zlacker

[parent] [thread] 47 comments
1. aeyes+(OP)[view] [source] 2023-07-01 22:09:35
This bug is very unlikely to be the reason. The rate limiter on the server side is cheap and the frontend bug only gets triggered with the rate limit active.

I have seen similar bugs in the systems I oversee because network libraries love to retry requests without sane limitations by default. But I never saw them make our rate limiters sweat. It's slightly more annoying when they hit an API which actually does some expensive work before returning an error but that's why we have rate limits on all public endpoints.

I also guess that the webapp is the least of Twitters traffic and the native apps probably don't have this problem.

replies(3): >>epista+l1 >>evan_+z4 >>reddit+xe
2. epista+l1[view] [source] 2023-07-01 22:17:55
>>aeyes+(OP)
One thing about having leadership that is known to lie about anything or everything, for any sort of imagined personal gain, is that the very concept of truth is destroyed.

I agree that this is probably not the bug at the root of it all. But I also don't believe the story that Musk is selling for why he's in effect shutting down the site. But both could be true and I'm still thinking about other potential reasons, a complete waste of my time, but it's a weird mental honeypot.

The book "Nothing is true and everything is possible" describes Putin's use of misinformation to maintain control of the populace and eliminate democratic types of politics, but it really feels like it applies here too. There will always be Musk fanbois who will parrot whatever he wants them to say, but most know it's just self-serving BS. And anybody trying to get to the root of everything gets easily sidetracked into narratives that feel right but have zero data backing them, like this bug.

Anyway, highly recommend this book if you want to see a likely path for the future of the US:

https://en.wikipedia.org/wiki/Nothing_Is_True_and_Everything...

replies(7): >>kristi+D3 >>concor+i4 >>Edward+Da >>hgsgm+9l >>sander+Rm >>Robotb+bq >>mrcwin+rs
◧◩
3. kristi+D3[view] [source] [discussion] 2023-07-01 22:34:18
>>epista+l1
That’s exactly it. Musk a week ago was telling us that there were a record number of user seconds on the site. Now he’s telling us they’re all content-scrapers. The very concept of truth is eroded.
replies(2): >>partia+C8 >>18pfsm+se
◧◩
4. concor+i4[view] [source] [discussion] 2023-07-01 22:41:19
>>epista+l1
> One thing about having leadership that is known to lie about anything or everything, for any sort of imagined persona gain

I will note that the few times I investigated claims of Elon lies they were not proper lies, either being misunderstood, misleading (which IS unethical, don't get me wrong), of indeterminate truth value (he said, she said type stuff), delusional optimism or actually true.

Like journalists, Musk rarely outright knowingly makes literally false statements, but this does not mean you should take what he says at face value.

replies(3): >>epista+w4 >>Negati+8g >>daniel+UI
◧◩◪
5. epista+w4[view] [source] [discussion] 2023-07-01 22:43:00
>>concor+i4
I don't want to quibble about semantics, but habitual behavior of this sort just falls under the category of "liar" for me.
replies(1): >>concor+E41
6. evan_+z4[view] [source] 2023-07-01 22:43:31
>>aeyes+(OP)
I don’t think it’s necessarily saying the self-inflicted DDoS has caused a technical issue that’s forced them to shut down access. I think it’s possible that shutting down anonymous access caused the DDoS, which led to giant spikes in some metric, which led them (Elon) to conclude that there was an uptick in scraping, so they imposed the 600/tweet/day limit to punish scrapers.

Seems like either my quota reset or they changed the policy because I’m able to access the site again.

replies(2): >>pschue+Gn >>olalon+jA
◧◩◪
7. partia+C8[view] [source] [discussion] 2023-07-01 23:17:48
>>kristi+D3
That's what makes it so perfect when he echos conspiracy theories, Russian talking points, and other whacky talking points and elevates untrustworthy sources by tweeting things like "interesting" catapulting them into everyone's feed.
replies(1): >>Freedo+sl
◧◩
8. Edward+Da[view] [source] [discussion] 2023-07-01 23:35:34
>>epista+l1
I feel like that book title is paraphrased from the maxim of the leader of the Assassins in that Alamut novel.

"Nothing is an absolute reality, all is permitted"

◧◩◪
9. 18pfsm+se[view] [source] [discussion] 2023-07-02 00:09:39
>>kristi+D3
His claim is that there are a few hundred scrapers causing issues, which makes sense given the recent changes to the paid API.

An assumptions of good faith was once a well-held principle on this website, and it's too bad legacy media has led so many astray.

Hatred for Musk has truly captured many otherwise very logical minds.

replies(2): >>Scalen+0g >>djur+wi
10. reddit+xe[view] [source] 2023-07-02 00:10:24
>>aeyes+(OP)
Depends on the scale of the overall system. I have personally seen and attempted to mitigate degenerate cases where these retries overwhelmed the backend so much that the servers were falling behind in simply rejecting the requests.

Infact it got so bad because of all those retries at multiple levels from upstream callers that requests were essentially timing out at the TCP buffer/queue before they could be processed by the application.

Don’t know if the Twitter homepage backend is at similar scale.

replies(1): >>aeyes+Yy
◧◩◪◨
11. Scalen+0g[view] [source] [discussion] 2023-07-02 00:24:55
>>18pfsm+se
To assume good faith in the face of a mountain of evidence to the contrary would be intellectually dishonest, something that is still, fortunately, held in disregard on this site.
replies(1): >>camjoh+Dk
◧◩◪
12. Negati+8g[view] [source] [discussion] 2023-07-02 00:26:07
>>concor+i4
There's a very well known case of Musk making the same claim multiple years in a row, with it being false every single time.

I prefer to assume the best out of people, but when someone is that obviously wrong that many times with that much personal gain to come from it, I can't believe that man as smart as he is would also be that misled.

replies(1): >>concor+R41
◧◩◪◨
13. djur+wi[view] [source] [discussion] 2023-07-02 00:48:50
>>18pfsm+se
The assumption of good faith for Elon Musk disappeared the moment that he baselessly called someone a pedophile for criticizing him.
replies(1): >>revsca+ln
◧◩◪◨⬒
14. camjoh+Dk[view] [source] [discussion] 2023-07-02 01:04:21
>>Scalen+0g
In fact there’s a giant list of reasons to assume bad faith with anything Musk says. https://elonmusk.today/
◧◩
15. hgsgm+9l[view] [source] [discussion] 2023-07-02 01:08:51
>>epista+l1
George Costanza said "it's not a lie of you believe it."

Is it a lie if you don't know or even care if it's true or what it means? That's where we are with Elon Musk.

replies(1): >>Projec+Ul
◧◩◪◨
16. Freedo+sl[view] [source] [discussion] 2023-07-02 01:10:39
>>partia+C8
As someone who doesn't use Twitter too much, what conspiracy theories has Musk echoed or platformed?
replies(2): >>minima+an >>Edward+2G
◧◩◪
17. Projec+Ul[view] [source] [discussion] 2023-07-02 01:14:38
>>hgsgm+9l
https://en.wikipedia.org/wiki/On_Bullshit
◧◩
18. sander+Rm[view] [source] [discussion] 2023-07-02 01:24:57
>>epista+l1
I know this is getting too political so I'll deserve the downvotes, but totally ignoring anything else, this is why I fundamentally could not and cannot stand Donald Trump. This is his playbook too, just constant willful dishonesty and distortion.
◧◩◪◨⬒
19. minima+an[view] [source] [discussion] 2023-07-02 01:28:55
>>Freedo+sl
The most notable one is the Paul Pelosi incident: https://www.cnn.com/2022/10/30/business/musk-tweet-pelosi-co...
◧◩◪◨⬒
20. revsca+ln[view] [source] [discussion] 2023-07-02 01:31:27
>>djur+wi
Same. Up until that moment he was still somewhat laudable, if weird, at least for casual observers. Then he labeled that guy a pedophile and it’s been pretty much all downhill from there.
◧◩
21. pschue+Gn[view] [source] [discussion] 2023-07-02 01:36:15
>>evan_+z4
This. I'd bet substantial amounts of money that the evil scraper idea is the result of a) another issue + b) paranoia + c) Musk thinking he understands better than anybody else.
replies(1): >>berkle+Lt
◧◩
22. Robotb+bq[view] [source] [discussion] 2023-07-02 02:03:12
>>epista+l1
Man, the really annoying thing is that Elon-haters are so dishonest (refusing to engage when obviously wrong claims are made, repeating things that have been long disproven, repeating hearsay as fact, etc), that it's hard to actually identify any actual lies Elon has made. Meaning, not just exaggerations or failed predictions, but straight up intentional lies.

My favorite is when things are claimed to be impossible, that Elon's lying because the thing is impossible, but then it happens anyway.

Where Elon gets in trouble is he's wildly over-optimistic on a few things, such as AI. He predicted an AGI would take over in 5 years around 2015 or so (so we're 3 years off), and I think he really believed it. That's why he's always claiming things like self-driving, and doing it without sensors or whatever. His paranoia of AGI and his over-confidence on self-driving have exactly the same root cause (believing AI will conquer all). Elon has had so many instances of overcoming status quo expert predictions (whether on solar energy, battery-electric vehicles, reusable launch vehicles, or whathaveyou), that I think it makes him increasingly unable to very effectively listen to experts.

And he's also incredibly gullible and easily taken in by all sorts of scammers, including rightists and just plain sycophants telling him what they think he wants to hear. Which is increasingly what he's left with as everyone else who is sick of his bull has left.

replies(1): >>the_gi+uR
◧◩
23. mrcwin+rs[view] [source] [discussion] 2023-07-02 02:29:36
>>epista+l1
This.

The media is flawed. Journalists are flawed. The Verge in particular has some atrocious coverage of tech.

But the problem with Elon is that in his world, we must rely on him for true information. As if the direct source is inherently true. He’s just as full of it as anyone else with interests.

◧◩◪
24. berkle+Lt[view] [source] [discussion] 2023-07-02 02:42:35
>>pschue+Gn
This is a really ignorant take to dismiss scrapers. LLMs operate by having petabytes of conversational training data. Scraping is how OpenAI trained GPT. It’s how all their copycats are trying to do the same.

Elon can be a monumental asshat, and he can be self-DDOS’ing, and can be accurate about scraping at the same time. It’s why every single social media platform is heading toward becoming a walled garden.

replies(3): >>pschue+Eu >>gmerc+8A >>dyno12+6I1
◧◩◪◨
25. pschue+Eu[view] [source] [discussion] 2023-07-02 02:49:35
>>berkle+Lt
I'm not denying that scrapers exist, I'm just highly suspicious of this explanation given that: a) he's proven time and time again how willing he is to say shit just to get attention b) he doesn't seem to understand software very well c) if shit was imploding for reasons related to decisions he made, this is precisely the kind of blame externalization I would expect.
replies(1): >>evan_+nv
◧◩◪◨⬒
26. evan_+nv[view] [source] [discussion] 2023-07-02 02:57:50
>>pschue+Eu
Yeah, scrapers have always existed and while their traffic is undoubtedly higher than it has been in the past, it can't possibly be any significant amount of traffic when compared to the rest of the traffic hitting the site.

A real scraper would be stopped by a rate limit set to, like, 100 tweets/minute. 600 tweets/day is a completely pointless, punitive limit.

replies(2): >>manque+A91 >>berkle+SJ1
◧◩
27. aeyes+Yy[view] [source] [discussion] 2023-07-02 03:36:28
>>reddit+xe
It is unlikely that a system with the scale of Twitter implements the API rate limiter in the backend. Usually you'd do this as early as possible together with other WAF stuff.

If IPs or IP ranges get really annoying we block them on the network level.

Big public sites like Twitter obviously need to have this technology. Due to their political content they probably also need sophisticated DDoS protection.

replies(1): >>gmerc+eA
◧◩◪◨
28. gmerc+8A[view] [source] [discussion] 2023-07-02 03:53:41
>>berkle+Lt
It’s quite ignorant to assume petabytes of garbage have any value at this point. See Chinchilla
replies(1): >>berkle+ZM
◧◩◪
29. gmerc+eA[view] [source] [discussion] 2023-07-02 03:54:46
>>aeyes+Yy
Sometimes people are cheap and don’t pay their cloudflare bill or their engineers.
◧◩
30. olalon+jA[view] [source] [discussion] 2023-07-02 03:56:49
>>evan_+z4
According to Elon, shutting down anonymous access was itself an emergency measure to deal with the DDoS[0]. Twitter did increase the quota significantly from earlier today[1].

[0] https://twitter.com/elonmusk/status/1674865731136020505

[1] https://twitter.com/elonmusk/status/1675260424109928449

replies(1): >>idiftl+3B
◧◩◪
31. idiftl+3B[view] [source] [discussion] 2023-07-02 04:07:10
>>olalon+jA
The links won’t load
replies(1): >>olalon+FB
◧◩◪◨
32. olalon+FB[view] [source] [discussion] 2023-07-02 04:14:17
>>idiftl+3B
[0] "Temporary emergency measure. We were getting data pillaged so much that it was degrading service for normal users!" (in reference to Twitter requiring users to be logged in)

[1] "Now to 10k, 1k & 0.5k" (in reference to rate limits which were originally 6K 0.6K and 0.3K)

And another tweet that confirms disabling anonymous access was an emergency measure: https://twitter.com/elonmusk/status/1674942336583757825

> This will be unlocked shortly. Per my earlier post, drastic & immediate action was necessary due to EXTREME levels of data scraping.

> Almost every company doing AI, from startups to some of the biggest corporations on Earth, was scraping vast amounts of data.

> It is rather galling to have to bring large numbers of servers online on an emergency basis just to facilitate some AI startup’s outrageous valuation.

◧◩◪◨⬒
33. Edward+2G[view] [source] [discussion] 2023-07-02 05:10:25
>>Freedo+sl
When Twitter's working, just search for any tweet where he says "Interesting".

Or Google "Elon Musk conspiracy" and click the top link. I don't know if you'll be able to read the tweets they reference though.

◧◩◪
34. daniel+UI[view] [source] [discussion] 2023-07-02 05:49:31
>>concor+i4
Which one of those did calling a cave-diving child rescuer a pedo fall under?
replies(1): >>concor+t41
◧◩◪◨⬒
35. berkle+ZM[view] [source] [discussion] 2023-07-02 06:36:13
>>gmerc+8A
I agree, but there are hundreds if not thousands of AI startups trying to make their own relevant LLM, and they're going to be scraping Twitter. The Onion called it many years ago [1]: "400 billion tweets and not one useful bit of data was ever transmitted".

[1] https://www.youtube.com/watch?v=cqggW08BWO0&t=138s

replies(1): >>rightb+7Y
◧◩◪
36. the_gi+uR[view] [source] [discussion] 2023-07-02 07:26:16
>>Robotb+bq
He has always been a rightists. He just dumped money on some cool stuff like self driving electric cars and rockets. Which is really cool, of course. But he wasn't "left with no other options". That's a lame excuse.
◧◩◪◨⬒⬓
37. rightb+7Y[view] [source] [discussion] 2023-07-02 08:39:22
>>berkle+ZM
I can't imagine worse training data than e.g. Twitter and Reddit posts. How about like, dunno, books?

Edit: Ah, nvm, if you are trying to do a chat bot it is essentially what you want.

◧◩◪◨
38. concor+t41[view] [source] [discussion] 2023-07-02 09:51:23
>>daniel+UI
Imo: jumping to conclusions based on stupid stereotypes.

He should have lost that libel case imo as I do think he meant it.

◧◩◪◨
39. concor+E41[view] [source] [discussion] 2023-07-02 09:54:00
>>epista+w4
I think it's an important distinction to make. Because it does means that you can actually infer true things from his statements (or a journalists statements, as news orgs do this kind of "lying" a lot), unlike with a habitual liar.
replies(1): >>epista+BS2
◧◩◪◨
40. concor+R41[view] [source] [discussion] 2023-07-02 09:57:02
>>Negati+8g
This is the guy who still thinks a Martian colony is a good idea and that the woke mind virus is one of the most dangerous threats to society. I'm not sure he's very grounded in reality at all when it comes to the future.
◧◩◪◨⬒⬓
41. manque+A91[view] [source] [discussion] 2023-07-02 10:57:49
>>evan_+nv
Depends on how they scrape .

100 tweets / minute is hardly deterrent for a botnet using comprised devices on non data center IPs

◧◩◪◨
42. dyno12+6I1[view] [source] [discussion] 2023-07-02 15:33:14
>>berkle+Lt
How do you feel about search engine indexing?
◧◩◪◨⬒⬓
43. berkle+SJ1[view] [source] [discussion] 2023-07-02 15:41:45
>>evan_+nv
> A real scraper would be stopped by a rate limit

I'm guessing you've never played an offensive or defensive role in scraping because what you've described is in no way a problem for a serious scraping effort. I agree the rate limits are stupid. They fuck over users, they stop amateur scrapers, and do nothing whatsoever to impede professional scraping.

If you want to stop most scraping, employ device attestation techniques and TLS fingerprinting.

replies(1): >>costco+OZ1
◧◩◪◨⬒⬓⬔
44. costco+OZ1[view] [source] [discussion] 2023-07-02 17:23:10
>>berkle+SJ1
But then you have to contend with this: https://github.com/bogdanfinn/tls-client... Just used this to bypass a Cloudflare check! I've never scraped Twitter but Elon said there was a large scraping operation from Oracle IPs. He could substantially raise the cost of scraping by just banning datacenter IPs. Something like p0f would probably help too. I pay for static residential proxies (basically servers running squid that somehow have IPs belonging to consumer ISPs) and with TCP fingerprinting these would be detected as Linux and expose my Windows or iPhone user-agents as inconsistent but I've never encountered a site that checks this. Although maybe sites are doing so silently but I don't notice because I don't otherwise meet the bot threshold.
replies(1): >>berkle+hd2
◧◩◪◨⬒⬓⬔⧯
45. berkle+hd2[view] [source] [discussion] 2023-07-02 18:39:36
>>costco+OZ1
for sure, using a custom TLS library like uTLS helps -- need to inject that GREASE cipher selection. I have a suspicion that private residential proxies are out of budget for many outfits, or the IPs are too few and then simple rate limiting kicks in? Who do you use if you're willing to share? I've not been happy with the, uhh, questionable ethics of Luminati/BrightData in the past.

There are definitely more and more sites doing TLS/TCP/etc fingerprinting or device attestation for mobile APIs, but it's still pretty rare. I mean Twitter is trying to limit requests by IP, so definitely amateur hour over there.

replies(1): >>costco+9n2
◧◩◪◨⬒⬓⬔⧯▣
46. costco+9n2[view] [source] [discussion] 2023-07-02 19:44:38
>>berkle+hd2
I use https://www.pingproxies.com/isp which is like $3/IP/month and unlimited bandwidth (I assume if you used a ridiculous amount they might charge you). Luminati pricing is extortionate. I have no idea how anyone doing anything at scale can afford $10/GB. I haven't investigated but I don't know if Twitter limits are per account or per IP.
replies(1): >>berkle+do2
◧◩◪◨⬒⬓⬔⧯▣▦
47. berkle+do2[view] [source] [discussion] 2023-07-02 19:52:36
>>costco+9n2
Seriously. I don't even consider a provider if they want to charge for bandwidth. I'm doing about 50 TB/mo atm.
◧◩◪◨⬒
48. epista+BS2[view] [source] [discussion] 2023-07-02 23:26:46
>>concor+E41
I don't agree. There is little to learn from Musk's statements. And lumping all of journalism as being as dishonest as Musk is just plain silly. Sure, if you're on the right the media which is biased towards yoi has been serving you dishonesty and lies for a long time. But that's not journalism as a whole.

Saying "he really believes his lies" is not excuse, because most habitual liars are exactly the same. They have prioritized their own narcissism so far above reality, but that doesn't make them any less of a liar. They are just lying to themselves and everyone else.

[go to top]