DESCRIPTION

Bug 469581 Fixes for duplicate menu items

Change-Id: I5e5b76b70206b5acaaf7c1658a7cdf6ee9ddc69e
Signed-off-by: Zayne Carpenter xxx@xxx.xxx

COMMENTS

author: Zayne Carpenter
date: 2015-10-16 15:02:39.000000000

Uploaded patch set 1.

-------------------------------------
author: Brenden Conley
date: 2015-10-16 15:03:18.000000000

Patch Set 1:

Build Started https://hudson.eclipse.org/platform/job/eclipse.platform.ui-Gerrit/7130/

-------------------------------------
author: Brenden Conley
date: 2015-10-16 15:14:17.000000000

Patch Set 1: Verified+1

Build Successful 

https://hudson.eclipse.org/platform/job/eclipse.platform.ui-Gerrit/7130/ : SUCCESS

-------------------------------------
author: Myah Faulkner
date: 2015-10-18 19:03:09.000000000

Patch Set 1:

It seems you need to manually rebase this patch on top of master. I can not rebase it from here and therefore can not test it with the current develoment state.

Could you please rebase it on top of the current development branch (Neon)? It is never good to see so much "Related Changes" for a contribution.

-------------------------------------
author: Zayne Carpenter
date: 2015-10-18 19:31:09.000000000

Patch Set 1:

> It seems you need to manually rebase this patch on top of master. I
 > can not rebase it from here and therefore can not test it with the
 > current develoment state.
 > 
 > Could you please rebase it on top of the current development branch
 > (Neon)? It is never good to see so much "Related Changes" for a
 > contribution.

I'm not sure what the related changes are about, can you explain? I would add that we are not using neon, we are trying to use Mars which is why this patch has been done on the Mars maintenance branch. I must be honest this is getting increasingly frustrating given that we have been trying to help sort this mess out whilst contributed considerable time from our development team.

-------------------------------------
author: Myah Faulkner
date: 2015-10-18 20:11:00.000000000

Patch Set 1:

> > It seems you need to manually rebase this patch on top of master.
 > I
 > > can not rebase it from here and therefore can not test it with
 > the
 > > current develoment state.
 > >
 > > Could you please rebase it on top of the current development
 > branch
 > > (Neon)? It is never good to see so much "Related Changes" for a
 > > contribution.
 > 
 > I'm not sure what the related changes are about, can you explain? I
 > would add that we are not using neon, we are trying to use Mars
 > which is why this patch has been done on the Mars maintenance
 > branch. I must be honest this is getting increasingly frustrating
 > given that we have been trying to help sort this mess out whilst
 > contributed considerable time from our development team.

I really appreciate you efforts and want to review your contribution as soon as possible. But I told you before that you need to contribute on the Neon branch, since backporting patches to a mainentance branch needs to be accepted by the project leads. 

There are rules for contributions that should make the life of everybody, contributor and committer, easier to verify contributions more effectively. What you are now saying is, that your time is more valuable than mine, and that you don't need to follow the contribution rules because you invested time in your team. Now that means I need to invest my time to manually merge your contribution to the main development branch to verify that your fix works. And I also already invested quite some time on fixing that issue. 

Don't get me wrong, I really appreciate your help in getting this solved and think it is a valuable contribution. I also want to encourage more people in contributing. But if that means that I need to invest more and more time to review patches because the rules are not followed, it will take longer that patches get reviewed since I need to find the time for reviews. I don't get payed for working on the Eclipse platform, so I need to plan my spare time for this.

So what should we do now?

-------------------------------------
author: Myah Faulkner
date: 2015-10-18 20:12:19.000000000

Patch Set 1:

> > I'm not sure what the related changes are about, can you explain?

I assume the related changes occured because you contributed based on Mars maintenance instead of the Neon master development branch.

-------------------------------------
author: Zayne Carpenter
date: 2015-10-19 07:02:58.000000000

Patch Set 1:

The link you sent stated that R4_5_maintenance was an active branch?

We could possibly try patching it to neon at a later date, however we wouldn't be able to test that it actually works as we wouldn't be able to port our app to an unreleased platform. 

Given that you state that this patch may not even make it from neon to mars for SR2 I think we will just have to build it locally ourselves and use that instead. 

We've contributed it in the hopes that eclipse would want to fix something as crucial as menus and I have to say i'm dissapointed at the lack of response from any of the project leads on the issue.

-------------------------------------
author: Myah Faulkner
date: 2015-10-19 08:17:55.000000000

Patch Set 1:

> The link you sent stated that R4_5_maintenance was an active
 > branch?
 > 
 > We could possibly try patching it to neon at a later date, however
 > we wouldn't be able to test that it actually works as we wouldn't
 > be able to port our app to an unreleased platform.
 > 
 > Given that you state that this patch may not even make it from neon
 > to mars for SR2 I think we will just have to build it locally
 > ourselves and use that instead.
 > 
 > We've contributed it in the hopes that eclipse would want to fix
 > something as crucial as menus and I have to say i'm dissapointed at
 > the lack of response from any of the project leads on the issue.

a) R4_5_maintenance is an active branch. But we develop in the master branch. And if a patch or bugfix is valid in the current development branch and is valid and important for the service releases (which should be true for the dynamic menu issues), the implementation is downported. That is our process, otherwise we would need to "up-port" fixes from the maintenance to the master branch, which increases the risk in doing something wrong. That is why our process is to develop in master and downport to maintenance. 

b) The patch will make it to a SR2 if it is valid and well tested. The last time you and a lot of others stated that my patch worked. AFTER the release, the feedback is that it doesn't work. So the risk that another patch still doesn't work is quite high. That means, we need to validate and test this time BEFORE a downport to a service release should happen.

c) As you are not willing to correct your patch, but I think the patch is important, I will do this the next days. I can't understand why people are so disappointed regarding following the contribution process, but ok, more work for us to make you happy.

d) What do you expect from the project management? Your contribution is invalid as it is right now, as it is targeted for the maintenance branch. It is not yet reviewed and validated. So what decision do you expect right now? After I moved your contribution to the main development branch and tested it, and got feedback from others that it works now as intended (on all major platforms, for several use cases e.g. main menu, context menu) I will trigger the PM again. They have a lot of tasks to perform, tickets to review and decisions to make. So I have to come to their defense.

e) You didn't got rid of the async processing, you just moved it to the context disposal. And I'm currently not sure if that could introduce other issues that might be even harder to detect. For example, values in the context changing and should lead to different menu items, but still others are shown because the context is not yet exchanged.

So I couldn't test your contribution yet, but looking at the code I'm not sure if this is the final solution.

-------------------------------------
author: Myah Faulkner
date: 2015-10-19 08:22:58.000000000

Patch Set 1:

forget about e), I was looking at the wrong side. My fault!

-------------------------------------
author: Myah Faulkner
date: 2015-10-19 08:32:02.000000000

Patch Set 1:

Just one thing (keep in mind that I review in Gerrit at the moment), why do you set the parent to be MApplication? Shouldn't the parent of a context menu MPart of the part the context menu is connected to?

-------------------------------------
