John Wang wrote:
[...]
I think it would be a useful addition since I imagine most people want it to
be autostarted and init.d tends to be more popular than daemontools.
Sure, but it's not an either/or choice. You can have SysV startup and
process supervision - that's how we do it in the SME Se
I was reading the wiki which says there's no good central repository for
plugins. Is there a reason why CPAN can't be used for qpsmtpd and its
plugins? Catalyst, CGI::Application and POE all have numerous plugins /
components on CPAN. SpamAssassin also uses CPAN. Just curious.
John
On Mon, 3 Apr 2006, John Wang wrote:
Per another thread, I'm hoping there will be more support for init.d/rc
script distribution so these DJB licensing issues can be avoided altogether.
If the license conditions for daemontools and ucspi-tcp bother you too
much, just use runit and ipsvd. In
John Wang wrote:
>
> Feel free to add to, modify, or remove from the wiki since you seem to
> really care about DJB's exact syntax and unreferenced / incompatible
> licensing requirements. In any event, the current instructions work and I've
> decided stay away, far away, from DJB code after all o
John Wang wrote:
[...]
Feel free to add to, modify, or remove from the wiki since you seem to
really care about DJB's exact syntax and unreferenced / incompatible
licensing requirements.
] [Charlie]
] It's not worth trying to find a logically consistent view of DJB's
license conditions.
] I'm
On 4/2/06, Robin Bowes <[EMAIL PROTECTED]> wrote:
>
> There are several ways to fix the problem - which one you choose depends
> on your "religious" leanings :)
I didn't fully appreciate the word "religious" earlier. I'm starting to
catch on :)
John
On 4/3/06, Charlie Brady wrote:
>
> It's there in the djbdns install doco, but hasn't yet been added to the
> daemontools (or ucspi-tcp, qmail, publicfile etc) install instructions:
>
> http://cr.yp.to/djbdns/install.html
>
> So if you are ultra cautious, you may feel there is no approved
> workaro
On Mon, 3 Apr 2006, John Wang wrote:
On 4/3/06, Charlie Brady wrote:
Because DJB tells us to tweak conf-cc, but says we may not modify source
code.
Is this mentioned somewhere on his website? If it's on his website and
referenceable, I'll add it to the wiki.
It's there in the djbdns insta
On 4/3/06, Charlie Brady wrote:
>
> Because DJB tells us to tweak conf-cc, but says we may not modify source
> code.
Is this mentioned somewhere on his website? If it's on his website and
referenceable, I'll add it to the wiki.
Just FYI, I put a second perl edit in the wiki that modifies conf-cc
On Apr 3, 2006, at 2:11 AM, John Wang wrote:
It's about 3 times faster than the qpsmtpd-forkserver that
ships with qpsmtpd.
Has anything significant changed with Apache::Qpsmtpd or forkserver
since
then? Was that test valid? According to Matt's 15 Sep 2005 O'Reilly
article,
qpsmtpd is be
On Mon, 3 Apr 2006, John Wang wrote:
I haven't seen any benchmarks but there's some anecdotal information. Back
on 19 Sep 2004, Matt posted on the mod_perl list re Apache::Qpsmtpd (
http://www.issociate.de/board/index.php?t=msg&th=79422&rid=0):
It's about 3 times faster than the qpsmtpd-forks
On Mon, 3 Apr 2006, John Wang wrote:
Granted, this only matters if you want to distribute DJB software, but I'm
curious why you think patching conf-cc is okay but error.h isn't?
Because DJB tells us to tweak conf-cc, but says we may not modify source
code.
I would assume the above prevent
On 4/3/06, Peter J. Holzer wrote:
>
> On 2006-04-03 03:35:17 -0700, John Wang wrote:
> > Would it make sense to include these (tailored to the tarball layout)
> > in the tarball distribution in addition to the .rpm and .deb packages?
> I think these would belong into the contrib directory we keep
On 4/3/06 4:11 AM, "John Wang" <[EMAIL PROTECTED]> wrote:
> Any other performance related information, anecdotal or otherwise?
I'll run metrics, raw or cooked. What's a good way to gather data?
peter
On 2006-04-03 03:35:17 -0700, John Wang wrote:
> Would it make sense to include these (tailored to the tarball layout) in the
> tarball distribution in addition to the .rpm and .deb packages? Some other
> tarball distributions come with init.d/rc scripts which I've found useful,
> e.g. lighttpd (RH
Would it make sense to include these (tailored to the tarball layout) in the
tarball distribution in addition to the .rpm and .deb packages? Some other
tarball distributions come with init.d/rc scripts which I've found useful,
e.g. lighttpd (RH, SuSE) and memcached (Debian).
John
On 2006-04-02 12:55:08 -0700, John Wang wrote:
> I did a brief update of several pages, primarily due to the install issues I
> mention on the tcpserver page:
Thanks.
> http://wiki.qpsmtpd.org/inetd
You seem to have mixed up inetd and init.d. I moved your comments from
http://wiki.qpsmtpd.org/in
On 4/2/06, Charlie Brady wrote:
>
> On Sun, 2 Apr 2006, Matt Sergeant wrote:
> > b) it should be significantly faster since there is no fork overhead
> > with Apache.
> It would be safer to benchmark than to trust speculation.
>
Are there any published benchmarks to comparing the various methods o
On 4/2/06, Charlie Brady wrote:
>
> Modifying conf-cc is the only way which is guaranteed to comply with DJB's
> license conditions. The perl edit script maybe considered as a source
> code patch, and binaries built from patched source code may not be
> distributed.
I put a distribution section o
19 matches
Mail list logo
