Site Update - December (Long)

I'm around and working, but as usual, everything is still "in development" and hasn't been released yet. Normally I release changes as soon as I make them, but this has resulted in some pretty nasty feedback along the lines of "don't release things that aren't done".

So, while I don't have anything new deployed this month (as indicated by the orange theme still being here, which is stuck in the deploy pipeline), I do have some spiffy new features that are so very close to being done.

Image Uploader

Several issues, including broken uploads, slow servers, NSFW not being flagged on upload, and other image-y problems trace back to the old uploader(s) being used on the site. While Character pages were the first feature made, image uploading was the second. Since then, the way images are handled on the server has improved a great deal, but the current image upload method still isn't using any of those features.

I implemented the new uploader for V2 profiles, but since those are delayed by the evil "holy hell this is a complex piece of code" demons, I figured I'd move the Image uploader to the site for everyone to use. One neat feature of this is that the image uploader wraps the entire site, so regardless of what page you're on, dropping an image will trigger the upload queue. From there, you can pick which character to upload each image to and send it on its way.

The image upload queue will show up as an icon on the top right next to messages/notifications that are currently there.

Image Lightbox

Some other issues are happening in regards to the image lightbox, which was the third feature implemented on the site. Currently, the lightbox is a bit of a mess, and some updates to other parts of the site broke that. So, I'm almost done re-writing this to use the newer, happier Javascript bundle that is significantly more robust and about two million times easier to maintain from a coding perspective. Yaaay, modular design.

The new lightbox works kinda like the old one with a few minor style changes. The big features here is the ability to navigate backward/forward through related images (news feed group, character galleries, etc). I'm also redesigning the metadata section to drop Markdown support and add #hashtags. This will also support multiple character and artist tags per image.

The Lightbox is almost done, I just need to implement the character tags (maybe), the new report modal, and the delete/download links in the dropdown menu. After I deploy this, SOME features (like background color settings) will be temporarily unavailable. I'm not happy at all with the color picker I'm using, and I would rather disable the feature until I can find a better solution than continue to provide something subpar.

Global Updates

I've done a lot of work cleaning up the source code and updating several out of date packages. The UI framework that I'm using, Materialize, has been finally updated to the modern 1.0.0 branch, up from the 0.100 fork that we've been using since the start. The problem with this is, it's broken a lot of the old site code in the other outdated Javascript bundle. I could continue migrating everything to the 2.0 Javascript pack, or I can apply dirty filthy hacks to keep the old site running while I implement the new stuff.

I'll go with dirty filthy hacks on the old site, thank you.

I've also given in to our "have some standard in your coding style" overlords, and enforced static code analysis so I can ensure my code is consistent and clean. Javascript uses Prettier, and Ruby uses Rubocop. Well--we've always had Rubocop installed but I never actually set it up. The Ruby code still needs some work to get up to standards, but the Javascript is all tidy and pretty and stuff.

I've also finally come up with some standard for interacting with our GraphQL API in Javascript. I'm probably the only person who is going to see or care about this, unless I hire anyone else to help develop this beast, but I think it's Pretty Slick(tm). I made a schpiffy compose method for exporting a chain of React HOCs that I really should release as an NPM module.

THAT'S NICE BUT V2 PROFILES WHEN

Honestly I'm lost. I forgot where I was because I've been changing tasks every time a new feature/support/bug request comes in, that I accidentally knocked V2 profiles off the back burner and haven't mopped up the mess. The good news is, every annoying detail that I have to figure out (Hey, Apollo cache, I'm glaring at you) has been solved in my Lightbox/Upload changes, so that's good!

Once I get this round of changes shipped, and the site doesn't break, I'll go back to V2 profiles. If I don't get distracted with...

Forums

Okay, folks, I've been doing A LOT and nobody knows about it but me. Did you know I'm redesigning the forums? That project started... last summer some time. It was sunny, and warm, and we had daylight and I had motivation and beer. Well, I still have beer, and motivation! That's where the new Markdown text editor came from, and I really want to release that but my current WIP pile is getting HYOOG. One person and all.

The new forums are still kinda in a sketchy figuring-out-how-it-works state, but anything is better than what we have now. Especially when it comes to load times. Oh gee, them load times.

Artist Profiles

AND ANOTHER THING. This is linked in to the artist tags on the new image thinger I talked about above, so that won't be released until this is, but I'm working on adding artist pages to give credit where credit is due. This paragraph is short, you can tell I'm getting to the end of my update.

As Always,

Thank you for your support toward this project, your kind understanding of my bugs and shortcomings, and your eternal patience that keeps things slowly moving forward, one month at a time.

This is it, you're done, you can stop reading.

Everything beyond this point is somewhat personal frustration that I want to sort out. Nothing toward y'alls. You're great.

Mobile App??

I killed it. I read the reviews, I got sad, I got frustrated, I realized I'm in over my head, and I pulled the plug on the Play store. One day, but today is not that day, not while the internet is a bunch of meanie beanies that yell at me for releasing some unfinished pile of free software. If someone wants to fund the venture, I'll update the app and release it again.

Advertising

THIS WEBSITE IS EXPENSIVE. Like, I could buy a new cosplay every month with the coin I'm still spending out of pocket to keep it alive. And I've been cutting costs (I'm so, so sorry Australia but your mirror is going to have to wait until I get more funding). It costs about $80-$100/mo to host the images alone, and that is on one region. I used to pay for 2 regions because redundancy...

I'm eternally thankful for all the support on Patreon, because honestly if I had to pay the full cost out of pocket this website would be yet another footnote in the history of character websites that never made it. Still, Patreon growth is slow (and directly proportional to my release of new features, which is equally slow). So, I need to contemplate some other way to acquire coinage.

I have an ad spot. I will have multiple ad spots. I'm still contemplating the specifics, and gathering data to establish pricing, etc, but if Other Art Websites that people have an Affinity for survive on some of their ad revenue, that's something worth investigating. Local, self-hosted ads only. None of that creepy thing where I'm on a website and see an Amazon listing for something I was going to buy but stopped myself from buying BUT THERE IT IS AGAIN.

Want to support the site and in turn get a little banner somewhere? Inquire within.

Marketplace

Yeah, it was a cool idea. You find an OC for sale you want, you pay for it. The creator gets paid. You get your OC. Want art for it? Insert plastic money card data, get art, artist gets coin.

I've refined the idea into an entire marketing brochure and business plan and everything, and then decided to gracefully let it die for now. You see, I like doing everything by the book. Like, I actually reviewed the terms and legalese of every payment processor option, and the end result is: There's dicks on this site. And because of that dick, Certain Pals of Payment and other Familiar(tm) processors said now and closed my applications.

I did, however, find a cool group of payment processors that would TOTALLY work for this operation, but that ran into Problem #2: The Court of Public Opinion.

I'm constantly worried about The Court of Public Opinion because I'm making this entire monster of a site for the Public(tm). And, at MFF I found a thread or sixteen talking about some commission website doing such horrible things as requiring bank account details, needing legal info about the sellers, accepting credit card numbers, and the most scandalous: Taking a service fee.

So I can build a marketplace and all that and skip the payment handling part, and just say "figure it out off-site, here's your quote". But what if something goes wrong in that transaction? You pay and the character is never transferred? I'm totally going to get a support ticket for that, and without the transaction info, I dunno what to say about it.

So, uh. Unless The Court of Public Opinion backs the Marketplace idea fully, and we as a whole decide how much of a tip lowly ol' Refsheet gets for providing the service, and we find a way to exchange financial data without freaking out, the marketplace is a no-go. Oh, and what is with the fear of a website "making a profit" from stuff like this? Honestly? I'm barely living on donations. I, personally, the human known as Mau, don't want to pad my pockets with coin from this. I just don't want it to be a money-loss venture.

Which is saying something, because nobody in this world should program for free.