So what kind of changes to hardware can we expect in the next 25 years?...
...We can certainly expect huge leaps in memory, storage, bandwidth and processing power as we saw in the last 25 years.
It's probably a bit optimistic to expect a similar level of performance gains. In the 90s and early 2000s, the generational improvements of computer hardware were advancing at a rapid pace, but that has since slowed considerably over the last decade or so. I would expect the advancements to slow even more, unless there is some huge breakthrough that really opens things up. Clock rates are barely improving, IPC isn't exactly advancing by leaps and bounds, and adding more cores is bound to become impractical past a certain point. Each new process shrink is becoming more expensive than the last, and eventually the gains from making things smaller are likely to dry up almost entirely. There will likely be big advancements in some areas, but others may see relatively little improvement.
Sadly, software has not always kept up with hardware's journey. Encarta used to be the best encyclopedia and now students have to comb the Internet trying to build their own.
I would disagree. Encarta might have been amazing back in the 90s, but even then, the amount of content available in it often tended to be limited compared to a set of physical encyclopedias, and both were arguably made obsolete by online resources like Wikipedia. Today, Wikipedia offers over 6.3 million English-language articles at no cost to users, whereas the online version of Encarta only ever got up to just under 1% of that article-count at the time it was shut down in 2009, and most articles contained less information and were locked behind a pay-wall. And while Wikipedia might be more prone to things like vandalism, most of those issues were largely addressed back in it's earlier years, and for the most part it seems to be pretty good at providing reliable information these days. And even traditional encyclopedias and software like Encarta were not immune from containing incorrect information and editor bias at times, and information about many topics tended to get obsolete rather quick. While Encarta might have been revolutionary for what it offered back then, Wikipedia took that premise to another level, and is arguably one of the most useful things to come out of the Internet over the last couple decades.
While there were only a few viruses and worms back in the day, today new ones are being introduced every week. Worst of all, back in 1996 it was easy to feel and be safe online, but today most pieces of software can be corrupted to steal information and create backdoors.
Malware from that era also tended to not be as destructive or profit-driven as a lot of the stuff out there today, even if computer systems were technically less secure. Most of it was just people experimenting and showing off what they could do, rather than holding systems for ransom and stealing financial details. Of course, back then, there wasn't really as much to steal, as consumers were generally not performing online transactions or holding large stores of data on their PCs, and those who were online tended to be more tech-savvy than the majority of the populace.
Hardware will continue to improve, but companies like Microsoft have held software back over the years more than they have pushed it forward to match the success of hardware.
People don't want their operating system and software interfaces to change drastically from one version to the next though, so I wouldn't necessarily say they have been "holding it back". If anything, big changes to software often tend to make it worse, as a lot of applications were refined to get the most out of them many years ago, and changes can potentially detract from what they already do well.
Maybe, just maybe, Arm's architecture and the transition to it will force major positive changes where software is concerned.
I doubt a wider adoption of ARM would improve much of anything on the software side of things. And if anything, most devices using ARM today have stripped down software that is in many ways worse than what's been available on PCs for decades. These devices also tend to greatly restrict what a user can do with them, often to allow a company to more heavily monetize a software, content and user-data ecosystem. The backward-compatibility of Windows and X86 has left it less affected by these trends.
Say what?! My first keyboard was from 1995, and I only replaced it in 2020. It has no problems whatsoever despite some heavy Mortal Kombat gaming in the early years and heavy use overall. It was one of the first membrane keyboards, as most keyboards before that were... mechanical! And I don't see how a keyboard that served 25 years of heavy daily use can be called horrible...
He's likely referring to the quality of the typing experience, which is generally considered to be "better" on mechanical keyboards. The move to membrane keyboards was mostly done to cut costs as computers became more mainstream and pricing became more competitive, as sheets of rubber and thin plastic with printed traces tend to be a lot less expensive to manufacture than a hundred or so individual switches. And of course, not all keyboards are created equal, and yours may have been better than the average membrane keyboard. I still often use a 15+ year old membrane keyboard, though it was one of the more premium models from Logitech at the time.
Facebook
Twitter
Instagram