Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
kbsletten
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
12 ms
·
91.
▲
by
kbsletten
9y ago
I'm not a full time Java developer, but unless you have a really hard time adjusting to the syntactic differences or will grumble when you do something you know would be "so much easier in X", you should be fine. You'll
92.
▲
by
kbsletten
9y ago
I've known people who took the "interfaces are good" to the cargo-cult where they would have a POCO and make an interface for it with getters and setters. No functionality and basically only one possible implementation becaus
93.
▲
by
kbsletten
9y ago
See, the thing is that even in this example he wasn't 100% implicit. Could rails magic up those has_many and belongs_to in method_missing? Absolutely. Do they? Of course not since that's "too implicit." For whatever reas
94.
▲
by
kbsletten
10y ago
This was a fun read! When I got married the invitations were very important to my wife, but I built a website to handle RSVP (with my mother as a fallback phone#). It was a fun project, I had my wife help me design it and I also found the t
95.
▲
Ask HN: Tech Jobs in Rural America
2 points
by
kbsletten
10y ago
|
0 comments
96.
▲
by
kbsletten
10y ago
Also, if we're talking about passing pointers to objects to C code, GC frequently moves objects in memory which will invalidate the pointer. It's not possible to reliably predict when this will happen, so you need extra language s
97.
▲
by
kbsletten
10y ago
I would say that in the specific case of this article (talking about passing pointers to managed objects to C) it is easier without a tracing GC simply because reference counting doesn't ever move the object and so pointers won'
98.
▲
by
kbsletten
10y ago
Well, passing raw pointers to garbage collected objects is impossible without language support for "pinning". This is because the GC will move objects around in memory breaking any pointers you're using in your C code. That m
99.
▲
by
kbsletten
10y ago
I don't know for sure, but I believe that memory addresses are generally represented as an unsigned integer of one of a few sizes. 24MB is well over the maximum value for a 16-bit integer (approx. 65K) so they simply used the next larg
100.
▲
by
kbsletten
11y ago
I believe the logic is that error handling (std::cerr) is optimized for being immediate because you need to see that last error message before your program went off the rails. Logging (std::clog) on the other hand is optimized for throughpu
101.
▲
by
kbsletten
11y ago
Genuinely curious, what's the use-case for this?
102.
▲
by
kbsletten
11y ago
> A type guard for a dotted name has no effect if any part of the dotted name is the target of an assignment within the guarded code. For example, a type guard for x.y.z will have no effect if the guarded code contains any assignments to
103.
▲
by
kbsletten
11y ago
In modern CSS we have even more ways to do this, specifically many values of "transform:" will also create a new context.