hello everyone! we have finally finished this hellish work, which took many years of our lives and an indescribable amount of nerves, money and other things. Yesterday we started transferring old data to a new server. This is a huge amount of data and it will take almost 2 days just to copy it, one of which has already passed. Yesterday, I was just completely unable to write even a few paragraphs of text about it. For the past few weeks, I've been sleeping in a "skip every other night of sleep" mode to finally end all this mess. It took an ugly long time, and we misjudged the scale and scope of everything we were going to implement. Over the past few months, it has been almost ready, but it contained tons of bugs and fixing one led to the appearance of several new bugs. The battle with the Hydra from ancient myths. But we did it.
As soon as we complete all this, I will write an additional statement here that you can already see what we have achieved.
Basically, it will become much nicer in appearance, it will work faster, and a number of new features will appear. But most of the really powerful changes won't be visible for the first time. For example, we have a system for creating multilingual content that can translate pages directly in the browser. And it won't be a bad machine translation, it can be done by both AI and humans - just by changing the text right inside the picture. Without the tedious tinkering with Photoshop, the endless pouring of different versions into the system and their complicated publication, which would inevitably create a ton of errors and problems, not to mention the fact that it would take even more effort from the artists. We have created a system that allows you to conduct multilingual projects with minimal effort. Tools for group work by artists, translators, and all that. We have created more advanced voting systems. Updated artist interfaces for publications and much, much more.
Yes, there are a lot of new things for subscribers too - for example, now you can receive an email notification about each new publication. And it's faster to create archives, and the pictures in your update feed will be created in advance, before you appear (excluding the first login that will launch this system). There are so many changes in general that you can write another 10 pages of text about it, but I'm still very tired and I need a rest.
I will also answer the most important question that worries everyone who sees the last page and tolerates our hideously prolonged pause with the publication of the video. We haven't completed this block yet, but we hope to complete it in the coming weeks. It's 95% done, but we still can't upload huge 4k files weighing 1-5 gigabytes via cloudflare. Therefore, we had to invent one sneaky cheat, but it will take some more time to implement it. In the system, inside the server, everything is ready to work with video. But the very method of stuffing huge files inside the system is something that we still have to finish and test. I hope it won't take long, and after all the tests and checks, we'll start publishing. Considering that the pause dragged on, and we learned some more cool things, we also updated the character models again in parallel with all this fuss to make them even better. I'll show you what happened soon. We bought personal lessons from a very cool specialist who works in a large professional studio. And he taught us some really cool things. Now part of the team is applying them to models, we will redo several fragments of renderers and finally resume regular publications. Well, there will still be pauses at first, because we want to dramatically redo the location, making it much more interesting in order to expand the number of situations justified by the plot. But it won't be as unpleasant or as long as all those agonizing months of tinkering. We also got an additional computer to speed up work and renders, and we're trying to find where to buy a 5090 for another computer. In our country, this is not for sale, and in places where it is sold, the margins amount to insane amounts. This will probably have to wait. But the 5090 has more memory (32 GB), which would make it possible not to split up mass scenes that do not fit into the memory of computers with the 3090 and 4090. In the case of pictures, this was not a problem, we could render a picture with 5-7 characters by dividing it into 2-3 parts. But with a video... It's not that simple. Therefore, we start with a scene where there are only two characters - Rose and Jay, and by the time the plot returns to the mass scenes, we hope to find a solution to that problem. There are several plugins that allow you to make a separate rendering... we have purchased them, but we have not yet studied whether it works well enough and how to use it. In general, work is underway, don't lose us and don't worry. After launching the 2.0 site, our team will have a very significant amount of energy and resources to finally return to a regular schedule.