The subject does sometimes come up in my casual conversations, since Robin Milner was my first CS lecturer.
He never actually spoke about type inference in my presence. He did teach me CCS (pi-calculus predecessor) a couple of years later, by which time I could appreciate him.
I did some work for Apricot at their Glenrothes factory around 1985-87.
In my memory they went heavier on GEM than Windows. I never saw an Apricot running Windows prior to the PC-compatible models.
They switched focus to Windows around 1988, I believe - it was around the same time they started using the MCA bus (I believe they ended up as most successful non-IBM MCA vendor), so perhaps they had been convinced by the hype around Windows being a sort of interim OS/2?
Whatever the reason, the Qi-386 (and its ISA-based derivative Xen-i) was often combined with the Deskside Environment Pack, consisting of a trackball, infrared smartcard reader, and Win/386.
My dad's small publishing company had a bunch of them, running Aldus PageMaker and FreeHand. Lovely machines, and about half the price of the equivalent Mac IIs!
Matt Godbolt was saying recently that using tail-calls for an interpreter suits the branch predictor inside the cpu. Compared to a single big switch / computed jump.
I would have thought it actually helps the branch target predictor rather than the branch predictor. If you assume a simple predictor where the predicted target is just the last taken one then it's going to be wrong almost every time for a single switch. It will only be right for repeats of the exact same instruction.
If you have a separate switch at the end of each instruction then it will be right any time an instruction is followed by the same instruction as last time, which can probably happen quite a lot for short loops.
You would think so, wouldn't you? Then why is the FX department at my bank based in London? As a mid market client they occasionally used to send one of them out to see me. They told me all about their office and team in London...sat in front of screens. Their system aggregated the banks position onto the screen of the master trader for each currency, sat at a screen in the next office. What utter waste
FX departments are in London because that's where the FX traders want to be. There's no government department telling them to be in London. It's kind of how free markets work.
Well it only needs to be Kafka-compatible, so personally I hope that things like RedPanda will turn out to be easier to run for this use case than actual Kafka.
Having an S3-compatible store was already a fairly heavy dependency in terms of something to run correctly in production, it's just that most people don't even consider running their own object store at any real scale, they just go to cloud. Whereas running your own Kafka is something more platform teams are already attempting.
Author here.
I know the old way still works and I respect that. Given the history I ask myself how long will it work, since ist not the default anymore.
He never actually spoke about type inference in my presence. He did teach me CCS (pi-calculus predecessor) a couple of years later, by which time I could appreciate him.