[yorgos_kats]

Thanks for this comment and taking the time, lot to unpack.

From a technical standpoint, you are correct that this is a massive challenge. However for us to eventually get something print production ready, we needed to build our own extraction software that preserves stylesheet, images, sequence etc. We will definitely face a lot of failures initially, but hopefully will be able to build interesting machine learning models that will make the service much better over time. We have already built tools that will enable us to fix most of the possible issues pre-production.

About the legal concerns, we will be printing content, we know 100% user already has access to, for its personal use. Won't be any different than creating a pdf file (eg from file print) and going to a print shop with a usb to print it out (or stitch).

Regarding the video, it is slow, but this is the best we have made it to snapshot html, send it to our servers and extract the relevant info. We aim to make it faster, but that's the reality and didn't want to show anything fake. Rendering could take even longer depending on size etc, obv tons of room for optimisation.

[boffinism]

Re the video, I get being honest, but no one wants to watch a video that's just endless loading animations. Instead how about:

Shot 1: an article Shot 2: "30 seconds later..." Shot 3: the finished product

That way I can see the product in action, know how long it actually takes to work, and not waste 30s of my life unnecessarily

[yorgos_kats]

Fair point and thanks for the feedback and suggestion!