SourceMage Grimoire Bug List

[bugzilla-daemon AT bugs.sourcemage.org]

http://bugs.sourcemage.org/show_bug.cgi?id=9636

           Summary: Master bug - all init scripts should have a consistent
           Product: Codex
           Version: unspecified
          Platform: Other
        OS/Version: other
            Status: NEW
          Severity: normal
          Priority: P2
         Component: Unknown
        AssignedTo: sm-grimoire-bugs AT lists.ibiblio.org
        ReportedBy: ruskie AT email.si


I was thinking for a while this and I can't get over the fact that some of 
our init scritps use smgl-somename and some just somename.

I think a good suggestion here would be if we used the following naming 
scheme for all init scripts:

sm-servicename - this could be used for generic services that can't run more 
than
1 instance at the time on the same box(e.g. gdm, slim, entrance)

sm-daemonname - this can be used for stuff that can be run side-by-side.

I wouldn't mind if everything would use a sm-daemonname though... since that 
would
kinda clear it up.


------- Additional Comments From jblosser-smgl AT firinn.org  2005-08-28 21:20 
-------
Why would we add a prefix to an init script name?  The only effect I can see
that this has is making it more difficult for admins to find an init script,
since the name is guaranteed to not match what any other distribution has.

We have a standards doc that says that we use standard names for anything 
found
in  http://www.lanana.org/lsbreg/init/init.txt, and prefix any others with
smgl-, but frankly don't even understand why we add the smgl- to any of them. 
Even if the scripts are home grown the services/daemons aren't, and from a
sysadmin perspective I don't care who wrote the init script, if I'm looking 
for
it I'm looking for it by the name of the service or daemon it provides. 
LANANA's argument is that all init scripts should use names from that list to
avoice conflicts, and that's somewhat valid, but I don't think the smgl- 
prefix
buys us anything there that we wouldn't get just being careful to avoid
conflicts.  Two spells with an smgl-foo script are going to conflict in any 
case.

-- 


------- Additional Comments From ruskie AT email.si  2005-08-29 05:39 -------
Well I just want uniform naming not some spells smgl- some without. But hey 
it was an idea only.

-- 


------- Additional Comments From dkowis+SMbugzilla AT shlrm.org  2005-08-29 
05:59 -------
I'm for renaming these things, smgl-metalog has always been annoying to me. 
I'd
like just metalog or whatever.

I also agree that prefixes are silly, for the same reasons already mentioned.

-- 


------- Additional Comments From sergey AT sourcemage.org  2005-08-29 12:38 
-------
I agree, "smgl-" is very confusing and unneeded.

-- 


------- Additional Comments From hgr AT vabo.cz  2005-08-29 12:43 -------
Yea, we should respect standards, no sm-


-- 


------- Additional Comments From v.merkatz AT gmx.net  2005-09-03 17:40 -------
If we want to respect standards we need the smgl- prefix. That's how lanana
defines it. But I agree that it doesn't buy us much as this is mostly done so
third party install packages don't conflict with distro-provided init scripts,
and I don't think that many third party packages with init scripts are 
installed
by smgl users (vmware is the only one I can think of right now).

-- 


------- Additional Comments From jblosser-smgl AT firinn.org  2005-09-16 06:15 
-------
So are we calling this INVALID?  Or deciding to go ahead and drop the smgl-
prefixes?

-- 


------- Additional Comments From dkowis+SMbugzilla AT shlrm.org  2005-09-16 
06:48 -------
I don't like the smgl-* naming scheme

my vote is for dropping smgl-* and just having it be the name of the daemon
(like bind:named) , or if it starts multiple things, the name of the spell 
(like
samba starts two things, I don't recall what they are at the moment)

-- 


------- Additional Comments From jblosser-smgl AT firinn.org  2005-09-16 21:42 
-------
samba starts smbd and nmbd, which are often separate init scripts (there are
reasons to run one and not the other).

I think the consensus is we don't like the prefixes, but do we need to keep 
them
anyone for standards-compliance?

-- 


------- Additional Comments From jblosser-smgl AT firinn.org  2005-10-26 21:57 
-------
This is still open...

I'm going to formally suggest we decide to add this to the list of LSB-derived
standards that we don't adhere to.  I don't see how it makes things any easier
for  users when they have to know for each init script if it's smgl- prefixed 
or
not.  Having to know prefixes across distros doesn't help them either.  The
names these scripts want to install are de-facto standard in most cases 
whether
they're on lanana's list or not, and we can protect against conflicting names
regardless.

Is anyone strongly opposed to this suggestion / strongly in favor of the 
lanana
scheme?

-- 
Configure bugmail: http://bugs.sourcemage.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.