macdice

9 ms·

61.
by macdice 6y ago
One phenomenon I am fascinated by is the advent of software that is entirely built out of open source components and open standards, that is closed off to people using non-mainstream open operating systems and hardware sort of by accident a
62.
by macdice 6y ago
I have done some work on parallelism in PostgreSQL, and I was fascinated to discover Wei Hong's work in this area from more than a quarter century ago: https://www2.eecs.berkeley.edu/Pubs/TechRpts/1992/ER
63.
by macdice 7y ago
Retry logic should not be complex! Model transactions as functions, and call them again with the same inputs.
64.
by macdice 7y ago
The Indy was a really important machine in my journey and I occasionally think I should pick one up for nostalgic use... but... meh, Irix is frozen (there is no OpenIRIX soldiering on...) and if I ran say FreeBSD on it, it'd just be a
65.
by macdice 7y ago
"There's no server process" vs "there's no server on your itemised bill".
66.
by macdice 7y ago
Will be interesting to see how kqueue systems respond.
67.
by macdice 7y ago
Care to elaborate?
68.
by macdice 7y ago
If the keys are in random order, and the hash table is larger than L3, I bet you can make the hash version faster by looking ahead N items and issuing __builtin_prefetch() on the hash table array. (There are papers on this for hash joins i
69.
by macdice 7y ago
In PostgreSQL we could do that in parallel: https://www.postgresql.org/message-id/CA+hUKGKzn3mjCAp=TDrji...
70.
by macdice 7y ago
One nice thing about the CTE inlining its that materialised CTEs were stopping parallelism from being used. WITH is useful in big OLAP queries, where parallelisms should really help.
71.
by macdice 7y ago
Another factor is that if you move countries some of your "purchased" movies disappear because you only "purchased" the right to watch them in territory X. While DVDs also made a half-hearted attempt at artificially pre
72.
by macdice 7y ago
I encountered this PICK stuff a couple of decades ago. It was really enlightening. I began to understand how the features of relational databases (transactions, isolation levels, constraints, normal form) fit together to avoid anomalies a
73.
by macdice 7y ago
Interesting to see the age distribution of committers! I wonder how it compares to other large established infrastructural projects (operating systems, compilers, databases etc).
74.
by macdice 7y ago
It has its origins in the use of Spanish dollars and various related bits of gold that could be reliably cut in half, in early American history. I think a lot of companies got away with using binary floating point maths without introducing
75.
by macdice 7y ago
That's the one, a great story!
76.
by macdice 7y ago
They used https://en.wikipedia.org/wiki/Interdata_7/32_and_8/32 and the Interdata OS/32 operating system. They wrote all the code in Fortran. There is a clue about the rationale for that in the Bloombe
77.
by macdice 7y ago
Cool. I was just reacting to the fact that you highlighted merge joins in parallel queries with a partial plan on the outer side only, which was a new feature back in 2017 when 10 came out (and was already supported for hash joins and nest
78.
by macdice 7y ago
The section on parallel joins is a bit out of date. Shameless plug: I've written and spoken about this subject extensively, if you are looking for more information, with further links and pointers to other blogs and academic papers.
79.
by macdice 7y ago
Also AIX has SIGDANGER (best signal name ever). There is also "oomd", I think like the iOS thing, for Linux, in user space. Not sure how useful any of these things really are but wondering if FreeBSD should get one...
80.
by macdice 7y ago
Well, there are/were a whole bunch of related problems and scenarios discussed in that and other threads, so it depends which bit we're discussing. (1) errors consumed by asynchronous kernel activity that never make it to user sp
81.
by macdice 7y ago
Totally agree. Speaking as a database hacker, just tell me it broke . I will deal with the rest.
82.
by macdice 7y ago
When a write-back error happens, Linux marks the block clean so it can be replaced by another block at any time. Additionally, some Linux filesystems (I don't recall which right now) proactively throw it away immediately.
83.
by macdice 7y ago
I think the problem is mostly ENOSPC from fsync() which jettisons data just like EIO on Linux. If you ran out of space, PostgreSQL would only learn about that while checkpointing, and then retry and carry on. Boom, data loss. Today Postg
84.
by macdice 7y ago
What about this: you write() some data, and then Linux throws it away because of a writeback error, and later you read() it, and Linux gives you an older version from before your write (because this error condition has somehow been fixed an
85.
by macdice 7y ago
The article uses "English" and "British" interchangeably when describing events in 1702. But the Acts of Union that united England and Scotland creating Great Britain didn't happen until 1706 and 1707. I guess tha
86.
by macdice 7y ago
Seems useful to me. Fixed sized data is very convenient for computers (fits in a register, pass-by-value), and being able to represent 0.1 and 0.01 is very convenient for humans. The hardware does seem to be of questionable value; I heard
87.
by macdice 7y ago
"A later revision of IEEE 754 attempted to remedy this, but the formats it recommended were so inefficient that it has not found much acceptance. DEC64 is a better alternative." This is just plain wrong. The IEEE decimal float ty
88.
by macdice 7y ago
Nice! Yeah, syscall interception may be a better way to do this than fuse. So, the thing I want is the ability to take a log (which probably also has checkpoints in it, I dunno) captured by normal operation of a database, and then run a su
89.
by macdice 7y ago
Database hacker here. I have dreamed of a filesystem fault injection system like this: a fuse filesystem that just passes through to a real filesystem, but also writes a log of syscalls. Then a test mode that can simulate crashes at vario
90.
by macdice 7y ago
Or use one of the CI providers that will test your stuff on macOS (alongside Windows, FreeBSD, ...). Possibly for free, if your stuff is open source. I use that for Windows, which I don't personally use at all, and it finds bugs and

More ›