Oregami-News

[MZ per X]

Server update is nearly finished!

Good news! It works so far.

@Mz per X: What do you think? can we switch to the jekyll site this evening? Any big changes you want to make to the website content?

It's online for so long. So yes, you can do the switch, and I will update the pages after the fact.

[gene]

Here my "official" update statement:

The server update for Oregami and Kultpower websites is finished.

Before the update the server operated on Ubuntu 12 and had one single apache web server, which had to be used for all websites.
From now on the new server runs Ubuntu 16 and no other software - except for docker. Every web site runs in an isolated docker container, or even on multiple small docker containers (web server & database). In front of all web servers we now use traefik, a reverse proxy, which distributes all webkits to the corresponding (docker) web server. A very neat feature of traefik is the automatic generation of SSL certificates for every web domain via letsencrypt. Yeah!

Another thing we updated is the public website of www.oregami.org : previously it needed a web server with PHP and a MySQL database to operate the content management system "Redaxo". The website was not designed for mobile devices, and the blog posts were written an blogger.com (by Google). From now on the whole website will be served from static HTML files. These are generated from a new GIT repository, which contains HTML templates and markdown text files for pages and blog posts. The generation is done with Jekyll, a great open source static website generator. And you should try to visit the site with your mobile device: works like a charm! (see attached picture: before - afterwards)

The update to docker based websites now gives us the opportunity to continue the development of our open game database: we needed docker e.g. to run "Keycloak", a nice piece of software for user and login management for our database.

Links:
New Oregami website: www.oregami.org
Oregami forum: https://forum.oregami.org
Git repository for new website: https://github.com/oregami/oregami.org-public-website
Traefik reverse proxy: https://traefik.io
Jekyll website generator: https://jekyllrb.com
Keykloak user management: http://www.keycloak.org

[MZ per X]

Hats off to you! The new website seems lightning-fast to me.

[MZ per X]

The website was not designed for mobile devices, and the blog posts were written an blogger.com (by Google).

Maybe we should publish farewell posts there for the people watching the blogger.com accounts.

[MZ per X]

The website was not designed for mobile devices, and the blog posts were written an blogger.com (by Google).

Maybe we should publish farewell posts there for the people watching the blogger.com accounts.

Done.

[gene]

Our public development site got an update.

When you look at the platform page or the screenshot below in this post, you should notice that the platform list (aka GamingEnvironments) now contains the concrete names of the connected titles (aka TransliteratedStrings). Previously it only contained the "IDs" of the titles.

This might seem "normal", "standard" and "easy to achieve" for you, but it wasn't. And that's because of the used technical instruments:

• I used two seperated bounded contexts according to Domain Driven Design principles
• all entered data for each bounded context is saved in events ("event sourcing")
• what you see on this page is a "read model", which is derived from all "events" that occured before when I entered data with "commands" (that is called CQRS)

These principles will build the foundations of our whole core domain application:

• the seperation of the bounded contexts make it possible to maintain the source code in the long term
• based on these principles it is possible to build multiple/different read models for the same data (e.g. one which contains all entered data, and one that contains only the data that is validated and proved by "mods" or so)
• we can attach additional information to all "data input events" (e.g. what was the source of the information, or how is it classified)
• other read models from all saved events can be created (e.g. a user based list of actions or anything that could be derived from the events)
• if we have some kind of bugs in the code that creates a read model, we could fix the bug and create all read models new from ground up based an all the saved events (or we could extend a read model after some time and simply regenerate it with the extension)

Read the above list again and think about it. Then you might understand why I am quite excited about this update.

[MZ per X]

I am excited, too! Thanks for all your work.

Should I turn this into a blog post?

[gene]

Should I turn this into a blog post?

No, not yet, thanks.

I will "polish" the code, the two bounded contexts are not perfectly seperated yet.
When that is done, we should write a blog post, yes.

[MZ per X]

I finished the update of all our static info pages on the website I dared to touch.

I didn't dare touching the developer pages. Would you overhaul these, gene, so we can then tell people to re-read the whole site?

Next project of mine is checking all the links in all the blog posts.

[gene]

The two sites https://forum.oregami.org and https://www.oregami.org just moved to a new server.
Let me know if something does not work as supposed.

[MZ per X]

No problems so far.

Thanks for your tireless work on all this!

[gene]

Please note that tis forum will temporarily be deactivated on May 24th so I have time to make it conform to the rules of GDPR.

[gene]

I am currently testing the updated forum installation.

[gene]

We're back online!
Jesus, that was much work

Every user has to accept the new privacy policy before he/she can post in the forum. New users have to accept the policy during the registration.

I also installed a new theme for the forum. Looks great even on mobile devices!

[MZ per X]

Thanks, gene, I can feel yor pain!

The forums seem to work without glitches for me.