Re: Latest on TMQL?

Dear Rani,

Thank you for your update!

Am 16.06.2011 um 09:41 schrieb Rani Pinchuk:

> However, later it turned out that we did not fully understand each other, and a main disagreement was uncovered (practically, Benjamin thinks that the path expression we developed should be *added* to the current draft, while I believe it should *replace* the path expression of the current draft).

Two path expression styles is one too many!

> Since Benjamin is unavailable in the current months, I am not sure how to proceed - an implementation of the suggested TMQL is not available (although we have most of the path expression implemented in house), and although the ideas are clear to me, they are not yet written in the draft standard.
> 
> Any suggestions, opinions are welcomed.

Here is one: 

1. Proceed alone and work with this community to get feedback.

2. Keep in mind who the target audience is and make it a useable technology for them. There is a majority of queries and that should be easy to throw down. For all other cases a tailorization mechanism should be present (like TMCL's user defined constraints).

3. Make it gel with TMDM and TMCL. Build upon the smartness of these technologies and live with they weaknesses.

4. Consider that the requirements may be faulty. If there is no agreement after 10 years of work, there will be no agreement after 20 years if you keep doing what you are doing. Committee and users need to come close together to define requirements that can be fulfilled.

5. Everybody should realize that the presence of a standardized Topic Maps Query Language is more important than an individual language feature.

Hope this helps!

Kind regards,

Robert