Just wanted to let you know that the bug was fixed in the trunk version.
Thanks for bearing with me.

On Mon, Mar 28, 2011 at 1:08 PM, Hadi Eskandari <[email protected]>wrote:

> Ah...that makes sense.
> If you check the same fixture, there was another culture related problem
> with the dates. Think I'll be able to reproduce and fix it now.
>
>
> On Mon, Mar 28, 2011 at 12:59 PM, Kristoffer Jansson <
> [email protected]> wrote:
>
>> Yes, it fail.
>>
>> I think I've found the problem. I thought I had tested all culture
>> related aspects of this, but I didn't dig deep enough. I'm running sv-
>> SE culture which uses a different decimal separator (comma instead of
>> dot). But changing the culture for the application thread is not
>> enough, you have to actually change your system wide settings and log
>> off/on first. When I did change my system settings to en-US and logged
>> off and on then ran the test, it passed.
>>
>> So it seems like a culture related bug after all. Could you verify
>> this?
>>
>> // Kristoffer
>>
>> On Mar 28, 10:13 am, Hadi Eskandari <[email protected]> wrote:
>> > Okay, my mistake, but the test still passes here. I basically run
>> > this<https://gist.github.com/890120>and it passes. Does this fail on
>> > your end?
>> >
>> > On Mon, Mar 28, 2011 at 12:10 PM, Kristoffer Jansson <
>> >
>> >
>> >
>> >
>> >
>> >
>> >
>> > [email protected]> wrote:
>> > > There's a bug in that test. You're testing against the original
>> > > entity, and not what you're getting from the database. If I test
>> > > against "readback", the test fail, see:https://gist.github.com/890120
>> >
>> > > // Kristoffer
>> >
>> > > On Mar 26, 10:56 am, Hadi Eskandari <[email protected]> wrote:
>> > > > Kristoffer,
>> >
>> > > > I have slightly modified and applied your test, to the trunk version
>> of
>> > > > JetDriver. It looks like this <https://gist.github.com/888143> now,
>> but
>> > > the
>> > > > test is passing without applying your patch. I can even see in the
>> Output
>> > > > window that the query ran on the db is like this:
>> >
>> > > > NHibernate: INSERT INTO DecimalEntity (SimpleDecimal, SimpleDouble,
>> > > > NullableDecimal, NullableDouble, Id) VALUES (?, ?, ?, ?, ?);@p0 =
>> 1.1,
>> > > @p1 =
>> > > > 1.2, @p2 = 1.3, @p3 = 1.4, @p4 = 0
>> >
>> > > > NHibernate: SELECT this_.Id as Id0_0_, this_.SimpleDecimal as
>> > > > SimpleDe2_0_0_, this_.SimpleDouble as SimpleDo3_0_0_,
>> > > this_.NullableDecimal
>> > > > as Nullable4_0_0_, this_.NullableDouble as Nullable5_0_0_ from
>> > > DecimalEntity
>> > > > this_ WHERE this_.Id = ?;@p0 = 0
>> >
>> > > > and as you can see decimal points are all there in case of Decimal
>> or
>> > > Double
>> > > > data types. I know you mentioned the repro is tricky but I need a
>> failing
>> > > > test here to make sure the bug is correctly patched. Could you help
>> out?
>> > > Did
>> > > > you run the test you provided before applying the patch and it
>> failed?
>> >
>> > > > Thanks,
>> > > > Hadi
>> >
>> > > > On Mon, Mar 21, 2011 at 8:12 PM, Kristoffer Jansson <
>> >
>> > > > [email protected]> wrote:
>> > > > > I ran into a problem with the JetDriver and decimals a while ago
>> when
>> > > > > working with MS Access databases, but I have no idea how to
>> > > > > contribute, so I'll just post it here.
>> >
>> > > > > The problem arises when you're trying to save a value of type
>> decimal
>> > > > > in the database. Regardless of whether the designated column is a
>> > > > > double, or an actual decimal doesn't seem to matter. What happens
>> is
>> > > > > that the decimal separator magically disappears. So saving '1.2'
>> would
>> > > > > result in the value '12' in the database.
>> >
>> > > > > The fix is making decimals into doubles before saving to database.
>> > > > > Similar as to how the code already makes Int64 into Int32. This
>> works
>> > > > > on both double and decimal columns.
>> >
>> > > > > I've no idea how to write an automated test to verify this, except
>> > > > > actually having a database lying around and try to store and read
>> back
>> > > > > the data. Even then I'm not sure if the test could be used safely
>> as
>> > > > > it's not very isolated. There may be some twisted environment of
>> > > > > drivers etc where the error wouldn't show up.
>> >
>> > > > > Anyway, I've seen this exact behavior on our systems as well as
>> > > > > customers, so it doesn't seem like it's only me :)
>> >
>> > > > > Here's a patch that fixes the problem:
>> > > > >http://www.itstod.se/files/jetdriver_decimal_to_double_fix.patch
>> >
>> > > > > --
>> > > > > You received this message because you are subscribed to the Google
>> > > Groups
>> > > > > "NHibernate Contrib - Development Group" 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/nhcdevs?hl=en.
>> >
>> > > --
>> > > You received this message because you are subscribed to the Google
>> Groups
>> > > "NHibernate Contrib - Development Group" 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/nhcdevs?hl=en.
>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "NHibernate Contrib - Development Group" 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/nhcdevs?hl=en.
>>
>>
>

-- 
You received this message because you are subscribed to the Google Groups 
"NHibernate Contrib - Development Group" 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/nhcdevs?hl=en.

Reply via email to