no-bugs

7 ms·

1.
by no-bugs 4mo ago
This is Chapter 4 of the book; Latency vs throughput is discussed in Chapter 1.
2.
by no-bugs 10y ago
> I recommend you stick a bone through your beard, pronounce yourself a performance guru, and make bank. Have fun. :-) :-) I LOVE when my opponent has to resort to personal insults :-). Leaving aside any sarcastic remarks in this regard:
3.
by no-bugs 10y ago
> made millions of dollars of revenue... I'm not 100% sure you are well situated to lecture me about the 'real world' because you wrote an academic paper. So, we're going to discuss millions of dollars instead, sigh..
4.
by no-bugs 10y ago
> Honestly, it's a lot more complicated than that, and the article itself presents a pretty complicated (and flawed) model. There's no substituting for benchmarks. Of course, all the models are inherently flawed, and in theory
5.
by no-bugs 10y ago
> may be a handful of microoperations merged in with a stream of other instructions. It may indeed in theory, but most of the time in real-world it won't. And BTW - let's not forget about implicit costs of being unable to inlin
6.
by no-bugs 10y ago
> My own benchmarks show costs that are nowhere near the ones cited. This is heavily dependent on the benchmark. In the OP, there is a ref to an academic research showing these numbers - and from my own real-world experience (NOT artific
7.
by no-bugs 10y ago
> The cost of virtual cals is usually due to missed optimizations and the possibility of misprediction. Yep (plus one L1 read to get that VMT pointer). And the OP is pretty much consistent with this (at least within declared "indica
8.
by no-bugs 10y ago
However, the address of this C call is calculated, which means that branch prediction (and the cost of potential misprediction) is involved. So additional cost we're talking about, is roughly a cost of L1 read (3-4 clocks by itself) -
9.
by no-bugs 10y ago
To be clear: I was NOT speaking of the first one :-). As for the second one: > the architect is the guy that can sketch that reporting engine that will be nurtured and improved by the team, Yes (in fact, I LOVE this definition ;-)). >
10.
by no-bugs 10y ago
> by creating a God-like super-architect. Interesting point of view, never thought along these lines. On the other hand, thinking a bit about it - I don't feel that this approach requires a super-architect; what it requires is a DEC
11.
by no-bugs 10y ago
> You can code and knowledge share also. The problem with it is that there are only 24 hours in a day.
12.
by no-bugs 10y ago
> how do you ensure that there are viable career paths for senior people who prefer to focus on code over "knowledge-sharing" kind of activities? From what I've seen, those Senior and Really Strong people who prefer to cod
13.
by no-bugs 10y ago
Added it, THANKS!
14.
by no-bugs 10y ago
THANKS! I've added it (not 100% sure how it would work in practice, but certainly looks interesting).
15.
by no-bugs 10y ago
Statistically - yep. See http://ithare.com/facelift-for-no-bugs/ - only 10 out of 160 said that they don't like the cartoons. BTW, there is a "very beta" way to turn them off - use http://itha
16.
by no-bugs 10y ago
> make a more efficient queue by using one-buffer-and-lock-per-thread. How the reader would block on such a distributed queue when it's empty (without creating one single mutex, which would re-establish the single contention point)?
17.
by no-bugs 10y ago
> If this happens to you, you're doing it wrong. Except for some corner cases - you're right, probably I should add a note about it...
18.
by no-bugs 10y ago
From what I've seen (YMMV), chances of it happening are MUCH smaller than chances of getting context switch under the lock (with lots of threads running into this lock and having their own context switches), because of spending more ti
19.
by no-bugs 10y ago
> Additionally, when you unlock a mutex before signalling the condition, another thread may be scheduled in between that may invalidate the condition that was signalled. This may introduce a race condition. And if you unlock the mutex af
20.
by no-bugs 10y ago
> Aren't most mutex/condition variable implementations optimized to avoid this case by deferring the condition signal to the mutex unlock? Yes, _some_ but not _all_ implementations are doing it; however, as it is not really gua
21.
by no-bugs 10y ago
Since IPv6 it is mandatory, but it still sucks :-(
22.
by no-bugs 10y ago
BTW, when I've wrote about those "undetected network errors", I've meant those-network-errors-which-arise-in-UNENCRYPTED-packets (as 16-bit UDP checksum is certainly not enough for this purpose). And DTLS etc. take care
23.
by no-bugs 10y ago
> Sorry to nitpick but UDP seems like the wrong protocol for authentication. While authentication over UDP is indeed a bit more difficult if implementing it yourself, fortunately there is DTLS which is a very official and well-recognised
24.
by no-bugs 10y ago
You're right, I forgot about it. But is there a ready-to-use library doing it?
25.
by no-bugs 10y ago
Last thing I've heard was that specifically CurveCP was separated from libsodium into libchlorium, and it is not maintained now.
26.
by no-bugs 10y ago
> Shouldn't you be verifying a MAC before decrypting anyways? Relying on decrypting scrambled data, especially for a bitflip, seems like a bad idea. Oh, we're going into the realm of MAC-then-encrypt vs Encrypt-then-MAC, which
27.
by no-bugs 11y ago
It is not only about development costs, but also about hosting costs. As a rule of thumb, when running an MMOG you won't have your own servers in your own office, but will rent them (apiece or as a part of the cloud) from an ISP/C
28.
by no-bugs 11y ago
Thanks! And BTW Nystrom's Game Programming Patterns is one of a few great books on game programming, so I take comparison as a compliment :-)
29.
by no-bugs 11y ago
> Is the entire book completely free or was this just some parts of it? Its too good to be true. "beta" of the entire book is going to be free. "Release" version is going to be paid (sorry). OTOH, if you make a "
30.
Beta Testing of the Book: “Development and Deployment of MMOG”(ithare.com)
2 points by no-bugs 11y ago|1 comments

More ›