VirtualBox Development At a Standstill

jones_supa writes: Phoronix notes how it has been a long time since last hearing of any major innovations or improvements to VirtualBox, the virtual machine software managed by Oracle. This comes while VMware is improving its products on all platforms, and KVM, Xen, Virt-Manager, and related Linux virtualization technologies continue to advance as well. Is there any hope left for a revitalized VirtualBox? It has been said that there are only four paid developers left on the VirtualBox team at the company, which is not enough manpower to significantly advance such a complex piece of software. The v4.3 series has been receiving some maintenance updates during the last two years, but that's about it.

Comments

Does It Matter?

[Anonymous Coward]

Legitimate question. I like VirtualBox and have used it for a long time, but as the summary said there are good alternatives available which are improving. As far as I know the only real "killer" feature of virtualBox is its OpenGL acceleration, and we'll probably see that in KVM and friends soon enough. Besides that, VirtualBox basically does what it's supposed to do at this point. Even if it stands still, it'll still be useful for awhile (I know I find no compelling reason to switch right now).

Are there some other core VirtualBox features I'm not aware of that keep people pinned to it? If not, I say let it stagnate and eventually be replaced.

Re:Does It Matter?

[TWX]

I only use vbox for local VMs, like when I need to emulate a Windows machine on my Linux box for some Windows-only software that I have to deal with from time to time. I'm not the VM guy at work, but there are lots of virtualized servers running headless on some big blade systems, does vbox do that or is that pretty much out of its scope?

I agree, for basic workstation stuff it works fine as-is.

Re:Does It Matter?

[kschendel]

I use VirtualBox to host linux and winders VM's on a Mac laptop. It's free, and my other alternatives aren't. All I care about is whether it works, and I'm not all that interested in graphics acceleration and the like. So I hope it sticks around, even if it "stagnates".

Re:Does It Matter?

[CodeReign]

I have vbox running a few hobby servers using its headless mode. But I do this from familiarity and a need for a user friendly cross platform service.

That said it's not a business worthy endeavor as its headless functionality is solid but there are 0 management tools that work WELL with it (phpvirualbox is fine but there are few bugs that cause major issues).

Oracle does have some of its own tools but if you're willing to pay oracle costs you are willing to pay VMWare costs too.

Re:Does It Matter?

[thedbp]

VirtualBox does have a headless mode, which is how I use it. Combine it with phpVirtualBox for a web-based front-end and you can admin from anywhere or any system.

Autostart, autosave, auto-snapshot, etc can be achieved with simple startup and cron scripts.

Re:Does It Matter?

[rwa2]

For basic workstation stuff it's fine.

It's also pretty heavily used for development and test of server deploys. A lot of DevOps types are trying to use VirtualBox to build disposable test clusters for their applications, and has been the default and one of the best supported engines for vagrant.

Unfortunately, a lot of app footprints are starting to rely on deploying other "appliance VMs" in your VM (yo dawg), and VirtualBox is still straggling behind the others on implementing some form of nested VM capability. https://www.virtualbox.org/tic... So it's kinda getting to a point of having a large and growing number of server apps that you won't be able to use VirtualBox to set up a local development and test environment for things that involve, say, using a Stackato PAAS, or a FEO appliance, or an Apigee API gateway appliance, etc. to pick a bunch of essential pieces from recent memory. At least not without a lot of work to host those VMs directly on VirtualBox and not looking or working at all like they would when they hit production.

Re:Does It Matter?

[jythie]

There is something to be said for 'fine as is'. Changes can cause bugs, changes can cause incompatibilities, changes can require updating skills to understand their impact or how configuration has been altered. When all you need is a tool for completing a task without heavy requirements, stable and predictable can be a real selling point.

One of the reasons I like VirtualBox is it changes so little. I have to worry very little about having to look up new things when all I need is a quick drop in solution for something small. Every time I go back to KVM I feel like I have to go find out 'ok, so how does it work NOW?' and then make sure I find documentation and forums talking about the KVM version in relation to the distribution and its version I am using.

Re: Does It Matter?

[AcerbusNoir]

Not many free options for devs on a mac or windows box.

vmware isn't free. And if you use vagrant, you not only need to pay for a vmware license, but also pay for a license to use the vagrant vmware plugin.

Re:Does It Matter?

[MachineShedFred]

Parallels really kinda sucks. Of the three major hypervisors available for OS X, it's the worst of them and that's with VirtualBox being stagnant for a year+. No support for OVAs whatsoever. If you virtualize OS X, you can't use keyboard shortcuts without the hypervisor thinking that Cmd+Q was meant for it, rather than an app in the guest OS. And yes, it doesn't do very nice things with thermal management on your hardware.

VMware Fusion works pretty good, but costs $. VirtualBox, for a time, was actually better than VMware Fusion and free. The guys at VMware have fixed that though.

Re:Does It Matter?

[Andy Dodd]

The only thing I use Vbox for is to perform firmware updates of devices where the manufacturer decided to only allow updating from Windows

Pretty much:
DJI Phantom 2 Vision+
Sony digital cameras

As long as USB passthrough works I'm golden.

Re:Does It Matter?

[Immerman]

The big features for me are the OpenGL and cross-platform support. Perhaps I'm wrong, but I've gotten the sense that Virtualbox is targeted more at the personal / local user, and is fairly mature in it's niche. Occasionally I'll encounter some OS that won't run in it (seems like recently the Mint live CD wouldn't get past th Grub stage, or maybe it was Xubuntu), but by and large it's far more convenient than many of the alternatives. Just migrated a friend's XP system to MS VirtualPC and discovered you can't even mount a folder as a virtual drive - a basic integration feature in most every virtualization/emulation program since... hell at least since the days when emulating C64s and Apple 68000s on the PC was cutting edge.

Honestly at this point is seems like the VirtualBox team has two options - accept that it's mature software in it's niche, and just needs a bit of maintenance here and there to fix the occasional bug and maintain compatibility with evolving OSes, or jump off the deep end and try to compete with VMWare, etc. in the corporate data center. I'm no virtualization expert, but frankly it seems to me that it would be sort of silly doing the latter - data-center virtualization has come a long way since Virtualbox was created - hypervisors, large-scale maintenance, etc. It seems like VBox would be hard pressed to be more than an also-ran in that market.

On the other hand for personal VMs, where compatibility, ease of use, and host-OS integration are of primary importance, I haven't found a better alternative. Though I'll admit I really wish it supported Virtual PC-style undo disks - that's a wonderful feature for experimenting with new/questionable software, especially stuff that may tinker with the OS.

Re:Does It Matter?

[dkman]

I'm not sure what you're asking by "auto deploy" but you can run a Win 8.1 guest or Win 10 preview guest in VirtualBox. Windows 10 preview broke the ability to run VirtualBox inside it after version x.12 (I think it was x.18 or 20 at the time), but that may be resolved now. You could uninstall and install x.12 to run like normal.

I restored my surface back to 8.1 so I'm not sure about later developments.

Re: Does It Matter?

[Guspaz]

How is vmware not free? They have free products for both baremetal and desktop virtualization. vmware player has been able to create new VMs for six years now.

I think the only feature missing from Player that any significant number of people would care about is snapshots.

Re:Does It Matter?

[BillAtHRST]

It would certainly be nice if they fixed the performance of shared folders. That would make it much more practical to run multiple VM's on a single machine. See http://mitchellh.com/comparing... for some interesting info. I tried this myself, and it's true -- read performance on shared folders is many times slower than virtual disks, making them fairly useless.

Re: Does It Matter?

[GlobalEcho]

Not many free options for devs on a mac or windows box. vmware isn't free

How is vmware not free? They have free products for both baremetal and desktop virtualization. vmware player has been able to create new VMs for six years now.

I think the only feature missing from Player that any significant number of people would care about is snapshots.

You are correct for Windows, but VMWare Player does not exist for OSX. They only publish Fusion.

Re:Does It Matter?

[mlts]

VirtualBox has one advantage now, and that is that it is licensed at no charge. On Linux, this isn't a big deal (as KVM and Xen are decent alternatives), but a hypervisor on Windows or OS X, this can be important.

However, if one can choose a non-free solution, the competition has lapped VirtualBox several times. VMWare is extremely strong, both with Workstation on Windows or Linux [1], as well as Fusion on Mac. For a dedicated box with a tier 1 hypervisor, both Hyper-V (can be downloaded separately from Windows) and ESXi are quite useful (although there are limitations without the commercial management tools.)

I've tried various VM products, and the main reason that I chose to just go with VMWare is the universal-ness, and because it is at least a generation past the competition with dealing with RAM overcommits, snapshots, clustering [2], and other features. Plus, if a company sells an appliance, it almost always will be distributed as an .ova file, and other hypervisor architectures come in second. The downside of VMWare is the price... it isn't cheap ($250 for Workstation, ~$70 for Fusion), but it does work well.

Hyper-V isn't bad, as the latest iteration auto-activates Windows VMs sitting on it (no need to worry about a KMS server accessible by all VMs... just the operating system instances running on bare metal). However, usually it is implemented with the full Windows Server OS underneath, making an attack surface, as well as a point of downtime. However, for a Windows shop, the price is right, and it does a good job. VMware is great... but you do pay a king's ransom for the features it brings with it.

[1]: If one needs a home machine to run VMWare stuff on, one might be better off running VMWare Workstation ontop of Linux because ESXi cannot use USB hard drives as backing stores, while VMWare Workstation really doesn't care since it is a type 2 hypervisor and lets the OS handle the disk stuff. Of course, don't expect vMotion or other stuff... but if one wants a dedicated box just for virtual machines, this is a usable alternative.

[2]: Clustering and fault tolerance is brain-dead easy, either using VMFS on a logical drive from a SAN or a NFS backing store.

Re:Does It Matter?

[xeoron]

Install WINE / CrossOver and run the Windows Programs directly. So far every program I have tried works great. I have WINE installed via Macports, but I mainly use it to run Foobar2000 on my Macs.

Re:Does It Matter?

[ncc74656]

Are there some other core VirtualBox features I'm not aware of that keep people pinned to it?

Its support for passing USB devices through to guests is pretty good. I have a Gentoo VM on a Win7 box for the sole purpose of continuing to use a scanner that the manufacturer doesn't support on Win7. The only area where it's let me down in the past was with trying to mess with iPhone firmware (such as for jailbreaking) from a Windows VM on a Linux host...don't know if it was something weird Apple was doing with USB or something else. Have other virtualization options caught up with this?

Also, VirtualBox console windows are less of a hassle to deal with than VMware console windows. Even with their respective guest addons installed and active, VMware is still enough of an annoyance that I'd rather RDP or SSH into the VM in question. (In fairness, VirtualBox is running locally, while the VMware VMs are on a couple of ESXi 5.x boxes accessed through vSphere...maybe their desktop virtualization tools, which I've not used in eons, are better.)

Re: Does It Matter?

[fnj]

So there is a free version of VMware (not on all platforms) that is dumbed down and by your own words not satisfactory for professional users and for some non-professional users, and the "real" VMware product is not free. Contrast with KVM, where the whole kit and kaboodle is free. Seems to me saying VMware is "not free" is roughly just as true as saying it "is free".

Oracle ...

[Anonymous Coward]

Where software goes to die

Re:Oracle ...

[Anonymous Coward]

Oracle is busy converting VirtualBox to run in Java.

Re:Oracle ...

[fnj]

Where software goes to die

Yeah, that is why Java is dead. NOT. Oracle may be destroying it to the best of their ability, but they have been so far unsuccessful. OpenJDK and Google Android are there.

And that is why ZFS is dead. NOT. Oracle has killed off OpenSolaris, severely cutting back on people using ZFS on Oracle products, and seems to have essentially halted all further development of ZFS features, but OpenZFS is flourishing in the form of ZFSonLinux (the real first class kernel driver, not just the FUSE toy), FreeBSD, PC-BSD, FreeNAS, NAS4Free, Illumos, et al. Significant improvements continue to be pursued with OpenZFS.

Once a project has been open sourced, even if further development is then closed by some asshole (I'm looking at you, One Real Asshole Called Larry Ellison), the genie is out of the bottle. You can't put it back in.

4 paid developers yes, but

[Anonymous Coward]

are you unaware that the majority of it is open source? Therefore there's far more than 4 people looking at the code

Re: 4 paid developers yes, but

[Anrego]

Indeed, and some software falls into a realm where you pretty much need paid developers working on it to get anywhere (due to complexity of the code base or lack of interest).

Re: 4 paid developers yes, but

[allquixotic]

This is a little story about four people named Everybody, Somebody, Anybody, and Nobody.
There was an important job to be done and Everybody was sure that Somebody would do it.
Anybody could have done it, but Nobody did it.
Somebody got angry about that because it was Everybody's job.
Everybody thought that Anybody could do it, but Nobody realized that Everybody wouldn't do it.
It ended up that Everybody blamed Somebody when Nobody did what Anybody could have done.

Basically, there needs to be a team of people (whether volunteers, paid employees, or a mix) who are dedicated to spending a specific number of hours explicitly assigned to working on security testing of a piece of software, and then have those hours held accountable. Meaning, if they have no results over a long period of time, or aren't putting in the hours, even if they're just volunteering, then their position on the team should be vacated for someone else willing to do the work.

Features are completely different, and most types of non-security bugs are also different. In general, people implement features because they find it genuinely fun to do so. Also, as long as the software has users, the absence of a feature will not normally cause millions of dollars in damage, loss of reputation, or identity theft. The consequence of the absence of a feature is usually annoyance or inconvenience, but is upper bounded by what that feature would provide if available, rather than being upper bounded by the limits of human cruelty and deviousness, which are MUCH higher bounds than even the most major features.

This is why it's OK to let features develop "organically" in a bazaar fashion. Even bugs can be developed this way: if nobody is encountering the bug, who cares if it's there? And bugs that are encountered frequently will get complained about and/or fixed directly by the core devs or a drive-by patch. Security, on the other hand, almost requires a deliberate, cathedral model to provide any guarantees.

Bringing small aspects of cathedral development philosophy -- the best parts of the cathedral only -- into projects that were once purely "bazaar-only" projects like OpenSSL, can only be a good thing.

Re:4 paid developers yes, but

[ShanghaiBill]

and are you aware the Oracle has a nasty habit of accepting patches and doing nothing with them ...

They did that with MySQL because they had an obvious vested interest in letting it stagnate and die. But what reason do they have to do the same to VirtualBox? Does Oracle have any reason to choke off VirtualBox development?

Re: 4 paid developers yes, but

[Austerity Empowers]

This is a little story about four people named Everybody, Somebody, Anybody, and Nobody.
There was an important job to be done and Everybody was sure that Somebody would do it.
Anybody could have done it, but Nobody did it.
Somebody got angry about that because it was Everybody's job.
Everybody thought that Anybody could do it, but Nobody realized that Everybody wouldn't do it.
It ended up that Everybody blamed Somebody when Nobody did what Anybody could have done.

I love this, I worked for a multibillion dollar gorilla that used this as a management principle. In that management thought that Anybody could do the work, but Nobody they hired could, and Everybody else wouldn't help. Eventually, to fix the dysfunction, they fired Somebody. Somebody told his friends what happened, and Anybody that heard quit. Nobody that was left could do the work, so Everybody was miserable.

The Ballad of Offshoring. It has a happy ending though, the CEO got fired (he's a Somebody that Everybody hated, which Anybody could replace) and things are slowly recovering. The funniest part is 6 months before he got fired he sent out a strange email to the company reminding them he's still the boss. We suspect that HR realized they forgot to tell him, and got around to fixing it shortly thereafter.

If it ain't broke...

[gabereiser]

don't fix it. I mean sure I'd like more features and stuff, but it works out of the box. No tweaking (other than to guest vm's) or anything necessary. It just works. Sure there are other (paid) alternatives out there but VirtualBox does it's job well for me.

Re:If it ain't broke...

[Anrego]

Generally agree. I use it for a handful of Windows apps I still need (like the updater for my GPS) and a few purpose specific Linux installs and it works fine for that. I'll probably keep using it as long as it still works. Worst case, KVM will probably do what I want just as well.

Sure there are other (paid) alternatives out there but VirtualBox does it's job well for me.

KVM is probably the closest alternative and is free (probably more so than VirtualBox is you go all church of Stallman mode).

Re:If it ain't broke...

[jellomizer]

Until the OS's that you want to virtualize will not operate well in it. Then you will need to switch.

Re:If it ain't broke...

[Wycliffe]

Until the OS's that you want to virtualize will not operate well in it. Then you will need to switch.

It's open source and still supported by at least 4 developers so when that time comes it should be simple enough to add support for the new OS.
4 developers seems plenty to support a stable software platform even if there is a new OS every few years that needs to be added.
It's probably not enough to do a major rewrite but that's not really needed at this point. The primary thing I use virtualbox for is to support
legacy OSes. As long as they can add support for new OSes before they become discontinued, virtualbox is fine for my use case.
Also, for my particular use case, because of moore's law, performance isn't a big deal either as by the time an OS is discontinued, the
current cpus are usually an order of magnitude faster than the cpu the OS was designed for so virtualbox is plenty fast.

Re:If it ain't broke...

[ihtoit]

that will be the point where I'll look for an alternative. As for right now, Virtualbox will, for me:

  - boot a native MS-DOS 6.22 image (forget DOSBOX, if you want DOS functionality use fucking DOS!).
  - boot a native Win32 image with complete Win16 compatibility - just like you got in Win9x. Oh hell, I use win9x when I want that kind of functionality. Virtualbox lets me do that.
  - do the above headless and feed a thin client or six, simultaneously, off a commodity desktop system.
  - let you export a disk image to a partition mounted via the host and thereafter, boot said exported image on a completely different piece of hardware with no further hacking required. I'm looking at you, DOSBOX.
  - let you merge snapshots from specified thin clients into the service image while the image is in use.
  - connect one remote session to another remote session from another server and directly collaborate between the two, migrating clipboard and keyboard events as you go, seamlessly between two completely different desktop environments as if you were hosting them both on the local system. Comes in handy on the odd occasion I'm moving bits of user data (eg user lists) between WAMP stacks that for some reason *have* to reside on the system partition and not the segregated data partition.

Re: If it ain't broke...

[gabereiser]

How about "Its working, dont fuck with it"

Wish someone had told Gnome that...

Yeah. I mean I totally buy into innovation for innovation's sake. But VirtBox just works. Sure when a new OS comes out there's work to be done to make it so it will boot in Virtual Box but still. It's worked well for me for years where when I upgrade my OS, VMWare Fusion refuses to work until I pay them (again). I also firmly believe that software which is currently working and working well for most, doesn't need constant attention and "updates" to keep it relevant. It's relevant by working. This is why we still have X11, why we still have Grub, etc. Get off your high horse about not working on something that isn't broken and join us who realize our time is better spent elsewhere unless it isn't.

Re:If it ain't broke...

[Rich0]

don't fix it. I mean sure I'd like more features and stuff, but it works out of the box. No tweaking (other than to guest vm's) or anything necessary. It just works. Sure there are other (paid) alternatives out there but VirtualBox does it's job well for me.

Meh, I abandoned it when it started refusing to run because there was a symlink in the path to its binary. It was less work to just move to virt-manager, which is just a wrapper around KVM which means I'm now running on a fully stock kernel as a bonus. Took a bit of effort to get networking working right, but it wasn't a big deal and the same setup works well for containers also.

Re:If it ain't broke...

[LateArthurDent]

boot a native MS-DOS 6.22 image (forget DOSBOX, if you want DOS functionality use fucking DOS!).

Well, depends on the use case. If you want to ensure your software will run on real DOS, you're right. However, in many cases, DOSBox will work better than native DOS. Run on DOSBox and never worry about not having enough conventional memory!

DOSBox will even let me install Win 3.11 drivers.

boot a native Win32 image with complete Win16 compatibility - just like you got in Win9x. Oh hell, I use win9x when I want that kind of functionality. Virtualbox lets me do that.

That's a good example of lagging development, actually. I have that need, but VirtualBox doesn't have Guest OS Additions for Win9x, which means incredibly slow and awkward performance. VMWare does have guest additions for Win9x, so I tend to use VMWare Player for that use.

do the above headless and feed a thin client or six, simultaneously, off a commodity desktop system.

Yeah, I suppose that's pretty nice. I can't vouch for it, because I haven't used that feature, but it sounds great.

let you export a disk image to a partition mounted via the host and thereafter, boot said exported image on a completely different piece of hardware with no further hacking required. I'm looking at you, DOSBOX.

Huh? DOSBox uses a folder on your box as it's C drive. Just copy that folder over to the new box, and you're done. No need to export or import anything. It's not like DOS has a registry to figure out what's installed, it just has config.sys and autoexec.bat, and whatever folders you installed things at. All of the DOSBox specific settings are really only about what hardware the DOS software sees, it has nothing to do with the host hardware (especially since the settings file now detects the CPU type you have and there's an auto setting for throttling cycles that works reasonably well). So you can copy the DOSBox settings file as well. If you use one of the many frontends, you can have a different configuration file for each game, which is another advantage over native DOS. I remember having an actual DOS Machine with a Turbo button because old games relied on clock cycles for their timing.

let you merge snapshots from specified thin clients into the service image while the image is in use.

Again, sounds impressive.

connect one remote session to another remote session from another server and directly collaborate between the two, migrating clipboard and keyboard events as you go, seamlessly between two completely different desktop environments as if you were hosting them both on the local system. Comes in handy on the odd occasion I'm moving bits of user data (eg user lists) between WAMP stacks that for some reason *have* to reside on the system partition and not the segregated data partition.

Can't vouch for it again, but sounds nice.

Re: If it ain't broke...

[rex.clts]

4013 open bugs against VirtualBox. Some of which make VirtualBox unusable for huge userbases, like the inability to use USB 2.0 devices behind USB 3.0 host controllers.

How about you take a read through that list before getting back on your high horse?

This is no surprise...

[petergriffinismyhero]

Who in their right mind would willingly submit to anything from Oracle? Have you ever been audited by them? Horrible company. They have some great products, but the company itself is a nasty evil entity that thinking people avoid like the plague unless they have absolutely no other choice.

Re:This is no surprise...

[lordmage]

Oracle licensing has some issues that cause smaller businesses to avoid VirtualBox like the plague. One cannot just buy one license for 100 bucks(as implied on website) instead have to pay 5k for a single license the way they work the "license magic". This type of cost for something small and simply pushes smaller companies to spend an extra day or two or more on development and use KVM, Xen, etc.

Oracle is shooting themselves in the foot over a good product.

Re:This is no surprise...

[dissy]

The core virtualbox is open source and free, but the guest tools extension pack is closed and commercial. Under restrictive use cases you are allowed to install the guest tools for personal use for yourself and not need to pay for a license. But even so much as installing it for someone else is a license violation and Oracle expects you to pay for that.

No guest tools extensions means you have no drivers for the guest VM, no shared folders, no mouse/window integration, no accelerated 2d or 3d graphics nor resolutions over 1024x768 vesa.

Whom ever installs the guest tools extension is the ONLY person legally allowed to run that copy of virtualbox afterwards (following the legal agreement when you downloaded it at least.)
If you install virtualbox and the guest extensions on a PC for your mom, mom isn't licensed to run it and Oracle wants a paid license in that case.
Installing virtualbox via scripts including the guest extensions requires a license for each install, even if you are the one using a copy.
(Academic use is somewhat excluded last I saw, but not being in academia I don't know any of those details)

There is an open source version of the guest tools, at least for Linux guests (maybe others by now.)
I'm not sure what features it lacks or differences in the drivers, but they are made by a different development team unrelated to sun/oracle.

VMWare is worth the money

[soft_guy]

After struggling with VirtualBox for a while, I broke down and bought VMWare. I use it for running Linux and running other versions of MacOS X on my Mac. I have found it to be well worth the money. In general, I like free software and I don't mind something that is a little harder to use if the non-free alternative is expensive, but at $79 VM Ware has saved me so much time its well worth it.

Re:VMWare is worth the money

[jedidiah]

VMware has better USB and SATA device support. It requires less resources to run multiple VMs (compared to virtualbox) and more readily supports virtual clusters.

Although I could certainly see how most other desktop VM users would be perfectly satisfied with Virtualbox.

Re:VMWare is worth the money

[Em Adespoton]

Same question from me -- VirtualBox is the basis for many niche solutions, such as http://www.cuckoosandbox.org/. The command-line toolset is great, and better than I've found for other similar products; you can easily do offline analysis of product runs, easily automate running test suites across multiple OSes, create a virtual network of VBox guests, and much more. VMWare does some of this, but is really aimed ad virtualized servers and desktops, not at testing and analysis. KVM could work, but is still maturing and hasn't quire reached the same level yet -- plus, it's nowhere near as portable to any host.

I also like that VBox inherits any improvements made in QEmu.

False story

[Anonymous Coward]

After looking at the release history, I don't see any changes in frequency of releases / updates.
https://www.virtualbox.org/pipermail/vbox-announce/

Add in test builds available showing future bits...
https://www.virtualbox.org/wiki/Testbuilds

Since Oracle spreads it's virtualization bits between products, talking about only VirtualBox paints an incomplete picture.

VirtualBox is akin to VMWare Workstation.
OVM SPARC / x64 is akin to VMWare vSphere (or whatever name they've selected this week).

VirtualBox coupled with kernel-zones and OVM (LDOMs) baked into the SPARC hardware and OVM for x86/x64 platforms - the entire gamut is covered.

Sorry, but Phoronix did not paint a complete picture. How much did they get from EMC for spreading FUD?

VirtualBox has been excellent, but needs QA

[DigitAl56K]

I user VirtualBox all day every day for fairly complex tasks, and it has performed admirably, yet it is sorely in need of QA help. Major releases happen with auto-update notifications and then you realize that your old snapshots can't be started, using a debugger blows up the VM, sometimes snapshots don't save properly even though it looks like they did, etc. etc. Then you have to dig out the last working version, which came out 6 months back, to get up and running again.

Aside from this "upgrade gamble", which I put squarely on a lack of beta releases, VirtualBox is fantastic. Hardware accelerated graphics with full Aero support, fast virtualization, shared clipboard and files, attaching USB devices - it's everything you need in a friendly UI that anyone can work with.

It'll be a tragedy IMO if it's left to rot.

For anyone interested, I find the last stable version to be 4.3.12 (on Windows).

Re:VirtualBox has been excellent, but needs QA

[DigitAl56K]

every software upgrade is a gamble

No. It is usually rare that a minor update version that is an official release will fundamentally stop working altogether. Sure, maybe some quirks are introduced, but generally the product has been tested enough that it is 95%+ working and most users either won't encounter or can work around the deficiencies.

On the other hand, official releases of VirtualBox can just flat out break to the point you can't even start some of your VMs, or crashing the entire VM is just the matter of running some common piece of software. The next release can be months away and when it comes, it may fix your original issue and introduce another equally as crippling to your ability to use the product.

NB: This isn't an attack on the VirtualBox authors, who obviously produce a great product used by many with few resources. But the lack of testing or beta releases literally mean I roll back more than I roll forward - not out of personal preference but because I am forced to just to use the product - and that is what I mean when I speak of the upgrade gamble.

Not at a standstill, just no major features

[tlhIngan]

Funny enough, Oracle updated Vbox with a new release just 2 weeks ago. That doesn't say "standstill" to be, but more "stable and fixing bugs".

Yeah, so what if they're not making big new feature requests? They're still supporting it with updates and bug fixes, and that's a sign of a mature stable product.

It's Oracle, what do you expect?

[ZorinLynx]

That company ruins everything it touches.

Look what happened to MySQL, leading to the need to fork to MariaDB.

Look what happened to ZFS; as soon as Oracle got its grubby mitts on it, it closed-sourced all future updates and made it incompatible with the open source version.

Do you use Solaris? If you do, I don't even have to write anything here. Support has gone absolutely to shit since the acquisition.

And now Virtualbox is stagnant and uncared for.

Why is anyone surprised? Oracle bought Sun and ruined everything awesome about the company. It was the absolute worst possible company that could have acquired Sun, and it shows in every way.

Fuck you, Oracle. With a turbo-charged chainsaw, sideways.

Re:usb3 support

[ihtoit]

there's a workaround that involves adding it as a raw hard disk image (works for UMS only I can confirm, streaming devices such as wireless, ask someone else)

Time for a fork?

[MrWin2kMan]

I've been using VirtualBox extensively as I've been taking some Linux classes at the local community college in an effort to brush up on my skills. I've found VB to be a great way to test out new Windws OS's and applications, and the BSD's as well. It would be a shame for this great, FREE, product to die. Sure, I have access to ESXi, and I could go out and buy VM Workstation, but VB does everything I need it to do. It can be a little difficult finding solutions to issues I have (such as multi-monitor support and screen resolutions, BSD wierdness, etc.) and could defintiely be better supported, but as another post pointed out, Oracle is the place where great software goes to die a lingering death...

Steady progress and under the hood improvements

[gnordli]

4.3 brought major changes in the vt-x code for stability and performance improvements.

You should look at the change log and source code commits.

https://www.virtualbox.org/wik...
https://www.virtualbox.org/tim...

It wouldn't surprise me if 4.4 gets released soon with a new batch of improvements. 4.3 will then get put into maintenance mode and 4.4 because "unstable". I normally don't deploy the current branch in production for several releases as they fix the issues.

Project development is far from a standstill. I don't need any more flashly features. I hope they continue to focus on stability and performance issues.

Re:I use it every day and need it

[drinkypoo]

AFAIK, there isn't any other free alternative for Windows systems,

Well, you get Virtual PC free with XP Mode for Windows 7. But egads, it's crap. Virtualbox is much better.

It moves...

[Doc Hopper]

Disclaimer: The views expressed in this post are my own and do not necessarily reflect the views of Oracle.

TL;DR: I am an Oracle employee. It's an awesome place to work with above market pay, superb benefits, and a demanding but rewarding engineering culture. Virtualbox is one project in a large and growing virtualization team, creating and improving some truly amazing cutting-edge technologies that make your virtualization life better.

I'm going to share some facts as I see them, and let you draw your own conclusions instead of drawing them for you.

1. The Oracle VM and Oracle VM Virtualbox teams are one and the same within Oracle. There's a lot of cross-pollination of ideas and effort, and the virtualization team is frakking huge: HUNDREDS of developers. Not "4", as some have asserted here!
2. There's a ton of stuff happening in virtualization at Oracle: https://blogs.oracle.com/virtu...
3. There's a substantial line-up of products that are demo'd to customers as part of "Virtualbox Appliances". Virtualbox demos are a key strategy for introducing many of our products to customers. http://www.oracle.com/technetw... .
Corrollary: I manage a lot of ZFS appliances. I like them; they make my job easier, particularly at the kind of scale at which one begins measuring one's storage in exabytes. You should download the Virtualbox-based Oracle ZFS Storage Simulator and check it out. Hint: Dig into the REST interfaces and ECMAscript workflows concepts. This kind of thing is Stored Procedures for enterprise-grade storage appliances with absolutely blistering scale, reliability, and performance, and if you don't yet understand how powerful that idea is, you might be insufficiently experienced in high-end storage and databases.
4. Wim Coekaerts is a smart, friendly, and communicative dude. He also happens to be SVP over our Linux & Virtualization efforts. If you're really interested in the details of virtualization development at Oracle, you should check out his blog: https://blogs.oracle.com/wim/

Next, my opinions. No longer facts!

VirtualBox is a mature, stable product that's doing its job and -- as a GPL project -- seems to me like more a vehicle for showcasing Oracle technology than a revenue generator in its own right. That doesn't mean development has ceased! It just means that, in general, Oracle engineering teams are laser-focused on how we can make money so we can stay employed so we can keep creating really unique and useful products for our customers. Responsibilities on teams shift as need demands, and with such an enormous knowledge base in virtualization on our Engineering staff, there's no question that if a product needs a feature to benefit customers, and a good case can be made that it'll pay off, it gets the engineering resources it needs to give it a try.

The Sun transition was tough for some employees. In advance of the merger, a lot of old-timers split. A lot of younger engineers went looking for somewhere hipper and younger to work than what would become a Fortune 500 company. Many Sun managers, sensing the change in the wind as Oracle's intensely results-oriented management team integrated with them, split for positions elsewhere.

I know and work with the survivors of the merger every day. And overwhelmingly, those who've integrated into Oracle culture, shown they belong here through their productivity and attitude, and produce results consistently have built success upon success, and are valued and rewarded.

They're also a bunch of brainiacs who routinely blow my mind with deep insights into operating systems, hardware, and performance optimization.

Those who don't deal well with rapid change, high expectations, and a dogged focu