> The implication that tweets - and by extension, history - should be mutable feels like 1984.
You've pointed out the problem: people in a democracy need access to the truest history possible; mutable history is a slippery slope.
But I disagree with
> Then don't tweet dumb stuff?
The challenges we're facing are new, and need new solutions. Political entities are acting on incomplete information and perspective; inevitably, they'll tweet dumb stuff (like I and many on HN have). I'm not saying that they should be able to delete tweets, I'm saying that politicians haven't had this ability before. Now that they have it, we (the democracy) have to decide what to do about it. "Don't tweet dumb stuff" is impractical and simplistic (sorry!).
Considering a democratic populace wants from its leaders intelligence, strategy, and (at least in the US) authenticity, American leaders will be struggling with a tension between authenticity and advertisement. This struggle leads to the deleting of tweets. I don't know the solution; knowing the solution is a different matter.
I don't know the solution; knowing the solution is a different matter.
The solution is to actually be intelligent, strategic, and authentic. On a really good day, a given politician is batting two of three there.
I think that part of the accountability of these folks is precisely that their gaffs are on easy public record. If we allow memory holes for the elite, they can keep fucking up and we're none the wiser.
Perhaps replace the ability to delete tweets with the ability for the tweeter to add notes the tweet. That way you cannot get rid of your twitter past, but you can attach context or a defense to it.
You can stop reading your comment after the first three words for the real lesson. Politicians will just pay anodyne spokedroids to manage their accounts for them.
I think this gets at the root of the decision at Twitter. They were faced with a problem: they want public figures to tweet, but tweeting can be risky. Killing Politwoops makes it just a little safer, and thus they hope more public figures will adopt the service.
Frankly, I think it's a bad call. If Twitter is infrastructure, it wouldn't work that way, but Twitter isn't infrastructure it's a company and I can see why it makes sense to do that as a company.
It's almost like capitalism doesn't make for good government watchdogs...?
I'm confused. Are you suggesting that we should follow twitter's TOS on this matter and be done with it because eventually politicians and the like will just work harder to filter what they say?
What's to stop someone from creating a service for themselves that has similar TOS (Anything published if later deleted MUST be removed from any data store containing it) then they use said service to communicate everything they want to say essentially giving them an undo/delete button for everything? This sounds terrible to me.
Yup. There is no undo on the Internet. Someone, somewhere is probably saving everything you publish on social media, be it the government, orgs with good intentions, spammers, etc.
Even better - the populace should develop a functioning brain and judge something on its merits. I would never hold it to someone about making a funny joke. Even if it is very crass.
For those unfamiliar with the Twitter API: the Streaming API pushes out notifications of when users who are eligible to show up in your timeline (i.e. users whom you follow) have deleted a message:
Politiwoops used monitored this to see which tweets it has previously collected were now deleted. You can peruse the Politwoops source code to see the details:
Twitter's complaint is that its terms of service forbids ignoring those deletion messages...but in the past, Twitter had turned a blind eye to what Politwoops was doing. Whatever the reason for the crackdown now, it doesn't mean you can't manually try to track these embarrassing tweets and then publish them under fair use protections...in fact, the Anthony Weiner scandal, in which a user noticed an accidentally posted DM, was what sparked Politwoops in the first place. It's just you can't do it on a wide-scale, automated basis via API.
I'm kind of surprised Politwoops informal waiver lasted this long...or rather, that no one else has applied its open source code to the tracking of celebrities...if that had ever happened (say, as an informal service of TMZ or Gawker), I have no doubt Twitter would shut that down, as celebrities would be spooked by having their mistakes be essentially non-deletable...and their protests would carry a lot more notice than politicians'.
edit: Another relevant link: the unitedstates/congress-legislators repo, which is a crowdsourced listing (augmented by a automated framework) of Congress social media accounts: https://github.com/unitedstates/congress-legislators
If you've ever worked with Twitter's API, you should know that not honoring deletes on your app is against the TOS. It doesn't matter what accounts are being monitored, when a tweet is deleted by the account, you have to delete it from your server.
Detecting another liar politician supersedes any API policy. You can also say that Snowden leaks are against the law but the importance of his relevations makes arguing about a ToS or law very naive.
I think there might be some mixing up of the morality of the situation and the legality of the situation. Were the Snowden leaks against the law? Yes. Were the Snowden leaks immoral? I'd say no, they were an incredibly brave thing to do.
Now a situation like that where the moral thing is against the law would require us reexamining and changing the law. Whether it's whistleblower laws, minimum sentences or sentences for minor drug offences, these things all require a good hard look, and most likely, a degree of retroactive pardoning. But believing one's cause to supercede the law is a dangerous road to be on.
Now ToS are completely different. ToS are set by private entities, and as a single user you have little power over their content, only whether you accept them or not. They're an offer, not a democracy. I personally think that if you choose to knowingly violate the ToS of a service, then you should face the consequences of that regardless of the moral standing you may have. If you were using the Twitter API to defeat african warlords and Twitter decides that you're violating their ToS doing that and bans your account for it, well...you're SoL.
> If you were using the Twitter API to defeat african warlords and Twitter decides that you're violating their ToS doing that and bans your account for it, well...you're SoL.
Not so fast, you can always use web scraping techniques as a defensive measure.
TOSes can't be illegal, and I'd argue that requiring deletion of tweets by public officials constitutes some sort of violation of open records / freedom of information laws.
You seem really confused about how this works.
If your argument is "twitter can't delete these due to it violating US law" that is an argument about twitter, and not it's TOS.
If your argument is "people don't have to follow the TOS because the TOS requires them to not comply with the law", that is simply wrong.
If the TOS asks you to do something illegal in your country, it does not make the TOS illegal per-se. It means you can't use the service, because you can't comply with the TOS.
There generally is no "well, i can do whatever i want and it's twitter's problem because i can't comply with the TOS".
The fact that you can't comply for legal reasons is your problem, not the TOS's or Twitters.
The "you can't use the service" because the TOS conflicts with the laws in you locality bit isn't necessarily true - it varies a lot by country. Plenty of places just invalidate the parts of your TOS (or any other contract) that conflict with the law. Which leaves it up to the company to decide whether they wish to pursue business in that locality.
IAAL :)
So, "conflict with the law" is a not-well-defined term.
There are 196 countries, so yes, you could say it varies. But i'm not sure which are the "plenty" you are thinking about, and i'd love to see specific examples of what you are thinking about.
Generally, most countries have pretty consistent law here.
If the contract is a contract for you to perform something illegal (like murder), that will not be enforced.
A contract that says "you must do X if you wish to use our service", in basically any reasonable country you can think of, is not going to be invalidated if you can't do X because it's illegal when done in very certain ways in very specific circumstances.
This is because, again, in general, you can't claim the benefit of the contract (using the service) but not have to have the burden.
<i'd love to see specific examples of what you are thinking about>
IANAL, but here goes:
In The Netherlands certain clauses in TOS ('algemene voorwaarden') are void by default when they are against a limited list written down in basis contract law. Another list of clauses are null when they are not unreasonably limiting. So any contract within Dutch jurisdiction is limited in that way.
A contract with X.com would not hit Dutch jurisdiction directly, although with a Dutch website / translation and perhaps a Dutch office you easily come into reach.
Examples: You cannot be held to TOS that are not willfully handed to you. (So no tiny signs in stores, or hidden on websites. Exact translation obviously ambiguous.) Changes in pricing, especially early in the contract, usually give the right to cancel the contract. Should be able to cancel contracts anyways. Should deliver timely, and cannot contract 'untimely delivery'.
Still, daresay that contract law on the internet in the Netherlands is not a big field of litigation. For one thing, our damages are usually only 'true (financial) damages'. So in the context of (nearly) free products, you hardly get any litigation. Class action is limited, too.
Example here: consider a webshop that cancels your order after a discount gets too popular. You could probably demand fullfillment, since they could easily deliver from other retailers at a loss. But considering the costs of litigation are many times the difference in price, most people resign upon ordering elsewhere. Can't even say I regret that injustice, since the benefits of not clogging the judicial seem large enough.
I kind of thought the generally consistent law was the way you describe it – however my not a lawyer understanding was if the contract requires X which violates the law in a very specific circumstance in a very specific way that particular clause under that particular circumstance, that particular way would be void, but the whole contract would not and would still stand - benefits and burdens combined. Including that clause, under "normal" circumstances.
Which is basically to say my understanding was that, in most countries, you can reasonably sign a legal contract and know you will never have to commit an illegal action in pursuit of it. And if your counterpart in the contract really wants you to commit an illegal action it's on them to decide whether they want to do business in a country where the action is it illegal. Not on you.
Once again, IANAL. And my understanding could be way wrong.
"Which is basically to say my understanding was that, in most countries, you can reasonably sign a legal contract and know you will never have to commit an illegal action in pursuit of it. "
If the only way to complete a contract is to perform an illegal act, you will be excused from performance.
But if you have an alternative, it's a valid contract for the legal alternative.
IE if i say "i'll pay you $1000, and you will deliver to me rice or smuggled opium", this is a valid contract to pay $1000 for rice.
Also as i mentioned, you will generally be required to pay restitution. That is, pay back any benefit and least stop using/doing whatever benefit you are accruing(in this case, i presume using the service).
In what way are you referring to?
I'm trying to understand what distinction you are trying to draw.
You are simply stating this bare as if it makes any sense, and i can tell you, it does not :)
Unilateral contracts are still contracts :)
(you can make some sane distinctions between licenses and contracts, but only in pretty simple cases, and twitters TOS is not a copyright license, but instead, an actual contract)
I would say that the argument should be for denying public officials the right to conduct any official business, campaigning, and communication with the public via non-transparent/non-auditable channels such as Twitter.
Hypothetically, if Yahoo! automatically bans a suspected spammer account, and it turns out that the account was actually H.Clinton's official address, it shouldn't be Yahoo!'s fault for not knowing the legal identity of the owner -- rather, it should be the owner's responsibility to use the appropriate channels to conduct official communication, and stick to the ToS otherwise.
I don't know that this is true, though. Not all of a politician's life must be open, just that through specific - often official - channels (read:protocols) of communication. I'm not sure that Twitter qualifies.
Well, then, it shouldn't be a problem to write a service that automates the capture of screenshots of tweets posted by a particular account or on a particular hashtag. That doesn't even require access to the Twitter API, or, for that matter, having a Twitter account.
1. Tweets wouldn't be stored. Screenshots of tweets would be stored.
2. Admittedly, that would be the most inconvenient part of the process. This would require regular scraping to validate what's been deleted and what hasn't.
> was extensively used by the media to investigate instances of deception, corruption and ineptitude.
Because getting an intern to hack up a 5 minute Python script to do the same is too much effort? OK maintaining a list of current politicians current Twitter accounts is probably where the work is, but if you are interested in a number of prominent accounts this is not that hard a task.
The twitter api allows to get a feed per author name, so if you have a list of users, that's pretty easy. All you need to do is use a basic http client lib in your prefered language to do the http query and then some basic html css skills to show the result there :)
It's a perl script that can hook into Twitter stream API. Sign up for an account, create an access token for ttytter, start subscribing to politicians, and it will suck down every tweet they make in realtime as they make them. If they make a tweet and delete it 30 seconds later, you'll have captured it. That is why whining about deletability of tweets is pointless, anybody with entry-level Linux skills can monitor thousands of politicians.
It can be easily done; but I think the problem here is that if you're using the Twitter API (as is the case with the link you mentioned), you are required to honor deletion requests. If you don't, then your access is revoked, which is what happened in this case.
Now, you could just archive the tweets internally and reveal them under some other name...
The choke point seemed to me to be that they end up notifying you of it via...twitter. You can go to their site and still see deleted tweets. If you weren't putting them back up on Twitter, you could easily monitor politician tweets and never get caught by Twitter.
How do archive sites and sites like Storify get away with storing deleted tweets? Storify even lets you repost your stories on Twitter.
Then again, with the thousands, if not tens of thousands of oauth credentials checked in on public github repos, I don't really see that revoking access to one credential as a major bottle neck.
Politwoops could circumvent this problem by just not publishing the tweets to twitter. They could just put them on a website, and maybe even aggregate the politicians other communications. No?
Though generally I feel that anyone at all being judged on an instant 140 character message is pretty sad state of affairs. Meaningful debate (political and otherwise) with evidence based arguments and so on seems like a naive fantasy these days.
No, alas - the Twitter terms of service are very clear. If you are storing a tweet, when someone deletes that tweet you have to delete all of your copies of it. Full stop.
This feels very "right to be forgotten" which was almost universally reviled here. I'm actually surprised there isn't more discussion about the similarities (and similar possible negative outcomes) in light of Twitter's TOS and actions.
So what? How do you connect that API key to some anonymous web site that saves deleted tweets?
The only way I can think of is for Twitter to publish different deleted tweets to different API keys and then see which one the web site surfaces. Not very hard but not trivial either.
And how would Twitter know which API key somesitewithtweets.com is using? There are ways, but if the API key user is careful and doesn't allow arbitrary searches (so Twitter can't try searching #someguid and then kill that account) it'd be pretty hard for them to figure it out. They could try adding fake deleted tweets on a per-account basis, then seeing who publishes those. A simple two-account double check or manual intervention would prevent that from leaking info.
Then just eschew the Twitter API and scrape the website, then tell Twitter to go pound sand. Doesn't do anything about getting hit with a frivolous lawsuit but that's a different topic.
Probably not Politwoops, because they were using the API. But perhaps a successor that did not use the API, but scraped and screenshoted only might have a good chance.
I can't blame them. These things are incredibly toxic to the Twitter community. Already, you have to warn everyone who joins the site not to delete anything because people run services that highlight things that get deleted.
I mean that descriptively, not normatively. I mean: no matter how you slice the public interest, these things are clearly harmful to Twitter itself.
Obviously, blocking them doesn't actually help the situation; once you publish something on the Internet, it's out there, and people can come up with 100 different ways to get around Twitter's restrictions.
I don't believe repeating this viewpoint is useful. It is a kneejerk reaction that in my opinion fails to take the actual situation into account.
This wasn't a service monitoring everyone, this was a service monitoring elected officials, whose statements, whatever they are, are part of the public record. This is especially important as particularly politicians derive a lot of their power from what they say (and often promise) to their constituents, regardless of whether they actually stick to what they say or not.
While I personally feel that Twitter taking down this service is a shame, and yet another sign of how reliance on these private companies for expression of free speech is never going to work, tptacek is right. When it comes down to it, Twitter as a private company is completely allowed to do this, and in fact it is in their interest to do so; these services harm Twitter as a whole, despite being (in my opinion) a net good for society. Of course Twitter would act in it's own interests here.
There's a lot that is legal that is morally reprehensible. Like bringing the world to the brink of financial collapse and then having governments bail out the institutions responsible while absolutely no one went to jail for the crime.
Sure Twitter is allowed to do this. It has nothing to do with whether the attempt to alter (or delete) history is morally reprehensible or not.
Calling someone else's viepoint not useful is itself counterproductive and not useful. Twitter doesn't in of itself owe anything to the public interest. It's a company; it needs to make money. It isn't funded by the government.
> Twitter doesn't in of itself owe anything to the public interest.
Twitter exists at the sufferance of the society that grants it its corporate charter. It most certainly does owe a great deal to the public interest--or it would, in a healthy society, be penalized, up to and including destruction of its convenient legal fiction, for acting against it.
What about turning the service to an events feed yous suscribe to? If it's just a log of "Politician X tweeted Y ... Politician X deleted tweet Y", would that be against the TOS too? You aren't storing tweets, just emailing or RSS'ing them, do you still have to delete them?
I find it odd that no one compares this to real life:
If I say "Tom said y", it is perfectly okay, why is this any different online? Why should what you say online be treated any different than what you say in front of other people?
I wonder what's going to happen with the RSS feeds of tweets. They usually get everything verbatim, and if readers read from it often enough, the tweet's done.
They'll likely hit $3.5 billion in annual revenue within two years. One year after that, they'll be the size of Yahoo (which is 12 years older than Twitter). They're already on a revenue ramp dramatically better than the one Netflix has had.
If that isn't a model that works, I can't wait to see the one that does.
Twitter's primary financial problem is an unnecessarily bloated cost structure.
They spend a ton on data centers and talent. Last quarter there was no user growth and profit per share was 7 cents. Shareholders definitely don't agree with you about this.
What's not to agree with exactly? The fact is they grew sales 61% - with zero user growth. For a company that large to be growing that fast, they're very clearly having no problems with growing the business.
They need 47% growth in the next four quarters, and then 36% growth in the four quarters after that, to hit $3.5 billion in sales. Given first quarter growth was 74%, then second quarter growth was 61%, it's not a far stretch to hit 47% over the next four.
Shareholders have sent the stock down on the basis of user growth, not sales growth.
Twitter has three problems.
1) They're not Facebook, it is not about users, and they should stop trying to be something other than a broadcast platform. It's about consumption.
2) They were extremely overvalued right out of the gate. There was a lot of delusional thinking around their potential. There's one Facebook, there isn't going to be another any time soon. Twitter will never be another Google type company.
3) They're carrying vastly more cost in their business than they need to. They have 36% the head count of Facebook, with 12% of the sales. They've stacked the business on the premise they were going to be a massive company like Facebook - they're not, it's time to adjust costs accordingly.
Yahoo is one of the largest consumer tech companies on earth with $4.6 billion in sales. Yahoo's problem isn't their size today, it's their lack of sales growth for ten years. Twitter doesn't have a lack of sales growth.
Then don't tweet dumb stuff? The implication that tweets - and by extension, history - should be mutable feels like 1984.