Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
br1
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
10 ms
·
31.
▲
by
br1
12y ago
Don't try to read the generated SQL. Debug the query plan directly.
32.
▲
by
br1
12y ago
- Joins that know about the foreign keys already declared in the table definitions. - Structured rows to kill all grouping special cases and the distinction between where and having. - Hierarchical rows so that join(Planet, Person) doesn&#x
33.
▲
by
br1
13y ago
As a LINQ user, I expected group('category').sample(3) to sample 3 categories, not to sample inside each category.
34.
▲
by
br1
13y ago
Visual C++ uses EDG only for the IDE. The compiler still has it's own frontend.
35.
▲
by
br1
13y ago
Macs have a similar issue. Unexpected programs activate the dedicated GPU. Skype and twitter used to do this. Power users run special utilities to force the dedicated GPU off, but the normal user has no idea that his Mac battery won'
36.
▲
by
br1
13y ago
I don't see how that part of standard mandates that all anomymous functions share the same closure. The example only has a single lambda.
37.
▲
by
br1
13y ago
lostoptimist surely meant running visual studio itself as a 64 bit process to use more than 4 gb of ram
38.
▲
by
br1
14y ago
Windows developers have it easier because debuggers are better. The same debugger can work on VB6 code calling C++ calling C#. Memory dumps also seems to be more common and useful in Windows than in Unix. If you get dump for all crashes,
39.
▲
by
br1
14y ago
I agree on (1). Typedef obscures more than it helps. On (2), I think the solution is intrusiveness. intrusive data structures and intrusive_ptr should be used more in C++. The dislike for intrusiveness is rooted on equating classes with r
40.
▲
by
br1
14y ago
To consider engine generations as separate and discrete entities stems from id software. The engines from Wolfestein, Doom, Quake, Q2 and Q3 were written from scratch each time. Once Carmack abandoned C for C++, throwing everything out st
41.
▲
by
br1
14y ago
Great idea. My in-laws have it backwards: the password is the street number... Changing it would break all devices that already store the password, so it stays.
42.
▲
by
br1
14y ago
Can you give a concrete example of a refactoring you had to do?
43.
▲
by
br1
15y ago
The insight of the article is that indexes and tables are the same thing underneath. Using an non covering index costs like a join. I'm a tokutek fan, but this article is the clearest report of the pros of tokudb so far.
44.
▲
by
br1
15y ago
"YOU CAN'T SOLVE A DEBT CRISIS BY ISSUING MORE DEBT" You actually can. The issue is who is the creditor and debtor. Krugman wrote about this.
45.
▲
by
br1
15y ago
In some RDBMSs, primary and secondary indexes are very different. MsSql clustered tables store the tuples themselves in the primary index B Tree. Traversing the data in the order of the primary index incurs no extra disk seeks. Also, you
46.
▲
by
br1
15y ago
Reference counts could be stored separately from objects and migrated to the thread that modified it. Or several reference counts for the same object could be used. Has this been tried?
47.
▲
by
br1
15y ago
It's more down to earth than SICP. O'Keefe is like Michael Abrash in Graphics Programming Black Book.
48.
▲
by
br1
15y ago
Another take on moving logic to the DB server: thehelsinkideclaration.blogspot.com
49.
▲
by
br1
15y ago
Algorithm efficiency is studied with generating functions, which are solved with complex calculus. That route includes most math stuff one normally thinks as unnecessary for CS.
50.
▲
by
br1
15y ago
Filesystems lose on a big opportunity by supporting hardlinks. If every file had only one name, files on the same directory could be stored closely, much improving locality.
51.
▲
by
br1
15y ago
The saddest thing is that computer science looks up to math and emulates the worst of it, such as cryptic, short variable names.
52.
▲
by
br1
15y ago
Microsoft actually went to great length to build an update mechanism that doesn't require reloading. It seems this is not so useful after all, and it's not being used: http://jpassing.com/2011/05/01/windows-hotpatching/
53.
▲
by
br1
15y ago
It's hard for the rest of the world but incremental for them... Let's see how they turn ego into code.
54.
▲
by
br1
15y ago
Exactly. Compilers manipulate symbols, so functional languages with pattern matching are a better match than C.
55.
▲
by
br1
15y ago
Not that I know of. Also, they have a patent, but that didn't stop Acunu from reimplementing and improving the algorithm.
56.
▲
by
br1
15y ago
Being in-kernel, acunu is very far from the lightness of LevelDB.
57.
▲
by
br1
15y ago
Read "Cache-Oblivious Streaming B-trees" http://supertech.csail.mit.edu/cacheObliviousBTree.html It' LSM with faster searches.
58.
▲
by
br1
15y ago
Inserts are much faster due to the log-structured merge-tree.
59.
▲
by
br1
15y ago
Implement COLA and leap frog the competition: http://supertech.csail.mit.edu/papers/sbtree.pdf
60.
▲
by
br1
16y ago
Great article. At first it seems to match the computer architecture course, but at least my professor didn't give fine detail like that hyper-threading had a 10% logic overhead.
More ›