Show Navigation
Notices by Verius (verius@community.highlandarrow.com), page 10
-
@maiyannah Yup. I suspect it doesn't happen for replies, though I'm not sure. But QVitter is weird for me anyway with things like stubbornly refusing to enter newlines (it just pretends to add vertical whitespace then immediately undoes it).
-
Come to think of it, a lot of codes of conduct could be summarized as "If you wouldn't do this when representing your workplace and if your boss would find out just don't do it.".
-
Btw, Qvitter is really unreliable for me lately, often silently not posting my notices with 500 errors.
-
https://www.w3.org/Consortium/cepc/ A thought on naming: code of professional conduct sounds much better than code of conduct. It implies both that the community is expected to consist of professionals (rather than idiot sociopaths who can't avoid harassing someone unless explicitly told not to) and that the rules are to behave professionally (rather than some arbitrary "don't hurt someone else's feelings" rule). It helps that the actual text of the code doesn't contain any red flags either.
-
@maiyannah Yeah, typical RPGs, even roguelikes have logic to it.
-
@maiyannah Honestly at this point I'm struggling to figure out what Go excels at that isn't done better by another language. Static compilation isn't exactly a new or uncommon feature. Green threads aren't either. Network programming has a range of options from .NET Core to C++ to Erlang/Elixir. Simplicity is better in Scheme. Design for idiots is better done in Java. Really the only plus point of Go for me at the moment is that it has good ecosystem support for container related networking due to everyone using it for that.
-
@maiyannah So an RPG is a gamified representation of office politics? Got it.
-
@maiyannah And of the tendency to see mostly the downsides of existing stuff but only the promises of new stuff.
-
@maiyannah Err, yeah, Go. Been there, done that, came to the conclusion that PHP is much better. And I know PHP's issues all too well.
-
@maiyannah Well opengl is pretty much crap in any language since it's pretty much its own special language.
-
@maiyannah Yeah, I don't get C# as a C++ replacement. The strength of C++ is performance and closeness to the C hardware ABI. C# on the other hand is about relative ease of use, relative safety and a solid base lib. I do think though that a lot of win32 applications that used to be in C++ make more sense in C#. Think business apps with lots of forms. Evangelists can be annoying though.
-
@maiyannah I don't really want to start a programming language flamewar but I'm a bit curious about the positioning of C and C++ as something that will crash and burn instead of doing something unexpected or triggering undefined behavior. From my understanding and experience of C and C++ one of the subtler points of those languages is that it is fairly easy to create a fault situation that does not lead to a crash. For example access to memory outside an object (in the C standard sense: a region of memory associated with a value) can and will lead to values being silently read/written from the wrong place unless you happen to be lucky enough to point to an invalid page (such as page 0, which is conventionally never mapped in order to make null pointer dereferences trigger access violations). And let's not get started on the fun that's called strict aliasing. ;)
-
@maiyannah That's a fair point. And I admit that C# can sometimes be a little annoying (no ability to call non-nullary constructors of a generic type without reflection is one of my favorites). But I honestly can't say I've experienced much hinder from C# as a language. In my experience if C# is blocking something there's usually a way to go around the runtime restrictions by moving down an abstraction level. Types can generally be overridden by casting, at the risk of causing null pointer exceptions if you're wrong. Managed code issues can be worked around with unsafe blocks and pointer level programming. As much as I like to moan about MS suckyness I do have a strong impression that most of that comes from higher level concepts like overcomplicated frameworks (hello SOAP), not so much from the compiler layer.
-
@maiyannah Err, compared to what? I know a lot of programming languages and from a design perspective C# is pretty decent for a practical programming language. The .NET Framework API's sometimes are awful but I don't really know any language in the space of object oriented and strong typing that's better designed. Java is much more limited and Kotlin is limited by the constraints of the JVM resulting in crappy generics. C++ has the downside of an incredibly complicated almost untyped template layer and a lot of subtleties. Typescript is limited by Javascript. D is a contender but I wouldn't say it's necessarily better designed (albeit not worse either), more like designed for different constraints. Swift I don't know enough to say anything about though it as I don't do Apple. Rust IMHO forces too much pain on the programmer in exchange for not having a GC and it massively complicated the language, though for its target environment the trade-off might be worth it.
-
@maiyannah Programming in C# is my day job. And yes, that involves applications with, ahem, interesting designs for historical reasons.
-
@maiyannah And marketing shenanigans. Because the problem isn't so much proprietary addons to a permissively licensed product but, being perhaps a bit cynical, an attempt to widen the definition of open source to include things that aren't. Interestingly as I understand the creator of Redis it's not a case where there are paid addons that would otherwise be in the core product, he seems actively opposed to adding those to Redis proper because they don't fit his, laudably narrow, vision of the product's feature set. But on the other hand Redis Labs sure wouldn't mind being seen as having the "full enhanced gold platinum game-of-the-year" version of Redis as that gets them moneys from dumb customers.
-
@maiyannah I have to disagree there. A language that's designed for .NET works wonderfully with it. However only one language is really fully and without compromise designed for .NET: C#. F# is awkward because there's an inherent conflict between the paradigms of the base lib and underlying type system and the functional style. C# doesn't have that mismatch and as a result works much better. VB.NET also doesn't have that mismatch but suffers from it's own problems (mainly design decisions for backward compatibility).
-
Ok, so if I understand it correctly Redis is open source but Redis Labs offering is open core with the core being Redis. Err, I think I understand antirez' position but his employer seems to be acting a bit against the interest of his project by making their relationship sound different than it actually is. As I now understand it the idea is that of a newspaper publisher where there's a journalist side and a business side and the journalist side is independent. But when I see a company named Redis Labs, the only named commercial support and employer of the primary creator of Redis without any extra knowledge it's only reasonable for me to assume that they are in fact running the show. And while I sympathize with antirez it's only natural given this situation that bad publicity for Redis Labs reflects upon the project.
-
I tried OCaml as well. But OCaml on Windows is pretty hard to get working. What I've seen of OCaml is good, it's slowly becoming a more modern language. But OCaml really is a unix language and I doubt that will change. On Windows you're pretty much stuck with Reason leftovers and Reason really is a very different thing than OCaml because of it's focus (webbrowsers / javascript compilation). All that said I'd absolutely use OCaml in a unix environment if the job requires it (i.e. the strictness of OCaml is a benefit over Haskell).
Verius
All Jonkman Microblog content and data are available under the