Frédéric Delaporte

> although maybe attempting rollback would still succeed in some cases, thus reducing the chances of ending up with DB inconsistencies Inconsistencies are not possible unless the database itself is...

> Ah, no, I'm referring to ending up with duplicated data, for example, say when re-running the same insert code, if assuming failure on the first attempt. Then we are...

If the connection is actually closed, the DbTransaction is rollbacked. (Note that with transaction scopes, that is another subject.) But it seems that connection pooling may not guard against ongoing...

I do not know if any NHibernate contributor is much knowledgeable about Ruby. I just know it was not running for me under Windows. Instead, I test it with a...

One second gap is still already a lot, especially for tests running entirely locally, without actual network access involved. I guess we could still try increasing it and run the...

We do not tend to force dependencies updates on new releases. You can opt-in to a newer version of the dependency by referencing it directly in your project. Is there...

Would it be worth it to add a new option, something like `sql_types.strict`, disabled by default, which, when enabled, would cause the basic types to assume the data type loaded...

The double boxing has been introduced by 990cd6af twenty years ago due to some deficiencies of a driver at that time. Prior to that, the code was using the typed...

The SQLite driver handles SQLite lax typing itself. It does perform the required conversion itself when a typed method is used, according to deAtog in #3530.

We do not tend to force dependencies updates on new releases. You can opt-in to a newer version of the dependency by referencing it directly in your project. Is there...