Hi Taher,
while this sounds like an interesting approach, I personally want to
avoid to face things more complex than they have to be.
At ecomify, we are shifting to Hugo to get rid of things like Wordpress
/ database maintenance, possible security issues etc.. We have limited
ressources in our project here so I'm interested to keep things simple
and effortless as much as we can.
Hugo is an infra supported solution, it seems to be the official
migration target away from Roller so it seems to be the given way for
the blog.
The question is if we also want to shift to Hugo for the website with a
- then integrated - blog.
Thanks for your input and greetings,
Michael
Am 08.02.23 um 11:13 schrieb Taher Alkhateeb:
Hello, just throwing another idea in the pile, which I used in the past.
It is possible is to deploy a headless CMS. A headless CMS gives you admin user
interface to control content, but the website is just a regular website (can be
the existing one) that communicates with that headless CMS. What you can do
with that is be able to deploy whatever exists and just replace the dynamic
parts with calls to the headless CMS. And you control the content from there.
So you don't even need to commit code, build and deploy, you just login to the
admin section and edit things.
Maybe the advantage of this approach is that you build on what exists
instead of re-writing. This disadvantage might be that you have more pieces
(website, CMS-admin, CMS-database) that should all be deployed and maintained.
Thank you for all the efforts on maintaining all this content either way.
On Wednesday, February 08, 2023 12:55 +03, Daniel Watford <d...@foomoo.co.uk>
wrote:
Hi Michael,
This sounds very interesting.
I have gone back and forth over the years between CMS driven (WordPress)
and statically generated sites (Jekyll) depending on the users that
would be responsible for authoring content. Mixing the two with tools like
Next.js was also interesting.
WordPress suited non-technical better, whereas technical users benefited
from being able to generate the exact site they wanted. Plus the static
sites have the benefit of fast load times without having to implement
caches (although caching at CDNs is still useful)
Since we are a techy bunch I would support building the website using a
static generation tool. I've heard good things about Hugo, and since there
is existing experience on this tool within our community it would seem a
reasonable direction to take.
I'm also interested to see if adding more data-driven pages to the site
would be easier with Hugo. I quite like the idea of a few utility pages to
help generate seed data for users, such as:
- Custom time period generator
- Chart of accounts generator
My only concern with changing the site is the amount of work involved since
we have limited resources, but since we are being forced to take action in
order to get the blog operational, perhaps widening the scope to the site
is reasonable. I expect we can keep the existing HTML pages, and replace
them with Hugo generated pages as and when needed.
+1 from me.
Dan.
On Tue, 7 Feb 2023 at 18:07, Michael Brohl <michael.br...@ecomify.de> wrote:
Dear community,
In the course of getting Wiebke to be able to help with our blog, it
turned out that Apache Roller will be discontinued as the platform for
blogs by infra [1].
There is an offer and already a pull request [2] to export the existing
blog posts for a migration to Hugo [3].
Hugo is a static site generator which will generate individual pages
from every markdown blog posts file which is provided in the PR.
That means we need to (at least)
- setup a Hugo project
- build up templates for the blog posts which also include header and
footer for full page rendering
(hence, we have to take over some of the templates/html
code/javascrip/CSS to Hugo as well)
- generate the post pages and include them in the website under /blog.
These are quite some steps we have to take anyway and I ask myself if we
should not switch over to Hugo for the whole website as well.
That should be not too difficult because we already have some kind of
templates (in php) and there is already some sectioning with header,
content, footer etc.
The advantages would be
- to truly separate content from the templates (it is now mixed
php/content)
- blog posts can be generated from OFBiz commits and, together with page
content, be send in by pull requests
- content automation (no need to change the copyright year manually,
automated taxonimies for the blog etc.)
- easy process for a git supported writing process for pages and blog
posts and a fast and automatable publishing process (git hooks or GitHub
integration).
Coincidence: I am on a journey to migrate the ecomify websites to Hugo
we have at least some knowledge there (maybe others as well?).
To sum up, I propose to migrate our blog and also the Apache OFBiz
website to be an integrated, Hugo based website.
Opinions?
Best regards,
Michael Brohl
ecomify GmbH - www.ecomify.de
[1] https://issues.apache.org/jira/browse/INFRA-22461
[2] https://github.com/apache/ofbiz-site/pull/8/files
[3] https://gohugo.io/
--
Daniel Watford
