vicaya

This is a must for AR/VR glasses. I thought that all android devices had this capability. Was disappointed to find out otherwise. I had to use a laptop to use the glasses on a plane.

Worker should really adopt Apache Arrow, which has a much bigger ecosystem.

https://github.com/apache/arrow

LI is an interesting data source for the pulse of economy (at least in my echo chamber :). I don't even need to log in, just a quick glance in my LI email folder would tell a lot.

colima pretty much solves dev experience for docker and k8s on mac, esp. for apple silicons (m1/m2), where you can build multi-arch containers with ease.

Some interesting caveats:

* By default, system packages don't persist, as the default alpine distribution runs on tmpfs and doesn't have a overlay. This is a reasonable default, as it keeps the default VM storage small.

* If you want to have additional system packages, you can turn on a ubuntu overlay that supports additional systemd services just fine. Of course, storage would balloon to a few GBs from a few hundred MBs.

Edit: typos.

BTW, the result of docker build is immediately available to the k8s (k3s) cluster without any insecure registry and/or side loading/caching steps, thanks to the seamless buildkit integration.

55+ vegan OMAD here. Experienced positive recomposition (17% BF to 11%) according to the DEXA scans: grew muscle and lost fat over the same period (1 year) with the same diet (no bulking and cutting phases). Didn't measure for the previous 6 months period, where I lost ~20kg (probably from 25+% to 17%) with the same diet.

My OMAD is technically 2 sub-meals over 4 hours: 4pm: protein shake pre-workout, 5-6pm, weighted calisthenics, 6-8pm, dinner and dessert (protein ice creams).

Could be newbie gains, as I never exercised consistently before 50s: from zero pull-ups to 20+ (+40kg 1RM) over a year; zero dip to weighted (+30kg) RTO dips. zero full (ATG) squat to 10+ pistol squats.

The vegan OMAD has been easy for me, but I've never done other diet and exercise regimens before for comparisons.

The original git approach is still the best, which I've been using for years:

  # source this to clone the .files repo and setup the command

  #git clone --bare https://<user>@bitbucket.org/<user>/dotfiles $HOME/.files
  alias .f='git --git-dir=$HOME/.files --work-tree=$HOME'
  .f config --local status.showUntrackedFiles no
  .f config --local remote.origin.fetch '+refs/heads/*:refs/remotes/origin/*'
  #.f config --local user.name '<User Name>'
  #.f config --local user.email '<user@emaildomain>'
  .f branch -a

No, `dirname` is correct. The goal of `ln -s dirname/ linkname` is that even if you accidentally do `cp file linkname` (note without slash) when dirname is gone, it will error out. Without the trailing slash in the ln command, a file (instead of directory) named dirname will be created.

The content needs major update to incorporate recent research on:

  - Macro nutrients (protein, carbs, fat)
  - Micro nutrients (vitamins and minerals, important if you do IF)
  - Intermittent fasting (IF) / time restricted feeding (TRF)
  - Strength training (including weighted calisthenics)
  - Mobility exercises
  - Sleep management
  - Supplements

BTW, plantuml has been storing the source uml dsl in the metadata of generated png for years. cf. https://plantuml.com/command-line#ce21470ab49d1d19

Fasted for 3 days and muscle mass increased by 2.4kg. Yeah, right.

Just like various chat bot incarnations, it's reasonably OK at generating generic sentences without real understanding. I tried several attempts with this leading fragment: "Is borrow checking unique to Rust? Probably not. But". The results are not pretty (the best one is as follows). Not impressed at all with this giant model :|

"Is borrow checking unique to Rust? Probably not. But the Rust library is written by and for Rust programmers, so if you're looking for a library that's designed for the language you already know, this might be your first choice.

The library is still in a fairly early stage. It does not support the standard libraries.

If you're interested in reading about how the library works, the docs are up on GitHub for reference!

How can I get this for my project?

Get it here: https://github.com/paulp/rustbinder

I've just read your blog post and want to use it, what can I do?

If you're looking for a project to use this library for, check out the docs for the "RustBinder project".

What are the current limitations and how will future changes affect this library?

This is still a work in progress. The library currently does not support the standard Rust library. There are a few work in progress"

3. "What's the best way to specify computation?" covered by programming language and compiler books.

Running 10k containers on one machine is totally realistic though. This is just yet another language vm (e.g. jvm) isolation all over again.

Hey OP, are you still looking? My company (a well funded AI startup with SV office) is definitely looking for eng roles from senior IC to VPE. PM if you're serious (email in profile). I can at least give you specific feedback on your resume.

To disable this class of exploits (bugs in unprivileged bpf) without installing a new kernel:

echo 1 | sudo tee /proc/sys/kernel/unprivileged_bpf_disabled

TLDR: two side-channel info leak attacks on CPUs: Meltdown that appears to affect Intel and latest ARM64 (Cortex-A75), and Spectre that was shown to affect Intel, older AMD and ARM64 and theoretically newer AMD processors.

Meltdown is more severe with read privilege escalation (javascript in browser could read kernel memory!), which has a kernel patch (KPTI) that can slow down many workloads by 30%.

Spectre also allows a user process to read memory of other processes/containers and potentially kernel, which was shown to work on Intel and ARM64 because their predictable branch prediction impl (Tomasulo). Newer AMD (Ryzen and Epyc) employs a hardware neural network for branch prediction, which is much less predictable but still theoretically possible to exploit. This is probably not the first time that the opacity of a neural network helped security :)

Chronos is just an example. There're many bugs in Mesos that don't get fixed for months/years. Mesos core is legacy (pre 11) C++ code nobody wants to maintain.

FTFA: "We’d previously been using Chronos (with Mesos) as a cron job scheduling system, but it was no longer meeting our reliability requirements and it’s mostly unmaintained (1 commit in the last 9 months, and the last time a pull request was merged was March 2016) Because Chronos is unmaintained, we decided it wasn’t worth continuing to invest in improving our existing cluster."

Though Chronos has a release recently with a bunch of fixes, Mesos is inevitably fading as a legacy platform.

I'd expect better write up from Daniel. std::unordered_set is implemented with chained buckets, which incurs extra memory references. Cursory google revealed that an open addressing hash set would be about 3x faster, which would handily beat the sort version.

Who has used both Airflow and Spinnaker? Feedbacks?

This is obviously data dependent. I suspect that the advantage of human is much higher in colloquial content vs written (esp. news) content. "Universal Adversarial Perturbations"[1] last year showed that you can easily generate reasonable (to human) perturbations to completely fool the state of the art DNNs for images. I suspect that the same is true for the current batch of NMTs as well. As a simple demo, I change the example Chinese a little (就要下班了。下班了再说吧。Only aux characters changes with the same meaning) and all NMTs failed spectacularly in different ways.

  Google: It is necessary to get off work. To say it again.
  OpenNMT: it's going to work. Go back to work again.
  Baidu: It's going to work. After work.

These NMTs seem tragically bad at translating simple straightforward Chinese e.g.: 我要下班了。下班了再说。

  Google: I have to get off work. Off to say.
  OpenNMT: I am going to work. After work, again.

    1. WordNet
    2. ImageNet
    3. SyntaxNet

    ...

    n. SkyNet

Hope this give google enough kick in the butt to improve mixed language voice to text.

My google voice inbox is full of hilarious examples...

Yay, I'm going to put "FAA registered pilot" in my profiles :)

NoScript (even when enabled for destination sites) does detect/filter the XSS attempt though.

Also interesting: it's like.

https://books.google.com/ngrams/graph?content=it%27s+like&ca...

Even more dramatic rise than "I feel like".

Storm is not even 1.0 yet. Since Heron is API compatible with Storm, why can't Heron be simply a code name for Storm 2.0?

Storm is a much better name than Heron, IMO.

You must be looking the median latencies. 99% latency was and still > 200ms. You can blame GC jitters for the much bigger variance. They should also show median and 95% latencies for this years number as well.

He should help his students understand how to acquire basic skills in a field, which is a prerequisite skill to do any research.

It start with a web search these days.

There is simply no excuse, when the knowledge of these basic-skill/bullshittery is openly available.

You know, teach a guy how to fish...