orf

5 ms·

61.
by orf 4mo ago
Was Rosa Parks unethical for sitting down on a bus? The point is that the context matters: both the users context and the context of the restriction. It’s not as clear cut as “ignoring restrictions = bad”. The restriction itself can be unet
62.
by orf 4mo ago
Are you making an honest attempt at stating your opinion that there is absolutely no reason to prefer a MacBook over your chosen device? Or are you just being glib because you’ve got a misplaced superiority complex?
63.
by orf 4mo ago
> how 99% of the population are not going to afford a setup like that > could or will? much fewer than 1% are going to spend their money on that. It’s ok to change your point, you don’t need to get combative. Not that it makes any
64.
by orf 4mo ago
That’s not at all what you said though, was it?
65.
by orf 4mo ago
99% of the population don’t code using models, local or remote. So that’s a useless metric. What % of developers could afford an older MacBook model, second hand? Far, far more than 1%.
66.
by orf 4mo ago
I wrote something similar with go, but MacOS only. Creating a worktree became instant, but the bottleneck shifted from that to git needing to build its index. Claude code runs `git status` in the background, meaning any speed gains are inst
67.
by orf 4mo ago
Why? Calling a reasonable thing grim without any follow-up isn’t the hallmark of a good comment either.
68.
by orf 4mo ago
> postgresql also does not synchronously replicate By default. Generally your primary database is in a completely different failure category than a kubernetes node running an ephemeral workflow pod.
69.
by orf 4mo ago
Yes, but directly using s3 as a key-value database is completely different from using SQLite + litestream.
70.
by orf 4mo ago
> The caveat is that Litestream replication is asynchronous. A restore can miss the newest local writes if the SQLite volume disappears before they are copied. That is fine for many AI and experimentation workflows In short: SQLite is n
71.
by orf 4mo ago
> Google followed in 2010 with Cloud to Device Messaging, then Google Cloud Messaging in 2012, then Firebase Cloud Messaging in 2016 Classic
72.
by orf 5mo ago
The essence of a program is all of those things. The omission of any one of them makes a program almost worthless.
73.
by orf 5mo ago
The release post for v2.8 is not yet published. Check GitHub releases page for the latest release status of Deno.
74.
by orf 5mo ago
Misc thoughts: Perhaps sharing your take home exercise might be a more useful avenue for feedback? I’m not sure how large the market is for ASP.NET developers, but the skills you’ve learned so far are more transferable than you think. Try c
75.
by orf 5mo ago
The entire executable isn’t loaded into RAM. And are laptops a primary deployment target?
76.
by orf 5mo ago
You’re right, I misstated. It’s not 10 million per exploitation , it instead limits the pool of people who can exploit you to those willing and have the ability to spend 10 million+ on an exploit. That is still quite a small pool, and ther
77.
by orf 5mo ago
> We have seen multiple software hacks resulting in >10 million dollar payouts This sets a nice price bar for exploitation. Is someone willing to pay 10+ million dollars to get access to your phone? The obvious caveat here is that for
78.
by orf 5mo ago
So you’re saying that if you append 1GB of junk data to the end of an executable it would be slower to execute? No? So it’s not purely a function of disk size.
79.
by orf 5mo ago
This makes sense if you’re a human-rights journalist working in a dangerous country, with the threat of state-level actors looking to compromise you. If you’re not then this seems quite paranoid, bordering on LARPing.
80.
by orf 5mo ago
SSDs can read at gigabytes per second, and only the used portions are paged into memory. Anything else?
81.
by orf 5mo ago
Why does the on-disk size matter for any of that?
82.
by orf 5mo ago
> You don't want that as an automatic update because it will break in production for anyone who is actually using it The problem with this take is that it’s stuck in the early 2000’s, where all servers are pets to be cared for and l
83.
by orf 5mo ago
Please no, that’s a terrible outcome.
84.
by orf 5mo ago
In what world is generating a UUID more expensive than checking for duplicates? at any scale ? Walk me through that please
85.
by orf 5mo ago
AKA centralising a decentralised identifier generator?
86.
by orf 5mo ago
Any company that does this doesn’t seem worth working for
87.
by orf 5mo ago
Please, share them.
88.
by orf 5mo ago
There are always trade-offs and there is never one best way to do something. Stack-based coroutines is one way to do it. A relevant trade-off here is overhead, requiring a runtime and narrowing the potential use-cases this can serve (i.e
89.
by orf 5mo ago
*some AI crawlers. Not many
90.
by orf 5mo ago
Not every change a developer makes increases revenue, and the changes that do often have a lag time.

More ›