[OpenAFS] Re: [OpenAFS-devel] Selecting a configuration file
format for OpenAFS Services
Michael Joyner ᏩᏯ
mjoyner@vbservices.net
Sat, 16 May 2009 09:17:50 -0400
This is an OpenPGP/MIME signed message (RFC 2440 and 3156)
--------------enigDB5D576820E782848F83B8B9
Content-Type: multipart/alternative;
boundary="------------060204000602030401040204"
This is a multi-part message in MIME format.
--------------060204000602030401040204
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Whatever format is chosen, *COMMON* arguments should be able to be
specified in a *COMMON* section, then overridden on a per daemon basis.
I was going nuts with the dots in principals thing, had to grep all the
binaries to figure out which ones needed it turned off in, had to find
the config file and add the args... And it doesn't help any when the man
pages don't have the correct spellings for available options either. :-P
FYI, not everyone who uses AFS is knowledgable of the KRB config file
format. Some people use W2K3/W2K8 for our KRB servers.
And speaking of feature requests... is there a way to add "hooks" into
the pts security database lookups?
Derrick Brashear wrote:
> On Sat, May 16, 2009 at 9:07 AM, <u+openafsdev-t07O@chalmers.se> wrote=
:
> =20
>> Hello Derrick,
>>
>> On Sat, May 16, 2009 at 08:43:40AM -0400, Derrick Brashear wrote:
>> =20
>>>> In an extreme case, if you would not be able to express some configu=
ration
>>>> choice without an elaborated syntax, then you may define an appropri=
ate
>>>> syntax and still place this data into a command line argument.
>>>> =20
>>> In some way which is arcane and not legible to the user? Why??
>>> =20
>> I did never suggest using an arcane and unintelligible way.
>> =20
>
> I had already composed a reply earlier, and junked it because I
> thought Simon made the point better, but...
>
> Consider the case which pushed us here. You have an array of numbers.
> It's a set of timeout values, which apply when you have between a
> certain number of callers.
>
> You can, of course, just pass an argument which is a comma separated
> list of timeouts. What do they mean? Well, you don't need to pass
> those numbers, so, well, just make sure you order things right.
>
> But wait. What happens if you want to also make the ranges the
> timeouts apply in tunable? Do you now pass 2 lists, one for the bucket
> ranges and one for the timeouts, or a set of pairs? How do you
> delineate the pairs?
>
> The work to allow passing arguments on the command line rapidly
> devolves the syntax of those arguments to absurd.
> _______________________________________________
> OpenAFS-info mailing list
> OpenAFS-info@openafs.org
> https://lists.openafs.org/mailman/listinfo/openafs-info
> =20
--=20
LyX: http://www.lyx.org/ OpenOffice: http://www.openoffice.org/
Inkscape: http://www.inkscape.org/ Scribus: http://www.scribus.net/
GIMP: http://www.gimp.org/ PDF: http://www.pdfforge.org/
--------------060204000602030401040204
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content=3D"text/html;charset=3DUTF-8" http-equiv=3D"Content-Type"=
>
</head>
<body bgcolor=3D"#ffffff" text=3D"#000000">
Whatever format is chosen, *COMMON* arguments should be able to be
specified in a *COMMON* section, then overridden on a per daemon basis.<b=
r>
<br>
I was going nuts with the dots in principals thing, had to grep all the
binaries to figure out which ones needed it turned off in, had to find
the config file and add the args... And it doesn't help any when the
man pages don't have the correct spellings for available options
either. <span class=3D"moz-smiley-s4"><span> :-P </span></span><br>
<br>
FYI, not everyone who uses AFS is knowledgable of the KRB config file
format. Some people use W2K3/W2K8 for our KRB servers.<br>
<br>
And speaking of feature requests... is there a way to add "hooks" into
the pts security database lookups?<br>
<br>
Derrick Brashear wrote:
<blockquote
cite=3D"mid:db6e3f110905160610i1c928753sbdd1509967d83915@mail.gmail.com"=
type=3D"cite">
<pre wrap=3D"">On Sat, May 16, 2009 at 9:07 AM, <a class=3D"moz-txt-li=
nk-rfc2396E" href=3D"mailto:u+openafsdev-t07O@chalmers.se"><u+openafsd=
ev-t07O@chalmers.se></a> wrote:
</pre>
<blockquote type=3D"cite">
<pre wrap=3D"">Hello Derrick,
On Sat, May 16, 2009 at 08:43:40AM -0400, Derrick Brashear wrote:
</pre>
<blockquote type=3D"cite">
<blockquote type=3D"cite">
<pre wrap=3D"">In an extreme case, if you would not be able to ex=
press some configuration
choice without an elaborated syntax, then you may define an appropriate
syntax and still place this data into a command line argument.
</pre>
</blockquote>
<pre wrap=3D"">In some way which is arcane and not legible to the u=
ser? Why??
</pre>
</blockquote>
<pre wrap=3D"">I did never suggest using an arcane and unintelligible=
way.
</pre>
</blockquote>
<pre wrap=3D""><!---->
I had already composed a reply earlier, and junked it because I
thought Simon made the point better, but...
Consider the case which pushed us here. You have an array of numbers.
It's a set of timeout values, which apply when you have between a
certain number of callers.
You can, of course, just pass an argument which is a comma separated
list of timeouts. What do they mean? Well, you don't need to pass
those numbers, so, well, just make sure you order things right.
But wait. What happens if you want to also make the ranges the
timeouts apply in tunable? Do you now pass 2 lists, one for the bucket
ranges and one for the timeouts, or a set of pairs? How do you
delineate the pairs?
The work to allow passing arguments on the command line rapidly
devolves the syntax of those arguments to absurd.
_______________________________________________
OpenAFS-info mailing list
<a class=3D"moz-txt-link-abbreviated" href=3D"mailto:OpenAFS-info@openafs=
=2Eorg">OpenAFS-info@openafs.org</a>
<a class=3D"moz-txt-link-freetext" href=3D"https://lists.openafs.org/mail=
man/listinfo/openafs-info">https://lists.openafs.org/mailman/listinfo/ope=
nafs-info</a>
</pre>
</blockquote>
<br>
<br>
<pre class=3D"moz-signature" cols=3D"72">--=20
LyX: <a class=3D"moz-txt-link-freetext" href=3D"http://www.lyx.org/">http=
://www.lyx.org/</a> OpenOffice: <a class=3D"moz-txt-link-freetext" href=3D=
"http://www.openoffice.org/">http://www.openoffice.org/</a>
Inkscape: <a class=3D"moz-txt-link-freetext" href=3D"http://www.inkscape.=
org/">http://www.inkscape.org/</a> Scribus: <a class=3D"moz-txt-link-free=
text" href=3D"http://www.scribus.net/">http://www.scribus.net/</a>
GIMP: <a class=3D"moz-txt-link-freetext" href=3D"http://www.gimp.org/">ht=
tp://www.gimp.org/</a> PDF: <a class=3D"moz-txt-link-freetext" href=3D"ht=
tp://www.pdfforge.org/">http://www.pdfforge.org/</a>
</pre>
</body>
</html>
--------------060204000602030401040204--
--------------enigDB5D576820E782848F83B8B9
Content-Type: application/pgp-signature; name="signature.asc"
Content-Description: OpenPGP digital signature
Content-Disposition: attachment; filename="signature.asc"
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iEYEARECAAYFAkoOvP4ACgkQtX52dpJWQ8eynACgtvG8Tzopu3V2Pwhkdy1NjAZU
E80AnRS4B7Bxg2LWxR+91lYg+4lu947n
=Ue+s
-----END PGP SIGNATURE-----
--------------enigDB5D576820E782848F83B8B9--