Seeing what happened this week to the Linus Tech Tips YouTube channel made me realise how well we have secured in transit data, password managers, etc (LastPass was also hacked via an end user session) but we appear to have the session data left wide open on our local machines.

I see that Firefox and Edge have profile logins, but mainly to protect the login passwords. Most Chromium based browsers do have profiles, but do not even appear to have any form of login attached to them.

Surely not just the logins can be protected, and we could have 1st party and session cookie access also protected behind a profile password? Whenever you start up your browser the first time, you are prompted for the profile primary password to unlock access to passwords, extension data, and cookies? In this way, if some bad (or good) actor stole your session data (the session data would be in use and unlocked), they'd still be prompted for a password before being able to actually use it on a freshly started browser elsewhere?

Maybe this is not the best way to do it, but clearly some improvement is needed to protect against this form of data hijacking.

#technology #security #sessionhijacking #browers

With an 8-core Rockchip RK3588S SoC, the Orange Pi 5 is leaps and bounds faster than the ageing Raspberry Pi 4. With up to 32GB of RAM, the Orange Pi 5 is also capable of serving for a more diverse user-base and even has enough potential for assembling a budget Arm Linux developer desktop. Phoronix have been testing out the Orange Pi 5 the past few weeks, and have found it's quite fast and nice for its low price point.

The Orange Pi 5 comes in four different versions with 4GB, 8GB, 16GB, or 32GB of RAM using LPDDR4 or LPDDR4X. There is a 26-pin header on this SBC for those considering adapting this SBC for other purposes, but what is unclear is what sort of compatibility there is (if any) with any existing Raspberry Pi modules.

Well not only is the Raspberry Pi 4 ageing a bit, it has also been pretty difficult to get hold of, and not been too cheap either.

See https://www.phoronix.com/review/orange-pi-5

#technology #OrangePi #RaspberryPi

In Python (and any language that stores variables), a variable is a reserved memory location that stores a value. They are names that can be assigned a value and used to reference it throughout your code. Using a variable makes a value accessible & gives values a context/meaning concerning your code.

Variables are at the root of manipulating/processing data input for any programming language. This is just such an easy-to-follow short tutorial that sets out the concepts.

Python itself is free to use, so can install it (also some links in this article about how to install it on Linux and Windows) and play around a bit with these concepts.

See https://www.debugpoint.com/python-variables/

#technology #Python #programming #variables

These are fully functional Sparkfun weather stations that require no soldering to assemble. They are based around the very cheap BBC micro:bit processor. The weather station can read temperature, humidity, pressure as well as light level, soil moisture and temperature, wind direction and speed, and also has a rain gauge.

Building a weather station is a practical project that helps unlock an interest in science, and the micro:bit processor also forms the heart of many more such DIY projects.

In South Africa I see the kits and parts are all available at Communica for purchase.

See https://learn.sparkfun.com/tutorials/microclimate-kit-experiment-guide

#technology #weatherstation #microbit #science #schools

This routine can trigger as an alarm, at time you specify, to wake you with a greeting, the time, the weather forecast, notify you of calendar appointments and reminders for the day, let you know what the commute traffic is like, and read the news headlines to you.

It can be triggered either by a Samsung Bixsby Routine, or as a time trigger in the Google Assistant routine.

You can use this instead of hitting the snooze function to lie and listen to this useful information for 5 or 6 minutes, or you can get up and just take the phone with you while it is telling you about your ahead.

Watch https://youtu.be/p0JoSpmGUk0

#technology #Samsung #Android #goodmorning #routines

As I've been testing various alternative social networks over the years, I've been picking up more accounts, with different followers, on each one. As at March 2023 I'm posting to 19 profiles daily.

Although I interact and reply directly on every network daily, I have tried as far as possible to automate the blogging creation process for two reasons:

1. My own WordPress blog gives permanency (I can back it up, move it {it was on Drupal previously}, etc myself), including backing it up to the Internet Archive. It acts as a single growing repository for searching through my posts (which I use myself to refresh on something I solved before, or something I read about 2+ years ago). Alternative networks do disappear along with all your posts, e.g. on Google Plus I had over 500,000 followers and a mass of posts I did, as did the YouMe social network, MySpace, and others).
2. It is also easier to compose once and replicate as far as possible as it takes time to paste titles, tags, images etc to each one.

So I've long been looking at replication services, and there are many free services, but usually the free services only allow posting to individual profiles (not groups or pages) or to a limited number of profiles. WordPress has a mass of plugin support, but the same goes really with many only posting to a single social network. One of the better ones I saw there, that would also post to LinkedIn Pages, starts at $12 per month. I don't have sponsorships or advertising for my blogging, so every cent I pay comes from my own pocket. I do it for fun, not income (is that a pun...).

Furthermore, I also started to realise that many blog to social services are limited to just the popular networks they support. So as Mastodon rolled around, nothing supported it.

I was eventually evaluating between Integromat, IFTTT, and similar. After weighing up costs, I actually found Integromat the cheapest at $5 per month back then, and it also had a lot of flexibility as you could set up your own webhooks and connections manually for something that did not already exist. For popular networks like Twitter, LinkedIn, Tumblr, etc it was just a matter of dropping a connector in, and then linking them, and choosing what fields appear where.

I should also mention that I've always tried to base my blogs automation around the blog RSS feed. So in this case Integromat would read my blog's RSS feed for new posts (as an RSS connector), and then I'd have connectors for Twitter, LinkedIn etc connected, each with the relevant title, description, hashtags, etc fields in the correct formatting. A bonus to this, is being able to also set filters on certain connections eg. my GadgeteerZA Twitter profile is only for technology based posts, so the connector will only push to that account if it recognises the technology tag as present. Similarly, I post only open source related posts to the SAOSS Twitter profile.

Integromat became Make about a year back, and accounts were migrated over. Fortunately, my existing account was grandfathered over still at $5 per month. The point though is the process flow works for me. Even though there is not yet any connector for Mastodon, I was able to configure a webhook using the Mastodon API (not a very easy task and I should do a video about it maybe), but it works very well. Make also has a lot more functionality than I'm using for things such as full programming statements for conditions, text parsing and processing, etc. So yes it can be relatively simple, but some more advanced stuff needs some learning. The point being, though, you have a lot of scope and flexibility.

So, for example, the formatting for the Mastodon post (image available in my WordPress blog post) will include a simple title field from the RSS feed, but because the description content may be longer than 280 characters or so, I either include the 280 if it is less, or I append the word "...continued" if it is longer. I test to see if there are WordPress tags, and if so, I prepend a # in front of each.

Make is perfectly usable for zero cost, and it was how I started out using it. I see the first paid tier is now $9 per month.

Disclaimer: I have an affiliate link https://www.make.com/en/register?pc=gadgeteerza that gets you a month free on the Pro plan. I know it is not a long term discount, but if you are going to use it anyway at the normal price, it gives me a $1 per month discount. I'm posting about how I'm using Make as a few have asked why I'm blogging from WordPress to other sites. My choice to use this platform long preceded it even becoming Make, and each person should first evaluate what they want to post to where, and decide on an option that works for them. Make is not going to be for everyone, as many will go for simper options (many free options will work), and businesses have no qualms about paying double the price for plain ease of use.

Full post at https://gadgeteer.co.za/how-im-using-make-integromat-to-automate-my-blog-posting-to-8-different-social-network-profiles/

#technology #withMake #blog #social #automation

I've seen many of the debates around the Samsung Moon photos, and am glad to see it has finally veered more towards the bigger "picture" or context, regarding photography.

The reason is, you have to see this in the context of all the other "fakery" that has been on the go for 130+ years. Even back in the day of film photography, I developed my own negatives and made prints, and used various tricks to alter lighting and effects during the development process, along with an array of creative lenses and other objects whilst shooting the photos. Yes, we'd even sandwich two or more separate images together (even placing the moon where it was not).

But certainly with smartphones, computational photography is absolutely the norm. From HDR, to night photography, to Beauty Mode, Portrait Mode, filters, blurry water photos, digital zoom, saturation, and more, are all altering the subject with corrections/fixes/effects.

Take say blurry water photos: The iPhone takes a short video and processes stills captured from the video (much like HDR is done), and on Android it is often a burst of still images taken which are also then stacked and processed. Beauty mode removes some face blemishes and wrinkles. Where do you draw a really objective line between what is real and what is fake? The moment you see the word "scene mode" you know there is some heavy processing going on.

If AI has learnt what a green leave should look like, and then sharpens that as a leaf versus say a picket fence, will that be fake? But that is clearly where computational photography is headed next, i.e. recognition of objects in nature and correcting accordingly (like beauty mode recognises a face, and "corrects" it already). AI is nothing but smart recognition.

So even film or DSLR purists have to admit that a plain raw photo, with zero improvement at all, is basically never what they do.

A line can be drawn when (like happened a year or two back) a company advertised photos being taken by their smartphone, and actually a completely different camera was used.

The fact is we're going to see AI more and more playing a role in sharpening digital zoom and other types of scenes. That does often mean recognising something, so it can fill in details. Much like ChatGPT, right now those are roads that are being explored by all.

See https://www.reviewgeek.com/148335/why-every-photo-you-take-is-fake/

#technology #photogrpahy #AI #fake

Google has tried to tap into the augmented reality world several times. Augmented reality, commonly known as AR, when used properly, can potentially be life-changing technology, but no tech company has yet managed to master it. Apple has taken the approach of using dedicated hardware, such as the LiDAR scanner included in iPads and iPhones, while other companies, like Google, are pulling everything off through software only. We're definitely getting closer to perfect AR, however, and Google is one of the companies leading the pack, with their augmented reality SDK and technology, which is dubbed 'ARCore'.

But what's ARCore all about? What phones are supported? Most importantly, should you even care about ARCore at all?

Yes, I do have to say that the iPhone's LiDAR sensor makes for pretty perfect AR imagery. Under the right conditions, Android can manage very similar performance, but generally the capturing can take a bit longer and require more patience. I used my iPhone to capture a car in 3D including the interior, but I still have to get that right on the Android phone.

Still, AR can be a lot of fun, and the linked article does give some examples of its use right at the end.

See https://www.xda-developers.com/arcore/

#technology #AR #Android #AugmentedReality

Just as the gold standard for multimeters and other instrumentation likely comes in a yellow package of some sort, there is a similar household name for thermal imaging. But, if they’re known for anything other than the highest quality thermal cameras, it’s the excessively high price. There are other options around, but if you want to make sure that the finished product has some sort of quality control, you might want to consider building your own thermal imaging device like Ruslan has done.

The pocket-sized thermal camera is built around a MLX90640 sensor from Melexis which can be obtained on its own, but can also be paired with an STM32F446 board with a USB connection in order to easily connect it to a computer.

It's probably worth reading the comments at the linked article though to also realise this is not going to be the high resolution accuracy of the excessively expensive thermal imager. So you get what you pay for, but many this may well do the trick.

See https://hackaday.com/2023/03/14/pocket-sized-thermal-imager/

#technology #opensource #thermalimager