On Tue, 2005-05-03 at 17:03, Wes Hardaker wrote: > Dave> /* > Dave> * NOTE NOTE NOTE: This helper isn't complete and is likely to > Dave> * change somewhat over time. Specifically, the way it stores > Dave> * data internally may change drastically. > Dave> */ > > Dave> I'm not > Dave> sure how reasonable it is to enforce this warning anyway. > > I'd argue we probably shouldn't enforce it... That'd be like changing > the AgentX internals and API because you've never taken it out of BETA
I don't think that's a fair comparison. It's more like the arguments about whether to change a MIB file as it progresses from Proposed to Draft to Internet Standard. (Or it'd be more like such arguments if anyone could actually be bothered to talk to me, but that's a different issue) Changing the AgentX internals would be fine, IMO. Changing the AgentX API wouldn't. In this case, I think we _could_ justify changing things (particularly since there's this warning that we might do so), but I can understand why you're reluctant to do so. But if we're not prepared to change it, then we should remove the comment that says we might! > Dave> Which brings us back to the earlier question - what API calls > Dave> would be necessary for clean access to a dataset-style helper? > ....the row extraction one is only easily useful in a request > environment. Being able to extract rows based on a set of index > values would be useful, but more of a pain to implement. I've been thinking about this over the last few days - both in relation to this, and my discussions with Robert over the table_container. I suspect that part of the problem is that we've never actually come to a clear view of what an (abstract) table API needs to cover. Most of the table-related helpers are concerned with handling the collection of rows that make up the table. The column values within a single row are typically outside the scope of these helpers (apart from the dataset and similar) I've come up with the following list of required tasks facing such a helper. What else (or different) would it need to provide? SNMP-request handling: - register (and unregister) the table - add or remove a row to/from the table - retrieve the row relating to a specific varbind request (this may possibly need to retrieve both the helper's representation of the row, and the data relating to it, if these are different) - retrieve the indexes relating to this row (?) External handling - retrieve a particular row (specified by index) - walk through the rows of the table (order not important) I *think* that should cover all requirements. What have I missed? Dave ------------------------------------------------------- This SF.Net email is sponsored by: NEC IT Guy Games. Get your fingers limbered up and give it your best shot. 4 great events, 4 opportunities to win big! Highest score wins.NEC IT Guy Games. Play to win an NEC 61 plasma display. Visit http://www.necitguy.com/?r=20 _______________________________________________ Net-snmp-coders mailing list Net-snmp-coders@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/net-snmp-coders