RFC: Breaking changes for collections in NHibernate 4.0
249 views
Skip to first unread message
Oskar Berggren
Sep 7, 2012, 4:41:40 AM9/7/12
to nhu...@googlegroups.com
Hi,
The plan for NHibernate 4.0 is to support building it on .Net 4.0, including support for .Net's own ISet<T> instead of using Iesi.Collections.
Today NHibernate 3.3 supports both non-generic and generic collections. For .Net's ISet<T>, there is no non-generic counterpart. Therefore, in NHibernate 4.0 support for non-generic ISet will be removed. Generic collections in general have been around since 2005, older than (almost?) all stable NHibernate versions. For the sake of consistency one could argue that support for the other non-generic collections should also be removed. Doing so would also reduce the amount of code somewhat.
Is anyone still using the support for non-generic persistent collections in classes mapped with NHibernate? To such an extent that their removal would be a major problem?
/Oskar
The plan for NHibernate 4.0 is to support building it on .Net 4.0, including support for .Net's own ISet<T> instead of using Iesi.Collections.
Today NHibernate 3.3 supports both non-generic and generic collections. For .Net's ISet<T>, there is no non-generic counterpart. Therefore, in NHibernate 4.0 support for non-generic ISet will be removed. Generic collections in general have been around since 2005, older than (almost?) all stable NHibernate versions. For the sake of consistency one could argue that support for the other non-generic collections should also be removed. Doing so would also reduce the amount of code somewhat.
Is anyone still using the support for non-generic persistent collections in classes mapped with NHibernate? To such an extent that their removal would be a major problem?
/Oskar
Ricardo Peres
Sep 7, 2012, 7:40:48 AM9/7/12
to nhu...@googlegroups.com
Go for it! :-)
Pete Appleton
Sep 10, 2012, 5:00:24 AM9/10/12
to nhu...@googlegroups.com
Hi,
We've not encountered a concrete use-case for non-generic collections, although there could be an edge-case somewhere that uses them (e.g. a particular set of semantics associated with one association resulting in a specific collection class) - it certainly won't cause a major problem for us. On balance I'd vote for removing non-generic support if it helps maintainability.
Pete
--
You received this message because you are subscribed to the Google Groups "nhusers" group.
To post to this group, send email to nhu...@googlegroups.com.
To unsubscribe from this group, send email to nhusers+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/nhusers?hl=en.
Maximilian Raditya
Sep 10, 2012, 6:10:17 AM9/10/12
to nhu...@googlegroups.com
Sounds good to me. It's time to leave the legacy and move on to the future... :D
Is there any JIRA report to vote for?
/Oskar
--
You received this message because you are subscribed to the Google Groups "nhusers" group.
To post to this group, send email to nhu...@googlegroups.com.
To unsubscribe from this group, send email to nhusers+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/nhusers?hl=en.
--
Regards,
Maximilian Haru Raditya
Asher Newcomer
Sep 11, 2012, 10:33:35 PM9/11/12
to nhu...@googlegroups.com
Agreed, please do this. Using the 4.0 ISet is currently more painful than it needs to be. This would be a good win.
Reply all
Reply to author
Forward
0 new messages