Dmitry Marakasov<p><span class="h-card" translate="no"><a href="https://fosstodon.org/@fosstodon" class="u-url mention">@<span>fosstodon</span></a></span> hey, it seems you've had a downtime yesterday, I've witnessed errors from varnish about first byte timeout or something for a few hours. That wasn't visible on <a href="https://status.fosstodon.org/" target="_blank" rel="nofollow noopener noreferrer" translate="no"><span class="invisible">https://</span><span class="">status.fosstodon.org/</span><span class="invisible"></span></a> though. Just FYI, probably monitoring could be improved.</p>
Recent searches
Search options
Dmitry Marakasov@AMDmi3@fosstodon.org
Free/open source software developer, 
#FreeBSD committer.
*You are what you contribute to*
Just implemented #FreeBSD support in `metrics-process` #rust crate (which collects generic process metrics such as runtime and used cpu/memory), hopefully it will be merged. Without this change, `metrics-process` won't build on FreeBSD, or, if `dummy` feature is enabled, will build but won't collect any data.
https://github.com/lambdalisue/rs-metrics-process/pull/53
PS. I'm also quite impressed by AI PR summary.
GitHubAdd FreeBSD support by AMDmi3 · Pull Request #53 · lambdalisue/rs-metrics-process
Continued thread
On the downside, I'll have to maintain ruleset (merging related projects, splitting unrelated ones, and blacklisting incorrect versions) in whole another ecosystem, and one which lacks good package naming (names are long unicode strings which need normalization) and versioning (`v` prefix and random suffixes all over the place) traditions.
So for now I'm hesitant about making Android repos first class citizens in #repology.
Continued thread
The intersection between #fdroid and #izzyondroid is not very big (222 projects of e.g 4k total in fdroid), but if Repology supports these, it will be able to show outdated versions in either repo within the intersection (and vulnerable versions everywhere if we add CPE bindings).
It would also be easier to merge projects present in both linux and android in Repology.
#Repology only has marginal support for F-Droid, only seeing a few handpicked packages for software which is also available in Linux:
https://repology.org/projects/?inrepo=fdroid
I've recently had a few PRs which improve F-Droid support and add #IzzyOnDroid, allowing full-fledged version comparison within Android ecosystem.
I wonder if any #fdroid maintainers or @IzzyOnDroid would be interested in that.
repology.orgProjects list - RepologyMultiple package repositories analyzer
Dmitry Marakasov boosted
OpenStreetMap data powers The Washington Post's story about the size and territory of supermarket chains in USA.
OSM is more than a map, we're a database. Everyone (incl you!) gets the same full access to the data, for free. And if it's incorrect, you can fix it immediately. We love to see new and surprising uses of OSM data like this.
What will you create with OpenStreetMap?
https://www.washingtonpost.com/business/interactive/2024/grocery-store-owners-map-kroger-albertsons-merger/
#OpenStreetMap #OSM
The Washington Post · Grocery chains are bigger than ever. See who owns the stores near you.
Dmitry Marakasov boosted
#!/bin/sh
set -euo pipefail
thanks, @b0rk - I've been writing shell for 25+ years and wish I'd been doing this 25+ years ago.
Continued thread
The comparison is not quite fair, as it's sync Python vs async Rust, and I could more or less transparently migrate from Flask to Quart and save a lot of memory and probably improve latency under load, still it's obvious that the languages are in completely different leagues in terms of performance. As the Rust code also feels cleaner to me, and Rust has better infrastructure, I will probably no longer consider Python ever for prototyping.
So after doing a few small projects in #rust, I've finally started rewriting #repology backend (currently in #python) in Rust.
First discovery - what I naively though of as a database-bound workload, e.g. where database RTT (for small queries) or response generation time (for complex queries) dominates, are still 5-10x faster (in terms of both max RPS and latency) in Rust!
Also rust process takes 3-10x less memory compared to single uwsgi worker, which there are 10-20 of.
Dmitry Marakasov boosted
#TIL You know how tape measures have the L-hook on the end and it's always a little loose? Turns out that's deliberate. It's designed to move an amount exactly the width of the hook, so the measurement is the same whether you're pushing it up against something or using it to hook on something.
@congusbongus hey, you interested in my OpenStrike? Can update it a bit if that's the case. I think I still remember some details on why I've stopped working on it.
Dmitry Marakasov boosted
Again, not FOSS related, but just finished reading "Invincible" by Stanisław Lem. And this book is... Ufff... It's not very exciting, I'd say that less than 0.5% of it is pure action, but it's very close to... Psychological horror, I'd say? If I would compare Lem to any other author, it would be Lovecraft. They have the same "nothing happens, but something is lurking in the shadows" vibe (and I'm a Lovecraft fan so...).
Overall - I recommend it a lot! And also the game Invincible is great!
Dmitry Marakasov boosted
HAPPY BIRTHDAY OpenStreetMap! Today is OSM's 20th Brithday! 
Here's SteveC's desktop at University College London, the original OSM "server" until 2006. You've come a long way in the last 20 years. 
https://birthday20.openstreetmap.org/
#20YearsOfOSM #OpenData #OpenSource #OpenStreetMap #OSM
#simutrans is a champion of inconsistency in how their tarballs are packaged, which makes packaging a pain
simutrans-src-123-0-1.zip: files in archive root
simutrans-src-124-0.zip: files under simutrans-Nightly/
simutrans-src-124-1.zip: files in archive root
simutrans-src-124-2.zip: files under simutrans-src-142-2/
simutrans-src-124-2-1.zip: files under trunk/
Ugh...