posted Mar 12, 2017, 7:10 AM by Sami Lehtinen
updated Mar 12, 2017, 7:11 AM
- Got to monitor one bathroom upgrade project. It's no different from IT. You can completely screw up things. What failed? Well only: Schedule, budget, planning, bad communication, resource scheduling, quality requirements, lack of key personnel during key decision points, etc. All totally messed up. Then fixing already fixed failures by using additional ad hoc stuff which leads to really messy outcome. When something is failed, make really quick and bad fix for it and hope it works. Then when customer complains, fix bit more. So when someone talks about ICT projects, it's just like "any other project", there's no difference. You can easily create a massive fail, just by doing random things without any plan and high level project overview management. - Funny, but also very true. - Just like the book of project failure said, there are all extremely classic examples. - Nothing new here, it's all obvious.
- Private Cloud can look cheap, if you've built it for specific capacity. But the problem is that capacity requirements aren't that static. When you exceed the designed capacity, cost for adding more capacity can be quite high. It's just like the blade servers and blade enclosures. Even if individual servers would be cheapish, the enclosure can be surprisingly expensive. In case of private cloud, the step is much higher. Especialy when dealing with quite small systems.
- Low Power Wide Area Network (LPWAN) - Also checked out Weightless and NB-IOT and DASH7. All the standard stuff like Forward Error Correction (FEC), Automatic Retransmission Request (ARQ) and Adaptive Channel Coding (ACC), very low idle power consumption. - I'm just wondering if IoT jammers will be lot of fun in future. Let's see what (everything?) breaks down when I ... It's guaranteed that NSA / TAO etc are going to absolutely love these devices. As well as hackers, just for lulz or for something more serious.
- Finnish tech websites are absolutely horrible nowadays. Most of articles are really short and very badly done translations from foreign news. Worst part is that usually it's clear that the translators are totally unaware about the subject they're translating about. Which leads to key facts being often utterly and absolutely wrong. I just dropped following the last Finnish IT news site. Just because it's so bad, I really can't stand it.
- Who Will Command the Robot Armies? - That's an excellent question. But as said earlier, it's coming, there's no way to stop it. Then time is ready, inventions and technologies will happen and that's it. If you won't do it, someone else will do it. IoT Robots will indirectly contribute to the surveillance, even if possibly not designed for it in the very first place. The list of Internet of Shit devices in the post is amazing. Who could survive without smart tampon? Hahaha. Or the TweetPee. Duh! Kisha seems to be also really valuable addition. More and more crap-apps, thank you for that too! And how many critical security issues those devices are going to have? Internet of Targets is here too! - I also agree with the point that Google knows way too much.
- Future seems really awesome and extremely scary at the same time. But isn't that the norm? Only change is persistent. - Ethical questions, justification. This justification thing reminded me about the Snowden movie. Just enter justification for this action. Let's make something up, nobody's going to verify it anyway.
- Also the e-receipt projects I've been working on. Do present very interesting privacy questions and so on. Should every purchase be tracked etc? People working in Bi and data analytics know that this can be used for so much more than just 'delivering receipt to the purchaser'. At least one bonus is immediately knowing who's buying and if not 'strong identity' is provided, there's at least pseudo-anonymous unique identifier, which can be linked to real identity using multiple methods later if required. Maybe it's just time to accept that normal daily privacy is absolutely gone. If there's something you want to keep actually private and secret, it requires extraordinary measures, extremely rigid determination and following set of key and privacy rules in absolutely solid way. - As we all know, people are very bad at this and it's highly likely they'll fail sooner or later, even if they were trying very hard. All that it takes, is one very minor fail and then it's all linked again.