So does this mean more custom themes can be added? In general, this is really cool and hope this allows for more eras and companies to have custom themes.
If So I think having a page for that would be cool.
‘Edit Page’ is being changed to hopefully have some in-site editing!
I will note that Decap is still fully accessible from the main page, and it + normal Github editing is still the main way to do stuff for now.
You’ll need to follow the instructions to make a GitHub token. From there you’ll be allowed to just edit the in-page text- and at the bottom add photos to the gallery. The editor also features macros for linking pages and citations. Expect things to not work and things to be broken- but hopefully as this goes well it should get added to and improved! It should soon be a much better experience over DecapCMS.
Also, the themes have been set up to not use resources when not selected (fnaf cameras, winter snow). The FNaF cameras have also been fixed to properly pan right.
So, are reviews being cut? With the removal of the contribute section, without using GitHub itself, people won’t be able to make reviews (even if not many have been done yet)
New setting for your contributor name. Any time you make a submission in the new editor it’ll add it automatically to the contributors, otherwise defaults to ‘Anonymous’.
Reviews have been added to the new system, they’re at the bottom of Edit Page once your token is added.
Removed the default option to ‘add images’ inside articles, don’t want that clutter.
Refactored a bit of the code, removed some older stuff, and finally consolidated more articles into the base ‘Article’ and ‘Infobox’ classes.
Removed location pages having the gigantic photos of the exteriors. I liked it, but @CyberSnout was bitching all the time about it so, unless anyone objects its the same across all articles.
Moving the token to your settings and having it last a bit longer per session.
Theory Web is now much more implemented, with its own search system, article inspector with OCR, and theming.
Themes have adjusted to default to ‘section based’, basically so TheoryWeb can use the FNaF theme (i could expand this to like, FNaF articles with that tag using the theme too, or the AAM page using the FunNet theme). Cheese-E-Pedia is now its own direct theme if you want that across all sections.
Today we have a new MMBB Intro. I’ve written the article up to Concept Unification, and I’d highly recommend anyone else do the rest up to modern day as I don’t have as encyclopedic knowledge on later stuff as I do the first years.
Today I have finally internally shifted all parameters to new names and types.
tags → type
categories → tags
‘Type’ is also now a string instead of an array, as its now supposed to dictate how the page is formed. Now finally ‘tags’ actually mean what the page will literally be force tagged with. I’ve looked around and tested what i can- but if you notice something not working right just let me know as it may be due to this large change.
Further steps to tagify everything have completed, now the ‘pages’ attribute in videos and photos have merged with the tags. The entire site is now fully tag-run.
One immediate downside is the Pizza Time Theatre and Chuck E. Cheese’s pages now have multi thousand image galleries. This was the intention all-along, just like how searching for those terms would also show you all images. I just need to implement the ‘show more’ system the search has soon.
Moving from Hugo?
I’ll note a Discord user suggested turning the site’s parameters into JSON data and having pages be built in real time. Stripes mentioned this would remove the ability for Google to index in-page data, but I don’t really give a fuck about Google- though its a valid point.
I think in the long term this could be an interesting direction to go. Compile time on the site is already down to about 4 minutes on my machine, so I mean there’s not like a need to reduce that any further. The more interesting path would be reducing the amount of compiled JSON blobs and trying to force the site to exclusively be just the article JSON and some JS rendering pages. Only issue is the search itself. Given the density of articles I’d imagine a client-ran search would pull a crazy amount of requests and wasted data trying to compile a usable list. There would still need to the compile_tags and compile_fuzzy_search scripts ran before a site upload.
A json only site would extremely reduce bloat though. Right now any minor change forces me to resend all the .html files for every page to my server after compile, but in this context I’d literally only be sending the changed pieces. That gives more room for the ever-growing selection of article text and especially images.
Cool Ideas
A split away from how Hugo runs things could also open up a bit of better sorting. Due to unchanging things having to be moved to /static/, I can’t just have the photos and articles live in the same folder, or have it be like /photos/myphoto/photo.json + photo.avif. It’d also let me have a /users/imauser/ where users could have reviews or comments all inside their folder.
Making the site just some base-level javascript and a collection of .jsons and images would mean no need to teach users how to install and use hugo. You just edit a file and see the results. That’d bring the site even closer to being very portable. I know people aren’t locally running CEP- but I always want it as an option. This change would mean as a novice user you could totally just get GitHub Desktop to download the entire site onto your PC and you view it locally- then any time there’s a change you just hit pull. I think the site currently is at what 8gb of space?
This would make it super easy to completely reformat how the data is pulled and aggregated. I know @stripes has wanted to change how CEP looks entirely as an alternate theme, and this would mean we’re not having to struggle within Hugo’s limitations to make that possible.
