Blog December 2025¶
LINKS
DeviantArt¶
03.Dec.2025
I've been a member for a week. My gallery is a mix of different ideas: period scenes and feminisation scenarios and even music. So far (11:35am) I have 47 'deviations', 1664 profile views and 171 watchers.
All in one week.
Un-freakin'-believable.
Of course, the tame images get only a few views: it's the 'feminise' stuff that seems most popular. And yet, when I look at other members' galleries, my stuff is truly tame. As in: nothing really objectionable (p*rn).
My aim is humour and enlightenment: the LGBTQ+ community are so demonised while at the same time eroticised that humour doesn't seem a thing. But, it is. We can laugh at ourselves. Even while we're crying, inside.
My goal is to, to the best of my ability and with the limited resources I have, amplify the lived experience of a woman to all the TG-attracted blokes on DA. No frilly maid's dresses, no Loctober or Cuck stuff. Too much is made of having a gynarchy rule the world, when most blokes can't even tolerate having their views questioned. It's all horseshit. Right now, 180-odd transgender-attracted or cross-dressing or fem-dreaming blokes (pretty sure it's mostly blokes...) are following me. I will eventually disappoint them. It will be a slow grind towards pointing out the true lived experience of an AFAB person, and I really think that cold, hard exposure to that reality will end up popping a lot of 'girly-fantasy' balloons. It's all in the interest of education, helping to cross the gender divide, of course. Will I succeed? Unlikely... once the male fantasy is no longer addressed, they move on. No dopamine hit. No erotic tingle. Bo-o-oring.
We'll see.
That Honeymoon Glow¶
09.Dec.2025
I was sort-of getting over DeviantArt by the 3rd. One 'lass' whose name trumpets 'her' proclivity for quaffing, well, I'd rather not say - just leave it to a repulsed imagination - keeps commenting on my images, making it all about 'her'. Poor thing. I wish I could say that those 244 Watchers were comprised of the finest and the best society has to offer, but I would be trifling with the truth. Some are, certainly.
The site has racked up 2.3k individual viewer views. One image in particular has been quite popular: 11.2k views... 130 have copied it to their "Favourite" folders. Of course I had to have a peek at their folders, their galleries. Which resulted in this feeling I was sort-of done with DeviantArt.
Still on the fence about that.
Set up AI-Toolkit in Windows 11: this will allow creating LoRAs for Z-Image-Turbo, which has become all the rage. Whilst there seems to be a new model coming out every few days which blows everything else out of the water, this one truly ticks a LOT of boxes. Will run on Low VRAM systems. Generates high-quality accurate images in almost no time at all. SDXL was the goto vision model for years: I reckon this one will take its place, if it hasn't already.
I tried running the .safetensor version on another PC using an 8gig VRAM card. Well, I don't think 8gig is actually available, based on what I'm reading in terminal. However, the .GGUFs do load, so there's that.
Z-Image, the New SDXL¶
15.Dec.2025
Absolutely! Super-powerful, highly accurate, easy to make LoRAs for, runs on the smell of an oily rag: what's not to like. Also, NSFW... which means: no restictions.
I posted (twice) a question I was having on the LoRA Hook nodes, which were central to my 2-character consistent face workflow, but with Z-Image, I rather think this is going to become a moot issue. I might try to find a solution in that arena instead.
The next challenge is going to be: build character LoRAs in Z-Image. Time-consuming but doable on my system, apparently.