24,701 karma · joined January 8, 2008
I'm sure most people here would agree patents stifle innovation, but if copyright doesn't work for companies then they will turn to a different tool.
> career-driven development
we don't have this and promote and reward as frequently for "I've done solid operations" as we do for "I've added this feature" (I'm on promotion committees and can state this confidently).
what we do have is high autonomy for engineers. This autonomy means it's a freedom that engineers have to identify problems they feel are important and to work on them, they do not need permission and leadership do not veto this. Some of the best features in the last few years have been a direct result of this autonomy, it's one of the things that makes working here so attractive to many of the engineers. But, with autonomy comes a little chaos, and not everything that is done is going to satisfy every end user of OSS or paid customer (of which these are a small percent of the whole).
a lot of the innovation speed is just in the DNA of the company, even the creation of Grafana can be traced to a desire to get things done; Torkel wanted Kibana to also work for Prometheus, Kibana declined to add this, Torkel didn't stand still and added things to a fork of Kibana now called Grafana and hasn't stopped adding things since.
> They also deprecated Angular within Grafana and switched to React for dashboards. This broke most existing dashboards.
we did, I think the entire journey was 7 years long, communicated many times, over at least 6 major releases. maintaining dashboards in two languages increased complexity, whilst reducing compatibility, and gave a very large security surface to be worried about. we communicated clearly, provided migration tools, put it in release notes, updated docs, repeated it at conferences and on community calls.
arguably we went too slow, and should've ripped the band-aid off, but we were sensitive to the fact that it was a breaking change and so we proceeded with extreme caution. it's done now, it was finally completed in the last version, only a very small number of users reported impact as a result of the time and care taken on this.
> I just hope OTEL settles, gets stable and boring fast
this is distinct from Grafana, but it's a good point... OTel is the product of virtually every vendor at this point, and a hell of a lot of engineers, it now has a lot of momentum and the pace is unlikely to ease up due to the sheer number of contributions and things that OTel as a community wishes to achieve.
the most likely eventuality is that enough stability emerges to allow vendors (including but not limited to Grafana Labs) to abstract away the pace of innovation occurring underneath, but this is in tension with providing the benefits of the innovation to the people that use it.
what I would say is that for most people the boring and slow path does still exist, and it's still good... just use Prometheus, a logging option of your choice, and simple Grafana dashboards and alerts. that combination hasn't varied in years, and those on it today are still immune from caring about the pace of innovation and change in OTel and across the Observability industry. OTel is being used in production at massive scale by lots of companies, but whether your project or company need move to it now reflects your priorities, many are adopting to gain independence from vendors, or just control over their telemetry, but many customers are also saying they're happy to stay on the slow and boring path and for everything to work predictably with low cost to keep pace... it works too.
Far less requires an actual app than most people imagine. It's the apps that leak so much.
Then Proton CEO made some statements I found offensive, so I re-activated my Bitwarden account, migrated back, and am now learning to love the changes.
The best I've got for tips are:
1. Settings > Appearance > Quick Copy
2. Settings > Appearance > Compact Mode
3. Settings > Appearance > Extension Width > Wide
I still don't love it, but it remains the best of the bunch.
Saka isn't all that either, he'd be nothing within Odegaard.
(this is convincing, and I know about as much as the IT Crowd did)
The app publishing didn't exclude the UK, it probably should.
Your site has young trans people, sexual content, and would also be a target for grooming from chasers. The risks, for you, are very high.
I've received magic links to my Gmail account that belong to other people, for accounts that have ordered flight tickets, or clothing, or digital services.
Those people, I guess they now have no way to access their online account, as they cannot password reset (if that was the fallback), or change their email (usually requiring confirmation), or receive their magic link.
There's nothing I can do here, except to delete the email, I don't have any indication as to what the correct email should be, and the person's name is the same as my legal name and there are a lot of people with that name in the World.
Few services verify an email during sign-up, because I'm sure data shows that added friction during sign-up results in fewer people signing up.
Hardly... the article links says that a 403 will cause Google to stop crawling and remove content... that's the desired outcome.
I'm not trying to rate limit, I'm telling them to go away.
I contacted the network team at Cloudflare to apologise and also to confirm whether Facebook did actually follow the redirect... it's hard for Cloudflare to see 2PB, that kind of number is too small on a global scale when it's occurred over a few hours, but given that it was only a single PoP that would've handled it, then it would've been visible.
It was not visible, which means we can conclude that Facebook were not following redirects, or if they were, they were just queuing it for later and would only hit it once and not multiple times.
In fact 2M requests arrived on December 23rd from Claude alone for a single site.
Average 25qps is definitely an issue, these are all long tail dynamic pages.
If a more specific UA hasn't been set, and the library doesn't force people to do so, then the library that has been the source of abusive behaviour is blocked.
No loss to me.
if ($http_user_agent ~* (list|of|case|insensitive|things|to|block)) {return 403;}
1. A proxy that looks at HTTP Headers and TLS cipher choices
2. An allowlist that records which browsers send which headers and selects which ciphers
3. A dynamic loading of the allowlist into the proxy at some given interval
New browser versions or updates to OSs would need the allowlist updating, but I'm not sure it's that inconvenient and could be done via GitHub so people could submit new combinations.
I'd rather just say "I trust real browsers" and dump the rest.
Also I noticed a far simpler block, just block almost every request whose UA claims to be "compatible".
I have data... 7d from a single platform with about 30 forums on this instance.
4.8M hits from Claude 390k from Amazon 261k from Data For SEO 148k from Chat GPT
That Claude one! Wowser.
Bots that match this (which is also the list I block on some other forums that are fully private by default):
(?i).(AhrefsBot|AI2Bot|AliyunSecBot|Amazonbot|Applebot|Awario|axios|Baiduspider|barkrowler|bingbot|BitSightBot|BLEXBot|Buck|Bytespider|CCBot|CensysInspect|ChatGPT-User|ClaudeBot|coccocbot|cohere-ai|DataForSeoBot|Diffbot|DotBot|ev-crawler|Expanse|FacebookBot|facebookexternalhit|FriendlyCrawler|Googlebot|GoogleOther|GPTBot|HeadlessChrome|ICC-Crawler|imagesift|img2dataset|InternetMeasurement|ISSCyberRiskCrawler|istellabot|magpie-crawler|Mediatoolkitbot|Meltwater|Meta-External|MJ12bot|moatbot|ModatScanner|MojeekBot|OAI-SearchBot|Odin|omgili|panscient|PanguBot|peer39_crawler|Perplexity|PetalBot|Pinterestbot|PiplBot|Protopage|scoop|Scrapy|Screaming|SeekportBot|Seekr|SemrushBot|SeznamBot|Sidetrade|Sogou|SurdotlyBot|Timpibot|trendictionbot|VelenPublicWebCrawler|WhatsApp|wpbot|xfa1|Yandex|Yeti|YouBot|zgrab|ZoominfoBot).
I am moving to just blocking them all, it's ridiculous.
Everything on this list got itself there by being abusive (either ignoring robots.txt, or not backing off when latency increased).
Oh I do... the link... HN must have a word based deny list
A forum that isn't proactively monitored (approval before publishing) is in the "Multi-Risk service" category (see page 77 of that link), and the "kinds of illegal harm" include things as obvious as "users encountering CSAM" and as nebulous as "users encountering Hate".
Does no-one recall Slashdot and the https://en.wikipedia.org/wiki/Gay_Nigger_Association_of_Amer... trolls? Such activity would make the site owner liable under this law.
You might glibly reply that we should moderate, take it down, etc... but we, is me... a single individual who likes to go hiking off-grid for a vacation and to look at stars at night. There are enough times when I could not respond in the timely way to moderate things.
This is what I mean by the Act providing a weapon to disgruntled users, trolls, those who have been moderated... a service providing user generated content in a user to user environment can trivially be weaponised, and it will be a very short amount of time before it happens.
Forum invasions by 4chan and others make this extremely obvious.
Then you will see that a forum that allows user generated content, and isn't proactively moderated (approval prior to publishing, which would never work for even a small moderately busy forum of 50 people chatting)... will fall under "All Services" and "Multi-Risk Services".
This means I would be required to do all the following:
1. Individual accountable for illegal content safety duties and reporting and complaints duties
2. Written statements of responsibilities
3. Internal monitoring and assurance
4. Tracking evidence of new and increasing illegal harm
5. Code of conduct regarding protection of users from illegal harm
6. Compliance training
7. Having a content moderation function to review and assess suspected illegal content
8. Having a content moderation function that allows for the swift take down of illegal content
9. Setting internal content policies
10. Provision of materials to volunteers
11. (Probably this because of file attachments) Using hash matching to detect and remove CSAM
12. (Probably this, but could implement Google Safe Browser) Detecting and removing content matching listed CSAM URLs
...
the list goes on.
It is technical work, extra time, the inability to not constantly be on-call when I'm on vacation, the need for extra volunteers, training materials for volunteers, appeals processes for moderation (in addition to the flak one already receives for moderating), somehow removing accounts of proscribed organisations (who has this list, and how would I know if an account is affiliated?), etc, etc.
Bear in mind I am a sole volunteer, and that I have a challenging and very enjoyable day job that is actually my primary focus.
Running the forums is an extra-curricular volunteer thing, it's a thing that I do for the good it does... I don't do it for the "fun" of learning how to become a compliance officer, and to spend my evenings implementing what I know will be technically flawed efforts to scan for CSAM, and then involve time correcting those mistakes.
I really do not think I am throwing the baby out with the bathwater, but I did stay awake last night dwelling on that very question, as the decision wasn't easily taken and I'm not at ease with it, it was a hard choice, but I believe it's the right one for what I can give to it... I've given over 28 years, there's a time to say that it's enough, the chilling effect of this legislation has changed the nature of what I was working on, and I don't accept these new conditions.
The vast majority of the risk can be realised by a single disgruntled user on a VPN from who knows where posting a lot of abuse material when I happen to not be paying attention (travelling for work and focusing on IRL things)... and then the consequences and liability comes. This isn't risk I'm in control of, that can be easily mitigated, the effort required is high, and everyone here knows you cannot solve social issues with technical solutions.
I don't know where you're seeing that as the site does not have such things. The only cookies present are essential and so nothing further was needed.
The site does not track you, sell your data, or otherwise test you as a source of monetisation. Without such things conforming with cookie laws is trivial... You are conformant by just connecting nothing that isn't essential to providing the service.
For most of the sites only a single cookie is set for the session, and for the few via cloudflare those cookies get set too.
the liability is very high, and whilst I would perceive the risk to be low if it were based on how we moderate... the real risk is what happens when one moderates another person.
as I outlined, whether it's attempts to revoke the domain names with ICANN, or fake DMCA reports to hosting companies, or stalkers, or pizzas being ordered to your door, or being signed up to porn sites, or being DOX'd, or being bombarded with emails... all of this stuff has happened, and happens.
but the new risk is that there is nothing about the Online Safety Act or Ofcom's communication that gives me confidence that this cannot be weaponised against myself, as the person who ultimately does the moderation and runs the site.
and that risk changes even more in the current culture war climate, given that I've come out, and that those attacks now take a personal aspect too.
the risk feels too high for me personally. it's, a lot.
What heats the water? What provides temperature control? How would I produce steam?
It is so single purpose that it does not feel useful by itself, it feels like the prototype for part of a whole.
I like the idea of it, and I like the idea of "part of the whole" being a composable coffee machine where one could put together components which were all independently maintainable and highly serviceable... this feels like a taster for that, but by itself is very expensive for a pump that claims to be an espresso machine but could not produce an espresso alone, and would need something else to make any espresso derived coffee.
What this replaces is a lever espresso machine, but I'm not sure anyone with a home coffee machine would've purchased a lever espresso machine without the integrated boiler... and if they would, then this is right there https://bellabarista.co.uk/collections/lever-machines/produc...
You would benefit greatly from a video that showed the workflow end-to-end of making an espresso... from bean to the final drink.
but... Android sucks because the default permissions and business model on Android all veer towards advertising, and this is heavy on network and bandwidth as well as background processing across lots of apps... and that drains battery and performance.
I switched to GrapheneOS and just disable network on many things like Camera, etc... the Cloud AI features are not worth it... I also run web apps instead of installed apps for the vast majority of things, if an app doesn't require some hardware capability that only an app can provide then it's staying as a web app, m.uber.com works, most news websites work best as web apps with the JS disabled... I routinely get multi-day battery and stellar performance by just not running apps that are always trying to continuously exfiltrate data about me.
I've found only 2 things that don't work the same on GrapheneOS: Revolution banking just does not work at all - so I closed my account with them, and AMEX forces 2FA on every sign-in - which I can tolerate. Nothing else was impacted, everything else is an improvement in performance and battery.
At which point we could just say all SQL should have a WHERE clause.