David Shadoff

Latch-Up 2026: May 1-3 in Waterloo, Ontario, Canada The Latch-Up conference is a weekend of presentations and networking dedicated to free and open source silicon. It's an event for the open source digital design community, much like its European siste...

I've been hoping for more conferences in Canada for a few years now, and I just realized Latch-Up will be in Waterloo Ontario this year. (Within driving distance from me).
I'll do what I can to attend.

fossi-foundation.org/latch-up/2026

As an estimate, I think most food and restaurant prices rose by about 10% since last February, which must be difficult for the people living there.

Plus, I couldn't even find the 5-year-aged version in stores while I was there, unfortunately.

加賀梅酒 五年熟成 - KAGAUMESHU 5 Years Aged｜萬歳楽 紅映梅を5年熟成させた、深い旨みとすっきりした余韻の加賀梅酒。まろやかな味わいで飲み方も自在。特別な日の一本やギフトにもおすすめです。

While in Japan, I noticed lots of prices rose since last year.
Example: my favourite plum liqueur - Kaga Umeshu - went from 1580 yen to 1980 yen.
But the sublime 5-year aged version rose from 3180 to 5500 yen, probably because it has been winning awards.
www.shop-manzairaku.jp/products/?id...

Happy end:
I got an email overnight, stating that they automatically updated my order to the new M5-based one; I checked the order, and the price was also the new lower price (they hadn't shipped yet, so they hadn't billed yet).

Knowing that:
-RAM and storage prices are soaring
-Apple hadn't increased their prices on these yet, and
-Apple was planning a release this week
...I ordered a Macbook to "beat the increase".
Instead, while they increased the base price, they also increased storage, making the config LESS expensive.

Not dead yet! Just been busy with other things :P
Here's a miniboss I'm working on for Special Ninja Squad. Featuring a bike chase because I listen to you guys :P

I should have most of the assets now. Just need to draw a few extra tiles and off to implementation!

#pcengine #gamedev #turbografx

Got a new piece here everyone can read on PC-FX exclusive Last Imperial Prince!

Screenshots are sure to prompt nostalgic feelings of time spent gazing at screenshots of niche, unattainable Japanese RPGs - just make sure you don't actually play it:

kimimithegameeatingshemonster.com/2026/03/02/l...

Starbucks Japan is an entirely different place than North America. I quite enjoy it.

Wrapping up the trip and heading home in a few hours.
I can’t believe after all these years I had never tried a proper Ichigo Daifuku. They’re delicious !

Yeah, Game Lab was still a great technical magazine long after all the ones in the west seemed to give up on detailed info.
Incidentally, I co-wrote an article in 2007 for an issue of Game Lab focused on PC Engine… (although translator friends did the difficult part)

this is a thing of beauty

Screenshot of the PicoIDE Crowd Supply campaign, showing $224,772 and 643% in funding.

Whoops, I missed my chance to make the “640% should be enough for everybody” joke. But it’s obviously not enough because the PicoIDE campaign keeps going. 6 days left! I’m blown away by how far it’s come. BTW the last update I posted brought even more people with weird systems out of the woodwork!

Uh, it was the future in 1996. But it’s still not perfect, and only deals with the actual gameplay (not the instructions or other materials).

I don't disagree, but Yahoo auctions is generally cheaper than the walk-in stores, despite the increased likelihood of competitive bidding increasing prices.
Today I went to Hard-Off in Taisho Osaka, and saw multiple foreign tourists inside - but none outside. Foot traffic may be the larger factor.

Thanks; I actually already knew of this store.
Actually, I already have a full set of PC Engine and PC-FX, and have made annual visits to Japan for about 20 years now. These days, I'm just watching for very obscure things, while observing price changes and what becomes rare over time.

Checking out used videogames in Japan, the prices are even higher than I expected for the old ones like PC Engine/PC-FX.

It seems to me that PS Vita is probably the "sweet spot" for pricing right now (though I'm not an expert on this system).

Capcom did a set of 5 or so arcade games based on SuperGrafx hardware. These were intended to be super short games like Jankenpon (single try). I don't have the list offhand, but I believe they included licensed characters too; maybe related ?

re @anthropic.com:

I've identified industrial-scale copyright violations on my content by Anthropic, OpenAI, Google, X, and more.

These companies created thousands of crawlers incorporating the text of all my blog posts, open source code, and books into their paid AI models to profit exorbitantly.

Hint: It'll be at least another 10-20 years...

The Challenges Of Simulating A Human Brain On A Supercomputer It’s quite the understatement to say that at this point in time we don’t quite understand how even the tiniest brain works exactly. Much of this is due to the sheer complexity and scale…

I feel like this type of research will lead us closer to the age of Artificial Intelligence that people keep claiming has arrived (every 10-20 years or so).

hackaday.com/2026/02/24/t...

And honestly, that is about the pattern-recognition type of AI which I feel has much more promise (when used within a protocol). In my earlier comments, I was referring to the more haphazard generative AI products which are being actively pushed aggressively.

Computer‐assisted diagnosis techniques (dermoscopy and spectroscopy‐based) for diagnosing skin cancer in adults - Ferrante di Ruffano, L - 2018 | Cochrane Library Select your preferred language for Cochrane reviews and other content. Sections without translation will be in English.

A single study is a data point. Check Cochrane meta-analysis when checking medical studies.
Below example states "Studies were incompletely reported and at unclear to high risk of bias across all domains" (and other deficiencies in reporting).

Example: www.cochranelibrary.com/cdsr/doi/10....

With respect to my earlier AI comments, it seems clear that there are people with strong views on both sides; my point was less about whether it’s an interesting technology in a laboratory setting and more about people selling it as something that it fairly obviously is not.

Those are all constrained in the degree to which they are random, and thus deterministic within well-defined parameters.

None taken. People invested in alchemy for quite a while, but it didn’t end up benefiting anybody (except quacks and “practitioners” patronized by the rich) until it turned into the study of chemistry several hundred years later.

I think we should ask the question of whether that is a result of good doctors improving (unlikely), or bad doctors being forced to review their initial analysis. Or whether the outcome of the study was predetermined by selecting only data which favoured the study’s sponsor.

It seems that AI is useful only for things that "you couldn't be arsed to do properly".
...Which arguably, is a non-zero set - but doesn't include production code.

I've been seeing more mania surrounding Claude and OpenClaw these past 2-3 weeks, and it just reminds me that all useful technological advancements since the beginnings of mathematics are based on predictable behaviours and repeatability.
I'll take a look again when it has those two attributes.

"Anthropic Accuses Chinese Firms of Stealing Claude AI Data".
I don't have any sympathy for companies who actively try to subvert copyright, argue (successfully) in courts that it doesn't afford such protection in their realm due to "transformation", and then cry that they aren't protected.