Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [cross-project-issues-dev] [WARNING] SimRel Headed Off the Tracks

Hi

I share Ed and Christian's concerns, and as I wrote earlier the original purpose of SimRel was to provide a coherent synchronized release across all active projects. The quarterly SimRel has totally undermined that because many projects cannot support the faster cadence. Personally, it means that I just chuck out my auto-tested builds without manual verification. Diligent checks that I used to perform annually have gone out the window. I used to use perhaps every other 6 weekly milestone for my personal development resulting in many regression bugs getting fixed pre-release. Now the regressions get detected after release; it is all much much too fast. The final platform RC is done almost before some projects have started on their first RC. The only things that can get fixed are panics and just look at the number of RC3 respin discussions we now have to have because of the unreasonable speed; in the past RC3 and RC4 were available.

Let's go back to annual SimRels that provide a coherent best endeavours release that users can be confident of enhancing by taking on board service releases. No longer a brand new different selection of novel bugs every quarter.

If projects want to do quarterly intermediate releases, fine, they are intermediate releases. Until such time as the high speed community steps up and commits the resources to support it, why does the rest of Eclipse have to suffer?

The timing of the annual release should be aligned to offer a useable rather than experimental latest-Java experience.

(When I advocate annual SimRels, that does not mean I endorse EPPs. Personally I only use the base Eclipse SDK ZIP, project ZIPs where available and the SimRel P2 as a last resort. I can easily live without EPPs, but a synchronized SimRel P2 is critical.)

(As a modelling developer I might be expected to use the Modeling EPP, but sadly I find that although ridiculously bloated, it lacks important projects and includes one project that I find detrimental to my UX.)

    Regards

        Ed Willink

On 29/01/2020 08:21, Dietrich, Christian wrote:
Hi all,

I personally share Ed's concerns about the direction the Eclipse
SimRel is thrifting. Maintaining 4 Releases a Year in a quite involved
project like Xtext with many dependencies was a lot of work. Also the
move to the new Cloud based Jenkins infrastructure ate a lot of
resources at the Xtext team last year. Resources that we won't have
this year. Thus is see this a burden for many smaller and less
resourced projects. I am not sure if a pull approach would solve this
problem as there still needs to be common ground for projects that
need to depend on each other. No EMF would mean no Platform, Xtext,
OCL etc.

~Christian
_______________________________________________
cross-project-issues-dev mailing list
cross-project-issues-dev@xxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://www.eclipse.org/mailman/listinfo/cross-project-issues-dev


Back to the top