5 ms·
From an engineering standpoint, this is definitely a cautionary tale about why you should be worried about vendor lock-in. If you don't own/maintain your entir
by jlkuester7 6y ago
From an engineering standpoint, this is definitely a cautionary tale about why you should be worried about vendor lock-in. If you don't own/maintain your entire deployment stack (right down to the bare-metal servers that your code is running on) you need to have some kind of plan for being able to migrate to a new vendor (whether because of a TOS violation or just the original vendor shuts down).
I don't know the technical details of the Parlor setup, but depending on how integrated they were with the AWS ecosystem, (e.g. using Lambda executors instead of just running services on EC2) moving off of AWS could mean huge changes to their architecture....
That being said, even in the best case scenario this kind of migration would be very expensive (S3 data egress fees??) and a 24 hour timeline seems absurd! Can anyone with experience in this area speak to if this kind of accelerated timeline is normal when it comes to web-hosting TOS violations? I just cannot imagine signing a contract with a hosting company that would allow them to unilaterally cancel my service with only 24 hours notice. (Admittedly, the NYT article references prior communication between Parlor and Amazon, so this is obviously not totally out-of-the-blue. Still, 24 hours from the final decision to account-deactivation seems insane for an enterprise offering....
(To be clear, I am not trying to argue for or against Amazon's decision to drop Parlor's account. I am just shocked that their AWS contract allows such a limited migration window for something like a TOS violation.)
- akvadrako 6y agoOwning your own hardware if not enough; if it was that easy it wouldn't be a big deal. You still need access to the network, meaning people who are willing to peer with you. It'll be a big problem if a major player says "we'll cut you off if you peer with Parlor, directly or indirectly". You also need access to domain names, payment providers, maybe advertising networks.
- jlkuester7 6y agoTrue! This is why I am interested in de-centralized solutions to some of these challenges. - Social networks: https://joinmastodon.org/ https://joinmastodon.org/ - DNS: https://handshake.org/ https://handshake.org/ - Payment Providers: https://en.wikipedia.org/wiki/Cryptocurrency https://en.wikipedia.org/wiki/Cryptocurrency (To be clear, it is not that I want to ensure that idiots always have a platform to spew their idiocy, but I would prefer that the decision on who is an idiot be distributed and not centralized.)
- nfRfqX5n 6y agoseems like they have been in talks for a little while at least, where they could've possible remediated the issues to buy some more time to avoid the TOS violation. curious to see if they had any effort to clean it up or immediately started working to migrate
- jlkuester7 6y agoYeah, based on what was said in the NYT article they did act to remove some of the posts in question. Ultimately it seems like Parlor's own moderation system was insufficient in dealing with the troublesome posts (either because the moderators declined to adequately enforce Parlor policies or because of a lack of moderators to deal with the sheer volume of content). Either way, it seems like Parlor has some (albeit lax) content policies and it has a moderation system in place (this was how it was approved for the Apple store, etc in the first place). But when these failed to deal with content that Amazon did not like within the timeline that Amazon set, Amazon chose to cancel the contract. I assume these actions are all within the bounds of Parlor's AWS contract (or at least Amazon's lawyers believe they can successfully argue such). However, I remain surprised that this is how enterprise AWS contracts would be structured (though maybe this is just because I don't have any personal experience with them).