May 12

Hi everyone! Sorry for taking a while with this post >_<

ArtCentral (long update):

I've been sorting out the CSS for the main site elements - still a fair bit to do on that front, but most of my available time has been spent working out what server architecture to go with. I'm honestly planning to work on the site layout and appearance last, at least in terms of getting things to a useable state, as I like to get the proof of concept for everything sorted out first.

If ArtCentral got big, it'd need more than one server, and doing that with my current webhosting provider could potentially be difficult, not to mention somewhat expensive. I've been experimenting with Amazon Lightsail, where it would be both a lot easier to scale up (as it offers pretty easy ways to create new webserver instances, load balancing and the like) and also quite a bit cheaper. Bonus points for the fact that all the other elements are already on Amazon anyway (S3 for object storage, CloudFront for CDN), and scaling up will mean using a dedicated database that Amazon RDS offers.

I've gotten Lightsail to work with S3, CloudFront and RDS, so all the infrastructure is in place for the site to work, minus an email server; I'll look into GSuite or something similar for that. With all that stuff mostly done, I can use AWS to host pretty much everything on the site, which should be cheaper and easier to scale. The next steps are thus:

• Write up documentation for all the AWS instance creation, etc (for reference later). Might post those here as free content if I write them in a sufficiently generic fashion. I have it done in a rough form for myself, but I'll take another look at it later.
  
  
• Work out the model for payments; Stripe's payment system and other payout systems are incredibly complex. I don't need to do all the coding now, but I want to make sure I have the concepts in my head correctly to avoid any big surprises later.
  
  
• Continue coding the site - first focus on the CSS and page layouts, then infrastructure to make the basic functionalities work, and then start implementing and testing Stripe payments and payouts. 

ArtCentral Business Model

I've also been taking a long look at the business model. As with most things, being a small business means higher costs - CloudFront CDN bandwidth in particular is a lot cheaper if you're a big customer (it's up to 4 times more expensive as a small business), so I have to figure out what maximum filesize to offer per product. A 3-4GB file for example would actually cost more to download once than ArtCentral's fee for selling it (and even more if the purchasing user downloads it more than once), which obviously isn't sustainable. I'll need to figure out whether to e.g. charge an additional fee for large files until the business gets larger, or adopt some kind of strategy to try and absorb those costs. I'm working it out, I think it can be managed properly, it's just something that needs some tweaking and planning and calculations.

I'm doing a survey of artists to find out what prices they are willing to accept, how many have a need for big file uploads, and so forth. I also need to do more research on Stripe's payment systems - there are a lot of different ways to make a sort of "digital marketplace" paymen model, each with lots of details, and I'm still getting my head around everything in that regard. Liability for chargebacks in particular differs a lot between platform types, and it's important to make sure I insulate artists from that.