5 ms·

Community Manager for MySQL here! A better overview link for what's new in 5.7.6 specifically is this one: http://mysqlserverteam.com/the-mysql-5-7-6-milestone

by morgo 12y ago

Community Manager for MySQL here! A better overview link for what's new in 5.7.6 specifically is this one: http://mysqlserverteam.com/the-mysql-5-7-6-milestone-release-is-available/ http://mysqlserverteam.com/the-mysql-5-7-6-milestone-release... To understand what's new in 5.7 overall: http://dev.mysql.com/doc/refman/5.7/en/mysql-nutshell.html http://dev.mysql.com/doc/refman/5.7/en/mysql-nutshell.html Specific to this blog post. Here is the page on upgrading: http://dev.mysql.com/doc/refman/5.7/en/upgrading-from-previous-series.html http://dev.mysql.com/doc/refman/5.7/en/upgrading-from-previo... I'm happy to answer any questions, please feel free to ask away.

jordigh 12y ago
> Community Manager for MySQL here! I always found this title a bit weird. How does one exactly manage a community? :-) https://plus.google.com/+AaronSeigo/posts/DHEx1tu5dSK https://plus.google.com/+AaronSeigo/posts/DHEx1tu5dSK
- morgo 12y ago
  Please see my comments on this thread: https://news.ycombinator.com/item?id=8318199 https://news.ycombinator.com/item?id=8318199 I agree it's a silly name, but it's not like we created it :)
  - ddorian43 12y ago
    Click on the datetime of the comment to get the correct comment-link: https://news.ycombinator.com/item?id=8318807 https://news.ycombinator.com/item?id=8318807
  - j_s 12y ago
    If you've not heard of the term 'tummler' this might be helpful: To build a strong community, stop “community managing”, be a Tummler instead. http://dangerouslyawesome.com/2014/04/community-management-tummling-a-tale-of-two-mindsets/ http://dangerouslyawesome.com/2014/04/community-management-t...
    - morgo 12y ago
      I like it - thanks for the link.
DonHopkins 12y ago
"The old syntax was meant to be only deprecated, but it was accidentally completely removed." How could Oracle have possibly "accidentally completely removed" the old syntax for setting a password even though it "was meant to be only deprecated", without being completely incompetent? I can not in my wildest imagination come up with a scenario in which a competent developer could possibly "accidentally remove" something like that. If they made such a huge "accident" with that security feature, what other terrible negligent "accidents" lurk just beneath the surface of this new release of MySQL? How can you "LOL" away such an incompetent "accident" like that?
- datacharmer 12y ago
  "LOL away?" What gave you the impression that I was making fun of it? I am the one who found the bug and reported it, and this is the explanation I got. Since the person I talked to is someone I trust, I don't have reason to think otherwise. I can elaborate further and tell you why this particular error was likely to happen: it's because since the beginning MySQL was built with root user without password, and most of the tests were running in such conditions. That was the culture of rapid growth that gave MySQL its popularity and the developers are paying now the technical debt that has left so many features with poor testing behind. What Oracle is doing now is trying to redress that situation by strengthening the defaults, for which I commend them. Being a QA developer, I am less pleased with the quality of tests that let such mistake pass, but I know that such things happen, and it does not necessarily mean that there is a can of worms behind this bug.
  - DonHopkins 12y ago
    It's not my impression that you're making fun of it, but that you're trying to excuse it as an inconsequential "accident" instead of incompetence. Accidents like that don't happen without causes like incompetence or negligence or malice. Accidents like that are a symptom that something is deeply wrong.
- morgo 12y ago
  This is noted in the release notes, and I have a blog post on it here: http://www.tocker.ca/2015/03/10/heads-up-mysql-5-7-dmr6-contains-a-small-known-bug.html http://www.tocker.ca/2015/03/10/heads-up-mysql-5-7-dmr6-cont... To describe our process: We have at least 2 code reviews, ~90% test coverage before we release a feature in a development milestone. It was then caught by at least 3 people that I know of prior to public release. Since this is not a GA release, it was decided to fix the bug in the next version.

by morgo 12y ago

jordigh 12y ago

> Community Manager for MySQL here! I always found this title a bit weird. How does one exactly manage a community? :-) https://plus.google.com/+AaronSeigo/posts/DHEx1tu5dSK https://plus.google.com/+AaronSeigo/posts/DHEx1tu5dSK

morgo 12y ago
Please see my comments on this thread: https://news.ycombinator.com/item?id=8318199 https://news.ycombinator.com/item?id=8318199 I agree it's a silly name, but it's not like we created it :)
- ddorian43 12y ago
  Click on the datetime of the comment to get the correct comment-link: https://news.ycombinator.com/item?id=8318807 https://news.ycombinator.com/item?id=8318807
- j_s 12y ago
  If you've not heard of the term 'tummler' this might be helpful: To build a strong community, stop “community managing”, be a Tummler instead. http://dangerouslyawesome.com/2014/04/community-management-tummling-a-tale-of-two-mindsets/ http://dangerouslyawesome.com/2014/04/community-management-t...
  - morgo 12y ago
    I like it - thanks for the link.

DonHopkins 12y ago

"The old syntax was meant to be only deprecated, but it was accidentally completely removed." How could Oracle have possibly "accidentally completely removed" the old syntax for setting a password even though it "was meant to be only deprecated", without being completely incompetent? I can not in my wildest imagination come up with a scenario in which a competent developer could possibly "accidentally remove" something like that. If they made such a huge "accident" with that security feature, what other terrible negligent "accidents" lurk just beneath the surface of this new release of MySQL? How can you "LOL" away such an incompetent "accident" like that?

datacharmer 12y ago
"LOL away?" What gave you the impression that I was making fun of it? I am the one who found the bug and reported it, and this is the explanation I got. Since the person I talked to is someone I trust, I don't have reason to think otherwise. I can elaborate further and tell you why this particular error was likely to happen: it's because since the beginning MySQL was built with root user without password, and most of the tests were running in such conditions. That was the culture of rapid growth that gave MySQL its popularity and the developers are paying now the technical debt that has left so many features with poor testing behind. What Oracle is doing now is trying to redress that situation by strengthening the defaults, for which I commend them. Being a QA developer, I am less pleased with the quality of tests that let such mistake pass, but I know that such things happen, and it does not necessarily mean that there is a can of worms behind this bug.
- DonHopkins 12y ago
  It's not my impression that you're making fun of it, but that you're trying to excuse it as an inconsequential "accident" instead of incompetence. Accidents like that don't happen without causes like incompetence or negligence or malice. Accidents like that are a symptom that something is deeply wrong.
morgo 12y ago
This is noted in the release notes, and I have a blog post on it here: http://www.tocker.ca/2015/03/10/heads-up-mysql-5-7-dmr6-contains-a-small-known-bug.html http://www.tocker.ca/2015/03/10/heads-up-mysql-5-7-dmr6-cont... To describe our process: We have at least 2 code reviews, ~90% test coverage before we release a feature in a development milestone. It was then caught by at least 3 people that I know of prior to public release. Since this is not a GA release, it was decided to fix the bug in the next version.