Re: [platform-releng-dev] What about dropping .pack.gz artifacts?

[Aleksandar Kurtakov <akurtako@xxxxxxxxxx>]

On Mon, Mar 15, 2021 at 12:18 PM Mickael Istria <mistria@xxxxxxxxxx> wrote:

Hi,

As there is a lot of thoughts and action going on about shortening build time and simplifying build process; I'm thinking maybe the Eclipse project should just stop building/shipping .pack.gz artifacts?

The rationale is that with JustJ, it's easy for RCP applications to use Java 14, most people who install EPP packages do use Java 14; and even without JustJ, Java 14 is spreading and the .pack.gz are thus becoming less and less useful, while they are actually a non negligible cost on disk space and build times.

The only blocker would be the SimRel requirement here: https://wiki.eclipse.org/SimRel/Simultaneous_Release_Requirements#Provide_optimized_p2_repository_.28partially_tested.29 , but at this stage, I think this requirement could be removed as well (especially if Platform says that it's not producing it anyway so it's either SimRel minus the requirement or SimRel minus Platform).

Let me bring this to Planning Council .

Another point is that it's no longer possible to compile Tycho and p2 using Java 16 (which I just got via Fedora update). IMHO, we should start from Tycho by deprecating the pack200 plugins so people start seeing that this is technology that lives its last days.

We can be a bit slower on platform side as we are not in a rush yet. But the day is not far away as being unable to use latest JVM to develop projects is a serious barrier to entry.

 

What do you think?

--

Mickael Istria

Eclipse IDE developer, for Red Hat Developers

_______________________________________________
platform-releng-dev mailing list
platform-releng-dev@xxxxxxxxxxx
To unsubscribe from this list, visit https://www.eclipse.org/mailman/listinfo/platform-releng-dev

--

Aleksandar Kurtakov

Red Hat Eclipse Team