Tag Archives: fall

​Windows Subsystem for Linux graduates in Windows 10 Fall Creators Update

More Windows 10

Interested in running Linux on Windows 10 with Windows Subsystem for Linux (WSL), but nervous about it being both a beta and only available in Windows 10 developer mode? Your worries are over. In the Windows 10 Fall Creators Update (WinFCU) WSL has graduated to being a Windows 10 feature that can be run by any user.

Tested for over a year, WSL on WinFCU is bringing many new features to this combination of the Linux Bash shell and Windows.

Besides WSL no longer being a beta or requiring users to be in developer mode, the new features include:

  • Install Linux distros via the Windows Store
  • WSL now runs multiple Linux distros
  • WSL comes to Windows Server & Microsoft Azure VMs
  • WSL now supports USB/serial comms
  • Miscellaneous fixes and improvements

Besides Ubuntu, the new WSL-supported Linux distros are SUSE‘s community openSUSE and its corporate SUSE Linux Enterprise Server (SLES). Fedora and other distros will arrive in the store shortly.

If you’ve previously installed WS, your existing “legacy” Ubuntu instance will continue to work, but it’s deprecated. To continue to receive support you should replace it with a new store-delivered instance. Without this, you won’t receive Canonical or Microsoft support.

To keep your old files, you should tar them and copy them to your Windows file system; for example: `/mnt/c/temp/backups` and then copy them back to your new instance.

In addition, instead of jumping through hoops to install Linux on Windows, you can install one or more — yes, you can have multiple distros on a single Windows 10 system — Linux distros from the Windows Store.

To do this, you must first enable the WSL feature in the “Turn Windows Features on or off” dialog and reboot. No, WSL is not active by default and yes, you must reboot.

After rebooting you simply search for “Linux” in the Windows Store, pick a version to install, hit install, and in a few minutes you’re good to go.

If you already have a Bash instance installed on WSL, you can start afresh with the lxrun /uninstall command. You run this command from the command prompt or PowerShell.

Besides being able to install multiple Linux distributions, you can simultaneously run one or more Linux distros. Each distro runs independently of one another. These are neither virtual machines (VMs) nor containers, and that means they need their usual system resources. I, for example, would only want them on systems with at least an additional 2GBs per instance of running WSL.

WSL itself requires only minimal system resources. Rich Turner, Microsoft’s senior program manager of WSL and Windows Console, wrote: “We don’t list [RAM requirements] because, frankly, we don’t have any of note! If you don’t install WSL, we add no RAM footprint. If you do enable WSL, there’s a tiny 850KB driver loaded briefly, and then it shuts down until you start a Linux instance. At that point, you load /init which launches /bin/bash. This causes the 850KB driver to load, and creates Pico Processes for init and bash. So, basically, WSL’s RAM requirements are pretty much whatever the RAM is that you need to run each Linux binary, plus around 1MB of working set in total.”

The Linux distros can also access Windows’ host filesystem, networking stack, etc. That means you should be cautious about changing files on the Windows filesystem.


You can now install Linux distros right from the Windows Store.

Why would you run multiple distros at once? Microsoft points out:

“This ability to run different Linux distros allows you to use the same tools, package manager/ecosystem, and environment that your production code will be running in. This results in less time wasted tracking down hard-to-find errors when it comes time to deploy your code. This allows you to, for example, use Edge/Chrome/Firefox on Windows, to view a website hosted on Apache on Ubuntu, that talks to a REST service running on openSUSE … without having to punch holes through the firewall when testing locally, because all these processes run above the firewall, alongside one another!”

Linux developers will be pleased to find that USB serial comms are now supported. This enables your shell scripts and apps to talk to serial ports.

WSL also now supports mounting of USB-attached storage devices and network shares. That’s the good news, The bad news is it only supports the NT filesystem IO infrastructure. In other words it only supports FAT/FAT32/NTFS formatted storage devices. Want *nix file systems? Microsoft encourages you to upvote and/or comment on the associated UserVoice ask.

Digging deeper into the new improvements, under the hood WSL on WinFCU now includes:

  • Improved TCP socket options inc. IP_OPTIONS, IP_ADD_MEMBERSHIP, IP_MULTICAST, etc
  • /etc/hosts will now inherit entries from the Windows hosts file
  • xattr related syscalls support
  • Fixed several filesystem features and capabilities
  • Improved PTRACE support
  • Improved FUTEX support
  • chsh, which enables you to change shells, now works. This enables you to use your favorite shell directly. Shell startup file other than “.bashrc” will now execute.

The following syscalls were added for the first time during the FCU cycle:

  • Prlimit64
  • getxattr, setxattr, listxattr, removexattr

As expected, WSL is also on its way to Windows Server and to Microsoft Azure Windows VM instances. This will make WSL even more useful for sysadmins.

All these improvements have made it even easier for developers and system administrators to run Linux shell commands on Windows. While this isn’t very useful for ordinary desktop users, for serious IT staff it’s a real step forward, making Windows more useful in a server and cloud world that’s increasingly dominated by Linux. Even on Azure, over a third of VMs are Linux.

With WSL, most Linux shell tools are at your command. These include: apt, ssh, find, grep, awk, sed, gpg, wget, tar, vim, emacs, diff, and patch. You can also run popular open-source programming languages such as python, perl, ruby, php, and gcc. In addition, WSL and Bash supports server programs such as the Apache web-server and Oracle’s MySQL database management system. In other words, you get a capable Linux development environment running on Windows.

While you can run Linux graphical interfaces and programs on WSL, it’s more of a stunt than a practical approach at this time. Of course, with a little work…

How does WSL work? Dustin Kirkland, a member of Canonical’s Ubuntu Product and Strategy executive team, explained: “We’re talking about bit-for-bit, checksum-for-checksum Ubuntu ELF binaries running directly in Windows. [WSL] basically perform real-time translation of Linux syscalls into Windows OS syscalls. Linux geeks can think of it sort of the inverse of ‘WINE‘ — Ubuntu binaries running natively in Windows.”

What matters now is that WSL works very, very well. If you want.

Related Stories:


Unicorns fall short on fighting patent trolls



When it comes to dealing with patent trolls, we’re seeing a big difference between how the tech giants (Apple, Google, etc.) handle the problem and how unicorns do. And it’s the unicorns that are playing a worse hand.

Although about a third of unicorns have been sued for patent infringement, only a couple of them have used a new United States Patent and Trademark Office (USPTO) process called Inter Partes Review (IPR) to deal with patent assertion entities (PAEs), or patent trolls. The technology giants, on the other hand, file hundreds of IPRs a year, most of them against patent trolls. Given that IPRs are often an order of magnitude less expensive than litigation and have a pretty good chance of halting a patent infringement suit in its track, according to the 2015 Patent Disputes report from UnifiedPatents, the fact that unicorns and other private companies have yet to catch on is a little bit surprising.

IPRs have become a go-to weapon against patent trolls, and the unique circumstances that unicorns and other private companies face arguably make IPRs even more attractive. Imagine a unicorn or other private company trying to fight a time-consuming, resource-draining patent infringement lawsuit against a patent troll while also preparing for an IPO!

The new IPR process was created in 2013 so that accused infringers could efficiently challenge the validity of a patent asserted in a lawsuit. An IPR petitioner (e.g., the accused infringer) can ask the court to stay the lawsuit pending the outcome of the IPR, which can pause litigation costs. The new IPR process has seen record growth since its institution, from 514 petitions filed in 2013 to over 1,700 in 2015, according to the UnifiedPatents report.

petitions by year

The report shows that in 2015, nearly 25 percent of all patent cases filed were IPRs. And about 40 percent of the IPRs filed in 2015 were directed at patent trolls. IPRs have quickly become a favored strategy among technology giants like Apple, Google, Samsung, Microsoft, Intel, and IBM because IPRs are often cheaper, faster, and more effective than litigation, according to the AIPLA 2015 Report of the Economic Survey.

Apple, for example, has filed 184 IPRs. The IPR process can be much less expensive (e.g., $ 300,000 v. $ 3 million) and at least twice as fast (an average duration of about 17 months from filing to final decision v. 2.5 to 5 years in court) as litigation, per the AIPLA report. In addition, for IPR petitions that are granted, the win rate is above 80 percent. Based on the above results, the Patent Trial and Appeal Board (PTAB) of the USPTO has become the second busiest patent docket in the country.

top venues

However, unexpectedly, unicorns appear slow to adopt IPRs as part of their patent defense strategy. To date, only two out of 24 unicorns sued by patent trolls have responded by filing an IPR. Those two unicorns have filed four IPRs against patent trolls out of 72 actions, for a ratio of one IPR for every 18 patent troll lawsuits (1:18), while the overall ratio is about 1:5 as of January 2016, according to USPTO Patent Trial and Appeal Board data.

Some possible reasons for unicorns’ slow adoption of IPRs are their technical nature, the complexity of the process, and unfamiliarity. IPRs differ considerably from litigation; the rules and procedures are more like other USPTO proceedings and often focus on specific technology details.

[Editor’s note: A version of this story originally appeared on Law360.]

Jeff Toler is the founder of Toler Law Group, PC and a registered patent attorney with over 20 years of experience in complex intellectual property matters. You can reach him at [email protected].

Robert Paladino is an attorney at Toler Law Group, PC.

Get more stories like this:  twitter  facebook