Description:
Drupal 7 uses ANSI,TRADITIONAL and that makes MyISAM impossible to use on MySQL 5.5 (unless you recompile your server I guess) because NO_ENGINE_SUBSTITUTION got added to the TRADITIONAL mode. Given that engines are MySQL specific, engine substitution happening or not has hardly to do with anything with a 'traditional' database. And InnoDB makes a minimal Drupal install take 6.5 second instead of 2.8 second with MyISAM.

How to repeat:
Start mysql with skip-innodb learn it's broken apply workaround from http://bugs.mysql.com/bug.php?id=55608 restart with skip-innodb default-storage-engine=myisam in my.cnf then try using TRADITIONAL with ENGINE=innodb, kaboom. In MySQL 5.0 and 5.1 it did not blow up.

Suggested fix:
Please restore TRADITIONAL what it was in 5.0 and 5.1 and add a new mode if you want to with this flag.

This change was made in the very first release of 5.5. See http://dev.mysql.com/doc/refman/5.5/en/news-5-5-0.html. See http://bugs.mysql.com/bug.php?id=21099 also.

Wait a second just because it got broken in 5.5.0 how is this not a bug? That's a non sequitur. A bug is a bug regardless when it got broken. 

You changed the semantics of something from 5.0 and 5.1 to 5.5. I say it's a regression. You say it's not a regression because it happened in 5.5.0. What?

As for the linked bug, I told you: create a *new* mode for *new* semantics. Sounds more logical to me than breaking the semantics of an existing mode.

I said it is not a regression, but a change in behavior requested by users (see bug #21099) in the next major version, 5.5, comparing to 5.1 and earlier versions. It was made at the early stage of 5.5 development, was explicitly documented since 2009 and nobody really questioned is it "OK" or not until this your report (as far as I can see).

I agree that any incompatible change is a problem for many users, but there is no way to avoid them in general when we fix bugs and solve other problems. What we can do is explicitly document these changes and explain them beforehand. Exactly the case here...

This whole train of thought here reminds me of http://bugs.mysql.com/bug.php?id=27877 btw. I really am curious which logic book should I study to get to the same place where you guys are.

a change in behavior requested by users (see bug #21099)

Really? Who asked for that? The first time traditional appears is the commit announcement. They asked for some solution. Why not a new mode?

New mode for new semantics sounds reasonable, but with the decision made long time ago and documented it will be just a feature request for 5.6+... Moreover, as 5.5 is GA for a long time already, we may have other users who rely on TRADITIONAL assuming NO_ENGINE_SUBSTITUTION already.

What do you have more, people running 5.5 expecting TRADITIONAL one way or people running 5.0 and 5.1 and expectign TRADITIONAL the other way? Your logic is still flawed on every argument.

I am reopening this until a non-flawed argument comes around which tells me how it's not a bug.

Thank you for taking the time to write to us, but this is not a bug. Please double-check the documentation available at http://dev.mysql.com/doc/ and the instructions on
how to report a bug at http://bugs.mysql.com/how-to-report.php

The hell it is not a bug, you broke working code, what is your definition of bug if not that?

it's intended behaviour, described in the documentation.