5 ms·

I think this is a bad idea. Standards are nice, but companies shouldn't be coerced into using them because there can be legitimate reasons to not use a standard

by hlandau 10y ago

I think this is a bad idea. Standards are nice, but companies shouldn't be coerced into using them because there can be legitimate reasons to not use a standard. It prevents innovation not envisioned by the standard. I really think the simplest way to make this sort of thing happen is to increase the minimum warranty period. If 8 year warranties become mandatory, companies will need to be able to stock replacement parts for much longer if they don't want to have to replace the machine outright with a newer model when it breaks. They may then find that standardising parts is in their interests, etc.

logfromblammo 10y ago
Standards can change when necessary, in a way that gives all participants in the market ample warning, so that no one goes bankrupt when the big companies suddenly change direction. When PCI was introduced, motherboard manufacturers gradually increased the number of available PCI slots and reduced the number of ISA slots until the point where most of the market really didn't want that last ISA slot as much as they needed another PCI slot. ISA is still a standard. Nobody uses it because PCI was better. Likewise, AGP appeared, and then went away, because PCIe is better. Wi-fi standards have likewise evolved, from A, to B, to G, to N, to AC. Many devices are backward-compatible. But an ATX motherboard manufactured last week will still fit in the ATX tower case I bought in 1999, with a power supply I bought in 2009. The original manufacturer of that case does not need to stock parts for it, or even still be in business, because if I want another 120mm case fan, 10 different companies can sell me a new one that will fit (and 20 more would be willing to buy a fan from one of the former 10, stencil a logo on it, and resell it to me at a markup). Perhaps it would be better to enforce a mandatory standard for 10 years, to establish it in the market, then make it voluntary again, so that innovation could occur. Once a standard exists, there has to be a really good reason for deviating from it, otherwise the market quickly allows the non-standard thing to fail. If it succeeds, it becomes a new de facto standard, and the official standard is likely to either adopt it outright or make the next version compatible. I think you absolutely do have to coerce a standard if one does not already exist. But after it exists, it is largely self-enforcing. To buck off the standard entirely, you literally have to be the size of Apple. And even then, I know for certain that there is at least one person alive that refuses to buy Apple hardware in part because of their proprietary connector shenanigans in a world where USB is a standard.
- hlandau 10y ago
  Non-standards can lead to standards. AMD's Mantle graphics API, for example, was donated to the GLnext effort, which was then renamed Vulkan. If not for this kick start, GLnext would probably still be figuring out a decent API design. Companies need to have the prerogative to do their own thing for a while, because it can lead to better standards. >I think you absolutely do have to coerce a standard if one does not already exist. I think I actually had this idea recently, with regard to print cartridges. The idea is that the government would mandate inkjet printers to adopt an open standard for print cartridges, but the law would expire after a number of years. The idea is that once it's established, no company is going to be able to get out of it without damaging their position in the market, but genuine innovation is allowed to appeal to consumers on its own merits, weighed against the downside of divergence from an established standard.
- Grishnakh 10y ago
  The ATX standard is the reason why no one bothers buying desktop computers any more except extremists and gamers. These computers are way too large and inefficient and ridiculously noisy too, because they aren't properly engineered for sound and ventilation (because they're held back by the 90s-era ATX standard). Everyone's just given up on it altogether and now they use laptops, or worse, tablets. Even corporate desktops abandoned ATX ages ago and went to proprietary SFF cases.
  - logfromblammo 10y ago
    I am moving from a laptop as my at-home computer to a quiet-build mini-ITX system literally this week, after the last parts arrive from Newegg. I think the ongoing market fragmentation resulting from the incompatible proprietary laptop form factors is pushing the whole industry towards consolidation and mergers, and will result in less consumer choice in the long run. I have been privately railing against this crap ever since Dell started shipping not-ATX-but-looks-like-it-at-first-glance power supplies in late 1990s. And Apple has always almost gratuitously used proprietary connectors and form factors, even when a perfectly usable standard already existed. They can all go to a Hell where nothing can interoperate without at least two adapters, and you never have both. I was driven to the custom build when the headphone jack on my laptop broke. I opened up the case and desoldered the broken one, and the microphone jack, and resoldered the unbroken jack into the headphone jack's place. That worked for a while. Then the laptop keyboard stopped responding to about 5 keys--all of them rather indispensable for typing words. So I replaced the keyboard, and the replacement is just a millimeter too small, so it constantly pops out of its recess whenever it shifts and the plastic clips disengage. Then the headphone-jack-formerly-microphone-jack broke like its predecessor. I already knew the component itself was impossible to find to replace it, so I looked to replace the motherboard. That would cost more than a newer, more capable laptop, if the part had even been in stock. So I'm building a system such that if anything breaks, I can just replace that one part. I can't get decent repairability without building it myself, and I can't build it myself if there aren't any rassafrassin' standards! I would have built my own clamshell computer (not even necessarily a "laptop"), but no one mass-produces an empty clamshell case for consumer sales, because there is no laptop-standard form factor. Replacement parts for laptops are always expensive, and frequently out of stock.

by hlandau 10y ago

logfromblammo 10y ago

Standards can change when necessary, in a way that gives all participants in the market ample warning, so that no one goes bankrupt when the big companies suddenly change direction. When PCI was introduced, motherboard manufacturers gradually increased the number of available PCI slots and reduced the number of ISA slots until the point where most of the market really didn't want that last ISA slot as much as they needed another PCI slot. ISA is still a standard. Nobody uses it because PCI was better. Likewise, AGP appeared, and then went away, because PCIe is better. Wi-fi standards have likewise evolved, from A, to B, to G, to N, to AC. Many devices are backward-compatible. But an ATX motherboard manufactured last week will still fit in the ATX tower case I bought in 1999, with a power supply I bought in 2009. The original manufacturer of that case does not need to stock parts for it, or even still be in business, because if I want another 120mm case fan, 10 different companies can sell me a new one that will fit (and 20 more would be willing to buy a fan from one of the former 10, stencil a logo on it, and resell it to me at a markup). Perhaps it would be better to enforce a mandatory standard for 10 years, to establish it in the market, then make it voluntary again, so that innovation could occur. Once a standard exists, there has to be a really good reason for deviating from it, otherwise the market quickly allows the non-standard thing to fail. If it succeeds, it becomes a new de facto standard, and the official standard is likely to either adopt it outright or make the next version compatible. I think you absolutely do have to coerce a standard if one does not already exist. But after it exists, it is largely self-enforcing. To buck off the standard entirely, you literally have to be the size of Apple. And even then, I know for certain that there is at least one person alive that refuses to buy Apple hardware in part because of their proprietary connector shenanigans in a world where USB is a standard.

hlandau 10y ago
Non-standards can lead to standards. AMD's Mantle graphics API, for example, was donated to the GLnext effort, which was then renamed Vulkan. If not for this kick start, GLnext would probably still be figuring out a decent API design. Companies need to have the prerogative to do their own thing for a while, because it can lead to better standards. >I think you absolutely do have to coerce a standard if one does not already exist. I think I actually had this idea recently, with regard to print cartridges. The idea is that the government would mandate inkjet printers to adopt an open standard for print cartridges, but the law would expire after a number of years. The idea is that once it's established, no company is going to be able to get out of it without damaging their position in the market, but genuine innovation is allowed to appeal to consumers on its own merits, weighed against the downside of divergence from an established standard.
Grishnakh 10y ago
The ATX standard is the reason why no one bothers buying desktop computers any more except extremists and gamers. These computers are way too large and inefficient and ridiculously noisy too, because they aren't properly engineered for sound and ventilation (because they're held back by the 90s-era ATX standard). Everyone's just given up on it altogether and now they use laptops, or worse, tablets. Even corporate desktops abandoned ATX ages ago and went to proprietary SFF cases.
- logfromblammo 10y ago
  I am moving from a laptop as my at-home computer to a quiet-build mini-ITX system literally this week, after the last parts arrive from Newegg. I think the ongoing market fragmentation resulting from the incompatible proprietary laptop form factors is pushing the whole industry towards consolidation and mergers, and will result in less consumer choice in the long run. I have been privately railing against this crap ever since Dell started shipping not-ATX-but-looks-like-it-at-first-glance power supplies in late 1990s. And Apple has always almost gratuitously used proprietary connectors and form factors, even when a perfectly usable standard already existed. They can all go to a Hell where nothing can interoperate without at least two adapters, and you never have both. I was driven to the custom build when the headphone jack on my laptop broke. I opened up the case and desoldered the broken one, and the microphone jack, and resoldered the unbroken jack into the headphone jack's place. That worked for a while. Then the laptop keyboard stopped responding to about 5 keys--all of them rather indispensable for typing words. So I replaced the keyboard, and the replacement is just a millimeter too small, so it constantly pops out of its recess whenever it shifts and the plastic clips disengage. Then the headphone-jack-formerly-microphone-jack broke like its predecessor. I already knew the component itself was impossible to find to replace it, so I looked to replace the motherboard. That would cost more than a newer, more capable laptop, if the part had even been in stock. So I'm building a system such that if anything breaks, I can just replace that one part. I can't get decent repairability without building it myself, and I can't build it myself if there aren't any rassafrassin' standards! I would have built my own clamshell computer (not even necessarily a "laptop"), but no one mass-produces an empty clamshell case for consumer sales, because there is no laptop-standard form factor. Replacement parts for laptops are always expensive, and frequently out of stock.