Thanks for picking this up, Fabio! On Apr 21, 1:32 pm, Neal Groothuis <[email protected]> wrote: > NH-2661. > > On Apr 21, 12:54 pm, Fabio Maulo <[email protected]> wrote: > > > > > > > > > The matter is what happen in the driver when we use SqlDbType.Timeand the > > server is SQL2000. > > which is the number of that issue ? > > > On Thu, Apr 21, 2011 at 12:46 PM, Neal Groothuis > > <[email protected]>wrote: > > > > Fabio, > > > > As I noted before, I don't have access to a SQL 2000 or 2005 > > > instance. I only have 2008. However, I'm not sure that it's > > > meaningful to do so; SQL 2000 and 2005 don't have aTIMEtype. > > > > If you'll reopen the issue, I'll attach the patch. > > > > - Neal > > > > On Apr 21, 10:48 am, Fabio Maulo <[email protected]> wrote: > > > > does it works also for MsSQL2000 and MsSQL2005 ? > > > > If yes, please attach the patch to your issue. > > > > > On Thu, Apr 21, 2011 at 11:37 AM, Neal Groothuis > > > > <[email protected]>wrote: > > > > > > I just verified that the unit test works with the change I suggested. > > > > > The exact code that I added to NHibernate.Driver.SqlClientDriver is: > > > > > > protected override void InitializeParameter(IDbDataParameter > > > > > dbParam, string name, SqlType sqlType) > > > > > { > > > > > base.InitializeParameter(dbParam, name, sqlType); > > > > > if (sqlType.DbType == DbType.Time) > > > > > ((SqlParameter) dbParam).SqlDbType = SqlDbType.Time; > > > > > } > > > > > > This seems like it's the ideal place to work around MS's error, since > > > > > the InitializeParameter() method is responsible for setting the name > > > > > and type of the parameter. > > > > > > On Apr 21, 7:50 am, Fabio Maulo <[email protected]> wrote: > > > > > > That ok too > > > > > > > On Wed, Apr 20, 2011 at 8:40 PM, Neal Groothuis < > > > > > [email protected]>wrote: > > > > > > > > I'm out for tonight, but hopefully I'll be able to do some more > > > work > > > > > > > on this tomorrow morning Easterntime(UTC-04). I've got a copy of > > > > > > > the NHibernate 3.1.0 source built and ready to hack on, and I'm > > > happy > > > > > > > to help test, but I'm not clear on what you're suggesting that I > > > do. > > > > > > > Where do you want me to copy the SqlClientDriver implementation > > > > > > > to, > > > > > > > and what does SetParameterSizes have to do with this? > > > > > > > > My proposed solution was to override InitializeParameter() in > > > > > > > NHibernate.Driver.SqlClientDriver, a la: > > > > > > > > protected override void InitializeParameter(IDbDataParameter > > > dbParam, > > > > > > > string name, SqlType sqlType) { > > > > > > > base.InitializeParameter(dbParam, name, sqlType); > > > > > > > if(sqlType ==Time) > > > > > > > ((SqlParameter)dbParam).SqlDbType =Time; > > > > > > > } > > > > > > > > Forgive me if this isn't quite right; I don't have VS on this > > > machine, > > > > > > > but that's the basic idea. > > > > > > > > On Apr 20, 6:31 pm, Neal Groothuis <[email protected]> > > > wrote: > > > > > > > > I don't have access to SQL Server 2005 or 2000. > > > > > > > > > On Apr 20, 6:28 pm, Fabio Maulo <[email protected]> wrote: > > > > > > > > > > copy&paste our SqlClientDriver implementation and change > > > > > > > > > the SetParameterSizes implementation setting the parameter > > > > > > > > > type > > > > > > > > > to SqlDbType.Time. We have to check if something wrong happen > > > for > > > > > > > MsSQL2005 > > > > > > > > > and MsSQL2000. > > > > > > > > > > On Wed, Apr 20, 2011 at 6:54 PM, Neal Groothuis < > > > > > > > [email protected]>wrote: > > > > > > > > > > > Fabio, > > > > > > > > > > > I'm not certain what you're asking me to check. > > > > > > > > > > > As for "pushing" Microsoft: what, specifically, are you > > > > > proposing > > > > > > > > > > that I do? > > > > > > > > > > > - Neal > > > > > > > > > > > On Apr 20, 5:16 pm, Fabio Maulo <[email protected]> > > > wrote: > > > > > > > > > > > So we have to change the SqlClientDrive... > > > > > > > > > > > Have youtimeto check something ? > > > > > > > > > > > ARG!!! somebody have changed the SqlClientDriver methods > > > > > > > > > > > to > > > > > > > static... > > > > > > > > > > > I'll fix it later... > > > > > > > > > > > Well... copy&paste our SqlClientDriver implementation and > > > > > change > > > > > > > > > > > the SetParameterSizes implementation setting the parameter > > > type > > > > > > > > > > > to SqlDbType.Time. We have to check if something wrong > > > happen > > > > > for > > > > > > > > > > MsSQL2005 > > > > > > > > > > > and MsSQL2000. > > > > > > > > > > > If nothing wrong happen we can maintain just one > > > > > implementation. > > > > > > > > > > > Were it fail, we have to implement another drive. > > > > > > > > > > > > btw, Neal, the fact remain: we have to push ppl in > > > Microsoft to > > > > > > > respect > > > > > > > > > > > their rules > > > > > > > > > > > When setting command parameters, the SqlDbType and > > > > > > > > > > > DbType< > > > >http://msdn.microsoft.com/en-us/library/system.data.dbtype(v=VS.90).aspx> > > > > > > > > > > > are > > > > > > > > > > > linked. Therefore, setting the DbType changes the > > > > > > > > > > > SqlDbType > > > to > > > > > a > > > > > > > > > > supporting > > > > > > > > > > > SqlDbType. > > > > > > > > > > > > On Wed, Apr 20, 2011 at 6:01 PM, Neal Groothuis < > > > > > > > > > > [email protected]>wrote: > > > > > > > > > > > > > The SqlDbType enum is System.Data.SqlDbType, in the > > > > > System.Data > > > > > > > > > > > > assembly. > > > > > > > > > > > > > On Apr 20, 4:56 pm, Fabio Maulo <[email protected]> > > > > > wrote: > > > > > > > > > > > > > In which assembly is the SqlType.Time? > > > > > > > > > > > > > > On Wed, Apr 20, 2011 at 5:54 PM, Neal Groothuis < > > > > > > > > > > > > [email protected]>wrote: > > > > > > > > > > > > > > Fabio, > > > > > > > > > > > > > > > I don't see a way to practically get "Microsoft to > > > follow > > > > > > > their > > > > > > > > > > > > > > rules", especially after they've said that, yes, > > > > > > > > > > > > > > it's > > > a > > > > > bug, > > > > > > > but > > > > > > > > > > they > > > > > > > > > > > > > > believe that fixing it in existing versions of .NET > > > would > > > > > > > cause > > > > > > > > > > more > > > > > > > > > > > > > > harm than good. > > > > > > > > > > > > > > > What's gained by forcing users to have to come up > > > with > > > > > ugly > > > > > > > hacks > > > > > > > > > > when > > > > > > > > > > > > > > the problem can be accommodated easily and cleanly > > > > > > > > > > > > > > in > > > the > > > > > > > core of > > > > > > > > > > > > > > NHibernate by overriding a method and adding two > > > lines? > > > > > > > > > > > > > > > - Neal > > > > > > > > > > > > > > > On Apr 20, 2:55 pm, Fabio Maulo < > > > [email protected]> > > > > > > > wrote: > > > > > > > > > > > > > > > In practice... > > > > > > > > > > > > > > > When an OSS RDBMS does not follow DbType rules we > > > ask > > > > > them > > > > > > > to fix > > > > > > > > > > the > > > > > > > > > > > > > > > problem. > > > > > > > > > > > > > > > Microsoft said "good bye" to Oracle data-provider > > > and > > > > > now > > > > > > > Oracle > > > > > > > > > > have > > > > > > > > > > > > to > > > > > > > > > > > > > > > follow all rules defined by Microsoft. > > > > > > > > > > > > > > > > We all should push Microsoft to follow his rules > > > > > > > > > > > > > > > at > > > > > first. > > > > > > > > > > > > > > > > NOTE: the same is for the DbProviderFactories its > > > > > famous > > > > > > > > > > "invariant" > > > > > > > > > > > > name > > > > > > > > > > > > > > > and the first one breaking the rule who was... > > > again > > > > > > > Microsoft > > > > > > > > > > for > > > > > > > > > > > > SQLCe > > > > > > > > > > > > > > > versions. > > > > > > > > > > > > > > > > On Wed, Apr 20, 2011 at 3:49 PM, Fabio Maulo < > > > > > > > > > > [email protected]> > > > > > > > > > > > > > > wrote: > > > > > > > > > > > > > > > > NH-2661 was closed. > > > > > > > > > > > > > > > > Microsoft said: > > > > > > > > > > > > > > > > "Thanks for reporting this issue. This is a bug > > > in > > > > > our > > > > > > > product, > > > > > > > > > > but > > > > > > > > > > > > > > after > > > > > > > > > > > > > > > > carefully investigating the issue, we have > > > decided > > > > > not to > > > > > > > fix > > > > > > > > > > the > > > > > > > > > > > > bug. > > > > > > > > > > > > > > " > > > > > > > > > > > > > > > > > We say: > > > > > > > > > > > > > > > > We are not the workaround fabric. If one of your > > > > > > > customer, who > > > > > > > > > > pay > > > > > > > > > > > > to > > > > > > > > > > > > > > have > > > > > > > > > > > > > > > > your product, say that you have a bug and then > > > you > > > > > admit > > > > > > > that > > > > > > > > > > you > > > > > > > > > > > > have > > > > > > > > > > > > > > a > > > > > > > > > > > > > > > > bug, then you have to fix it. > > > > > > > > > > > > > > > > > On Wed, Apr 20, 2011 at 3:36 PM, Neal Groothuis > > > > > > > > > > > > > > > > < > > > > > > > > > > > > > > [email protected]>wrote: > > > > > > > > > > > > > > > > >> Hi Fabio, > > > > > > > > > > > > > > > > >> I've raised two: NH-2660 and NH-2661, with > > > failing > > > > > tests > > > > > > > > > > attached. > > > > > > > > > > > > > > > >> I've also got potential solutions for both. > > > > > > > > > > > > > > > >> The > > > > > Set() > > > > > > > method > > > > > > > > > > on > > > > > > > > > > > > the > > > > > > > > > > > > > > > >> custom TimeType in the test for NH-2661 is one > > > > > possible > > > > > > > fix > > > > > > > > > > for > > > > > > > > > > > > > > > >> NH-2660, and I've included a possible solution > > > to > > > > > > > NH-2661 in > > > > > > > > > > the > > > > > > > > > > > > > > > >> comments on the issue. > > > > > > > > > > > > > > > > >> Please let me know if you have questions or if > > > > > > > > > > > > > > > >> I > > > can > > > > > > > help > > > > > > > > > > further! > > > > > > > > > > > > > > > > >> - Neal > > > > > > > > > > > > > > > > >> On Apr 19, 8:21 am, Fabio Maulo < > > > > > [email protected]> > > > > > > > wrote: > > > > > > > > > > > > > > > >> > mmmm perhaps yes. > > > > > > > > > > > > > > > >> > There is a > > ... > > read more »
-- You received this message because you are subscribed to the Google Groups "nhusers" 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/nhusers?hl=en.
