วีดีโอ

Left of the text

I guess it mostly depends on your use case. It is rare that I use the battery iron for work that would require more power than this gives. Ideally I'd like a slightly longer working time, and the second revision that has recently been released gives this. The graph can be useful for power usage more than heat, the heat is going to be almost instant and constant on the smaller cartridges anyway. If I wanted small tip, power for large ground planes, etc... I'd probably reach for my TS-80P, but I'm finding it rare that I need to use that. Usually my regular JBC T245 iron/tips cover everything else.

@@linuxjedivideo That's right. In today's electronics industry, which is advancing by leaps and bounds, tools like soldering irons are becoming more and more popular and intelligent. The only regret is that it is not possible to truly get rid of the shackles of wires. Of course, there are many varieties of soldering irons now, but in the final analysis it is just an ordinary heating tool. One aspect depends on how users use them. I agree with this. I am currently using a Bakon BK210 battery (18650, 3200mAh) iron, and also use the C210 soldering iron bit(≈2Ω), and to sleep and wake up at any time to make the battery more durable. Its only disadvantage is that the battery cannot be replaced more conveniently and the maximum power is only 12W[You see, the actual maximum power may be smaller P=U²/R (≈7W)]. It would be even better if it could reach 20W like JBC's B-Iron 210.

Sure, the trapdoor STL can be found here: www.thingiverse.com/thing:5980829 The Pi from the PiStorm32-Lite adds RAM, disk and HDMI (via an RTG driver) for now. A prototype pass-through video device exists right now to put the Amiga AGA graphics into the Pi's camera port to show on the HDMI too. No WiFi or USB yet, but I do know WiFi is being worked on. It also loads any Kickstart ROM file you want from the SD card to boot from. Michal's Patreon feed covers the software/driver development side quite well: www.patreon.com/michal_schulz

@@linuxjedivideo Thank you very much. I like the way you presented. I hope you will do further video for significant updates

I suspect the biggest issues are: 1. It might need an MMU (Amiga Linux does), and neither PiStorm implementation provides this. Although Musashi can have it with some minor code changes is. 2. It likely needs Unix specific SCSI drivers to talk to the SD card.

It's not really that surprising - the 68k series stopped evolving architecturally back in 1984 when Motorola decided to refocus their efforts on PowerPC. If we lived in some alternate universe where the 68k architecture got another 3 decades of development as happened with x86 then by now we would have some incredibly fast 68k parts. I wonder what the 68k 64 bit extensions in this parallel universe would look like?

Then don't 🙂 It is not for everyone, and that is OK.

Unfortunately the way FOSDEM works is that a topic gets a room for either half a day or a full day, they can fit as many or as few talks as they want in it. But to make it fair to everyone, 20-25 minutes is pretty typical for a talk. To be fair, every other talk I saw in the Retrocomputing Devroom was great. Amiga OS is still developed today, which is great. My Amiga 1200 was running the latest 3.2 release.

Thank you. This is actually a cut-down version of a much longer talk I had planned to give at a museum last year (and would be on TH-cam), but things didn't go to plan. Maybe later this year.

I didn't know DOSBox_RTG was a thing, but taking a quick look, it should run extremely well. As for CAMD, I don't think there are plans right now. That isn't to say there is opposition to doing it, but things like WiFi and USB drivers are a higher priority. It is open source, so you are welcome to contribute patches to support this.

That can probably be done using Musashi. We do something similar to access the Pi's Linux CLI from the Amiga.

They were all coded to the lowest common denomination for easy porting. Which made them a bit clunky. I have another video which showed that due to the tight loop memory copy they do with sprites, using a 68010 with Xenon II is much smoother than a 68000.

@@linuxjedivideoInteresting find. As the 68010 is always said to deliver no significant improvement. I actually have an 68010 laying around and planned to put into my A600 using an adapter socket (which I bought for maybe a tiny bit better performance and especially being able to use a quit-key in WHDLoad. I think this adpater would also overclockd to 14MHz IIRC), but could not get around to do so, yet. Looking forward do one day see Xenon II as it should have been :)

@@elmariachi5133 for the most things, the improvement is minimal. There is a tight loop optimisation that is about 50% faster, but it is only in very specific circumstances. Luckily the Bitmap Brothers games hit that state often.

There is a very long answer to that which in summary is "it depends, but that isn't always the case". But, does it really matter? There are very few things the current chip RAM performance impacts. Compatibility has been achieved with pretty much every game and demo that exists now.

I can only imagine. The last 4 years in particular have been a nightmare for anyone working for the NHS unfortunately. The longer it takes to fix, the more everyone suffers. People who can do those jobs with a smile and care for the patients despite everything going on behind the scenes are real heroes.

This is a cut down talk of a much longer one I planned to give at the RMC Cave last year, but it didn't work out. Hopefully we will make that happen soon.

It is definitely possible to do so. But the effort required is high, and the amount of things that really need it is very low (anything other than Linux?). It would also incur a performance hit, but that is true of 68000 MMUs anyway. I doubt it is a high priority for Michal.

Lol! No speaker on-board. Maybe for v3.0, although that may be a distraction for when I'm speaking at conferences 😁

I have no experience of the 2.3mm, so I can't really advise. But I do really like the 1.9mm one. I also primarily use a 2.0mm spoon T210 tip on my JBC B-IRON.

@@linuxjedivideo Thanks for advice keep up the great work - from Australia

You would probably want parallel or QSPI for higher resolutions with a nice frame rate. A quick back of the envelope shows that over normal SPI 320*240 16bpp will be 16fps on normal 20MHz SPI. 640*480 16bpp will be 4fps.

I don't think so, I've considered making a Zorro based version before. Maybe I should.

@@linuxjedivideo Would be nice, I hate ripping out the kickstart chips all the time switching to Diagrom or upgrading kickstart

Yes, there is an RGBtoHDMI board (my design) in the Denise socket. At the time of recording, it was missing the Pi Zero that would normally be connected to it.

Information here: linuxjedi.co.uk/2023/02/10/interfacing-a-raspberry-pi-pico-to-an-ssd1351-oled-display/

...or just faster?

You can sort of see in the intro that the 68000 gets a few frames behind (these were frame synched at the beginning of the video). But more importantly, in game Xenon II starts to slow the whole game down when there are lots of sprites on the screen, this doesn't happen with the 68010. This section of the first level shows it quite well due to the snake like enemies that use many sprites.

@@linuxjedivideo There's an small improvement for what i can see. However, the overall performance boost for the 010 was around a 9% due the cache inclusion. I'm not 100% sure about this statement though.

@@linuxjedivideo Oh, maybe longer videou would be better to see or notice the difference. I was playing Xenon on PI4 as set on A1200 performance, but I did not see any significant speedup.

@@MartinDurik the A1200 is about 2x faster than the CDTV, so it should run quite smoothly on that. When you are playing it makes a big difference. Main reason for a short video is I just didn't have time to capture a long one and I was asked about it. I'm sure many can reproduce this with their hardware.

@@davidv776 the performance boost due to the cache was about 50%, but only in code that could be cached, which are single instruction loops, like the kind to move memory around. These are used a lot in CPU heavy games such as the Bitmap Brothers games, due to Atari ST being targeted as the common denominator. On average it works out to a 9-10% performance increase in benchmarks, but benchmarks don't show the whole story.

Unfortunately I am not. I do get asked that quite a bit, turns out my name is popular amongst people working in tech 🙂