@Joe - You are jumping ahead --- the point of appropriate abstractions is to specify what _kind of_ thing you are doinng, so in the future, when new uses arise, users do not need to adjust (e.g. site writers) - only the underlying implementation need change.
@Massimo - while you "can do this" using integer, it misses the point of appropriate abstraction. It is not at all about what "web2py" knows; it is what the programmers using the framework know and read. (integers are implementation specific, and as an abstraction for this use are useless - integers are useful for MANY _kinds of_ things, and for this use, their use if obtuse in that it masks what intent and use is). If you want web2py to "look" (to programmers) like it supports generic references (it does) then the language should represent that. Otherwise, you are relegated to implementation details and questions of "how do I do _this_ _kind of thing_ in this framework - you do not want to be replicating implementation details at every programmer - whenever possible, it should be self-evident; this is one of thse cases where that improvement could be made. - Yarko On Sun, Sep 6, 2009 at 5:43 PM, mdipierro <[email protected]> wrote: > > There are oranges and there are apples. > > oranges are arbitrary foreign keys. The problem here is not > implementing them in the DAL. this could be done in a couple of lines > of code. The problem is 1) is this good practice? I think no. 2) how > should SQLFORMs handle this? > > apples are "polymorphic references". You can do this using "integer". > I do not see the need for defining another keyword unless you want to > make web2py aware that this integer contains a polymorphic reference. > Assuming this were possible (and it is not in an easy way), what would > you like the DAL to do with it? > > Massimo > > On Sep 6, 5:02 pm, Joe Barnhart <[email protected]> wrote: > > OK, I'm confused. Are you suggesting that "reference generic" should be > an > > alias for "integer" or that it should be smarter and permit UUIDs and > other > > structures? Is this a beginning of a "foreign key" concept for web2py? > > That might have some pretty extensive usefulness... > > > > On Sun, Sep 6, 2009 at 12:49 PM, Yarko Tymciurak <[email protected]> > wrote: > > > > > On Sun, Sep 6, 2009 at 2:45 PM, Joe Barnhart <[email protected] > >wrote: > > > > >> I am not against sugary goodness in general, but this seems like > something > > >> that might be better handled with education rather than overlaying the > name > > >> "integer" with "reference genric". (Besides, then what would you do > with > > >> people who wanted to use UUIDs for their "reference generics" instead > of > > >> integers?) > > > > > ... that is my point - why educate on integers, when that's not good > > > education; hide the implementation into a logical abstraction (generic > > > reference term just being borrowed from Matthew's post). > > > > > Your point about UUIDs sounds more like an agrument _against_ teaching > > > about integers than anything else, to me at least... > > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "web2py-users" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/web2py?hl=en -~----------~----~----~----~------~----~------~--~---
