Hacker News new | ask | show | jobs
by pestaa 5245 days ago
All your questions are fair, but I guess this site is for users who already know the difference between stripe.js and the API.

Also, your worries are rightful but pointless. The API provided 4 nines and stripe.js is close to 5 nines in the last three months. There were no 10+ minutes downtimes in either (still not counting the website.) I think that is impressive uptime.

2 comments

No. The site is a selling document, to potential developer-users. It sells the uptime and the honesty/integrity/trustworthiness of stripe.

It should also be selling the simplicity, transparent communication and ease-of-use of stripe.

I understand the difference between the two. The point I had is that stripe.js depends on the api. I don't understand why there is a separation in the status output. If the api is down, stripe.js is also down, you can see the reference to the api in the .js file.

The way that my business works is that we take a large number of orders in a short amount of time and then it tapers off. If my site was down because their api is down for 4 minutes (and I don't know when it is going to come back up), that would really affect my sales.

If you are at a cash register and pull out your credit card and the swiper doesn't work, chances are you can just hand the person cash. The internet doesn't work that way. 4 9's for a payment processor really isn't good enough. Also, that stat is just across 90 days, not 365.

I really like Stripe and what they are trying to do. Now that I can visually see it, I just worry about their uptime performance.

Update: Some one just sent me email calling me an asshole, which seems rather undeserving. This is their ip address pool-173-49-2-60.phlapa.fios.verizon.net.

> 4 9's for a payment processor really isn't good enough

If 99.99% isn't good enough for you, then you should just store the payment data yourself so that you can re-send it if your gateway is unavailable during those critical 4 minutes.

Payment processors of all sizes experience downtime. Authorizenet was down for almost a day last year despite having redundancy in every layer of their operations -- up to having an entire hot standby data center to fail over to. None of them lost all their customers and went out of business, so 4 9's IS good enough.

They haven't lost customers because there is nowhere else to go.