Shreyas Ragavan: 2020-01-19 A quick Summary on website setup plans

I stumbled across Fisseha Berhane's wonderful website today and was stunned to see the quantity of work he has accrued over 10 years. He has every certification that I would consider to be a decent value addition and a body of work that I immensely envy. With respect to work - the way ahead is clear and a definite challenge to match.

I was struck by the simple and efficient design of his website. All the content was displayed more or less flawlessly. There were no annoying advertisements or pop ups or anything that detracted from the content.

I was suddenly reviewing my entire website setups and efforts at converting OddmuseWiki into the sole a blogging platform that I rely on. The irritation was that no single platform, whether hugo or wordpress or oddmuse or even Emacs for that matter offer a complete environment to get everything done. The important thing is to keep it simple and efficient.

This is a summary.

Oddmuse as a wiki has unique ways of organising information in a very quick manner. It is a light engine and can serve atleast 2000 documents without issues as mentioned in #oddmuse by CapnDan
The aesthetics of oddmuse have much left to be desired, but it still is efficient in conveying information and I believe even now - the reading experience is quite passable. This is something that can be customised and worked on and it is a worthy pursuit to understand CSS. AS a side note - I want to have a standard CSS template that I use for all my projects. I need to work on this.
The aesthetics of hugo are good. The convenience of ox-hugo for exports is clear and it is an established workflow that is fast enough. While tags can be collected and custom pages and widgets can be made in hugo, it still falls short in conveying inter-connected information in the capability that a wiki can.
The articles and technical notes that I write - will always have a scope for improvement. This can be easily done and tracked in a wiki. While it is possible still to generate an RSS feed for each page or section in Hugo, the RSS combined with the page collection filters (in Regex none the less) is excellent.
Somebody comfortable with the wiki (almost nobody in my circle) or atleast somebody familiar with regex would be able to conduct interesting searches of the information and I believe personally, this is a small capability abundantly available in Emacs that I desire in the wiki.
Content like Rmd documents, and interactive plots and datascience results should be conveyed in HTML. I could directly link a code notebook, that may even be hosted on my own server to convey the information. The wiki should contain all or atleast most of the documentaiton. Eventually I will find a way to host my own git server and control the documentation via a Software wiki similar to AlexSchroeder. In fact that is the direction my efforts should lie in. I think such a wiki is a mark of a developer who cares.
I've used oddmuse curl only for a day, but the setup was about 5-10 lines of elisp and involves no password. The pages load instantly and the commit and revision is also instant. The search is also actually quite handy. The only aspect to be solved is the export from Org mode to the Wiki syntax and handling tagging. The number of blog posts published speak speaks for itself. It is extremely convenient to edit and create posts and pages on an oddmuse wiki via Emacs.
In fact combined with efficient CSS and other efforts, the wiki has better chances of becoming like Fisseha's website.

All in all - the point is that I will continue to build my blog on Oddmuse and will make plans of shifting all my publishing only on Oddmuse.