Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
andrewchambers
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
9 ms
·
91.
▲
by
andrewchambers
4y ago
This sounds like a terrible way to exist.
92.
▲
by
andrewchambers
4y ago
You can already do this by using rqlite to redirect writes.
93.
▲
by
andrewchambers
4y ago
This is not really accurate, it stores keys as securely as it is able to. Also the heap is not inherently insecure - it depends on other factors such as network isolation, user permissions, containers and other kernel protections. Many thin
94.
▲
Memory Safeish Hare
(acha.ninja)
2 points
by
andrewchambers
4y ago
|
0 comments
95.
▲
by
andrewchambers
4y ago
You are unfairly comparing ancient C with modern rust - Try comparing modern C with modern rust to avoid making a straw man argument. Like I said, I don't even disagree, I still think you are just doing everyone a disservice with this
96.
▲
by
andrewchambers
4y ago
Or he could do the normal thing and run valgrind to instantly see exactly where the problem is. I don't even disagree with the premise, but these are straw man arguments.
97.
▲
by
andrewchambers
4y ago
This is not a good anecdote - You hit a 10 year old bug because your sysadmin was too lazy to update gcc and then formed a bad impression of the software? Guess what, if you install a 10 year old version rust none of your code would not com
98.
▲
by
andrewchambers
5y ago
A simple solution might be to just look for name aliases that match a type and report that to the user. I don't see why you can't name the shape of a structural type without having to use nominal types.
99.
▲
by
andrewchambers
5y ago
SSH certificates can sign a key with an expiration, thats a bit different to a password.
100.
▲
by
andrewchambers
5y ago
More great things: - https://c9x.me/compile/ - https://github.com/vnmakarov/mir
101.
▲
by
andrewchambers
5y ago
I updated the readme to better explain what I meant.
102.
▲
by
andrewchambers
5y ago
It can assembly all the assembly generated by cproc - which can compile the tools it depends on including cproc itself. I will update the readme.
103.
▲
by
andrewchambers
5y ago
Author here, that's exactly what I meant. It can also assemble the cproc C compiler when it compiles itself - thus creating a loop. I updated the readme.
104.
▲
by
andrewchambers
5y ago
First there is perhaps some slight refactoring that could be done. Though it does sound nice to reduce the dependency tree. One thing is I like the declarative nature of the peg, so I considered trying to make a tiny peg parser generator to
105.
▲
by
andrewchambers
5y ago
> Examples, image editors need to load new formats Alternatively, they could be designed to support plugins.
106.
▲
by
andrewchambers
5y ago
The benefit is mainly that it is interesting and easy to learn/experiment because it has a very small and tidy implementation.. Just an example, building cproc from source takes a few seconds - building gcc from source takes like 15 mi
107.
▲
by
andrewchambers
5y ago
I honestly don't mind going without C const (compared to C++ where it is more useful).
108.
▲
by
andrewchambers
5y ago
There is one, see pp.c.
109.
▲
by
andrewchambers
5y ago
Really once the secure channel between two processes running on each host is setup, there is no difference between the two methods. A pull backup can be tricked just as easily by presenting a false filesystem or backdooring the 'pull p
110.
▲
by
andrewchambers
5y ago
Full heap scans still happen, just not as often, then you get the alternative 'cringe' of having huge amounts of old data sitting idle for nondeterministic periods of time.
111.
▲
by
andrewchambers
5y ago
Well, a (more) compact bytecode representation is one of the design decisions made for web assembly... That is one reason why it is a stack machine.
112.
▲
by
andrewchambers
5y ago
I started from beginner to about 1800 on lichess in about a year of casual play and watching some grandmasters on youtube. I am 30 years old, i think younger kids can do better and we know many prodigies do. I guess this is different to off
113.
▲
by
andrewchambers
5y ago
Humans can gain hundreds or thousands of points per year at the start, but then tend to plateau afaik, it's actually easy to check this out on peoples public lichess accounts. Here is the world champion: https://lichess.org&
114.
▲
by
andrewchambers
5y ago
For me it originally worked flawlessly. Then I tried to use it again and lots of the menus the support pages referenced were no longer present. It was an endless cycle of broken links and features being hidden deeper and deeper in menus. If
115.
▲
by
andrewchambers
5y ago
Yes, but google could easily support per application passwords. They just don't want to... in fact, they used to and removed it. They could also invest in other ways to make existing workflows work securely instead of imposing their ow
116.
▲
by
andrewchambers
5y ago
This is google imposed FUD to exercise control over users. They didn't need to design like that, but they want to make it hard to use things that are not google controlled.
117.
▲
by
andrewchambers
5y ago
Not everything that isn't perfect needs to be 'fixed'.
118.
▲
by
andrewchambers
5y ago
They might if they were seeking approval to show as later evidence.
119.
▲
by
andrewchambers
5y ago
People who did not need the iron lung still had other problems, including things like weakened lungs. Some conditions reappeared later in life afaik.
120.
▲
by
andrewchambers
5y ago
Yes, but more often that means one of your users protects themselves by fixing the vulnerability and sending the fix to you.
More ›