On Mon, 16 Nov 2009 21:49:58 -0800, Roger Binns <[email protected]> wrote:
>-----BEGIN PGP SIGNED MESSAGE----- >Hash: SHA1 > >Bogdan Ureche wrote: >> I just noticed the ticket was closed without being fixed, with the following >> remark: > >I was the one who closed it and added that remark. > >> Now, I have to admit that English is not my first language, nor the second, >> but the way I see it this remark means something like: "This is not a bug, >> and if you want to find out why please ask this question in the mailing >> list. " > >Not quite :-) The CVSTRAC tracker has been closed and the replacement >mechanism is described at http://www.sqlite.org/src/wiki?name=Bug+Reports > >In cases where it was obvious that the issue was a bug then I copied it over >to the new one http://www.sqlite.org/src/timeline?n=200&y=t > >In cases like this one it is not necessarily a bug and I don't have enough >knowledge to make the call. That is why the suggestion to take it up with >this group where there are people familiar with the standard and the >behaviour of other databases. > >> So here I am asking the question here: is this a bug or not? > >Which is exactly what we want to know :-) If the SQL standard specifies it >then it is a very strong candidate for fixing. If all other databases do it >then that is also strong. When only a subset do then that is not good. If >it alters the behaviour compared to previous SQLite versions then breaking >that backwards compatibility would require exceptionally compelling evidence. If this BNF is correct (and I assume it is), the construct appears to be in the SQL92 standard: SELECT http://sqlzoo.net/sql92.html#query%20specification http://sqlzoo.net/sql92.html#select%20list http://sqlzoo.net/sql92.html#select%20sublist http://sqlzoo.net/sql92.html#qualifier http://sqlzoo.net/sql92.html#table%20name http://sqlzoo.net/sql92.html#qualified%20name http://sqlzoo.net/sql92.html#qualified%20identifier FROM http://sqlzoo.net/sql92.html#table%20expression http://sqlzoo.net/sql92.html#from%20clause http://sqlzoo.net/sql92.html#table%20reference http://sqlzoo.net/sql92.html#table%20name http://sqlzoo.net/sql92.html#qualified%20name http://sqlzoo.net/sql92.html#qualified%20identifier (where schema name would be synonymous with a database name in the SQLite naming convention). It may have been a design decision not to implement it, in which case it would become a RFE instead of a bug ;) I can imagine query generators tend to use more, possibly redundant, qualifiers than a human programmer would. -- ( Kees Nuyt ) c[_] _______________________________________________ sqlite-users mailing list [email protected] http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
