Bug 494504 - Resurrect GdbDebugServicesFactory single argument constructor
Summary: Resurrect GdbDebugServicesFactory single argument constructor
Status: RESOLVED FIXED
Alias: None
Product: CDT
Classification: Tools
Component: cdt-debug-dsf-gdb (show other bugs)
Version: 9.0.0   Edit
Hardware: PC Linux
: P3 major (vote)
Target Milestone: 9.0.0   Edit
Assignee: Jonah Graham CLA
QA Contact: Marc Khouzam CLA
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2016-05-25 04:34 EDT by Jonah Graham CLA
Modified: 2016-05-25 16:29 EDT (History)
3 users (show)

See Also:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Jonah Graham CLA 2016-05-25 04:34:42 EDT
Bug 488909 improved the code base by merging GdbDebugServicesFactory and GdbDebugServicesFactoryNS. However it left an unusually hard to cope with API change due to removing GdbDebugServicesFactoryNS class and the GdbDebugServicesFactory(String) constructor.

This bug is to leave the code base merged, but provide the API to construct the improved class.
Comment 1 Jonah Graham CLA 2016-05-25 04:35:56 EDT
Cross referencing the the mailing list discussion on this: https://dev.eclipse.org/mhonarc/lists/cdt-dev/msg30724.html
Comment 2 Eclipse Genie CLA 2016-05-25 04:44:31 EDT
New Gerrit change created: https://git.eclipse.org/r/73559
Comment 3 Eclipse Genie CLA 2016-05-25 07:08:31 EDT
New Gerrit change created: https://git.eclipse.org/r/73574
Comment 4 Marc Khouzam CLA 2016-05-25 08:10:13 EDT
I haven't looked at the Gerrit proposals yet but my hope was to find a (incomplete but) sufficient solution without reintroducing the GdbDebugServicesFactoryNS.

Maybe that's not possible?
Comment 5 Jonah Graham CLA 2016-05-25 08:38:01 EDT
(In reply to Marc Khouzam from comment #4)
> I haven't looked at the Gerrit proposals yet but my hope was to find a
> (incomplete but) sufficient solution without reintroducing the
> GdbDebugServicesFactoryNS.
> 
> Maybe that's not possible?

Sorry, we don't need GdbDebugServicesFactoryNS for Liviu's case. I was reverting the entire API for the cleanup, without losing the "goodness" of the cleanup. We don't have to resurrect GdbDebugServicesFactoryNS.
Comment 6 Marc Khouzam CLA 2016-05-25 11:14:15 EDT
The proposed solution is even better than I expected.

Looking at the most likely way extenders would have been using this API,
I believe the proposed new API will allow them to continue working
with all-stop and non-stop even if they don't migrate to the new constructor.

I verified this with the GdbExtendedLaunchDelegate as it was written for the old,
API and confirmed that both all-stop and non-stop are still handled properly.

Thanks Jonah, this change is how I should have done it from the start.
Comment 9 Jonah Graham CLA 2016-05-25 16:29:31 EDT
Updated N&N API Modifications section for this change. (Note change is in a moderation queue so may not be visible yet.)