matrss

21 ms·

181.
by matrss 3y ago
> I would say that containers are the foundation of the practice of reproducible builds. If you mean containers as in the isolation features that are utilised by docker et al. to provide their flavor of compartmentalization then yes, tho
182.
by matrss 3y ago
I agree with this being an issue, although certainly less pronounced with go compared to other languages. I don't think containers solve the issue though: none of the mainstream approaches to building containers actually make the build
183.
by matrss 3y ago
If you have your project packaged with nix (which shouldn't be too complicated if your build process is not too special) it is trivial to turn a nix derivation into a docker image: https://nixos.org/manual/nixpkgs&
184.
by matrss 3y ago
In my experience, not using docker to build docker images is a good idea. E.g. buildah[0] with chroot isolation can build images in a GitLab pipeline, where docker would fail. It can still use the same Dockerfile though. If you want to get
185.
by matrss 3y ago
> [...], but ideally each snippet of shell script would be entirely independent and run effectively reproducibly. Sure, if you design your scripts carefully (e.g. make them idempotent, reproducible, make all inputs explicit, etc.) you ca
186.
by matrss 3y ago
So, only use the CI system for it's dispatch/scheduling part and put the imperative steps into a dedicated script? Reasonable, and what I have done with some GitLab Pipelines as well, but I don't see how that would be conside
187.
by matrss 3y ago
> I want my pipelines to be declarative anyway. Is there any CI system that allows one to write declarative pipelines? What would that even mean? GitHub Actions, GitLab Pipelines, etc. are all effectively just shell scripts disguised as
188.
by matrss 3y ago
> I 100% agree that managing large datasets by moving them around is not practical, and definitely not in LFS/DVC-style. There should be a level of indirection if reproducibility is needed (pointers are versioned to files, not the d
189.
by matrss 3y ago
Hmm, that is an interesting combination. I forgot that I had tried that in the past as well. It had its own issues, because it deletes all cookies by default. Is there any way to keep the convenience of staying logged in with that, e.g. a c
190.
by matrss 3y ago
I've used that in the past. Unfortunately, it had some issues; some sites completely broke when consent-o-matic did its thing (e.g. page remained greyed-out and/or scrolling was broken from the cookie popup). Not their fault reall
191.
by matrss 3y ago
If it is not then it must be unnecessary, since you could get the same information from the access logs already.
192.
by matrss 3y ago
That is just an automatic opt-in to tracking, the opposite of what was asked for.
193.
by matrss 3y ago
Why wouldn't it make sense? If a chinese user e.g. specifies that they speak chinese and, say, french, all characters in those two language would show up as unicode and everything else as punycode. A malicious domain using e.g. cyrilli
194.
by matrss 3y ago
What you mention here does not contradict my assumption that there is a finite set of valid characters in every language, unless there is a rule in chinese that lets you assemble an infinite set of meaningful characters/symbols (unicod
195.
by matrss 3y ago
> Opensuse with snapshots and some repo magic, can do 99% of functionality, [...] I doubt that. Does Opensuse have an integrated declarative configuration management system? How would I apply a patch to a program on Opensuse and build th
196.
by matrss 3y ago
As a german I wouldn't go to göogle.com. If my native language didn't include ö? Then that might be a speck of dust to me as well. A safer approach would be to only ever show a user the characters they expect to see (and are famil
197.
by matrss 3y ago
That or whitelist specific characters based on the users locale. E.g. on a system set to german I could expect umlauts and ß to be rendered properly, but anything else should just show the punycode. Basically any legitimate use case should
198.
by matrss 3y ago
Windows still exists. Or does that have a package repository now?
199.
by matrss 3y ago
Google could do something good and test the license. They have the resources to fight it out.
200.
by matrss 3y ago
Meanwhile, the first sentence in their readme is this: > JAX is Autograd and XLA, brought together for high-performance machine learning research. That does not really convey the generality of it that well.
201.
by matrss 3y ago
I see which email thread you mean now. AFAICS OP submitted the patch with a proper sign-off again approximately 6 hours after they were told the sign off was missing. So I don't see how that could have been an issue. The maintainers pa
202.
by matrss 3y ago
I did not find that yet. Can I read that anywhere? My understanding is that if the initial patch was not signed off then the maintainer could not sign off his patch, unless he wanted to take on the liability of certifying that the contribut
203.
by matrss 3y ago
The patch I found was signed off. If it wasn't I fail to see how the maintainer could have legitemately signed off his own patch considering it was modified from or inspired by the original patch.
204.
by matrss 3y ago
Reported-by should go to the person who reported the issue 6 years ago. Finding the issue (again), investigating, debugging, fixing and testing that fix sounds more like a hypothetical "Fixed-by", or co-authorship, depending on ho
205.
by matrss 3y ago
To be fair, the color is misleading. I associate red text with dead links and that is the only reason I checked what they point at.
206.
by matrss 3y ago
It contains a http link to "ssss-0.5.tar.gz", which I would assume is meant to be a source code archive. That really looks suspicious.
207.
by matrss 3y ago
That thing is almost as big as my microwave and could replace halve of an oven. But I guess it does toastiness. I wish someone would just build a new toaster that works like the sunbeam radiant control ones, which would be useable on 230V (
208.
by matrss 3y ago
Honestly, it would be enough if toasters could operate by "toastiness", instead of time, at all. We had that decades ago and then just stopped, I guess? I have watched too much Technology Connections... /rant
209.
by matrss 3y ago
True, nix is also a package manager, and that is the crucial step necessary to actually provide a declarative interface to system state. Without it you can only get the leaky abstraction that is ansible. To illustrate my point, imagine this
210.
by matrss 3y ago
So, the last time I used ansible, which was quite a while ago to be fair, there was a builtin way to install packages with apt on the target system. You could add packages to a list to be installed and that would work. But removing them fro

More ›