[
https://issues.jboss.org/browse/JBIDE-13835?page=com.atlassian.jira.plugi...
]
Nick Boldt commented on JBIDE-13835:
------------------------------------
"We are already moving source zips around for every build "
No, we're not. We don't produce source zips for the components. Surely you've
noticed we only produce update sites w/ source BUNDLES in them, but no source-dump zips?
--
"If someone does a push force in between. It will mean the source gathering will fail
most likely though."
Actually, I think even when you force push the old SHAs persist in a detached state --
I've been able to access old commits via the github website even after force pushing
to replace them in a topic branch. So the tip of the branch moves to a new SHA, but that
old one will still be resolvable and therefore the source fetch should still work. But...
I thought force pushes to origin were not allowed, so this shouldn't even matter as
it'll never happen.
+0.75 for PR 185. Comments on the PR: can we have less boilerplate and more inheritance w/
variable overrides? Also you forgot about the central and earlyaccess sites, published by
the child-sites job [1].
[1]
https://jenkins.mw.lab.eng.bos.redhat.com/hudson/view/DevStudio/view/DevS...
Improve publish script (split? Move to maven?)
----------------------------------------------
Key: JBIDE-13835
URL:
https://issues.jboss.org/browse/JBIDE-13835
Project: Tools (JBoss Tools)
Issue Type: Enhancement
Components: build
Reporter: Mickael Istria
Assignee: Nick Boldt
Fix For: 4.3.x
Publish script deals with a lot of different and not-always-related things. It makes it
difficult to maintain it. We should think about some improvements such as
* different conventions from different stream
* components vs aggregate
* devstudio vs jbosstools.
Or,
* Make publishing part of a "mvn deploy" step.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)