Subject: | Re: Need Advice on Secondary Indices
| Date: | Thu, 19 Apr 2018 09:41:20 -0700
| From: | Peter <peter@whiteknight.email>
| Newsgroups: | pnews.paradox-programming
|
Setrange and setGenFilters, yes this is the method I have employed for a
few years, works very well.
In this application the users will need to see the data in many
different ways. There seems to be about 3 fields to use in setrange, but
there are another 3-4 fields that need to be included in the SI for
alternative sorts.
Peter
On 04/18/2018 09:25 PM, Tom Krieg wrote:
> The "many more" statement worries me. Too many 2'ary indexes and your
> tables will throw themselves on the corruption sword all by themselves.
> If, however, you have 3 or 4 major views, you could have 3 or 4 2'ary
> indexes (and do setranges on these) and then do a setGenFilter to refine
> your views as users enter preferences.
>
> Peter wrote:
>> I am making a kind of ticketing system. It will require several views
>> (setranges) such as viewing tickets assigned to certain individuals or
>> everybody, open/close tickets, by due date or priority and many more.
>>
>> The easiest thing is to just define all the secondary indices that are
>> required. The problem is that I worry about corruption.
>>
>> As an alternative I was thinking of having multiple tables with a few
>> SI each and keep them in sync. On the form the user would click the
>> view they want and the fields would be redefined to reflect the table
>> with the required SI. That sounds like a headache.
>>
>> I am interested in hearing about other possibilities. Thanks.
>>
>> Peter
|