anarazel

45 ms·

541.
by anarazel 7y ago
> as remote from one another as Postgres procedures and custom Lisps must be in every other respect. Just a fun fact: Parts of postgres (before the SQL days, I think) used to be in lisp. To this day we still have some weird function&#x2F
542.
by anarazel 7y ago
That sounds like you could have hit transparent hugepage / compaction related issues. They, IME, tend to hit more often with lots of long running processes, than when there's a lot of churn. It has gotten a lot better in more rece
543.
by anarazel 7y ago
I find pause/play to be super unreliable. Triggers accidentally (even just a warm sweater sometimes is enough), and other times not at all. Not really important, but standing around and tapping on my headphone to trying and failing to
544.
by anarazel 7y ago
> Anyone who puts non-tactile touch controls on _headphones_ shouldn't be allowed to design electronics interfaces ever again. My QC35's died - probably not Bose's fault. Too old to be repaired, apparently. Got the 700&#x2
545.
by anarazel 7y ago
> with Postgres you will need a LOT of RAM to sustain the same The per-connection memory overhead is not actually that high. If you configure huge_pages, it's on the order of ~1.5MB-2MB, even with a large shared_buffers setting. Unf
546.
by anarazel 7y ago
One more thing: There's definitely a significant overhead implied by the process-per-connection model - I don't want to deny that. In my opinion it's at the moment not the most urgent issue wrt connection scalability (the sna
547.
by anarazel 7y ago
> I'll trust postgresql more when it can support a few thousand connections without resorting to running middleware (pgbouncer) all over. That was his point. PostgreSQL is just abysmally bad in this area. Depending on your workload
548.
by anarazel 7y ago
Postgres does have ldap based auth, and also can authenticate against AD using sspi/gssapi. The problem with that is that it requires users to have been created inside postgres first, and that you can't manage group membership ins
549.
by anarazel 7y ago
That is correct. And neither is a full page write logged if the page is initialized from scratch. And even without WAL compression, the "hole" in the middle of the page if the page is not full, is "compressed" out. That&
550.
by anarazel 7y ago
> Do you have an example query? I mean, just about any geospatial query? E.g. finding all polygons in a table that overlap with one specified in a query. https://postgis.net/docs/ST_Overlaps.html > If you're
551.
by anarazel 7y ago
You need different index structures. Yes, it matters. Hard to believe that spatial queries over geometries more complex than points will meaningfully benefit from "vectorized executions on columnstores".
552.
by anarazel 7y ago
> How would you feel about Postgres listening over QUIC instead of/in addition to TCP? At some point? Probably not a bad idea. Right now? Doubtful it's worth it - the stack doesn't seem ready yet. > and enable clients t
553.
by anarazel 7y ago
We should allow the client, on the protocol level, to add (and presumably remove) types to a server side list of types that can automatically be sent in binary if the result format is specified as 'auto'. That way a driver can add
554.
by anarazel 7y ago
> Logical changeset extraction itself does not in itself provide a replication feature It is however a building block for logical replication, and used by https://www.postgresql.org/docs/current/logical-replicat
555.
by anarazel 7y ago
> The driver knows which types it can parse, but it doesn't know what types a given query will return. [...] Postgres could know the data types, but that means that the driver would need to first prepare the query (which is sometime
556.
by anarazel 7y ago
> * Having two protocols/data types that two the same thing (text / binary) isn't the end of the world, but it just adds confusion. Also it adds complexity for a different server/proxy/pool implementation. Recomm
557.
by anarazel 7y ago
> Even if they didn't, there are plenty of algorithms based on lat/lon which uses numeric/float data types with simple indexes. Not that trivial if you're not just dealing with points.
558.
by anarazel 7y ago
> The one thing I wish they would do, though, is improve the error messages when a query fails. Yea, we definitely need to do better there. But it's harder than one might think initially :(. My suspicion is that we (the PG developer
559.
by anarazel 7y ago
Hm. It was more recent than that. Perhaps I am just remembering that it was not possible to get it to trace user defined allocators / wrappers around malloc?
560.
by anarazel 7y ago
> - Postgres' version of upsert allowed any command to be run without checking the permissions. So the vaunted "software engineering" behind Postgres is not that solid. That never was the case (see evidence in initial comm
561.
by anarazel 7y ago
Implementation defined behaviour is not the same as undefined behavior.
562.
by anarazel 7y ago
Most passenger airplanes also takes cargo containers (of a size specific to a group of plane types) below the passengers. Not that many of course, compared to a cargo plane. E.g. even a A321 takes 10 LD3-45 containers. Depending on the dens
563.
by anarazel 7y ago
There's a few pure python drivers. Check https://wiki.postgresql.org/wiki/Python and https://wiki.python.org/moin/PostgreSQL E.g. pg8000 looks like it's being maintained?
564.
by anarazel 7y ago
Several schools, after requests, are graduating students early, precisely because of covid. And I don't think normal times and these are the same. Having younger, less likely to be infected staff is not nothing. And the variety of task
565.
by anarazel 7y ago
Last time I checked - I'm not sure how long ago that is - llvm didn't detect double frees etc statically. There's an annotation framework for locking though, which I hope to play with more.
566.
by anarazel 7y ago
How do you think hiding the possibility of doing calls in the browser fit into that picture? Surely doesn't make things easier for the people that don't want to / aren't allowed to install things.
567.
by anarazel 7y ago
FWIW, the README.md says "the Phoenix server listens to PostgreSQL's replication functionality (streaming WAL)" - that makes it sound like you're actually parsing the WAL itself. But you seem to be looking at the already
568.
by anarazel 7y ago
(not not an epidemiologist, just interested) Once testing becomes widespread enough that it's feasible to test a lot more people it's quite plausible to relax the restrictions. One big problem before the recent "shelter in pl
569.
by anarazel 7y ago
There has to be more to the situation than the aove. Unless you configured join_collapse_limit to be way lower than the default, the JOIN query just gets transformed into the former. Any chance you either had a lot more tables joined togeth
570.
by anarazel 7y ago
And relevant to the roundtrip issue at hand here: You can send bind/execute in a pipelined manner. I'd benchmarked batch upserting at some point, and at that time for large amounts of data the fastest approach was somewhat unintui

More ›