| Current worldwide Internet problems and how they might effect us Posted by grahame at 12:52, 21st November 2026 | ![]() ![]() ![]() |
When the Internet (and the Worldwide Web part of it) was created, it was hailed as a great step forward for communications - for us understanding and co-operating with each other no matter where we lived, and for us to humanise one another so that we had more compassion for the whole world and not just locally. Alas, whilst we have made so many good-side gains, there have also been negatives made possible by the technology. The less than morally honest will take advantage, those wanting to profit by business or war means have a tool which aids them in being more effective in those endeavours, and those who hold authority over us have a wonderful tool which too often tempts them to add rules, regulations and controls which are over the top and make like for the ordinary person ever more complex. Don't get me wrong - I wouldn't be without it - but my goodness there's bad as well as good.
Artificial Intelligence is now - 30 years after the internet became mainstream - the next revolution. It can, and already is, hailed for and doing great things; I am using aspects in my day to day life - 'work' and I hope those aspects are used very much for the better rather than the worst. However, again there is a dark side too and where most of us take advantage of the tools provided for what is good, there are those elements who may look to push it riskily into areas where it can do unintended harm, or indeed into areas where harm is intended to suit there own agenda. And the compute power, algorithms and potential consequences are so new that technicians, laws, and enforcers have to be reactive at times rathe than proactive.
In recent weeks, the power of AI tools have discovered flaws in systems and software - some of which have existed for years or even decades, and we have seen a number of issues. There's been a big concern about PHP. Outlook email went down, as did the USA's air traffic control. A bank transaction that I put through and usually takes 2 hours took ten days. And some commercial freezers thawed out - which might sound funny until you realise that there might have been a serious risk to the food supply chain. It's not my purpose in this posting to start listing other potential problem areas - suffice to say that our government is suggesting we stockpile some food and other items such that we could cope with an outage. Indeed, yesterday Lisa and I reviewed where we stand though a couple of "what - if" scenarios.
Some of the issues with new AI techniques is that it's accelerated by an order of magnitude the speed with with flaws can be found, and the speed with which they can be shared on the darker side of the web. Malicious people, organisations and governments can take them up very quickly. And as this happens and as flaws are found they are fixed - except that there are not enough skilled people around to develop the fixes, and they are often implemented quickly and with minimal testing, so that at times a fix actually introduces a worse problem. Oops.
So where do we stand with our "Coffee Shop" forum?
Starting point: We don't hold critical or personal data beyond registered email addresses and visiting and posting histories. We have a small number of areas - accounting for under 10% of our content - tat's not public available, and a mirror of lots of documents which can be searched but are not on public release due to their copyright not being with us. At worst, a leak might reveal messages between moderators concerning a thread and how and whether an intervention should be made, or revealing personal plans a member happily shared with members but did not make public (and is probably out of date by now anyway. We do NOT hold credit card details, medical records, nor do we have any data which is safety critical or controlling of any real world operations. We have no ability more than any server would to mess up other online resources is something malignant was injected, and whilst some members would be very unhappy if the Coffee shop just stopped functioning, that would not put lives, health, wellbeing or personal data at risk.
And so
1. Both of our servers run "heartbeat" scripts so that failures significant performance changes are noted and come to my attention quickly. Heartbeat parameters get logged, so that smaller changes get noticed and recorded, and can be overviewed through a graphic web interface if there are concerns. Individual accesses to the web servers and other external interfaces are logged and available to look through if necessary.
2. Our servers back up all the websites hosted on the machines twice a week, and the full databases twice a day. The worker server also backs up the dynamic Coffee Shop databases every 4 hour, including posts and personal messages. These backups are help on the servers to allow for quick restoration if need be. About once a week, I download a set of backups and server configuration files onto my laptop so that if push comes to shove, I can re-upload them for a fresh onto brand new machines. A couple of mature backups are also kept so that in the event of a latent problem showing up we can backtrack the main setup significantly, and with a plan to have a much more recent update
3. Should the whole go tits up, I can redirect the name service to other computers ... and if the name server is down, try the domain passenger.chat, wellho.net and vcrp.uk which are hosted in different places.














