ec109685

20 ms·

301.
by ec109685 2y ago
php with lambdas eliminates a bunch of the syntax and can let map/reduce style code be colocated while still achieving the functional style benefits. A big issue though (at least for some apps), is there is no attempt to make that code
302.
by ec109685 2y ago
IO has a very different cost than wasting CPU cycles, so shouldn’t necessarily be lumped together when analyzing tradeoffs. Your point still stands that fundamentally changing the design of code should limited to the hottest parts. Probably
303.
by ec109685 2y ago
php has the opposite problem where if a script hangs on io, you can be cooked (assuming other requests blocked similarly). Different trade offs for sure, but given IO is often the bottleneck, having it force you to think about running code
304.
by ec109685 2y ago
The attack requires getting yourself CC’d on a support ticket. In this case to show how bad that is, it was a support ticket that had an oauth ticket to log into slack as “support@company.com”.
305.
by ec109685 2y ago
The researcher showed how they could hop onto any Zendesk support ticket thread with zero authentication, so that should have been enough given Zendesk was exposing customer data via that attack path. Clearly Zendesk needs to change things
306.
by ec109685 2y ago
Not only is Apple non-directory, they are non-discretionary as well, so foisted on services and handled poorly as a result.
307.
by ec109685 2y ago
It’s from 2016, so probably lost its mojo.
308.
by ec109685 2y ago
While obviously Zendesk leaving such a huge hole was the main reason for this exploit (you should obviously fail closed when email signals suggest it’s an unauthenticated address), a contributing factor is that Apple forced themselves to be
309.
by ec109685 2y ago
I don’t agree. Bug bounties are taken seriously by at least some companies. Where I have worked, we received very useful reports, some very severe, via HackerOne. The company even ran special sessions where engineers and hackers were brough
310.
by ec109685 2y ago
They can use “we’re working on the chassis” as an excuse for why it’s not ready yet.
311.
by ec109685 2y ago
They are Jedi’s at jq though.
312.
by ec109685 2y ago
How does this compare with Cloudflare’s new offering SQLite it durable objects: https://blog.cloudflare.com/sqlite-in-durable-objects
313.
by ec109685 2y ago
At least we know a human wrote it :)
314.
Christian Selig's unofficial YouTube app for the Vision Pro just got taken down(theverge.com)
2 points by ec109685 2y ago|1 comments
315.
by ec109685 2y ago
Yes, in that scenario of 500ms of 200% CPU for a request / response type workload, 50% of responses will have an extra 25ms response time tacked on as the system is sleeping during the remaining portion of each scheduling period. This
316.
by ec109685 2y ago
If you don’t let people burst, you lose a benefit of multi-tenancy. Each workload stays conservative ensuring they never throttle, and your nodes end up very underutilized since you can’t share that buffer amongst workloads. With auto scali
317.
by ec109685 2y ago
In an ideal world, it’s far better to not use Limits but instead have applications set their CPU requests. That way, if the system has CPU available, applications can use more than their requested CPU (and won’t get throttled), but if CPU
318.
by ec109685 2y ago
It’d be helpful to update the blog post to make that more clear. The example focused on “constructing the response” in parallel with making the write durable, so confusing it with a response gate is an easy mistake to make.
319.
by ec109685 2y ago
The parent’s point was that you don’t have to care about white space when composing code in other languages since the LSP can auto format. So you could theoretically never press return or space more than once, and always have perfectly corr
320.
by ec109685 2y ago
Given how slow AI inference is (and for training it doesn't matter at all), the advantage of it being a few milliseconds closer to the user is greatly diminished. The latency to egress to a regional data center is inconsequential. Good
321.
by ec109685 2y ago
Would having a VM inbetween help in that case? It seems like protecting against malicious GPU workloads requires the GPU to off virtualization to avoid this exploit. This is helpful in explaining why AWS hasn't been excited to ship thi
322.
by ec109685 2y ago
If the calls first pass through a memory safe language as what gvisor does, isn’t the attack surface greatly reduced? It does seem however that Firecracker + GPU support (or https://github.com/cloud-hypervisor/cloud-hyp
323.
by ec109685 2y ago
Their solution isn’t GA yet. For headless browsers, the latency benefits of “container anywhere” seems high. For things like AI inference, running on the edge seems way less beneficial than running on the cheapest location possible which wo
324.
by ec109685 2y ago
I am seeing questions about why this is a big deal? The issue is that CNN is using their domain authority to boost this drivel affiliate Money content, and hiding the fact that they are doing it by using layer 7 routing to cloak the site. I
325.
by ec109685 2y ago
Ironic given the app’s purpose.
326.
by ec109685 2y ago
Given caffeine has a half life in the body, couldn’t you have less coffee but drink it for longer in the day and have the same result?
327.
by ec109685 2y ago
Allergies are a thing. Why is it hard to believe that certain types of food affect people in specific ways?
328.
by ec109685 2y ago
It’s quite the saga, including the designer putting in their portfolio that they actually completed the project as they were ghosting their client: https://x.com/Shpigford/status/1807802947394588842
329.
by ec109685 2y ago
Microsoft has made git work for massive repositories through a lot of blood sweat and tears: https://github.blog/open-source/git/the-story-of-scalar/
330.
by ec109685 2y ago
Microsoft has all of windows in a single repository.

More ›