There’s an LLM called Grok? What the fuck. Seriously. Tech bros understand the absolute most obvious thing from your favorite piece of science fiction, please.
The meaning of grok is in the plain text! We’re not even talking about obvious metaphor, it’s just a made-up word for truly deep understanding and you’re going to name your idiot, fabulist, word generator after it?! #grok#LLMs#TormentNexus#StochasticParrot#Tech
@c0dec0dec0de like it was built by Musk's team. "Compromised land" is a play on promised land, which is what Twitter was expected to be until the whole EM thing. All to say that the people who built it [and named it] are not representative at all of the ppl working in the LLM space and their work. I would just suggest to not let that model influence how you see this field
@redscroll okay.
LLMs seem neat, but they’re being framed as something they are not and can never be. It’s fun to get one to spit out prose in the style of some dead writer talking about Batman or poems about git.
But most of the coverage and hype over them positions them as stepping stones to artificial sapience or as an interface to the knowledge of the world. And those are impossible for it.
Fundamentally, an LLM doesn’t know things. So, to call it grok is wrong and almost offensively so.
I have a stupid idea, but I'm not quite sure how to make it happen. Modern linux systems seem to be eschewing virtual consoles, but assuming that you wanted to keep using them and also had to meet a requirement to display a classification banner on-screen at all times, where would be the best place to try to insert a banner, agetty? #linux#tty#security
@htotw@argv_minus_one that threw me a bit. I remember the days 1-6 were consoles and 7 was X, when I ran it. Now I seem to have 1 and 2 for Wayland, 3-6 as consoles, but they’re on-demand.
Relatedly, I recently discovered bear, a tool that’ll generate a JSON compilation database for arbitrary builds (which enables vim-ALE to properly process the sources).
Since we’re all onboard with telling off Linux evangelists in this moment, Linux has privacy issues too! If your screen goes to sleep, when you come back you’re greeted by your screen contents - not the Lock Screen for a solid two seconds or so (at least in GNOME Shell, the default desktop environment for many distros). #gnome#linux#security
@visone and GNOME is a Linux desktop environment. It’s the default on the most widespread workstation and desktop distributions.
But, fine, Linux doesn’t actually do SecureBoot attestation. The other major OSes do. (Yes, you can load a machine-owner key and do it yourself, but that surely doesn’t count as the OS providing it or supporting it)
@c0dec0dec0de
Sure, but it's a gnome-wayaland problem. That's the thing with linux, you can choose.
So you want to move to linux cause microsoft, but still using microsoft keys in your secure boot?? ....
automated wireless intrusion in the setup environment. Don’t have an internet connection? Don’t worry; we’re sure we can break into one of these networks! #Windows#Windows11#Security#joke#satire
A spokesperson for Microsoft denies that the Windows installer hacks into people’s WiFi networks saying:
“It doesn’t do that. Any wireless decryption or cracking routines only come into play if the password spraying attempts all fail - like the one that we were breached by last week. So, really, it almost never happens.”
@c0dec0dec0de It shows each commit as a Diff, so the match will come up on its own line; you can then scroll up to find which file was changed, and further up to find which commit is being Diff'd.
Bitwarden’s been throwing warnings on my phone telling me to scale back my hashing parameters because they might fail on this device.
Of course, now that I post about it, it’s not doing it so I can’t screenshot it… #Bitwarden#PasswordHashing#Infosec
I have used a git subtree, and while I kinda hate it. I still think it was the right call given the constraints.
This is not a reply because I don’t want to argue with the take that brought it to mind.
If, for some reason, you want to use git subtree, think very carefully about doing so. Reasons follow.
I guess that’s all I’ve got for general reasons. There’s also fuckery with JIRA issue tags in commit messages if you track those or use pre-receive hooks that reject commits with mentions of closed issues. Recall that you can’t rebase. You can filter-branch, but that is also painful and dangerous to the extent that it tells you to never use it.
5/5
This also ignores completely the concept of contributing back to the source repository. Generating synthetic trees that cleanly partition changes between subtree and non-subtree and then moving those changes to the subtree’s branch. Then I guess adding the subtree repo as a remote(?) and actually pushing to it. We ignored ALL of that as far too fragile, esoteric, and just damn weird. Changes flowed in one direction and we still mucked it up several times in the last couple years.
Current, unpaid-for projects I want to work on:
A tool to strip out nondeterminism from RPMs
A tool to run only CTest executables affected by a list of changed files
CMake scripting to automatically add all the necessary subpackages as dependencies of the devel subpackage
My headcanon version of git-grep
A requirements tracker website à la UCF, but free; or OpenSSF badge app, but customizable and can be self-hosted
Terminal classification banner
Virtual console classification banner
So, I'm reading about my stupid brain and this is fucking with me. (see attached excerpt from Taking Charge of Adult ADHD) Other people have detailed sensory recollection? My recollection of sensory information seems more like metadata than replay to me. Sweet and sour chicken is good and tangy, but I can't like taste it in my head.
I think this is the first time I've felt really like there's something wrong with me. #ADHD#neurodivergent#neuroscience#qualia
Mid-life “crisis”: as emphatically not-a-car-guy, I am possessed of a desire to build a car. This is facially absurd. I don’t possess the necessary skills, don’t have a garage to store a partially-constructed vehicle, and definitely don’t have the time or funds to spend on such a venture.
I’m not building a fiberglass facsimile of an ‘89 MR2. But I kinda want to. Maybe make the lines just a touch sharper - and the cabin a touch bigger to accommodate my height better…
No. Silly. Frivolous. Nope.
@scudderfish oh, I watched a video about making fiberglass car bodies. It looks like an insane amount of work - although I could probably make if less work because my desired shape includes a lot of planes and fewer swooping curves. https://youtube.com/watch?v=xhtIVXgEt18
Ugh. I want to write a blog post to share something but I don't have a blog and I don't really feel like I have enough things to say often enough to warrant really starting one.
@boki unexpected highlight of GitHub’s refusal to accept HTTP authentication for pushing code: it’s that much harder to make and push commits using the wrong account/identity. (I was a little worried about this)
Reading (and absolutely loving) Ferrett Steinmetz’ Flex. I would swear that @EAPodcasts did a read of one of the chapters a ways back, and it’s what got me to put it in my list to buy.
@bruxisma looks like ctest’s JSON output isn’t useful until after the test executables are built. This even though I’m avoiding using gtest_discover_tests so it can enumerate the tests pre-build. I guess since the commands aren’t ready, they’re not shown?
@c0dec0dec0de it's output is based on whatever the CTestConfig.cmake file contains, technically speaking, so it should just require the add_test calls to have been made 🤔