c0l0

I've been using Gadgetbridge, installed via F-Droid, to interface with my Xiaomi Mi Band 4 for years now and can only recommend it. For my device, it's 100% feature-complete, it's VERY polished design- and UX-wise, and it comes without any ties to any vendor's data-siphoning cloud services. You can even set it up so that it works in parallel with your vendor-provided app, by not deleting any data from the device it fetches, if you're yet uncertain you actually want to use it as a replacement for your current app.

Free software at its very best, really!

I disagree. All of our effort should go towards all (relatively) easily attainable reductions of emissions in parallel. Substituting consuming meat with consuming plants is such a thing for me, so I readily do it.

Maybe check out https://ourworldindata.org/emissions-by-sector

Since we had grasped the implications of this graph[0] a few months back, the wife and I finally managed to actually make the jump to drastically reduce our meat consumption, and also mostly replaced dairy products with soy- and oats-based alternatives. I've even had the McDonald's "McPlant" three or so times since then, and must say that it tastes about the same as I recall other burgers' taste with similar toppings. Turns out I'm totally fine with "fake" meat, and I would encourage everyone to give it a shot.

[0]: https://coloss.us.to/food_co2.jpg

It does exactly as much as using `sysupgrade` with any other suitable image will or would do :) So yes, limitations apply - but for most users with a bunch of custom installed packages, this is enough of a game changer regardless.

With 80MHz channels configured (160Mhz are possible, but I do not need/want that) on the 5GHz band, I can get about 480Mbps TCP bandwidth over the air using iperf3. Client is a laptop with Intel Wireless 8260 running GNU/Linux.

Reports on the OpenWrt forums I've seen range up to ~800Mbps.

There is a VERY elegant solution to this problem, called "attended sysupgrade", that is not as widely known as it should be.

It's probably best explained in a video on what may be the best youtube channel on OpenWrt-related things: https://www.youtube.com/watch?v=FFTPA6GkJjg

Or, if you prefer textual information: https://openwrt.org/docs/guide-user/installation/attended.sy... (OpenWrt's infra is kinda slow today for me, must be many people checking out the new release :))

You get frequent software updates with meaningful changelogs and detailed lists of bugs/vulnerabilities fixed.

Also, you can be quite sure that there are no software components active that serve the interestest of the device's manufacturer more than those of its owner and users.

Since you specify US$, I'm assuming you are based in the US. Walmart had special offers of the at the moment probably best-suited OpenWrt device there is only a few days ago, where they sold the Belkin RT3200/Linksys E8450 for 60US$. If you can get that, it's definitely worth the 10US$ premium. Check https://forum.openwrt.org/t/belkin-rt3200-currently-60-at-wa... for pointers.

This is a pretty significant release - it marks the end of iptables in the main distribution; the OpenWrt-specific translation layer (previously firewall3, now firewall4) for netfilter-related configuration now targets nftables. Also, yet more devices have made the switch from swconfig to DSA for configuring Ethernet switches.

I've been using 22.03 release candidates on all my networking gear (router, switch, access point) for several weeks without problems, and can only recommend giving OpenWrt a chance :)

I never really "googled" it in the first place.

Since I started using a smartphone/Android, I've been on CyanogenMod/LineageOS without Play services installed. I use FOSS apps from f-droid (Firefox, K-9 Mail, osmand~, DAVx5, Gadgetbridge, Signal) exclusively, and self-host the server-side of all my email (postfix/dovecot/amavis/opendkim) and CalDAV/CardDAV (radicale) stuff. I even set up my own public-ish DNS recursor that all networks I take care of use.

Never regretted any of it, but I'm aware few will find it enjoyable taking care of some of this infrastructure. I am lucky that I do :)

This - the people involved actually giving a shit - is exactly why FOSS projects' volunteer support (on IRC, for example) regularly trumps the crud experience you get from huge, established, and deep-pocketed vendors of alternative products that are astonishingly expensive. Ironically, often the most extreme in the "support" contract department.

Being able to specify a "super-secret-master-password" on the argument vector of the program when invoking it is a Really Bad Idea, since, at least for a brief window of time, all logged-in users on a UNIX-ish box will be able to see it in the clear in the process table...

I've played it for some 30 hours together with my much younger brother-in-law at the beginning of the pandemic, and despite (and sometimes even because of) the glitches involved, we had a great time.

I think it would be a mistake to go in and expect (or demand) a perfectly polished experience. That they even managed to retrofit a multiplayer mode onto a strictly single player game is a completely mind-blowing achievement in my book - esp. given how well it works, all things considered.

Ryzen APUs (i.e., Socket AM4 CPUs with with integrated AMD graphics) do _not_ support ECC UDIMM, _unless_ they carry a "Pro" in their name.

Ryzen CPUs without an iGPU _will_ support ECC UDIMM, _unless_ the mainboard specs specifically tell you that it won't support ECC, or simply omit mentioning ECC UDIMM at all.

Fwiw, I've enjoyed proper ECC on an ASRock Fatal1ty B450 Gaming-ITX/ac with a Ryzen 5 3600 for more than two years now (using GNU/Linux; I am not sure how Windows would fare).

"Pride month's over anyway, folks - let's pack it in!"

This is kinda naive to wish for - tar has applications that reach beyond stuffing a BLOB containing other BLOBs and some metadata into a seekable file somewhere, such as archiving data to magnetic tape/LTO.

And even if you develop the perfect archive format that happens to nail every possible use-case 100% (which you won't, because there just are too many), you will STILL have to deal with compressed and archived artifacts which accumulated over the last six or so decades in various places.

Many (most?) toilets around where I live do let you exert "analog" control over the flushing process, at least when initiating it. On the one in my apartment, If I begin to press the flush button very gently, water will start flowing in at a very reduced rate (and carrying material with it down the drain without causing a splash), compared to when just pushing down the button without any consideration.

Full-on Baader-Meinhof here, seeing this posted while I am at home/in the hospital (for two days about a week ago) for the second week because of a C. diff infection I got from an antibiotics treatment for the tonsillitis prior.

I do know for a fact that I never want to feel this sick again. If that requires me to keep the toilet lid closed whenever flushing for all eternity, I'll gladly do so - and I think you should, too. (I was actually advised to do that and flush 'carefully'/slowly, with the lid down, by my doctors, and took care to comply.)

I've built a few perls over the past decade, and I am not sure I would want to use this build as per its linked Dockerfile as a (universal) starting point.

It does not verify the integrity and/or authenticity of the 5.36 release tarball, specify -fPIC twice (seems sloppy, or is there a reason for this?), and unconditionally compiles threading support (which will cause some workloads to exhibit reduced performance).

Disabling the test suite on a Linux distro that does not use glibc might be a bad idea, too...

I always loved the in-depth nature and clear writing style of the man page of the hwclock(8) program distributed with util-linux(-ng), and happily refer to it as an example of great documentation whenever I get a chance to: https://www.man7.org/linux/man-pages/man8/hwclock.8.html

Actually, everything is a routing problem.

Zotac ZBOX Series have always been absolutely solid for me, but they are not that cheap any more...

I bought my latest x86-based router from AliExpress, and never regretted it! Review available here: https://johannes.truschnigg.info/reviews/2021-01_fwbox/

It would be __very__ interesting to learn how much bandwidth this beautiful beast can saturate with cake managing SQM (under Linux, esp. OpenWrt), and if proper support for ECC UDIMM is implemented. What an awesome piece of networking machinery!

This post nicely sums up why I just cannot get excited about anyone's announcement of some new and (on paper) awesome SBC with specs much better than those of a contemporary Raspberry Pi.

Software ecosystem support is THE killer feature for this class of device. And the RPI has, despite its fundamental shortcomings (needing the proprietary GPU to manage bootup, etc.), the best story in that particular department by so much that it's not even a contest.

I find it exceedingly hard to tell whether headlines like these lead to articles that are in fact satirical, or not...

Please continue using "loading spinners", as they are a design language element that is, in this day and age, practically universally understood.

Yeah, I am disappointed that fanless devices have been on the retreat these past several years. My wife's previous laptop was a Lenovo Yoga 710, with a fanless Intel Core m3-6Y30 (Skylake, 2 cores with HT). Performance was more than adequate for her work as a scientist. My current laptop is a Xiaomi Mi Air 12.5, with the same CPU installed. It's powerful enough for everything I do on the go. If I were to replace it, there's just no up to date fanless device available except for the Macbook Air - which will have to have decent GNU/Linux support before I can/will consider buying it.

But it's not just laptops that lose their fanlessness (which I appreciate both due to noise and reliability concerns), but also NUCs - almost every single NUC-like device that has entered the market and managed to come to my attention over the last months has had a fan installed. I wonder why that is - TDPs do not seem to have gone up, at least not when looking at data sheets... maybe it's just cheaper to make stuff that way.

Does this pass the vt100 torture test (http://www.inwap.com/pdp10/vt100test.txt)?

Feels like the right time to repost this aperiodic reminder: https://coloss.us.to/chrome_truth.png