It's apparent SE doesn't care that non-JP servers are encountering 2002 issues

[Yahallo]

sry to burst your bubble, my current work is it related but i also worked as a coder for almost 8 years (changed careers after i felt that i wasted 3 years mastering ruby for it to fall flat) and there is no excuse on this word for when a dude with a open source packet sniffer is doing a better job than your entire networking team

It also sounds like this bug only occurs at rather extreme conditions since if it has been around since 1.0, nothing really triggered it until now. If a bug doesn't appear, it can be pretty difficult to find it.

[VelKallor]

It also sounds like this bug only occurs at rather extreme conditions since if it has been around since 1.0, nothing really triggered it until now. If a bug doesn't appear, it can be pretty difficult to find it.

If we ever meet, remind me to tell you the story about a bug that brought down a multimillion dollar company network for three days.

What was it? A zero instead of a 1 in a line of code that hadnt been used in AGES. To quote him "The more complex a system is, the easier it is for one tiny error to wreak havoc"

[Yahallo]

If we ever meet, remind me to tell you the story abut a bug that brought down a multimillion dollar company network for three days.

What was it? A zero instead of a 1 in a line of code that hadnt been used in AGES. To quote him "The more complex a system is, the easier it is for one tiny error to wreak havoc"

Honestly, legacy code is the worst to deal with. Often times it is not even documented, and the people who wrote it are long gone.

[Saefinn]

If we ever meet, remind me to tell you the story about a bug that brought down a multimillion dollar company network for three days.

What was it? A zero instead of a 1 in a line of code that hadnt been used in AGES. To quote him "The more complex a system is, the easier it is for one tiny error to wreak havoc"

Speaking as a developer...this feels so relatable. But this is the right of it.

But to add my own story to yours, which isn't as bad, but you reminded me of one of my own bugs.

One of our customers had an issue in their payment gateway. I found the fix required a lot of work and parts of it to be re-written. Bear in mind, in the development process there's a lot of different checks along the way to ensure the solution isn't flawed, the code isn't flawed and that it passes all testing, and testing can be thorough because people will try to break it.

Within a few days of going live it turned out that under very specific circumstances if somebody went to pay towards their product it would give them a full refund. It was in these circumstances that it was triggering a refund flag. It is just the nature of development.

And given the nature of the 2002 error, I can see why they'd suspect connection issues/packet loss and when the evidence showed contrary, they investigated it.

The level of detail and transparency is not something I am used to seeing companies give because so often they'll save face and provide less detail so it cannot be scrutinized or used against them. But it's an approach I appreciate, because it cuts the BS and means people can give more meaningful feedback.