supervision-scripts 2015-06

From: Avery Payne <>
Date: Wed, 08 Jul 2015 18:07:43 -0700

I wish I could say I have new definitions but work commitments are just
now slowing down. I do have some vacation time in the near future, and
perhaps a few new definitions will emerge from that. Right now, some
housekeeping and documentation is being addressed.

  - - - -
+ A small comparison table has been added to the wiki, showing 6
different frameworks and their various "related concepts". It is
incomplete but somewhat usable, and contains various links to sites,
downloads, etc. Because there is still a lot of missing information,
any additional hints, comments, or links sent to me are gladly
accepted. It can be viewed at

+ A new document has been started outlining the thought process behind
the use of template scripts, and what decisions have been made. It is
incomplete as well but provides enough information to understand most of
the structure of the project. I am ashamed that it can be viewed at ; I
ask you hold critical comments until it is sufficiently complete. For
instance, the sequence in which concepts are introduced needs to be
revised, as there is a bit of "but first we jump over here to discuss
this, then come back later".

+ Bugfix: for reasons I haven't delved into, mingetty-tty* wouldn't
launch. The addition of pgrphack to each definition resolved the issue.

In Progress:
  - - - - - - -
+ Renaming /sv to /svcdef to better reflect the meaning of the "service
definitions" it contains (here service is loosely used as a meaning for
"daemon"). This is because some installations will be needing to copy
definitions into place, rather than linking directly to them. The new
name should be fairly agnostic to any existing frameworks or code, and
removes a potential naming conflict with runit's use of /etc/sv as a
directory name.

+ Adding support for x2go

To Do:
  - - - -
+ Preliminary support for versioned daemon options, as discussed on the
mailing list. I think I have settled on a format that is portable,
lightweight, self-contained, and still supports "template switching" as
a feature. This will solve a few issues I've encountered with porting
between environments (Environment A uses version 1.23 but Environment B
uses older version 1.2 which is missing the newer option --foobar, you
can't have it both ways in one script.)

+ Addition of support for set-up/tear-down of a named socket. There are
several definitions that are held up for only this reason. Example:
clamd needs a socket that must exist prior to launch, or it will fail.
Separate UID/GID environment variables will be used to convey ownership,
with "inheritance" from the existing state directory UID/GID as a default.

+ Slight restructure to remove non-critical code from the common
template script. Specifically, peer-based dependency launching; yes,
this idea was based on prior discussion in the mailing list, Thank You,
you know who you are. :) This will make "melding" with other
environments much easier and cleaner.

+ Everything else from last month. Ugh.
Received on Thu Jul 09 2015 - 01:07:43 UTC

This archive was generated by hypermail 2.3.0 : Sun May 09 2021 - 19:44:19 UTC