3 ms·

Does that make him wrong?

by gcv 13y ago

rimantas 13y ago
As much as cherry-picking makes anyone wrong. Replication is one of the main reasons some big project choose MySQL and the guy ignores that point and only mentions replication in passing. I am sure you can have as good a selection of reasons to avoid PostgreSQL or any other. That does not really matter though. Some guy who never bothered to read MySQL documentation is now switching to PGSQL and will run it without bothering to read any documentation. It does not matter because his project is so trivial it would run on anything, but he is happy because he uses "the real database". So be it. And the guys who really do know what's going on and run systems with hundreds of millions of users don't waste their time writing bullshit articles like this. Guy recomends MongoDB as an alternative, what can I say.
- julie1 13y ago
  replication in mysql? Well, kind of random, I remember @work people not getting it right for 3 years... because they forgot to read the documentation. We may have too much clueless developer in the wild, imitating for security. Like the more linux is spreading, the more the sect of the post devil spread, you recognize them with their dark incatation: chmod 777 (the number of the post devil) and chown root:root.
- sergiosgc 13y ago
  I see this explanation a lot. However, from what I can gather: - pgsql supports synchronous master-slave replication natively, with excellent performance - mysql supposedly does multi-master replication, but does not solve the hard problems. You can't write on the same row on two masters simultaneously or replication stops; sequential ids are not synchronized between masters; you can't get serial transaction isolation between masters. All in all, I'd rate replication support between both databases as nearly identical. Pgsql does master-slave better, mysql has a half-hearted attempt at multi master. Given that on all else pgsql is markedly better, I can't see replication as the reason for picking mysql.
- epo 13y ago
  What use is replication if the data itself has been mangled by MySQL? The significant arguments against MySQL revolve around it doing a poor job of enforcing data integrity. You go on to claim he hasn't bothered to read the documentation but give no evidence other than wild fanboy accusations that you have read it either. There are plenty of assertions in that blog post, none of which you refute; from which I can conclude you have no arguemnts against what he actually said.

Does that make him wrong?

by gcv 13y ago

Does that make him wrong?

As much as cherry-picking makes anyone wrong. Replication is one of the main reasons some big project choose MySQL and the guy ignores that point and only mentions replication in passing. I am sure you can have as good a selection of reasons to avoid PostgreSQL or any other. That does not really matter though. Some guy who never bothered to read MySQL documentation is now switching to PGSQL and will run it without bothering to read any documentation. It does not matter because his project is so trivial it would run on anything, but he is happy because he uses "the real database". So be it. And the guys who really do know what's going on and run systems with hundreds of millions of users don't waste their time writing bullshit articles like this. Guy recomends MongoDB as an alternative, what can I say.

julie1 13y ago
replication in mysql? Well, kind of random, I remember @work people not getting it right for 3 years... because they forgot to read the documentation. We may have too much clueless developer in the wild, imitating for security. Like the more linux is spreading, the more the sect of the post devil spread, you recognize them with their dark incatation: chmod 777 (the number of the post devil) and chown root:root.
sergiosgc 13y ago
I see this explanation a lot. However, from what I can gather: - pgsql supports synchronous master-slave replication natively, with excellent performance - mysql supposedly does multi-master replication, but does not solve the hard problems. You can't write on the same row on two masters simultaneously or replication stops; sequential ids are not synchronized between masters; you can't get serial transaction isolation between masters. All in all, I'd rate replication support between both databases as nearly identical. Pgsql does master-slave better, mysql has a half-hearted attempt at multi master. Given that on all else pgsql is markedly better, I can't see replication as the reason for picking mysql.
epo 13y ago
What use is replication if the data itself has been mangled by MySQL? The significant arguments against MySQL revolve around it doing a poor job of enforcing data integrity. You go on to claim he hasn't bothered to read the documentation but give no evidence other than wild fanboy accusations that you have read it either. There are plenty of assertions in that blog post, none of which you refute; from which I can conclude you have no arguemnts against what he actually said.