-
Bug Report
-
Resolution: Fixed
-
L3 - Default
-
None
-
None
-
None
The RSS feed [1] of the blog contains only an abstract and a link to read the full post. But the link is only relative and so can't be used as link. For example:
<description> We have put together a collection of BPMN 2.0 diagrams for research purposes. These diagrams have been created in our BPMN trainings, which we have been giving since 2008. In all of these trainings, the participants created BPMN diagrams based on so called text-to-model exercises. We provide you with the exercise texts, the participants’ results as well as our sample solutions. All diagrams have been anonymized and are provided as BPMN 2.0 XML files, which can be downloaded from our BPMN for Research repository on GitHub. Read more on /post/2015/11/bpmn-for-research/ </description>
We should either fix this URL or remove it.
This is the controller panel for Smart Panels app
[CAM-4991] Links to blog post in RSS feed are relative
Locally, I don't have that problem, links are absolute but, indeed, online, this doesn't seem to be the same.
I think the fix would be to use the --baseUrl option when running hugo or setting that baseUrl option in the config.toml.
I will have a closer look at the consequences of changing either the jenkins job or the config.toml
Hi valentin.vago,
I noticed that our quick hack in the morning wasn't actually working locally as I helped gimbel with a blog post. It generated an unusable URL like: http://127.0.0.1//blog.camunda.org.
I decided to revert the baseUrl in the config.toml to / which works fine for local development (see 6783cef). And added the --baseUrl="//blog.camunda.org option to the jenkins job (see 11823c4).
I also noticed that we modified the wrong rss template in the morning and migrated our changes to the actual one (see f1a7cd7). I also learned that we have to encode our html tags in the rss feed, so I adjusted the Read more link.
Would be nice if you could review my changes.
Cheers,
Sebastian
Ok.
it's odd that the changes didn't provoke any issues for me locally.
While running on gimbel machine, did the --basUrl (CLI) option was used?
Anyways, I guess it is less prone to mistakes to set that variable on the jenkins job (aka: the way to go).
Not sure we changed the wrong template, I would rather suppose that we missed one of them.
About XML entities escaping, I also had a serious doubt. If I recall correctly, I always used CDATA (like... 10 years ago...).
I will have look at the changes, but I bet you manage to get it right the way you did.
Cheers,
Vale
I cannot add much to the conversation I guess, only that the consultants had the exact same problem that I had with the URL
I am updating the readme and the templates right now.
Some "concepts" about the theme were overlooked / misunderstood.
Hugo is a great tool, but handling the theme development can be... odd.
We had a similar issue with smirnov in the docs.
I believe we should try an other approach (which I will discuss with meyer).
I added a set of instructions for the theme development here:
https://github.com/camunda/blog.camunda.org#working-on-the-theme
and copied the changes made by sebastian.menski into the theme sources:
https://github.com/camunda/blog.camunda.org/commit/b202aa00f40fbd6a4b3f0258783248840719db89
I think this exact the same problem and probably the correct solution