Twitter rate-limits itself into a weekend of chaos It's been a few weeks since the chaos at Twitter rose to a level worth noting, but that changed this weekend when owner and CTO Elon Musk announced the imposition of limits on how many tweets users can see each day. It began on the afternoon of Saturday, July 1 with a tweet from Musk informing tweeps that harsh rate limits …
A couple of years as a monk as portrayed in Monty Python's Search for the Holy Grail would be better.
Heck, I'd even donate the plank.
:)
Here is a Californian who did just that...
Ann Russell Miller/Sister Mary Joseph
(And... just like Musk, she had 10 children - though, unlike him, she carried them to term and then some)
Ah, but the article doesn't say that the adverts are being rate limited.
The further you scroll, the greater the ad:tweet ratio becomes - it'll work because by the time you get to 80% plus ads you have sunk into the state of infinite scroll hypnosis.
With appropriate scaling, they can guarantee that you will never actually hit the tweet limit within any 24 hour period. You noticed Musk adjusting the limit up a couple of times? That was just the result of live testing: a few die-hard fanatics were scrolling faster than the in-house testers (both of him) managed. Probably the International Thumb War champion amongst them (not in the Musk Time Zone or "Elon Sols" as it is known).
Sound like another case of him trying to save a few $ by rate limiting for him to not understand how the Twitter site works and that it might end up costing more in the long run.
Social media relies on engagement and if people get to their allotted 600 tweets viewed, I doubt they will all be rushing to pay for a subscription to see more, more likely they will just open another app such as Instagram and carry on doom scrolling on there, so Zuckerberg will be getting more ad views and therefore revenue and not Twitter.
Sound like another case of him trying to save a few $ by rate limiting for him to not understand how the Twitter site works and that it might end up costing more in the long run.Social media relies on engagement..
.. And social conditioning relies on disengagement. I'm kinda curious how many Twitterati will actually hit their quotas, and if they do, should perhaps consider their life choices. But in the wider world of politics, governments are looking for ways to censor 'social' media. We've seen this already with accounts being cancelled for offending TPTB by not blindly following various narratives, or expressing unpopular views. Assorted AI and 'fact checkers' are hoping to make big money providing censorship tools, and have been training their bots using Twitter & Facebook content. If bot herders are making money, or weaponising Twitter, doesn't some countermeasure seem reasonable?
"trying to save a few $ by rate limiting for him to not understand how the Twitter site works"
I'm not sure that's right. 600 tweets at half a minute per tweet is 5 hours. Even the most hardcore of 'free' Twitter users would be hard pressed to push those limits,
And even the fastest speed reader / speed scroller reading tweets at 6/minute would take up all almost 17 hours to read 6000 tweets, so that higher limit is clearly designed to limit bots, not real human users (as, in fact, stated by Twitter).
I'm pretty sure that for all the screw-ups, Twitter know how many tweets/day the vast majority of real human users can read and set their limits accordingly after a couple of tweaks.
> I'm not sure that's right. 600 tweets at half a minute per tweet is 5 hours. Even the most hardcore of 'free' Twitter users would be hard pressed to push those limits,
Tap into one of those tweets.
Oh no, it has 1000 replies - goodbye today's rate limits.
As someone mentioned below, there have even been cases of people hitting the rate limit because of the number of replies on a tweet *they* posted.
Not to mention, those that have found that - once you've hit the limit - you can't even view your own tweets.
Is it possible to build rate-limiting that works the way it is in your head? Certainly. Is that what Twitter have done? No.
"Is it possible to build rate-limiting that works the way it is in your head? Certainly. Is that what Twitter have done? No."
Aha! seems like I was missing an essential piece of information there! As to who would even think that reading a tweet would automatically count reading all the replies.... well I guess all the best engineers quit as soon as the dumpster fire became apparent
When you come to this site, how many articles do you read and how many comments? Did you read the tens of comments that were on this page? I'm guessing you did when you replied to them. It doesn't take half a minute to read each comment and most are significantly longer than a tweet would be.
It would be a rash thing to take anything Musk says at face value given he's a compulsive liar.
There might be a rate limit to stop scraping (doubt it, it was never a problem before), it might just be an effort to cut cloud bills by putting everything behind a login screen that backfired as clients not expecting the login screen started DDOSing it, or there just might just be a system disappearing up its own arse with nobody around who can fix it.
We'll never know until until Canada takes in some of Musk's H1B slaves and they start talking.
"Your turn to hold the Twitter Control Panel[1]; each button is a policy you can enact[2]"
[1] TWP, proudly made by Vtech
[2] No, no, it is just a fancy word that means "do". Now, where did you put your drink[3]?
[3] Tommee Tippee's latest "CEO proof" model.
I thought that was Prime Minister of the UK...
There was a period in the late 90s when a succession of Conservative Party leaders seemed to last about half the time of their predecessors. Had that continued everybody in the world would have been Tory leader by about 2003, the last for about 3 nanoseconds.
I am being punished in the new regime because I have a quick brain. I keep being put in the bot cage for liking tweets ‘too quickly’. No matter what stratagems I use to slow it down, to attempt to measure the time (highly variable) etc have come cropper of the capricious new rules. Anything seems to be the rule. I am currently in the bot cage for the fourth time today. Yesterday I spent most of the day in it from the first ‘infringement’.
The whole thing is both arbitrary and capricious. You are given no time limit for your punishment and no notification for when it has ended.
I have seen a suggestion this is actually because Musk has extended his ‘get paid only if you sue us’ attitude to office rent to paying for hosting, server farms etc. So Twitter itself is being rate limited in response.
Advertisers need to know there are fewer eyes on twitter as we are all serving time outs.
Back in pre COVID, pre WFH days when I regularly took the train as part of work commute (train had lots of college students too) I used to be impressed at the lightning speed students scrolled and liked posts on various social media apps: I could see why any attempt to flag "bots" could easily trap people like those students (I'm a great speed reader & I could not have read stuff at the speed they did, either they had phenomenal reading speed or they were "skimming" posts & "liking" on general gist of post (or maybe just liking as a social nicety) )
I don't like embedded tweets: it's passing my IP-address to another provider without my permission (this is the whole reason behind it, a sane system would provide an API).
Mining Twitter for AI? I think not. The days of "sentiment analysis" based on Twitter passed as soon as researches realised that it's full of bots and the few users who aren't bots are from a very skewed dynamics. Pity the journalists never got this note, but it was just so easy to copy and paste a few quotes as "representative" of opinion.
I now heard several reports that rate limiting was enforced as part of contractual negotiations…
There are other, less actively hostile, alternatives.
Unfortunately none of the have Twitter's reach. Mastodon has its own quiet version of culture warfare going on, with many sites rejecting federation with other sites, not because of their content, but because they don't like the software being used FFS!(*) Bluesky is overwhelmed with users already because it's still a beta but many people are trying to get on board. Spoutible may be a scam (the "developer" keeps asking for more funding to add extra features but the software seems to be off the shelf PHP with those features already provided). And now we've got Meta's Threads with all the lack of personal data privacy associated anything Zuckerberg does.
Bring back Usenet!(**)
(*) Seriously. The argument being basically "this software is used on sites I don't like, therefore I don't like sites using this software".
(**) Yes, I know it actually never went away.
"Twitter tech lead also decided last week to block all users not signed in to Twitter from viewing any posts whatsoever, inadvertently killing tweet embeds in other apps. This, Musk claimed on Friday, would also be temporary."
He made the claim in a tweet that, unless you were logged in, you couldn't read!
1. Twitter gets its contents at no charge from its posters.
2. Twitter re-sells that content via advertisement views.
3. Twitter sells subscriptions.
4. Twitter fails to honor its financial obligations to other companies.
5. Twitter (or, at least, Mr. Musk) is unhappy that some other people and companies are obtaining the content which was given gratis to Twitter.
My empathy for Twitter's "problem" is best expressed in femto-give-a-shits.
I said back when he started breaking things that the line in the sand for when Twitter would enter a death spiral would be when it is no longer possible to see tweets without being logged in. As of a couple days ago, that's now the case.
That means no more ability to click on Twitter content/link on a page and be taken to the full tweet including replies, unless you login. Since most people don't have a Twitter account, and aren't likely to sign up just to view tweets, that makes Twitter less relevant to the world today than it was last week.
It is also breaking a lot of existing web pages - if you had previously embedded a tweet that contained media like a picture or video that's now gone. A logged in user could click on it to see that, but for most people it will just look like a partially broken link and they'll ignore it. I'm not even sure it is still possible to embed tweets anymore...if it is you'd have to be logged in to do so, which adds another hassle and between that and most of your readership being unable/unwilling to view the full tweet I imagine the number of embedded tweets in newly published pages / newly written posts going forward will take a serious dive. Again making Twitter less relevant.
I guess there is a reason Bluesky reported record activity this past weekend, and had to suspend new signups to keep from falling over - and that's despite it still being invitation only. Hopefully they can get the resources to go all-in and allow open signups. Twitter's activity will drop off a cliff once Bluesky is fully ready, and Musk will be 100% solely to blame.
Brilliant stuff, Mr CTO, He hasn't realised that a cloud hosted website, where you pay per processor cycle, will actually use more processor cycles determining whether or not someone is logged in or not and serving up 404 page not found responses hundreds of times, than actually just serving up the requested content once.
But if he's getting miserly with his cloud processor cycles, it must mean that the advertising revenue has gone completely down the pan, because any cloud hosted website usually uses ads to finance that quite easily.
This is quite beside the fact that a "pay to read" site requires very good content to get people to pay, and shutting people out is going to reduce the amount of content available, further reducing the number of advertising and subscriptions they will get. Call it the twitter death spiral.
Seems the real issue was the "hundreds of times" part, which was just sloppy coding. The webapp could have been patched before the switch was thrown, so it'd stop trying to pull content when it wasn't logged-in. Apparently this didn't occur to anyone (who still worked there).
The webapp could have been patched before the switch was thrown, so it'd stop trying to pull content when it wasn't logged-in. Apparently this didn't occur to anyone (who still worked there).
Maybe it did but they have a classical appreciation of hubris.
While I generally agree with your comment, the idea that a pay-to-view site needs to provide high-quality content to attract paying customers is clearly not true. If you want proof, check out the Daily Telegraph, for instance.
“Blue Bird In the Emerald Mine: 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."
The whole quote is going to be the subtitle. That statement is, without a doubt, the most out of touch thing I’ve ever heard a tech leader say.
《Am I the only one wondering what the heck an AI Agent trained on zillions of Twitter posts could possibly be good for?》
Only thing that comes to mind is a replacement bionic brain for Trump or indeed Twitmeister himself.
I recall an old joke where some codger explaining the name of the then new fangled twitter to another codger - to put it more politely basically he held its was a portmanteau that could be a synomym for the perineum. Quite clearly twitter has now well and truly plummeted up the back passage.
"is Musk now claiming he can reverse gravity?"
Claiming he can reverse gravity? That doesn't sound like Elon. Now if he were claiming that SpaceX or Tesla (or the Boring Company for that matter) were working on reversing gravity and confidently expected to release a product in six to nine months ... THAT sounds like Elon.
For all their faults, one thing reported about the LLMs is that they generally output good English: excellent grammar, large vocabulary, correct use of punctuation and they even have a grasp on essay/story/script structure.
I even have hopes that, seeing a lot of generated text all over the place, the general public will start to pick up these traits.
Who knows, maybe we will once again have a generation that knows the word "take" exists and serves a purpose, instead of continually replacing it with "bring"!
Please, please, don't ruin the one decent thing about LLMs by scraping Twitter!
Would they cry or laugh like a maniac when seeing all links redirect to a blank paywall landing page?
This is how you reduce new customer acquisition to zero. This is how you drive away extremely high-value accounts that were using Twitter for customer service and important public announcements. When there's a list of social media contacts on a website, Twitter will be the one that doesn't work.
I'll guess laugh like a maniac.
Wasn't the Muskrat all about spotting bots while trying to get out of buying Twitter? How come now it looks like that he and his team now cannot spot scrapers and has to rely on draconian rate limiting for everyone? Surely he wouldn't be lying?
Last night one of the news magazine programs in Japan covered this latest disaster, and random people on the street were saying they were burning their unverified quota in just 10 minutes. Also, most every local and national government department uses Twitter for emergency notifications, etc, and with the country right in the middle of serious flooding and evacuation notices, with Twitter pining for the fjords word may not be getting out properly.
I deleted Twitter back before Musk put the deal together.
I actually gave up “life online” once already, prior to the era of the World Wide Web.
I had my own computer, paid for my own telephone line, programmed my own bbs.
But then the “scene” exploded, there was practically a bbs phone book (just for the local numbers), technology was evolving. But really, I found engagement to be less & less rewarding.
Then I graduated high school, sold ALL of that stuff, and did not hang around to witness the demise of bbs culture.
…although I did recognize the worst parts of bbs culture in Twitter.
About once a year I accidentally sight a twitter post because I mistakenly read a pretend journalist who instead of writing a story, takes a screen grab of a twitter message and uses it as a reference.
Usually, I am quite good at spotting this, and once I see the by-line of a twitter cut and paste artist, I simply back pedal and read a different story.
Twitter messages should be limited to verified users only and they should not be able to repost them anywhere else. A nice walled garden for special people. It is okay to be special, as long as it does not hurt others.
Elon is doing a good thing by destroying this horrible platform, and of all people on earth, he has the means to do it and not care about the costs.
600 views isn’t as much as you may think. Tap a political tweet, and scroll the replies to see what people are saying… Your daily quota is used up in like 15 minutes, and if you have participated in a thread, you’ll have to wait until the next day to follow up.
Long story short, Twitter is currently unusable for anything but very light use.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
