If I generate bad HQL, then it will likely give a nasty exception. The ANTLR error messages aren't the best, and since I'm injecting at the second phase of the parser it has even less context to work from.
However, I really shouldn't be generating bad HQL - if there are queries that, for whatever reason can't be translated properly, they should be failing well before the HQL stage. Should also be easy to take these as bug reports during the alpha / beta stages and sort out any bad reporting that may exist. Cheers, Steve On Sun, May 2, 2010 at 12:36 AM, Tuna Toksoz <[email protected]> wrote: > Just one point that I would like to ask: what i've seen is that Linq > exceptions are pretty cryptic due to ANTLR stuff. Did you make any work on > it? (been a loooong time since i checked out NH code and read it) > > > Tuna Toksöz > Eternal sunshine of the open source mind. > > http://devlicio.us/blogs/tuna_toksoz > http://tunatoksoz.com > http://twitter.com/tehlike > > > > > > On Sun, May 2, 2010 at 1:32 AM, Steve Strong <[email protected]> wrote: > >> I'm happy with the Linq provider. There are some queries that will cause >> it to fail, but I think they are largely edge cases for NH (and you can >> always revert to HQL, so it's never going to be a major issue). >> >> The key is that for all the queries that do run, I'm pretty happy that >> they are doing the right thing. >> >> Plus I've applied the provided patch for 2169 (silly bug, quite >> embarrasing), and upgraded to the latest releases of ANTLR (both code >> generation and runtime libraries). >> >> I'm happy for it to become an official alpha :) >> >> Cheers, >> >> Steve >> >> >> On Mon, Apr 26, 2010 at 5:35 PM, Fabio Maulo <[email protected]>wrote: >> >>> friends... >>> probably you are new in NH. >>> To release Alpha1 is not the end of the world. >>> Have sit, hold on, take a coffee and wait. >>> >>> Between Alpha1 and GA release we will fix a lot of issues as usual. >>> >>> >>> On Mon, Apr 26, 2010 at 12:22 PM, nadav s <[email protected]> wrote: >>> >>>> i'm not sure if its the right post to say it but, >>>> i know that this patch: >>>> http://216.121.112.228/browse/NH-2170 >>>> >>>> <http://216.121.112.228/browse/NH-2170>might not be a problem to many >>>> people >>>> but for my teams, and i'm sure many others (although did not vote for >>>> it) >>>> would very very much appreciate if it will be applied to the trunk so we >>>> could use it and won't have to edit the code (again) to make it available >>>> >>>> >>>> On Mon, Apr 26, 2010 at 4:46 PM, Diego Mijelshon < >>>> [email protected]> wrote: >>>> >>>>> LOL... true. >>>>> Anyway, it's just an example. My only concern is that many (important?) >>>>> bugs remain open... I believe it would be important to know the plan for >>>>> each (like, "won't fix"/"will fix before GA"/etc) >>>>> >>>>> Diego >>>>> >>>>> >>>>> >>>>> On Mon, Apr 26, 2010 at 10:07, Fabio Maulo <[email protected]>wrote: >>>>> >>>>>> On Mon, Apr 26, 2010 at 9:59 AM, Diego Mijelshon < >>>>>> [email protected]> wrote: >>>>>> >>>>>>> And what about bugs like 2169<http://216.121.112.228/browse/NH-2169> >>>>>>> ? >>>>>>> >>>>>> >>>>>> It is an open ticket as various others. >>>>>> >>>>>> Using Naples's dialect: >>>>>> Ogni scarrafone é bello a' mamma soia. >>>>>> >>>>>> Translated in Italian: >>>>>> Ogni scrafaggio é bello, per la sua mamma. >>>>>> >>>>>> Translated in English + Information Technology: >>>>>> Each bug is the most important for its reporter. >>>>>> >>>>>> -- >>>>>> Fabio Maulo >>>>>> >>>>>> >>>>> >>>> >>> >>> >>> -- >>> Fabio Maulo >>> >>> >> >
