Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
parhamn
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
22 ms
·
301.
▲
by
parhamn
5y ago
Confession. This vaguely reminded me of something (now admittedly, terrible) I did way back in highschool to give myself a bit more time on an assignment: submit a corrupted docx by putting in some random bytes in the file. This would give
302.
▲
by
parhamn
5y ago
They're called airplanes and we’ve managed to outsource half the shipping and handling to the cargo itself.
303.
▲
by
parhamn
5y ago
I find this comment surprising. I didn't say anything too controversial. Investors regularly refuse (and it's often the morally correct thing to do) investments in competitors. It's also often in their financial incentive too
304.
▲
by
parhamn
5y ago
Sure, I also, still, feel some of the kinetics need to be explained to avoid confusion. Like the homepage doesnt make much sense to me after all these years of reading HN. As I write this there are 160 upvotes in one hour on this post and i
305.
▲
by
parhamn
5y ago
I think if you get over the drama here and implied "dirtiness", the main result is still a bit interesting. Even if you exclude the implied carteling and take everything he said with a grain of salt, the outcomes of this incentive
306.
▲
by
parhamn
5y ago
Be careful with this sort of reasoning for the sake of false negatives. Under this lens only 0.001% of all actual incidences of what he's describing would be deemed as enough proof (your bar effectively requires testimony under oath an
307.
▲
by
parhamn
5y ago
As a 27mb video: https://imgur.com/a/dqc2Gii I think it's super noisy so it doesn't compress well.
308.
▲
by
parhamn
5y ago
It's not really different. If you were given the sums in the first operation to 5-6 digits and it was acceptable error margin, the comparison is within that acceptable error margin too, it's just this time the error led to a wrong
309.
▲
by
parhamn
5y ago
Well, for one, nonsense like this is why I set out to build a browser in the first place :) More generally, the user experience is, naturally, worlds apart from extensions.
310.
▲
by
parhamn
5y ago
Awesome. To keep focus on the main topic, feel free to email me to chat more (FWIW I've done the 50 extension patchwork thing and generally find the extension experience to fractured and suboptimal for me).
311.
▲
by
parhamn
5y ago
It's not just the user-agent, it is definitely doing non-trivial fingerprinting (both linked projects also had UA mitigations before). We don't have an easy workaround (besides a sketchy cookie hack that took hours to reverse engi
312.
▲
by
parhamn
5y ago
They also do this thing now where they block [1] smaller browsers (even ones using the latest version of chromium) under the guise of security. According to their docs they're fighting MITMs by generally disallowing any browser they ca
313.
▲
by
parhamn
5y ago
> I'll never be comfortable with any tool for that automatically generates schema changes, as I'm just never sure at what point it decides to delete parts of my prod db. To be fair, they analogized themselves to Terraform, whic
314.
▲
by
parhamn
5y ago
Not per the linked wiki [1]. But it is still being designed it seems. [1] https://github.com/curl/curl/wiki/JSON#--jp-part
315.
▲
by
parhamn
5y ago
After this change the equivalent would be: curl -XPOST --jp username=YOUR.USERNAME password=YOUR.PASSWORD https://api.ctl.io/v2/authentication/login Which isn't too far from your desired outcome (notab
316.
▲
by
parhamn
5y ago
It's very interesting seeing exploits like this from this organization. On one hand their service fingerprints users and offers extended metadata like whether the user is in incognito via sketchy web apis. On the other hand they report
317.
▲
by
parhamn
5y ago
It's cool to see how these posts are becoming less and less important in the wake of today's frameworks/tools protecting devs by default. From ORMs escaping SQL, to FE frameworks escaping html/js, to browsers starting to
318.
▲
by
parhamn
5y ago
At the risk of being downvoted and called out for my potential biases, I don't really buy the 'last bastion of hope' argument re: FF. I spend most my day trying to create a competitive alternative to chrome. Heres my main rea
319.
▲
by
parhamn
5y ago
Doesnt Aurora effectively have a PG bouncer in front?
320.
▲
by
parhamn
5y ago
> The best part was the emails I'd get. Wow. I've never been more curious. What sorts?
321.
▲
by
parhamn
5y ago
This is a great point. I think last I measured a basic electron browser/webview is about ~20-36mb which is not far from an about:blank chrome tab. The 300mb+ memory usage you see is mostly the javascript from the apps themselves. Eithe
322.
▲
by
parhamn
5y ago
I recently saw a go generic data structures library in the wild: https://github.com/zyedidia/generic . Anyone have any more? I'm curious what people come up with for goroutines/channels.
323.
▲
by
parhamn
5y ago
Glad you like it! Email me directly at parham@synth.app so I make sure I don't miss it :)
324.
▲
by
parhamn
5y ago
I took some notes through the first part here: https://synth.app/s/JeJYQgwxey8 if you're on the go and can't watch.
325.
▲
by
parhamn
5y ago
It's a federal crime. The US Government took him to court. As for who tipped them off, thats a different question.
326.
▲
by
parhamn
5y ago
Some of the vendors mentioned: - VistaraIT, LLC - Platfora, Inc. - Sumo Logic, Inc - ElasticBox, Inc - Numerify, Inc. - Docurated, Inc - Maginatics, Inc. Is what they did illegal too? Presumably with zero chance of being charged. I haven&#x
327.
▲
by
parhamn
5y ago
I appreciate the links!
328.
▲
by
parhamn
5y ago
Are there any editors that let you edit code in a structure closer to the way the language handles it (e.g. graphs/stacks)? You could achieve a lot more than the conventional file-per-module approach. Off the top of my head benefits co
329.
▲
by
parhamn
5y ago
When I saw this a few weeks ago it set off a lot of thinking about compilers and ML. Even the unoptimized C version is 10x slower than the one that is painstakingly optimized. Some of this performance gain will certain be achievable by ML&#
330.
▲
by
parhamn
5y ago
> LLVM IR that Clang would produce You most certainly can’t unless your code is unsafe and in that case the comparison isn’t so interesting anymore.
More ›