tomiko_nakamura

6 ms·

1.
by tomiko_nakamura 12y ago
MVCC means "Multiversion Concurrency Control" and yes, that's pretty much exactly how it works (creating copies of the rows, so that different snapshots see different rows). In practice it's more complicated of course. A
2.
by tomiko_nakamura 12y ago
I don't think XML was all that widely used, and the implementation looked more "looks interesting, let's develop something and people will use that". With JSONB it's more like "people are already using JSON dat
3.
by tomiko_nakamura 12y ago
I think that if you're using JSONB for data that can be easily stored in a proper tabular form (aka relation), you're doing it wrong. No doubt there will be people using it this way, though, and maybe there are cases where it real
4.
by tomiko_nakamura 12y ago
People are working on other storage types for PostgreSQL. That's all I can say at this moment.
5.
by tomiko_nakamura 12y ago
Yeah, IRC is a good place to ask too. The obvious problem is that not all developers hang there all the time, so for longer discussions the mailing lists are better (and you can just attach test cases, for example).
6.
by tomiko_nakamura 12y ago
PostgreSQL does not have in-place update in the first place, on every UPDATE it does a DELETE+INSERT. So "partial update" does not make much sense, because you'll create a new copy of the row anyway.
7.
by tomiko_nakamura 12y ago
I really wonder what you mean by UPSERTS. Because the traditional meaning for UPSERT is "UPDATE if exists, INSERT otherwise", aka MERGE. And you're right PostgreSQL is missing that, but hopefully we'll get something like
8.
by tomiko_nakamura 12y ago
It doesn't have MySQL-like pluggable storage, and frankly I'm thankful for that. How many of the MySQL storage engines actually work, including transactions for queries working with tables using different storage engines and such?
9.
by tomiko_nakamura 12y ago
I'm not saying "plan locking" is not a useful feature in some cases, and maybe it would really solve this particular issue. But my experience is that when people say "I need it" in most cases it's "I'
10.
by tomiko_nakamura 12y ago
Expression indexes are nice, and you can still do that. It however requires you to know what you'll be looking for (and create index on that particular expression). The stuff I find really awesome is indexing for arbitrary expressions:
11.
by tomiko_nakamura 12y ago
Why? What mistake? I mean, many people won't use that, but for many it's very handy. And the features delivered with JSONB (e.g. the indexing) is way better than what was available with XML.
12.
by tomiko_nakamura 12y ago
JSONB had issue with compression, because the compression is decided on the fly by checking if first part of the serialized data, and in the original format the first part was incompressible (thus no compression was used). But this was fixe
13.
by tomiko_nakamura 12y ago
BTW I wonder what you mean by "accumulated stats" ... Regarding your senior engineers - maybe they're competent, maybe not. But most importantly, the experience from one database is often difficult to transfer to a different
14.
by tomiko_nakamura 12y ago
The only response I have to that is "talk to developers on the mailing list" (either pgsql-performance or pgsql-hackers). Maybe it's possible to improve the planning - maybe your queries are uncommon / difficult to estim
15.
by tomiko_nakamura 12y ago
IANAL, of course, but ... 1. There's no suitable legal entity that could receive the copyright/rights on behalf of the PostgreSQL project. There are entities handling domains/... but that's not really useful. 2. The fact
16.
by tomiko_nakamura 12y ago
I don't really see how that matters. This only shows how utterly broken the concept of "patch Tuesdays" is ... If you plan your internal deployment updates based on the belief that the schedule will never change, then I'
17.
by tomiko_nakamura 12y ago
Well, they thought so too. And it turned not to be true. And now they're blaming the others.
18.
by tomiko_nakamura 12y ago
If this is used just as an AP, how much harm can a baseband backdoor do? I mean, there are no interesting sensors, the other side already knows the (approximate) location ... assuming the user uses proper crypto (not on the same device, of
19.
by tomiko_nakamura 12y ago
If you start making exceptions, everyone will want one.
20.
by tomiko_nakamura 12y ago
You're just citing the press release (or a blog post, or whatever). How does that respond to the original question? The issue mentioned in the article [ https://code.google.com/p/google-security-research/issues
21.
by tomiko_nakamura 12y ago
If you define a limit for disclosure, and then not stick to it, why to define a limit in the first place? 90 days is more than enough - if MS has a lot of internal overhead, you should probably complain to them, not to google.
22.
by tomiko_nakamura 12y ago
So, who forces Microsoft to stick to (so called) patch Tuesdays? No one, actually - it's Microsoft internal schedule, and clearly there are cases when it's absolutely unreasonable - e.g. when there's a 0day in the wild. So th
23.
by tomiko_nakamura 12y ago
That really depends on what version you're using now, and what exactly you mean by scalability. If you're using 9.1 or older, you may see a significant improvement in OLTP workloads on many-core machines (making it linearly scalab
24.
by tomiko_nakamura 12y ago
I'm not a big fan of MongoDB, but I don't think the introduction of JSONB in PostgreSQL 9.4 makes it a dud (which does not mean MongoDB is not a dud for other reasons). JSONB allows you to do a lot of things that people are often
25.
by tomiko_nakamura 12y ago
We certainly are. We're operating an analytical service operated on PostgreSQL - tens of TBs of data, hundreds of machines, tens of thousands of clients. Initially it was running on MySQL, but because of various reasons we migrated to
26.
by tomiko_nakamura 12y ago
AFAIK the default serializer does not guarantee the order to be stable, but I haven't participated in the development of this feature that closely. So maybe I'm wrong. Anyway, it should be possible to write a simple serializer on
27.
by tomiko_nakamura 12y ago
Yeah, it's old. Most basic principles described by Bruce are still valid, though. But if you're looking for fresh info on PostgreSQL tuning, buy this: https://www.packtpub.com/big-data-and-business-intelligence&#x2
28.
by tomiko_nakamura 12y ago
This is the way prosecution works in the US - charging with heaps of bullshit felonies with the aim to scare the defendant, forcing him to plead guilty in exchange for minimum sentence. The defendants have to consider the risk that some of
29.
by tomiko_nakamura 12y ago
"Travel twice as fast" is clearly not what's going on. The article is about full-duplex mode, i.e. doubling the throughput. The data are travelling just as fast as before.
30.
by tomiko_nakamura 12y ago
I wonder how long they've been looking for a background image distracting enough ...

More ›