anarazel

89 ms·

781.
by anarazel 8y ago
Marketing / Evangelism / Outreach. I'm not kidding, and I don't predominantly mean that in a bad way. I think it's something that the postgres community / companies really aren't good at.
782.
by anarazel 8y ago
Postgres maintains 5 years worth of major releases, not more. After that no more fixes for data corruption, security issues etc. IOW, using that old a version of postgres is careless, and might even bring legal liability after an intrusion
783.
by anarazel 8y ago
Looks like it already happened a few months back. Earlier today Seamus Hughes noticed an apparent copy&paste mistake in an independent court filing: https://twitter.com/SeamusHughes/status/1063232297674162176
784.
by anarazel 8y ago
> I'm pretty sure that you and I both agree that we shouldn't refer to someone else's hard work in that way. I only said they're not large changes. small != worthless.
785.
by anarazel 8y ago
My point wasn't to belittle the contributions, but to point out that he wasn't a "key figure".
786.
by anarazel 8y ago
> What? I would be proud to have got this into postgres core, but perhaps you operate on a different plane of excellence. https://git.postgresql.org/gitweb/?p=postgresql.git;a=blobdi... . It's just a minor rewri
787.
by anarazel 8y ago
"key figure" as in: A few small changes. https://git.postgresql.org/gitweb/?p=postgresql.git&a=search... https://git.postgresql.org/gitweb/?p=postgresql.git;a=commit... Edit: Added t
788.
by anarazel 8y ago
Most databases don't have "root level snapshots of the state" in the way you appear to imagine them. Leaving trivial per-transaction information ('did this transaction id commit', ~a few bits per xact) aside, usual
789.
by anarazel 8y ago
> Thus, a backend concurrent connection (i.e. transaction) cap of N, translates to O(N) concurrent MVCC "world states" that the DB engine has to keep track of. No matter how efficient the DB engine is at doing that, that's
790.
by anarazel 8y ago
> Are large numbers of concurrent transactions really an issue in practice though? There may be some per-transaction overhead, but I haven't seen any stern warnings about it, e.g. in the Postgres docs for max_connections. There&#x27
791.
by anarazel 8y ago
After enough of these experiences I nearly always carry a bunch of fruit bars and such around. Both improves the groups mood, and usually allows to find better food afterwards.
792.
by anarazel 8y ago
He does not maintain the postgresql planner. He's a PostgreSQL developer, but he has not focused on the planner for at least a decade. Edit: Grammar.
793.
by anarazel 8y ago
> Timescale is just an automatic partitioning extension for Postgres. That, and a good PR operation.
794.
by anarazel 8y ago
> Not just any service. For instance, Apple refused to assist the FBI in unlocking the iPhone belonging to one of the late prime suspects in the San Bernadino massacre. Wikileaks is refusing to co-operate with the CIA. Microsoft is resis
795.
by anarazel 8y ago
> Meanwhile, Gab was able to provide a complete profile of this individual and confirmation of his identity to the authorities. This repeatedly has been presented as an achievement. How on earth isn't that the absolute absolute basi
796.
by anarazel 8y ago
It kind of still works like that...
797.
by anarazel 8y ago
Oh, if that's true, I'm sorry. They did tweet a lot of stuff around that time, but so much of it is deleted, that it's hard to remember exactly what is what. Edited post to note uncertainty. Not that I think that makes the p
798.
by anarazel 8y ago
Comeon, you're cherry-picking. The comment about inferring a position on free-speech from Krassenstein's name clearly was a play on his name sounding Jewish. [EDIT: WRONG: And the retweet in the other link was just after the shoo
799.
by anarazel 8y ago
The profile had been discovered before they removed the contents. A few people / journos downloaded the contents. So no, they couldn't have covered this up.
800.
by anarazel 8y ago
Meh. Look at the official PR of gab, and tell me again they're solely concerned with free speech: https://twitter.com/Popehat/status/1056266704148262912 https://twitter.com/ClenchedFisk/s
801.
by anarazel 8y ago
It's not realistically possible to challenge them. If you do, you'll be brigaded for weeks (hopefully just online). And a lot of it is not exposed to anybody challenging their views, but largely just to people at least vulnerabl
802.
by anarazel 8y ago
It's a significant difference whether you can become an infowars listener by watching a video on youtube that has infowars content as the next suggested title, or whether you need to explicitly search it out.
803.
by anarazel 8y ago
Because the criticism by mainstream media obviously isn't that they're cooperating with law-enforcement (if they are), but that they're very willing to host a lot of the awful stuff. They're just trying to stoke the fl
804.
by anarazel 8y ago
> > If it is, why don't Conservative just create their own tech stuff? > That's literally what Gab is. You're doing "conservatives" a disservice if you claim that anything with a conservative bent would ha
805.
by anarazel 8y ago
Have you read Gab's twitter feed? The whole "We're just for free speech" spiel would be a whole lot more believable if they weren't also constantly spewing reprehensible rhetoric. https://twitter.com&#x2F
806.
by anarazel 8y ago
> Gab is benign, they started based on people getting kicked off Twitter Read the official twitter profile (what they write and reteweet). And then tell me again that they're not explicitly reaching for the racist fringe.
807.
by anarazel 8y ago
As I commented in another post, I don't think this is an actual problem in postgres. But assuming it would be, the difference is that control of central decision making structures allows control over other companies work, whereas hav
808.
by anarazel 8y ago
See https://news.ycombinator.com/item?id=18296064
809.
by anarazel 8y ago
Disclaimer: I used to work for Citus and have options. > Equity, however, creates a conflict of interest. If the non-profit Postgres organizations make money when Citus makes money, then they have an incentive to favor Citus over other c
810.
by anarazel 8y ago
I think "will be making" unfortunately includes a bit of wishful thinking. It's far from clear we'll have a patch that's mature enough by the time the feature freeze is there and that also has the necessary suppor

More ›