<br><br><div class="gmail_quote">On Thu, Apr 28, 2011 at 7:31 PM, Alex Mizrahi <span dir="ltr"><<a href="mailto:killerstorm@newmail.ru" target="_blank">killerstorm@newmail.ru</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div>> Why? I am by no means an expert and never looked thoroughly the code,<br>
> but ,abstractly speaking, doesnt it boil down to just creating<br>
> more indexes automatically at class definition time?<br>
<br>
</div>Code assumes that there is only one index it needs to update at time, so it<br>
needs to be revised to handle many of them.<br>
Nothing particularly hard, but it requires some code rewriting and thus<br>
<br>
While another alternative is pretty much trivial.<br>
<div><br>
> Independently of the whole subject, I think this should be done anyway,<br>
> and it can be an optional param with a default for old behaviour for those<br>
> who want the whole pie.<br>
> I think many people subtypep the results of get-instance family funcs<br>
> early in their projects.<br></div></blockquote><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div>
</div>I don't think that 'old behaviour' was intentional, people should not depend<br>
on query for one sub-class returning instances of another sub-class.<br>
And it is trivial to fix it -- just pass super-class instead of sub-class to<br>
this function if you want all instances.<br>
<br>
So I think get-instances-by-... should unconditionally do the filtering.</blockquote><div><br></div><div>You are right then. </div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
</blockquote><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
While map-inverted-index probably shouldn't do the filtering. I consider it<br>
a lower-level API which works with index directly, without a promise to<br>
return instances of certain class.<br>
And also implementing it at that level is somewhat harder.<br>
<div><div></div><div><br></div></div></blockquote><div>Totally agree with that. From the index point of view, class hierarchy relationships are irrelevant and even non existent, since at that level the only distinguishable relationship is the key-value one. It would be wrong to inject such semantics at that level.</div>
<div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div><div>
<br>
<br>
_______________________________________________<br>
elephant-devel site list<br>
<a href="mailto:elephant-devel@common-lisp.net" target="_blank">elephant-devel@common-lisp.net</a><br>
<a href="http://common-lisp.net/mailman/listinfo/elephant-devel" target="_blank">http://common-lisp.net/mailman/listinfo/elephant-devel</a><br>
</div></div></blockquote></div><br>