I'm not worried about how it's set up now. I just wanted to add some
documentation to make it clear for folks who are unaware of how it works.
I think there are more "convenient" solutions out there, but it's also
not at the top of my priorities list :)
On 6/6/19 12:23 PM, Elek, Marton wrote:
We can change this strange behaviour.
It turned out that it's possible to use just one branch which contains
both the source and rendered version (see the hadoop site) Would you
prefer this approach?
It would be even better to generate the side during the deploy time.
AFAIK the apache infra doesn't support it.
3rd party provides can do it for free (for example
https://app.netlify.com). I can't see any risk as the source of the web
page would be under apache control anyway, but it may require more logistic.
Or (if we can get git credentials with committer permission on apache
jenkins) we can do it on apache jenkins.
m.
On 6/4/19 5:51 PM, Josh Elser wrote:
Oh, I had filed RATIS-582 as well.
I'll take a look at what you put on 583. Thanks!
On 6/4/19 6:43 AM, Tsz Wo Sze wrote:
Filed https://issues.apache.org/jira/browse/RATIS-583 and posted a
patch. Thanks.
Tsz-Wo
On Tue, Jun 4, 2019 at 3:48 PM Tsz Wo Sze <[email protected]> wrote:
Sorry about that I somehow did find the asf-site-source branch earlier
when I made the website change.
Let me file a JIRA fixing it.
Tsz-Wo
On Tue, Jun 4, 2019 at 5:17 AM Josh Elser <[email protected]> wrote:
Folks,
In the future, please do not make changes to the asf-site branch unless
you have corresponding changes to the asf-site-source. Right now,
asf-site-source is lagging behind changes that were made by hand to
asf-site.
I'll have to spend some time tomorrow to fix this, but please,
again, do
not make changes to the generated content in the branch asf-site by
hand.
Thanks.
- Josh
