From: Eric J. <er...@su...> - 2004-12-02 22:30:44
|
Greetings, We've set up a Roundup tracker with a mysql backend and everything is working fine. As this is beginning to catch on, our supervisor wants me to separate issues into different "bins"; meaning (mostly): certain users only see certain issues (all issues, regardless of bin, have the same schema) I'm reluctant to set up separate trackers, for all the reasons that you can probably guess. Mostly because it balkanizes our database unecessarily (they will all have the same schema anyway). The logical alternative is to keep everything in the same database. Browsing the documentation and the mailing list archives, it seems that there are two main ways I could go about this: 1) I could either keep all issues in the same issues class and just add a "category" property; I would then add different Roles, assign them and update the default queries to enforce the use of the appropriate value in the issues query. or 2) I could make additional variants of the issues class (as described in the documentation) and separate the issues that way. My question (finally!) is: what are the main drawbacks to #2? I'm concerned about being able to generate single queries on ALL issues regardless of category; i.e. we want to be able to query all the issues tables with one query to pick up issues that are of interest. Clearly this works with method #1, but its not immediately clear how easy that would be in method #2. Finally, if I did go with method #1, how easy would it be to maintain separate keyword tables for each category of issue? Thanks for your consideration. We're Python fans here and this is one of the nicest python apps I've seen. Extremely clean design, nicely coded, good abstraction everywhere and great documentation! -- Eric Jeschke Software Engineer Subaru Telescope/National Astronomical Observatory of Japan er...@na... http://www.naoj.org/staff/eric |