A study has come out recently, confirming what we already know. Children spending more than an hour a day engaged in social media can make them less happy. Take something like Facebook for example. You post something, people read it then they give it a like and sometimes make a comment. Now imagine you are a young impressionable child somewhere under 18. You post something and get 50 likes. Sometime later you post something else and get 65 likes and feel better. Then the third time you only get five likes and some comments about how lame it was. Now you feel worse. Multiply this by a few hundred times and the emotional roller coaster can have someone with a developing emotional platform spiralling into their first depression.
The best way to stop this is to reduce the amount of time spent in front of a screen and get them out into public with some playmates. Another problem is the growing gap between generations where the last generation is not using social media in the same way that the current one is. This reduces the ability to talk about any problems like cyber-bullying and instantaneous disappointment. As the world changes, so do the problems faced by the current generation and for some it is all they can do to keep up with it all.
Project Veritas recently published a series of undercover exposés on Twitter confirming what I'd written about in the last article. The reading of private direct messages by a phalanx of Twitter staff, including those that had been deleted by their users. Shadow banning, where someone could still send tweets, but no one sees them. Banning of people who didn't fit in with the politics of those running the organisation, including by individual staff members following their own ideals. All in all, a picture of an organisation that was anything but open. I predict that Twitter will not be around for the long term.
Intel continues to be beset by the problem of trying to tidy up the mess that is Spectre. It will be years before chips are produced that are immune to the branch target injection design flaw. In what can only be called a marketing ploy, Chipzilla is touting their Spectre mitigation strategy as a "feature" rather than what it really is, a design flaw.
Reactions from the industry have been less than supportive. Linus Torvalds, the man who looks after the Linux kernel code, declared: "As it is, the patches are COMPLETE AND UTTER GARBAGE." He then went on with more unflattering colourful language I'll let the readers look up if they are interested. The issue is that for a while future processors will include a boot time flag that will, Intel claims, along with Single Thread Indirect Branch Predictors (STIBP) and Indirect Branch Predictor Barrier (IBPB), prevent a potential attacker or malware from abusing branch prediction to read memory it shouldn't. Or in simpler terms a flag that will activate some protection against Spectre v2 attacks. Linus' objection is that this will be an opt-in process rather than the default as it should be; Intel has it backwards. It also means that operating system updates will need to set the flag. This will slow down the CPU's performance, by up to 20% in some cases. By making the bit setting opt-in any speed degradation becomes the "fault" of the OS i.e. not Intel directly, or at least initially, a very fine legal distinction.
At the time of writing, to date, there have been many problems with the various patches with some machines crashing. At this time, it is better to wait until the microcode bugs have been sorted out. As far as the roadmaps for a Spectre- and Meltdown-free future, the manufacturers have been quiet. All future generation processors will need to go through a redesign process that could significantly delay some roadmaps. One other impact has been the death of Moore's law which will now go into reverse for a while. It has had an excellent run since first posited in April 19, 1965. There may still be a bit of life left in the law for graphics processing but as far as the humble CPU goes it is the end of the road.
I'm still thinking about a upgrading my home PC. The first design decision is a motherboard that has a dedicated USB 3 system, including at least one USB C connector front and back. USB is one of those technologies that has never been quite right so no USB 2 on any new motherboard is better for both reliability and performance. CPU and graphics cards can be middle of the price point but it's worth knowing that standard SSD drives will slow down over time, so you need to get a corporate version for the system drive. They cost a bit more but have a much longer life. More on this as I flesh out the specs.
James Hein is an IT professional of over 30 years' standing. You can contact him at jclhein@gmail.com.
.jpg?w=600)
