CIM Provider HOWTO Document

Stephen Gallagher sgallagh at redhat.com
Fri Apr 5 12:57:03 UTC 2013


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 04/04/2013 05:15 PM, John Dennis wrote:
> When I started development of the realmd CIM provider I was told I 
> was being a guinea pig. Take a developer who knows nothing about 
> CIM or OpenLMI and see what issues a novice might encounter when 
> asked to write a CIM provider. Then use that experience to write
> up documentation to pave the way for other developers to follow.
> 
> When I first started a few months ago the material on the OpenLMI 
> wiki was sparse. I see that it has since been enhanced.
> 
> Attached is a 24 page PDF. It is what I would have liked to have 
> started with as a developer.
> 
> The content is written in pandoc format, a universal format that 
> can be formatted into HTML, PDF, mediawiki, reStructuredText, and 
> many other document formats. As such this content can be put up on 
> the web easily. I'm distributing this initial draft as a PDF 
> because in my opinion the PDF formatting yields the most readable 
> presentation. The document contains many clickable links to other 
> material.
> 
> Those of you in the OpenLMI group will no doubt have corrections 
> and suggestions on the technical content which I welcome, I'm sure 
> as a neophyte I've made some errors along the way. Others may want 
> to add their own suggestions and comments.
> 
> It's been an interesting ride. I hope this serves as a good 
> foundation to build upon.
> 
> John
> 


Thank you, John. You've obviously put a lot of time and effort into
this document, and it shows. I hope you will forgive my nitpicks :)

Here are my first thoughts (I haven't read through it all yet).


Issues I see:

In the "WBEM Components" on page three, you describe "only one
protocol is in wide use, CIM Operations over HTTP". This is not
actually strictly true. There are a number of companies (Dell,
Microsoft) using the WS-MAN protocol instead. We're currently
investigating whether it makes sense for the OpenLMI project to offer
both protocols, but at the moment we've chosen WBEM because we believe
it to be technically superior due to its introspection capabilities.

First line of page 6: s/breath/breadth/

Last paragraph of page 6:
s/what KonkretCMPI gives to is is incomplete/what KonkretCMPI give you
is incomplete/

When introducing the two CIMOMs, it might be useful to note their
respective licenses (MIT for OpenPegasus and MPL for SFCB). This can
be a deciding factor in selecting which one to use or which license to
use for your own provider.


Formatting:

Using boldface fonts to denote high-importance content isn't clear
enough, since it's too close to the section header appearance. Perhaps
italics (or boldfaced italics) would be more readable? Case in point,
under "CIM Schema and MOF Syntax", the boldfaced note at the end looks
exactly like the header of the "Models and Profiles" section.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.13 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

iEYEARECAAYFAlFeyh8ACgkQeiVVYja6o6OkrQCcDYSOq7upd84ykeZIt10cRWIc
+igAnikYqKXBEXoClJRfX1dDCrJNQPgz
=/Tl6
-----END PGP SIGNATURE-----


More information about the openlmi-devel mailing list