> I am not sure exactly sure what you mean by this, but I am certainly not 
requesting nor expecting you to support any fallout of this PR. Our team 
will obviously step up if something bad arises 

Sorry for confusion, I have no doubt that your team will provide good 
support for this change. What I meant is that I am not ready to put my +1 
for merging, but I do not block others from proceeding with the merge.

> What I'm trying to avoid here is stalling this work. We created this PR 
on the 1st of March.  

I do not want to stall it either. This is a good technical debt reduction 
work, and " I definitely do not want the PR to miss the LTS merge window" 
like stated above. I'd love to see it in the September LTS release. We are 
also interested to get it in weekly rather sooner than later so that we can 
address any regressions if reported. 


>> and do the better effort in reaching out to plugin maintainers and 
getting releases or explicit up-for-adoption where possible.
> Care to elaborate please? IIUC you mean sending an email to the last 
known maintainers for plugins that are going to be broken when we merge the 
Core PR.
 
Yes, I mean sending emails offering to either merge/release the fix or to 
put the plugin for adoption. It would be great to finalize Basil's plugin 
EOL JEP so that we have this process more or less automated. But now yes, 
emails. They can be retrieved from the plugin metadata, Jira or Git commit 
history. I can help with these emails as a board member.

> TBH after this, I fear a bit the Guava upgrade that we want to help on 
next... 

Topic for a contributor summit? I also expect difficulties with Guava 
upgrade and then Groovy update needed for Java 17 and housekeeping.
Contributor summit could be a good venue to develop strategy for such 
massively used components.

> What, specifically, is the threshold of usage that would cause you to 
express concern? 

I do not have a specific threshold at the moment. And I do not think the 
installation numbers represent the community importance well. Big Jenkins 
setups at enterprises tend to use "exotic" plugins with low installation 
numbers, just because they have specific requirements to their environment 
and scalability. Admins of these instances also tend to disable sending 
usage stats when they discover this option. So I just use personal 
expertise which might be biased due to my Hardware/Embedded background. Ask 
Daniel or Jesse how often they do a facepalm after hearing about my 
use-cases and hacks :)

Speaking seriously, we could definitely think about defining our criteria 
about what we care during such upgrades. Stale plugins and plugins waiting 
for adoption for years should not be a blocker for changes we need as the 
community.


On Monday, May 31, 2021 at 11:57:12 PM UTC+2 [email protected] wrote:

> Dear Oleg,
>
> On Sun, May 30, 2021 at 11:55 AM Oleg Nenashev <[email protected]> 
> wrote:
> > I have commented about the plugins removal in another thread.
>
> In particular, you wrote: "From the list, I am particularly concerned
> about Code Coverage plugins which seemed to be actively used."
>
> You expressed concern about Emma, a code coverage plugin with 3,148
> installations. You did not express concern about BlameSubversion, a
> plugin not related to code coverage with 825 installations.
>
> What, specifically, is the threshold of usage that would cause you to
> express concern?
>
> Thanks,
> Basil
>

-- 
You received this message because you are subscribed to the Google Groups 
"Jenkins Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/jenkinsci-dev/7486685b-b93e-4151-b952-8b927bf7f7d0n%40googlegroups.com.

Reply via email to