kevinsimper

A lot of articles mentions FTX development team consisting of 10-25 developers. In the Bloomberg documentary it says they all lived together. In court Gary Wang has also been showcased as one of the primary developers.

How did the FTX platform not fail technically? How were they able to build a trading platform that people were happy with and could support all the transactions?

In the court presentation screenshot of code from GitHub is shown. How has the source code not been leaked with such a small team?

The incentive to infiltrate a trading platform is enormous. Without dedicated security team and air tight deployment, I would have estimated that a technical failure would have been much higher than fraud.

There are examples like WhatsApp and Instagram where a small team made some big, but those were not integrating with money, which leaves a lot more room for error. A image or message not send or save, no big deal. A trade on a trading platform lost? Trust immediately lost and never recovered.

Terminals are great, text input -> text output.

Terminals are not so visual, and any advance input can be complicated to remember or even type (see input to ffmpeg).

However I feel websites that are used heavily like for customer support teams, that continuously does the same actions, but a wide range of them, could benefit of being able to control more of a website with keyboards.

I think the best example of a keyboard focus website is linear.app, but even that has two types of search (shift + / and CMD + K) And it shows keyboard shortcuts on mouse over.

However, clicking the right and left arrow does not move between the main area and the sidebar. Only clicking up and down arrow changes between the active row currently selected.

I know of Vimium for Chrome, which allows you to press F and then the letter for clicking a link.

However, what happend to those terminals that were focus heavily on the arrow keys to quickly navigate? Those old terminals? Like the current "raspi-config" we have.

Obviously it is pretty hard to navigate with a keyboard on mobile, but "mobile first" does dictate a design that is very narrow and would fit a keyboard navigation. It is funny to see how all cloud providers has a Cloud Shell.

Have we decided that navigating with mouse is the best interface?

I am using Google Cloud and others and I am amazed by how many graphs there are everywhere, showcasing you all kind of metrics. And the crazy part is that you can define any time periode and it is equally fast.

Does somebody know how they designed that?

It is not bigquery, because bigquery is fast but multiple seconds to respond to queries.

It could be bigtable, but it just seems expensive and keeping this many timeseries up to date all the time and few of them are looked at that often. Bigtable starts at $500 usd for a single node and you need multiple.

I understand that there is a lot of money in Cloud. I also tried to search and could not find any resources on how this is done on large scale.

How can I do something like that in my own applications?

I have looked at Postgres TimescaleDB and other timeseries databases, Prometheus, but you quickly end up with a lot of timeseries that takes a lot of memory to compute (Prometheus running out of memory).

They have made an article here talking about OpenTSDB https://cloud.google.com/architecture/monitoring-time-series-data-opentsdb