cle

33 ms·

511.
by cle 5y ago
Every significant Java codebase I've worked in has had incredible amounts of magic. Between AOP, DI frameworks, APT, reflection, proxies, SPI, Lombok compiler hacks, dynamic classloaders, etc., there is a LOT of magic flying around and
512.
by cle 5y ago
If you take a black-and-white interpretation like that, then democracy is not accessible to everyone either. Not even the citizens living under a democracy, who for many reasons may be unable to vote, including financial reasons. If a servi
513.
by cle 5y ago
If you don't agree on definitions, then there is no discussion.
514.
by cle 5y ago
Oxford Dictionary has a second definition of "democratize" as "make (something) accessible to everyone", which is the sense in which this word is being used.
515.
by cle 6y ago
Your tests would need to exercise all your transitive dependencies' code paths that are used in production. I very rarely see tests written that way.
516.
by cle 6y ago
This is the most important point, IMO. Amazon's value add is not the software itself, it's the operation of the software. That includes a LOT of stuff, not just making sure it's running. It's security modeling and patchi
517.
by cle 6y ago
Graal has incomplete support for Java. Last time I tried a few months ago, ObjectOutputStream was not supported (I think it is now?), which was a critical part of a library I was using, forcing me to abandon Graal. Dealing with reflection i
518.
by cle 6y ago
They do but the Java ecosystem takes a hit here, with many libraries being slow to initialize. The slower it takes to startup, the longer it takes to absorb the new load, to the point where you either keep some headroom (IOW, waste $$$), or
519.
by cle 6y ago
And yet here we are, discussing a paper by principled ones who are refusing and building systems to reclaim control of that power--and they are just a small sliver of the people working on this. The circumstances surrounding this discussion
520.
by cle 6y ago
I think it is, which is why it hasn't happened. We shouldn't couple ourselves to the GUI itself (widgets, layouts, etc.). When you decouple the data and core operations from the presentation, you get APIs. Slap a flag parser aroun
521.
by cle 6y ago
I completely agree, and I do the same thing too. If there's something that people are scared to touch, I immediately want to change something fundamental in it so we can get un-scared of it. I'd rather have a controlled demolition
522.
by cle 6y ago
This fear is well-founded. Changing a system is a lot riskier than leaving it alone and building things around the periphery. Changing it requires a deep understanding of the implications of the changes, the impact on consumers/clients
523.
by cle 6y ago
This is solid advice. VCs are just as susceptible to the same biases as everyone else. Learning to navigate that minefield is unfortunately part of the massive extra effort it takes to start a SV company as a woman. My minority wife has rai
524.
by cle 6y ago
I stuck that in /usr/bin/s, and now I present to you an HTTPS server in one character: s
525.
by cle 6y ago
What are you talking about? TypeScript has the best Language Server libraries, if you want maximum productivity when making a Language Server, you run with TypeScript. I don't even like TypeScript but I can at least acknowledge that it
526.
by cle 6y ago
Node and TypeScript have the best Language Server tooling, because that is what Microsoft uses for VS Code.
527.
by cle 6y ago
> The “best possible syntax” is a universe of possibilities - environment variables are, thankfully, limited to strings. The same can be said for command line arguments. I usually prefer those instead of environment variables, because th
528.
by cle 6y ago
Laziness? I can just set it up with a couple of clicks and pay almost nothing (it runs on a t2).
529.
by cle 6y ago
This is what I do. I host a controller in AWS on an EC2 instance in my account. It works great.
530.
by cle 6y ago
The memory layout of a pointer is quite different from a full-blown object. E.g. Optionals are only efficient in Java because the HotSpot compiler optimizes them at runtime. And it further obfuscates the memory layout--is it an object or no
531.
by cle 6y ago
Yes this is basically what I was talking about. It becomes pretty tricky to understand memory layout if you try to masquerade as a regular type (hence my apparently controversial reference to JIT compilers). I guess my point is that Go poin
532.
by cle 6y ago
The only apt comparison IMO is Rust, the other languages with JIT compiling runtimes aren't really useful comparisons. To me this looks like a discussion about syntax and ergonomics. Go provides the same mechanisms: * for safe derefere
533.
by cle 6y ago
It’s easy to criticize, but the reality is that making decision with and motivating large groups of people involves tradeoffs that lead to theoretically suboptimal outcomes.
534.
by cle 6y ago
Yep exactly. If you go into a meeting and you haven't talked to at least a few of the people there about the topic beforehand, IME it's not going to be a productive meeting. I usually start with getting feedback from one or two su
535.
by cle 6y ago
That's not the sense that I was using, and I don't think it's a relevant point either. What's important IMO is that the platform and the data are not controlled by elected officials.
536.
by cle 6y ago
I'm not. There's a natural scarcity of space for physical cameras and reporters. As far as I understand, the press themselves are largely responsible for figuring out these logistics via the WHCA. (Someone please correct me if I&#
537.
by cle 6y ago
To me, the sentiment seems to be that Rust is explicitly tailored for code that cares about no runtime overhead. If you don't have those requirements, then don't use Rust. This is why I reach for Go by default when writing CLI too
538.
by cle 6y ago
Now we're just quibbling. No it doesn't improve the speed at which the Elisp interpreter can churn through Elisp bytecode. We can however dynamically improve the speed at which some Elisp functions return, which would have the net
539.
by cle 6y ago
Instructing users who want high-performance code in Emacs to write their code in C and rebuild their Emacs binary, or figure out how to make it a dynamic module, is about the most user-hostile way I can imagine to improve Emacs performance.
540.
by cle 6y ago
While technically true, you can still improve performance of existing Elisp code by redefining certain commonly-used Elisp functions perform better in Deno. Sure the Elisp bytecode itself doesn't evaluate faster, but that distinction i

More ›