I used to work with Django 10 years ago, and I loved it.
I totally agree it looks totally outdated nowadays and I wouldn't pick it over the other alternatives. It's incredibly how bad it looks if you're going to need anything else than an API.
TLD "owners" own TLDs in much the same way that we own domains, and it's very possible that ICANN phases out the .io domain when the British Indian Ocean Territory ceases to exist. From what we gathered in the other thread it somewhat depends on what ISO decides to do with its codes.
At a minimum I expect that control over the .io domain will go to Mauritius and they'll be able to reassign it as desired (since they never contracted with the hedge fund). But the typical path for a code when its country goes defunct is to get phased out.
The more I think about it, the more I agree that this is the likely outcome.
IO has been in the ISO standard forever, so there's plenty of historical precedent (like UK). Furthermore, it continues to be descriptive of a specific part of the world (like SU). The easy move here is for the ISO committee to mark IO as exceptionally reserved, for ICANN to declare that this of course makes it a special historical case which sets no precedent, and for everything to continue mostly as usual.
This assumes, of course, that ICANN aren't looking to make some kind of example/statement about misuse of ccTLDs. If they are, things may be different.
TL;DR: ICANN policy forces deletion if CC disappears from the ISO list of countries, with one famous exception (.su); but Mauritius could cut a Tuvalu-style deal to maintain it.
50k or 100k would be far more appropriate given the severity of this issue. But overall, this makes me think there's probably a lot more vulnerabilities in Arc that are undiscovered/unpatched.
Also, there's the whole notion of every URL you visit being sent to Firebase -- were these logged? Awful for a browser.
Ya this is fair! Honestly this was our first bounty ever awarded and we could have been more thoughtful. We’re currently setting up a proper program and based on that rubric will adjust accordingly.
Hey! Author here. Do you mean whether the cycle changes the original test requirements for the _evaluation test_, or for the test our agent is running for the customer's application?
If you are asking about the requirements for the evaluation test, that's possible for sure. We'll need to refine and maintain our eval asserts as we go as this happens. For example, I'm currently working on revising how our test steps are generated (from batched generation to a more iterative process), which changes the structure of the output value we expect from the LLM. This will impact some of our deterministic (i.e., non-model-graded) prompt eval asserts.
If you are asking about requirements for the _customer test_ itself: the customer test goal is generated _once_ based on testable actions we detect in the application or, optionally, manual input. Tweaking the test generation prompt that generates this goal could/would change that goal, but after it is created the goal is essentially 'set in stone' for each test (unless manually modified). So in our agent's test result evaluation stage after a test is run, it will always evaluate the result against an immutable goal point.
_However_, tweaking the _test evaluation_ stage's prompt could definitely impact how the result is evaluated even if the goal stays the same! This is why we have prompt eval tests running against tagged data points for that stage as well - to make sure our prompt tweaks to this stage produce expected results. And if the requirements for that output change as we go, we just tag more up-to-date data points (or add evals on manually-defined inputs if desired) to make sure we're testing on the right thing.
I hope that answers your question - let me know if that doesn't quite cover it!
220m users. Let’s imagine 50m are streaming concurrently. That’s 100TB an hour in logs lol. They could be storing an entire petabyte of logs a day. My friend did some data center stuff for the large hadron collider and wasn’t hitting these data ingestion states, and these are just to record me binging the office.
Responses like this from C-levels in crypto make me glad I moved from the space after 4 years. Everyone is so egotistical. Blaming it on mental patterns of the employees without accepting it could be some really shitty management from your colleagues instead is toxic. Have a good hard look at the people named in this petition.
This is my issue too!