[ 
https://issues.apache.org/jira/browse/LUCENE-7543?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15664580#comment-15664580
 ] 

Hoss Man commented on LUCENE-7543:
----------------------------------

Thanks for filing this jira steve, I'd forgotten about this conversation...

bq. So, what uses doap.rdf and how is it kept up to date? And if it is moved to 
GitHub, do we abandon the original location/processes?

The doap.rdf files are published publicly on the lucene website for consumption 
by any system (internal or external) that wants to know about the history of 
solr/lucene releases -- DOAP is an RDF standard for this purpose, much like 
FOAF.

Notably the doap.rdf files power projects.apache.org...

* https://projects.apache.org/doap.html
* https://projects.apache.org/committee.html?lucene
* https://projects.apache.org/project.html?lucene-core
* https://projects.apache.org/project.html?lucene-solr

In theory the DOAP files could be auto generated by some system like Jira, but 
it's a good idea to edit them manually for the express purpose of being able to 
correct mistakes (ie: they are living documents)

bq. An extra release step will be required if we move it to the Git repo, so 
that the files will continue to be present on the website, similarly to how 
javadocs are now copied to the website.

that would not be neccessary...

from an external perspective, the doap files just have to be available at some 
public / well-known URL.  traditionally that has been...

* http://lucene.apache.org/core/doap.rdf
* http://lucene.apache.org/solr/doap.rdf

...but there is no reason the well known & public URLs can't be things like 
"https://git-wip-us.apache.org/repos/asf?p=lucene-solr.git;f=solr/doap.rdf;a=blob_plain;hb=HEAD";
 that are served directly form source control (the projects.apache.org 
instructions actually suggest that approach) and we could add some .htaccess 
rules to the website to make the existing URLs redirect to new URLs served 
directly from GIT.

>From an internal perspective we just have to update this list to point 
>wherever we want...
https://svn.apache.org/repos/asf/comdev/projects.apache.org/data/projects.xml

bq. 2. Put them in place in a cloned Lucene/Solr Git repo (probably at lucene/ 
and solr/).

I would suggest they might make more sense in dev-tools?  just a gut feeling.

> Make changes-to-html target an offline operation
> ------------------------------------------------
>
>                 Key: LUCENE-7543
>                 URL: https://issues.apache.org/jira/browse/LUCENE-7543
>             Project: Lucene - Core
>          Issue Type: Task
>            Reporter: Steve Rowe
>
> Currently changes-to-html pulls release dates from JIRA, and so fails when 
> JIRA is inaccessible (e.g. from behind a firewall).
> SOLR-9711 advocates adding a build sysprop to ignore JIRA connection 
> failures, but I'd rather make the operation always offline.
> In an offline discussion, [~hossman] advocated moving Lucene's and Solr's 
> {{doap.rdf}} files, which contain all of the release dates that the 
> changes-to-html now pulls from JIRA, from the CMS Subversion repository 
> (downloadable from the website at http://lucene.apache.org/core/doap.rdf and 
> http://lucene.apache.org/solr/doap.rdf) to the Lucene/Solr git repository. If 
> we did that, then the process could be entirely offline if release dates were 
> taken from the local {{doap.rdf}} files instead of downloaded from JIRA.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org

Reply via email to