秋田预测市场 · Discord 审计

Bumbling Bayesian参与话题 33消息总数 209

该用户的聊天记录定位。

Reset

聊天记录

共 209 条,显示第 151-200 条
Bumbling Bayesian 2026-06-13 04:02:47 Kalshi

Yeah, I was torn between being helpful and being adversarial.

Yeah, I was torn between being helpful and being adversarial.

Bumbling Bayesian 2026-06-13 03:29:22 Kalshi

You have a queue of infinite capacity?

You have a queue of infinite capacity?

Bumbling Bayesian 2026-06-13 03:21:10 Kalshi

You still have to have the same amount of compute to process the messages.

You still have to have the same amount of compute to process the messages.

Bumbling Bayesian 2026-06-13 00:34:58 Kalshi

I'm at ~27k and no errors.

I'm at ~27k and no errors.

Bumbling Bayesian 2026-06-12 22:48:19 Kalshi

That is, by not very strict, I mean that they could easily add some more constraints to the property definitions that would match what the documentation says.

That is, by not very strict, I mean that they could easily add some more constraints to the property definitions that would match what the documentation says.

Bumbling Bayesian 2026-06-12 22:47:43 Kalshi

Their schema isn't very strict. But it would catch at least some problems if they validated requests and responses against it.

Their schema isn't very strict. But it would catch at least some problems if they validated requests and responses against it.

Bumbling Bayesian 2026-06-12 22:23:24 Kalshi

That makes sense and is a smaller ask.

That makes sense and is a smaller ask.

Bumbling Bayesian 2026-06-12 22:18:09 Kalshi

They need a way for `add_markets` to accept a market filter predicate rather than simply explicit tickers.

They need a way for `add_markets` to accept a market filter predicate rather than simply explicit tickers.

Bumbling Bayesian 2026-06-12 07:24:12 Kalshi

Nice!

Nice!

Bumbling Bayesian 2026-06-12 03:59:28 Kalshi

I took a semester off in 1993. Never went back.

I took a semester off in 1993. Never went back.

Bumbling Bayesian 2026-06-12 00:54:46 Kalshi API 新增 200,000 resting 订单上限引发讨论

I figured that someone had already figured this out as a way to DoS the exchange given the number of problems we have seen since.

I figured that someone had already figured this out as a way to DoS the exchange given the number of problems we have seen since.

Bumbling Bayesian 2026-06-12 00:53:24 Kalshi API 新增 200,000 resting 订单上限引发讨论

On July 16, 2025, Noah reached out to me. I had a bug that was creating dumb orders. Noah said I had the majority of resting orders on the exchange and it was causing problems.

On July 16, 2025, Noah reached out to me. I had a bug that was creating dumb orders. Noah said I had the majority of resting orders on the exchange and it was causing problems.

Bumbling Bayesian 2026-06-12 00:48:58 Kalshi API 新增 200,000 resting 订单上限引发讨论

Looks like the 200,000 order limit was added no later than 2025-12-02, which is when I first committed a version of openapi.yaml which had that note.

Looks like the 200,000 order limit was added no later than 2025-12-02, which is when I first committed a version of openapi.yaml which had that note.

Bumbling Bayesian 2026-06-12 00:27:21 Kalshi

I do not remember seeing a limit before.

I do not remember seeing a limit before.

Bumbling Bayesian 2026-06-12 00:27:04 Kalshi

No, v2 is v2. There is now a v2 of v2.

No, v2 is v2. There is now a v2 of v2.

Bumbling Bayesian 2026-06-12 00:26:33 Kalshi

There was also that time period when v2 provided an endpoint for yet another version number.

There was also that time period when v2 provided an endpoint for yet another version number.

Bumbling Bayesian 2026-06-10 01:18:47 Kalshi

And Kalshi has a history of remaining silent on questions explicitly asked. It is entirely reasonable that engineers and support have been instructed to not be honest or candid on some topics.

And Kalshi has a history of remaining silent on questions explicitly asked. It is entirely reasonable that engineers and support have been instructed to not be honest or candid on some topics.

Bumbling Bayesian 2026-06-10 01:13:06 Kalshi

Some days I feel a little mischievous and consider encouraging people to do dumb things.

Some days I feel a little mischievous and consider encouraging people to do dumb things.

Bumbling Bayesian 2026-06-09 22:25:56 Kalshi

Is Trump Jr. not helping these companies he advises on compliance?

Is Trump Jr. not helping these companies he advises on compliance?

Bumbling Bayesian 2026-06-09 22:09:36 Kalshi

That makes sense. Maybe not so much thinking about international users specifically as just needing an SSN for reporting.

That makes sense. Maybe not so much thinking about international users specifically as just needing an SSN for reporting.

Bumbling Bayesian 2026-06-09 21:37:47 Kalshi

You should bring this up directly with support. There is no guarantee they will see this conversation.

You should bring this up directly with support. There is no guarantee they will see this conversation.

Bumbling Bayesian 2026-06-09 21:32:35 Kalshi

Only via API, as far as I know.

Only via API, as far as I know.

Bumbling Bayesian 2026-06-09 21:32:16 Kalshi 国际用户在 Kalshi 无法获得利息收益的困惑

But in Kalshi’s defense, I would have no idea how to handle reporting requirements for international users.

But in Kalshi’s defense, I would have no idea how to handle reporting requirements for international users.

Bumbling Bayesian 2026-06-09 21:29:44 Kalshi

Just a guess. I have no idea. But don’t be surprised that Kalshi is behind on what is expected of fintech companies in general.

Just a guess. I have no idea. But don’t be surprised that Kalshi is behind on what is expected of fintech companies in general.

Bumbling Bayesian 2026-06-09 21:28:55 Kalshi

The web UI.

The web UI.

Bumbling Bayesian 2026-06-09 21:21:20 Kalshi 国际用户在 Kalshi 无法获得利息收益的困惑

Interesting. I am wondering if this is because of compliance/KYC requirements. If they cannot figure out who your interest should be reported to, maybe they don’t pay it.

Interesting. I am wondering if this is because of compliance/KYC requirements. If they cannot figure out who your interest should be reported to, maybe they don’t pay it.

Bumbling Bayesian 2026-06-09 21:19:54 Kalshi

It makes sense to me also that it should be deposited to subaccount 0 because maybe it is hard to tell which subaccount(s) provided the liquidity. But it shouldn’t show up on the history of other subaccounts.

It makes sense to me also that it should be deposited to subaccount 0 because maybe it is hard to tell which subaccount(s) provided the liquidity. But it shouldn’t show up on the history of other subaccounts.

Bumbling Bayesian 2026-06-09 20:40:24 Kalshi 部分市场 API 端点返回 404 错误的临时解决

Try https://api.elections.kalshi.com/trade-api/v2/historical/markets/KXGOVSHUT-26JAN31. Looks like it has been archived.

Try https://api.elections.kalshi.com/trade-api/v2/historical/markets/KXGOVSHUT-26JAN31. Looks like it has been archived.

Bumbling Bayesian 2026-06-09 20:38:36 Kalshi

Are liquidity incentives just going to subaccount 0? I see the same values reported under Credit for every subaccount.

Are liquidity incentives just going to subaccount 0? I see the same values reported under Credit for every subaccount.

Bumbling Bayesian 2026-06-09 18:43:11 Kalshi

I got one at 2026-06-09T08:48:27.586097Z.

I got one at 2026-06-09T08:48:27.586097Z.

Bumbling Bayesian 2026-06-06 21:59:08 Kalshi

U.S. fintech is so dumb. It feels like the banks I work with are deliberately dragging their heels on FedNow. So much duct tape and Rube Goldberg plumbing could go away.

U.S. fintech is so dumb. It feels like the banks I work with are deliberately dragging their heels on FedNow. So much duct tape and Rube Goldberg plumbing could go away.

Bumbling Bayesian 2026-06-06 21:16:43 Kalshi

I have code that works with my business bank to pay vendors via ACH and Nacha formatted files SFTP'd to my bank. Some businesses use Plaid for that. It is just table stakes for any modern business of scale. Kalshi will eventually support this if they want more institutional users.

I have code that works with my business bank to pay vendors via ACH and Nacha formatted files SFTP'd to my bank. Some businesses use Plaid for that. It is just table stakes for any modern business of scale. Kalshi will eventually support this if they want more institutional users.

Bumbling Bayesian 2026-06-05 21:07:46 Kalshi

I think it makes sense that from Kalshi's perspective they might think of CloudFront as a shared resource versus having their own load balancer.

I think it makes sense that from Kalshi's perspective they might think of CloudFront as a shared resource versus having their own load balancer.

Bumbling Bayesian 2026-06-05 21:05:26 Kalshi

I think this is just imprecise terminology for CloudFront (shared) versus the load balancer they are now using.

I think this is just imprecise terminology for CloudFront (shared) versus the load balancer they are now using.

Bumbling Bayesian 2026-06-05 21:01:38 Kalshi

I am not answering for James or in any official capacity, but will try to help: The API has been hosted on AWS CloudFront. That is the legacy/deprecated https://elections.api.kalshi.com/. The API now is also now being served up at https://external-api.kalshi.com/ and that is where we are directed to go now. So "dedicated external trade api hosts" is the new host, external-api.kalshi.com (and related), and "existing shared hosts" is the deprecated hostname(s), which might stop working at some point.

I am not answering for James or in any official capacity, but will try to help: The API has been hosted on AWS CloudFront. That is the legacy/deprecated https://elections.api.kalshi.com/. The API now is also now being served up at https://external-api.kalshi.com/ and that is where we are directed to go now. So "dedicated external trade api hosts" is the new host, external-api.kalshi.com (and related), and "existing shared hosts" is the deprecated hostname(s), which might stop working at some point.

Bumbling Bayesian 2026-06-04 00:23:46 Kalshi Kalshi计划移除批量订单端点引发开发者讨论

From the POV of my client SDK, I pass in a collection of orders and it decides how to batch them. I agree that everything should just be either one or a collection.

From the POV of my client SDK, I pass in a collection of orders and it decides how to batch them. I agree that everything should just be either one or a collection.

Bumbling Bayesian 2026-06-04 00:17:34 Kalshi Kalshi计划移除批量订单端点引发开发者讨论

When I first started with the API, it seemed promising to minimize latency and there was the decreased cost against the rate limit. I have since discovered it is not that simple, so wouldn't care if batch was deprecated.

When I first started with the API, it seemed promising to minimize latency and there was the decreased cost against the rate limit. I have since discovered it is not that simple, so wouldn't care if batch was deprecated.

Bumbling Bayesian 2026-06-03 02:46:03 Kalshi

My system noticed this morning that `category` disappeared from the event object. (https://discord.com/channels/871819895443189862/927686720990892032/1511333667803238470)

My system noticed this morning that `category` disappeared from the event object. (https://discord.com/channels/871819895443189862/927686720990892032/1511333667803238470)

Bumbling Bayesian 2026-06-02 23:23:02 Kalshi

If you have never had an auth request work before, you have likely not yet correctly implemented the signing algorithm.

If you have never had an auth request work before, you have likely not yet correctly implemented the signing algorithm.

Bumbling Bayesian 2026-06-02 22:58:00 Kalshi

Got another one of these at 2026-06-02T13:11:36.156316Z.

Got another one of these at 2026-06-02T13:11:36.156316Z.

Bumbling Bayesian 2026-06-02 22:55:21 Kalshi

"single market" or "single event"?

"single market" or "single event"?

Bumbling Bayesian 2026-06-02 22:51:46 Kalshi

It seems pretty clear to me after the last round of collateralization checks was implemented.

It seems pretty clear to me after the last round of collateralization checks was implemented.

Bumbling Bayesian 2026-06-02 22:50:13 Kalshi

I do. And I think it would be great to have a feature to check that my model of how that works matches Kalshi's.

I do. And I think it would be great to have a feature to check that my model of how that works matches Kalshi's.

Bumbling Bayesian 2026-06-02 22:49:01 Kalshi

It is not as simple as "available balance less all resting orders" when taking MECNET/DIRECNET into account, right?

It is not as simple as "available balance less all resting orders" when taking MECNET/DIRECNET into account, right?

Bumbling Bayesian 2026-06-02 22:48:02 Kalshi

Yes. There is the blunt feature. It sounds to me like Melon wants another feature that surfaces the result of the collateralization checks.

Yes. There is the blunt feature. It sounds to me like Melon wants another feature that surfaces the result of the collateralization checks.

Bumbling Bayesian 2026-06-02 22:42:56 Kalshi

Yes, the exchange keeps track of how much you can actually spend. It is reasonable to want a way to retrieve this.

Yes, the exchange keeps track of how much you can actually spend. It is reasonable to want a way to retrieve this.