I could not imagine the pain being being the head of a Libre/Open project nowadays.
It used to suck when there was almost nobody contributing because it felt isolated. Worse was if your project did become popular, you would be managing loads of code AND spending loads of time managing the community than the code.
Nowadays, you probably don't have the community to push back and converse with, just a flood of code with few people you can't actually question it with. So back to isolation but with a million lines of code.
It feels like you would have to have a project with an direct artificial limit to increase the quality of submissions. Like, all submissions must be able to run on a 486 25Mhz and 4MB of RAM, this would keep the code small and manageable. Don't do that, but you get the idea.
The old hardware approach is still king when it comes to effortlessly catching someone "optimizing" something in a way that will break again in just a few generations of whatever Intel is doing. Unfortunately 486 is now gone from linux upstream (better/worse: all TSC-less or CX8-less cpus are dropped).
I get why project like Linux do that, it allows them to drop some old code and provide optimizations for more modern processors without having to worry about 30 year+ legacy chips, but you do lose something with that.
At least, others usually take up the job and keep it going with a fork.
> Like, all submissions must be able to run on a 486 25Mhz and 4MB of RAM, this would keep the code small and manageable. Don't do that, but you get the idea.
> Nowadays, you probably don't have the community to push back and converse with
Why do you think this? People still use open source software and most of them don't immediately turn to a LLM if things aren't working for them, or they want additional features. They have to ask.
What has changed immensely is support. Stack Overflow sites are dead. Forums still get traffic but the impetus to sign up used to be "help me!". LLMs are doing all this work now.
The drive is going remain engaging with people, getting donations, taking about stuff. But I suspect the route into these communities is going to have to evolve.
The last couple of months I've found it easier to just open up the Hermes app on my phone and just tell it to clone a project and build my features into it and then build the app and push it to my phone. 99% of the time I don't bother upstreaming my changes. If it's a project I use regularly I typicallt send the creator $25-50 for all the work they put into the project and for making it open source.
I think this will become commonplace over the next decade, the ability to customize something to your own needs within minutes is immensely powerful once you learn that it's possible to do.
Why not? I've been at workplaces where a PR needed to be less than X lines of code without approval from some skip level contributor vouching for it. It's usually an absurdly high limit that you'd realistically never hit in a legacy codebase, but maybe that limit would work fine to filter out massive vibecode PRs.
The goal wasn't to add bureaucracy, but to stop this precise problem: stopping a potentially rushed engineer from pushing in a massive change without oversight. Even large, quality PRs can cause all kinds of long term maintenance issues down the line, so I have no clue why vibecode is getting such grace.
I quite like the idea though... I would love to see a real resurgence in retro computing app development that actually brings it all past the "niche of a niche" stage. Even if it's just "make this fast on a XX MHz CPU" and not even targeting a retro OS... embedded could use a lot of extra talent there too.
After all, restrictions are exactly what sparks creativity in people.
I also think the Linux desktop could definitely do with a lot more restriction, which would also help with the decision paralysis people are often faced with.
The paradox of choice in action. Sometimes a limited possibility is better as you can only choose and benefit from a limited amount, but with unlimited choices opportunity losses that can stack indefinitely.
I've talked to many deceased estate cleaners over the years, I have seen them in action many times. The amount of stuff people want to keep from relatives is a rounding error of 0%.
Whole lives of collecting sentimental goods all end up going in a giant bin, a whole house cleaned out in less than 2 hours. Photo albums are rarely kept, they just go in with it all and off to the tip usually in a single morning.
What you think is important to you, is basically just trash to your relatives/kids. They will probably have want they want before you are gone.
This has been my system for a while now. The funny side is that, I have yet to have a drive failure so the amount of local backups is getting silly.
The other thing I did was I ended up collecting HDD's out of old DVR's people were throwing out. Those drives are used for cold storage but not for anything important because they already have tens of thousands of hours on them and in not the best temperature conditions. Again, non have failed yet.
Funnily enough, the only HDD I have failing is a 2.1GB Quantum Fireball from 1997 in my Pentium 1! Really should swap that for an SD card via IDE.
I've had drive failures now and then. I once dropped a 10T drive on the floor. I thoroughly examined it, and it was not only merely dead, it was really most sincerely dead!
A couple of years ago, I repaired one of a pair of drives (I think they were 750GB) that had the connectors mysteriously snapped off[1] while I was poking around in a client's system.
It was a special system, Linux-based, running proprietary software for expensive proprietary hardware.
I apologized profusely. I got them some new hard drives coming right away. I found a forgettable donor drive in the boneyard that had a close-enough footprint for the connector, harvested it with hot air, and put its connector onto one of the broken client drives.
A little bit of dd and some mdadm magic later, and they were back in business with that RAID-1 array like it never happened (other than just a few days of downtime on a system they only used occasionally).
---
A different time, 25 or more years ago now, I somehow broke a 4-pin Molex[2] power connector off of one of my own IDE drives. I don't remember the details, but I do remember that I fixed it by soldering a Y cable power splitter into the holes on the board. This gave me a drive that worked, and that had power pass-through. It was kind of nice to have (and it worked for years and years after that).
Anyway, I guess shit happens sometimes.
---
[1] I wasn't rough with anything; I was just moving wiring out of the way, as one does, to attack the dust bison that had accumulated over the previous decade or more with some canned air.
I somehow broke the connectors of both drives, anyway. I can't explain it and I don't have to, so I guess I won't bother trying. They were fine when I started and they were broken when I finished.
[2]: Those aren't Molex connectors. Everyone calls 'em that, like everyone calls adhesive bandages Band-Aids, but the old drives used a design characterized by the Amphenol Mate-N-Lok connector design which had nothing to do with Molex.
I've been programming for 50 years now. I've had all sorts of failures, including two incidents of my desktop computer burst into flames.
I now set the desktop on a piece of sheet steel. I also only use steel enclosures for them.
I have a Carver amp that runs 24/7 in my office, for the last 40 years. It's now emitting a 60 cycle hum, which I learned is due to decaying capacitors. I guess I need to get a cap kit and fix it.
This is going to be the good side of the RAM supply crunch, it is forcing developers to be a bit more mindful of how all this stuff gets used.
We could be about to have a Windows Vista to Win 7 moment. Under the hood Win 7 was largely just Vista cleaned up to be more processor and memory efficient. And it is considered one of the best MS ever shipped. Hopefully, they go back to their past and take this route.
If they had resisted the temptation to muck around with the GUI and permit people to bring their Vista era email tool into the modern view with less pain, A lot of suspicion around 7->10->11 would have gone.
And look what you did to my kid (Outlook, 10 to 11)
So yes to the memory stuff. But no to "we improved things because our design people needed a kpi"
Windows 10 was the most recent big push for memory reduction actually, back when MS still harbored ambitions in the smartphone space and wanted a single OS across smartphone, desktop, and server. The minimum RAM requirements for the 32-bit version of Windows 10 was 1GB.
Now we have smartwatches with 4GB of RAM (rolls eyes). WTF do they even do with all that memory?
I do wonder if Huawei will eventually just pivot to their own ISA so they can cut even more royalties?
5 years ago I wouldn't have considered it but after Huawei build a total swap over OS via Harmony OS and the likes of Loongson pushing their Loongarch ISA, it doesn't seem so unlikely anymore.
China's domestic CPU tech is getting pretty interesting, their GPU's are still a long way behind however. The Lisuan 7G100 looks to be the best so far and it is fine but it is still years behind Nvidia/AMD/Intel.
Probably, I see a lot of RISC-V chips coming out of China and they are fine for what they are. Progress is pretty good nowadays but they have a while to go.
As for the software, this is why I mentioned Lisuan and not Moore Threads. Lisuan's drivers are mediocre but they work. Moore Threads software is a total disaster to the point were the cards tend to perform WAY below what, on paper, their cards should be able to do.
It is an uphill battle when it comes to software because without that, your hardware is useless. Just ask Intel with their iAPX and Itanium processors. Intel couldn't overcome the software mote despite being THE processor manufacturers at those times.
Perhaps outdated but when I was hunting a Loongson laptop (just recently deprecated by OpenBSD) I got the vibe most Chinese nationals wanted x86 and pirated Windows to use software like anyone else.
Pretty much. It is an uphill battle for them in terms of mind share. It doesn't help that their top end chips are still a few years behind most new x86 chips on general work loads, and they are WAY behind on more specific things like encryption/decryption stuff due to lack of dedicated ASIC's for them.
That sadly matches the conclusion I reached after years of following that processor arch online. It just doesn't seem anyone is willing to do the hard work of maintaining something different. Other than Apple and that has its own issues...
Even current apple is not the apple that gained the big initial momentum that got it where it is. No revolutionary products or services for decade+, just milking user base and heavily investing into marketing around supposed privacy. You can take ie their phone and compare it with 5 year old model and struggle to find differences (which is valid for most but not all manufacturers at this point).
Market is simply too mature and saturated and moving phones into foldable phablets/tablets is an evolution at best. Also it started and was perfected elsewhere so nothing novel or shocking and 0 WOW factor.
Interesting. LogicFolding is something that seems so obvious in retrospect and yet somebody has to be the first to do it.
I did find it neat that it ends up actually reducing overall heat despite the multiple layer wafers as signals have less overall distance to travel as they move in layer space rather than across the chip.
I suppose you could say it's an obvious thing to try but not to implement. I remember proposing something along those lines here >10 years ago but I didn't have a good comeback for how you'd address the heat dissipation problems, or enough know-how to justify trying to research it more seriously.
Pretty much, that was why I was surprised that power use declined doing this. The reduced of signal resistance has a greater off sets than the additional thermal layers.
But my shoot from the hip thoughts are that this might end up being used more for mid range chips than high end. higher clock frequencies would probably compound the thermal laying beyond standard resistance and buffers.
It seems fundamentally easier to do the alignment in the lithography step as you only have get it correct once in the mask instead of correct for every chip every time.
It is probably like a lot of patents, some very specific details on design and the technology of interconnects between layers.
Chips have been multiple layers since the beginning practically but it was also laid out in a somewhat 2D fashion of ALU here, FPU there, cache somewhere else, but now they are stacking functionality so that is isn't so discrete, in a 2D space it would get a little fuzzy.
It isn't mind blowing ground breaking stuff but it is very cool. Evolutionary not revolutionary.
It feels like so much software was just anticipating storage to keep getting cheaper and processors to keep increasing in performance forever. Now that we have hit a significant road block on that path, we are seeing the big disconnect.
I realized that in 2005(ish) my work station was a fully decked out Dual 2Ghz G5 Powermac that had 8GB of RAM. Yes, that was OBSCENE at the time but it was being used for generating light maps on 3D scenes and we were going to use all the memory we could get and 64bit processors with their memory addressing was a nice to have. But it was 8GB, a size that is still common today.
New laptops and even desktop still come with 8GB some 21 years later, who could have seen that coming? It would be like selling a PC in 2005 with 2MB of RAM.
yes, it's totally bizarre. And RAM used to be a commodity, well sort of prior to the past two years. It definitely didn't drive the bulk of the bill of materials, however manufacturers honed in on it as a way of segmenting the market.
Apple is going down the same path as Microsoft only at a slower pace and a decade behind. I fear they will turn OSX into something like Win 11 but with slightly better design in the next decade.
I also realized that I still call it OSX even though it is macOS nowadays.
Apple Intelligence and even more so Siri isnt worth the disk space. Regular user of chatGPT's Voice/Conversational AI (while driving) and comparably the new Siri is training wheels Voice AI. You can not have a seamless, back and forth conversation with it, so I wont call it Conversational AI.
Now you may ask who talks to AI? Well this dork does in time I bet more will. Especially, for giving me directions like "That new sub shop is across the street from your favorite McDonalds." No need to fiddle with a phone or look at a screen just drive to a familiar place. I believe talking to AI is safer while driving then us on our phones or messing with screens.
> Now you may ask who talks to AI? Well this dork does in time I bet more will. Especially, for giving me directions like "That new sub shop is across the street from your favorite McDonalds." No need to fiddle with a phone or look at a screen just drive to a familiar place. I believe talking to AI is safer while driving then us on our phones or messing with screens.
I've been driving for some decades, and probably still have some decades remaining, and these are just never scenarios I've needed, or will ever need. When I get in my car, if I don't know how to get to my destination, I set it on my nav, and then I drive there. My phone stays in my pocket. I never need to use my phone or mess with my screen to find some unknown destination mid-trip.
That's cool everyone is different and I have used my phone in my car since my first iPhone for streaming music and following Google map directions. Now personally myself I dont have to look at my phone as I am now just talking something similar to another person in the car - one who knows it all.
I am surprised at how many I have funded and have made it across the finish line.
But similar to the mineserver, I lost $65 on this thing as it went quiet in 2020. It seemed like they were moving forwards but a few dud manufacturing runs combined with Covid problems just meant the whole project just folded. But I have had like a 90% success ratio so I will take the occasional loss. I am still amazed I got my Ouya!
It used to suck when there was almost nobody contributing because it felt isolated. Worse was if your project did become popular, you would be managing loads of code AND spending loads of time managing the community than the code.
Nowadays, you probably don't have the community to push back and converse with, just a flood of code with few people you can't actually question it with. So back to isolation but with a million lines of code.
It feels like you would have to have a project with an direct artificial limit to increase the quality of submissions. Like, all submissions must be able to run on a 486 25Mhz and 4MB of RAM, this would keep the code small and manageable. Don't do that, but you get the idea.