edwintorok

10 ms·

61.
by edwintorok 6y ago
I think FUSE limitations: fuse is served by a single userspace process, which is limited the same way as any other userspace process by ulimit. It is not a fundamental limitation of podman, just of podman's default rootless behaviour.
62.
by edwintorok 6y ago
If you run it as non-root it is significantly slower than docker as root. Docker can us the overlay2 kernel driver, whereas podkan would use fuse-overlayfs in userspace. This has a high CPU overhead (e.g. don't try to run AFL inside po
63.
by edwintorok 6y ago
I don't know what is with all the negativity here in the toplevel comments. I've downloaded the Android app and bookmarked quite a few talks that I'm looking forward to.
64.
by edwintorok 6y ago
A better test is to use dieharder, see here: https://github.com/systemd/systemd/issues/18184#issuecomment...
65.
by edwintorok 6y ago
Could you clarify the "one of five" statement please? Are the other 4 vulnerabilities still unfixed, or they are fixed but a write-up is still pending? If there are still 4 unfixed RCE bugs in Teams I'd rather people uninstal
66.
by edwintorok 6y ago
If you are running podman rootless it will use the fuse-overlayfs driver which comes with various limitations and is much slower than Docker: * you are limited to 1024 FDs for your entire container. So running make -j40 usually ends up with
67.
by edwintorok 6y ago
IIUC the `scp` cmdline will still exist it'll just use the sftp protocol behind the scenes instead of the eco protocol. Note that the deprecation is about the protocol not the cli.
68.
by edwintorok 6y ago
Worth pointing out that the bug is still present in latest glibcs (I can reproduce it on Fedora 33), and affects programs that use pthread_cond_signal, in particular language runtimes like OCaml, Python and .NET that have a global runtime l
69.
by edwintorok 6y ago
I think it is a common misconception that banks use database transactions the old fashioned way. They seem to be more like eventual consistency. I initiate a transfer and there is some check about available balance (although with various ov
70.
by edwintorok 6y ago
As mentioned below it can deteriorate over time. A while ago I had a RAM stick fail. I noticed because GCC started segfaulting randomly and I haven't upgraded GCC recently. Sure enough a memtest revealed the fault. I think it didn&#x27
71.
by edwintorok 6y ago
Since we're onto recommendations, I've been using 'Go Mono' for a while: https://fontlibrary.org/en/font/go-mono https://blog.golang.org/go-fonts Wish the 'Recursive'
72.
by edwintorok 6y ago
Gnome's Evolution is quite good as well and the built-in exchange support for both email and calendar via EWS works better than the thunderbird plugin. I used to use thunderbird, but after several troubles with the exchange plugin (e.g
73.
by edwintorok 6y ago
DNS rebinding can be fixed at the DNS server level. OpenWRT has an option for it. But this websocket thing in browsers can't easily be turned off/mitigated AFAICT.
74.
by edwintorok 6y ago
If both sides of the comparison run inside Docker that is still a valid comparison. With Docker the benchmark should be a bit more reproducible by anyone.
75.
by edwintorok 6y ago
I had the opposite experience. Jemalloc uses a lot less (virtual) memory, especially with multithreaded applications. The glibc allocator wastes quite a lot of memory mappings from a thread and never cleans them up even if a thread only ha
76.
by edwintorok 6y ago
One day almost all the phones started ringing in the office one by one. Turns out someone/some automated system started dialing our phone numbers one by one with the above Chinese message scam. At that time I realized I never needed my
77.
by edwintorok 6y ago
Mesa still hasn't implemented OpenCL 2.x ... now they don't have to. Although applications like Folding@home don't really work with OpenCL as provided by Mesa, so it'll likely have to be updated to work with the 3.0 ~= 1
78.
by edwintorok 6y ago
Once your code grows and you end up with too many repos you'll be yearning to get back to monorepos. It allows to make changes in one go instead of having 3-4 PRs that all have to be merged at same time or otherwise the build breaks. H
79.
by edwintorok 6y ago
It used to work, with the latest update audio doesn't work anymore in Chromium (although it does see the device), I'll have to report a bug. (snap connections does show audio-record connected)
80.
by edwintorok 6y ago
Worth mentioning other alternative small backends: http://c9x.me/compile/
81.
by edwintorok 6y ago
The semantics of C aren't very well defined, there is a lot of ambiguity in the form of undefined and implementation defined behaviour. This ambiguity is often needed to build an efficient optimizing compiler. When you have a higher le
82.
by edwintorok 6y ago
I've noticed some issues a while ago with TCP syn cookies breaking DLM cluster setup of 64 hosts. (63 hosts all trying to join the same cluster generate enough TCP traffic that the kernel thinks it is a flood and starts sending syn coo
83.
by edwintorok 6y ago
There is an alternative method called Flowtime: https://zapier.com/blog/flowtime-technique/ where you focus on a task until you feel you need a break (I think it is still crucial that you don't let yourself g
84.
by edwintorok 7y ago
Wouldn't something like Rust meet your performance concerns?
85.
by edwintorok 7y ago
At least you can sandbox and run webassembly everywhere with an open interpreter/JIT. With plugins you have to trust and install some binary code on your system, no sandboxing, no idea what it does and what it can access, etc. Much wor
86.
by edwintorok 7y ago
Exactly my thoughts when I first read of WASI: conceptually you could now have a WASI runtime ("OS") and an HTML.wasm potentially independently developed and swappable. And since WASI is a lot smaller surface, it is easier to audi
87.
How to Upgrade from Windows 7 to Ubuntu(ubuntu.com)
20 points by edwintorok 7y ago|8 comments
88.
RGB/XYZ Conversion Matrix Accuracy(terathon.com)
2 points by edwintorok 7y ago|0 comments
89.
by edwintorok 7y ago
> security level 2 (defined as 112-bit security) in the latest release (Debian Buster); this already prevents dangerous usage of SHA-1 FWIW this doesn't apply to Fedora currently, because it has a patch that re-enables SHA-1 in secu
90.
by edwintorok 7y ago
Note that by default kernel.sched_rt_runtime_us limits RT usage to 95% of the whole system, not an individual core. If you have 1 RT process using 100% of a core on a multicore system it won't trigger the limiting, which can cause all

More ›