SourceForge : Post

Project: Editor Discussion > REC: CDDLM SmartFrog-Based Lan > SF-CDL spec for community practice track > List of Posts

Forum Topic - SF-CDL spec for community practice track: (3 Items)

View: as

Update

Hiro Kishimoto

08/22/2004 3:13 PM

post4435

SF-CDL spec for community practice track

Jun's presentation at GGF11 (page 15 of http://tinyurl.com/49wkv )said that the purpose of XML-CDL is composability and 
interoperability and the minutes underpins the characteristics of the two CDLs as follows;

> Note that the SF CDL is intended to be more user-friendly /
> user-tractable, while the XML CDL is more standardised, is the interop
> format (i.e. we can translate from various front ends into the XML
> form), but is less user-friendly to use directly.

The BizGrid system (platform) and other grid systems (platforms) may decide to adopt XML-CDL as their description 
language. And a CDL translator would enable these systems (platforms) to accept a description written in SF-CDL. (
Similarly they could choose to implement other proprietary languages and use XML-CDL for interoperability.) So I am not 
sure how much interest there would be in implementing SF-CDL. Keeping in mind that Proposed Recommendations need two 
interoperable implementations within 2 years to become Recommendations I am doubtful that SF-CDL would make it to the 
Recommendation stage.

So if I understand the roles of the two CDLs correctly, I think SF-CDL specification should be "community practice" 
track instead of "recommendation" track. Community Practice Documents (GFD-C), inform and influence the community 
regarding an approach or process that is considered to be widely accepted by consensus and practice in the Grid 
community. I think in this sense SF-CDL is excellent evidence and will guide XML-CDL and related platforms.

Hiro Kishimoto

08/22/2004 3:15 PM

post4436

Dejan's reply to Hiro's comment

1. SF is not proprietary because it is open source and therefore accessible to anyone to use it.

2. I am not sure that there is a definite answer, but I put a lot of importance on the format in which the configuration
 is specified, not just an intermediary XML format which is better suited for the computers to read it. That is SF 
offers semantics in addition to syntax offered by XML.

3. I recommend we take this to the whole CDDLM group because there is much more wisdom among the whole group than only 
me.

Hiro Kishimoto

08/22/2004 3:17 PM

post4437

My additional comments to Dejan

1. I agree that SF is open source. However open source and standard are two different things. (e.g. Globus Toolkit is 
also open source and is based on several standard specs but GT itself is not standard.)

2. SF-CDL is no doubt very good practice. Its language semantics give human readability. I still believe SF-CDL should 
be community practice track.

Return