Hacker News Reader

Comment by cowmix | Hacker News Reader

VirtualBox has done quite well under Oracle.

MySQL has done well also.

Disclosure: I work on this team. Take a look at new features in 5.7 for example: http://www.thecompletelistoffeatures.com/

Really great to see native JSON support in MYSQL finally.

Also, "Random password generated by default on installation"

This has been a long time coming. I can't tell you how many mysql servers I've cracked into with "root"

This is fantastic. It'd be good see more products list new features (and changes) so nicely, especially those links to the manual. Thanks for the list.

Couldn't help posting it: https://news.ycombinator.com/item?id=12181637

Are MySQL and MariaDB going to diverge or stay parallel?

They started diverging at MySQL 5.5.

I wish the 3rd party tool makers would keep up though - the change in the authentication mechanism in the mysql.user table means that tools like Navicat tend to make a real mess of things.

We had to downgrade a project here from 5.7 back to 5.6 because of the issues that 3rd party tools were having. And we were using latest releases of the tools too!

Or perhaps 5.7 should have some sort of fallback or wrapper functions for tools that insist on doing a SET on the password column?

You can do something like this for doing a cross version approach:

  SELECT /*!50706 authentication_string as */ password FROM mysql.user;

I wrote about the rationale to some of the password changes here: http://www.tocker.ca/2014/09/05/proposed-changes-to-user-man...

That's an awesome list of features. It knocks a few entries out of my list of "Features I wish MySQL had". (Yes, I actually have such a list.)

I am curious as to how features get prioritized, though. I would think certain things would be higher priority, such as supporting a full outer join and ensuring that cascaded deletes/updates fire triggers.

Prioritization is based on a few things (customer demand, trends in the market in general). Besides prioritization, there is a resourcing component too (the server is split into several teams with specific expertise).

In general I would say that we like to understand the use case more than the feature :) If the workaround viability is low, the priority goes up.

Yeah... if the AWS team ported them all to Aurora that would be great.

.... Oracle bought a domain name just to list the new features in a minor version release?

I know MySQL doesn't use Semver at all, and I know the domain probably cost 3/5 of fuck all, but still...

I bought the domain name.

It's not a minor release.

MySQL predates semver. It is not alone here (see Mac OS X; PostgreSQL etc.). But you will be happy to hear that the next version will be called 8.0.

5.7... to 8..

Oh right, Oracle. Like how Java 1.5 became Java 5, eh.

Is >=8 going to be Semver?

6.0 was a canceled release.

7.x. is used by Cluster.

8.0 is the next number along, and makes it easy to follow as "it just drops the 5".

Dude, Oracle is bad but when they bought Sun, they were already on Java 6. The "1.4 to 5" move was all Sun, back in 2004 -- and to be honest, it wasn't even a bad move.

Looking at WHOIS, I think morgo bought the domain as an individual.

The upgrade experience could use some polish.

It seems like every time I fire vbox up locally it's prompting me to upgrade, which is a very clunky experience given the state of other software patching.

This happens to me too but I thought I was going crazy!

It's good to know I'm not alone in this. Also, since I don't remember the exact version I'm currently using it feels like I'm always using an old installation and that I ended up opening the wrong version... well that's some therapist money saved right there, thanks!

I'm pretty sure that the market for VB, like VMWare is being eaten very quickly by lxc-based alternatives like Docker. Probably a bad thing for VB in the medium to long term.

VirtualBox and VMWare do very different things than lxc/Docker.

VirtualBox and VMWare, for example, allow you to run Windows on a Mac, or FreeBSD on Linux, or Linux on Windows. lxc/Docker does none of this, it's simply a different technology with a different purpose.

Yes they are different technologies and have different strengths. Before Docker, a VM was just about the only way to decouple a software environment from specific hardware. Now some people are migrating from a VM to Docker if that's all they need.

I have no idea how big the impact of Docker is on the various VM vendors, but it isn't zero.

Only in the "tech fanboy who doesn't actually understand VMs vs Containers" space, which admittedly is quite a big space right now.

For OS X/Windows hosts the only thing keeping VirtualBox around is the price tag. No one is going to pay for the commercial version with the terrible performance it has.

If you're willing to part with $ you have much better options.

On OS X I think new players like Veertu will further displace VirtualBox even in the $0 category.

Veertu's main obstacle is cross-platform support. The point of providing virtualbox images is that they will work everywhere.

Look at the number of opensource projects distributing Hyper-V images, which has the exact same problem; and the OSX market is orders of magnitude smaller.

I like the Veertu concept and I think it has its advantages, but realistically, it has no chance to encroach on VB's market share.

This is why basically every virtualisation tool has "import from..." functionality

Veertu doesn't seem like a contender in the $0 category either, considering that the website has statements like "with free tier to run Linux VMs". Since no actual price is mentioned I'd assume that it's in the "if you have to ask..." category.

Huh? Veertu Premium is a $40 IAP.

https://itunes.apple.com/app/veertu/id1024069033

Good luck finding that on their website.

They link to their Mac App Store page, which shows the IAP pricing for the upgrade.

Reply on news.ycombinator.com