example.com.br"
server.errorlog = "/var/log/lighttpd/wiki/error.log"
accesslog.filename = "/var/log/lighttpd/wiki/access.log"
server.error-handler-404 = "/e404.php"

url.rewrite-if-not-file = (
    "^/wiki/(mw-)?config/?" => "$0",
    "^/wiki/([^?]*)(?:\?(.*))?" => "/wiki/index.php?title=$1&$2",
    "^/wiki/([^?]*)" => "/wiki/index.php?title=$1"

fastcgi.server = ( ".php" =>
                   ( "php-local" =>
                       "socket" => socket_dir + "/php-fastcgi-1.socket",
                       "bin-path" => "/usr/bin/php-cgi",

                       "broken-scriptfilename" => "enable",

The rest is the default.

I even tried some options inside fastcgi.server vector, found in google, but didn't change anything:
                       "max-procs" => 1,
                                   "max-procs"         => 4,
                                   "min-procs"         => 1,
                                   "max-load-per-proc" => 10,
                                   "idle-timeout"      => 20,
                                   "bin-environment" => (
                                                "PHP_FCGI_CHILDREN" => "2",
                                                "PHP_FCGI_MAX_REQUESTS" => "500"
                                   "bin-copy-environment" => (""),

Thank you in advance.



Davide Baldini | 28 Feb 05:52 2012

CGI - carriage-return for new lines?

I'm writing Bash script to process a stdin caming from lighttpd's CGI. I
noticed that the caming lines carry a trailing carriage-return for
breaking lines, instead of the expected line-feed, and obviously bash
fails to recognize new lines.

The MIME type is "multipart/form-data";
lighttpd -v
lighttpd/1.4.28 (ssl) - a light and fast webserver
Build-Date: Dec 20 2011 22:59:43

Christian Kujau | 15 Dec 12:12 2011

ssl.ciphers vs. ssl.cipher-list


the config example added in r2814 (for the lighttpd-1.4.x branch) 
mentioned that "ssl.ciphers" could be used to specify allowed ciphers. But 
when configured, error.log says:

 2011-12-15 10:03:32: (server.c.968) WARNING: unknown config-key: 
 ssl.ciphers (ignored)

The documented[0] "ssl.cipher-list" works though. Is "ssl.ciphers" the 
new "ssl.cipher-list", even in lighttpd-1.4 and soon to be supported? Or 
is the config example wrong?


[0] http://redmine.lighttpd.net/wiki/1/Docs:SSL

BOFH excuse #317:

Internet exceeded Luser level, please wait until a luser logs off before attempting to log back on.