Category: Enterprise Technology

12
Oct
2020

Microsoft Linux: Maybe Someday

Jack Wallen over at TechRepublic has a new though provoking article out about why Microsoft should replace it’s Windows core with Linux. Basically, the argument goes that if Microsoft is intent on making open source a priority and investing in Windows Subsystem for Linux (WSL), why not go all in and move Windows over to the Linux Kernel? After all, that would unify their cloud strategy, desktop strategy, development strategy, and server strategy around one software stack.

While I think this makes a lot of sense from a IT and infrastructure perspective, it may not be all that easy in practicality. The investment Microsoft has in the Windows Kernel, between desktop and servers, is significant and runs deep. To simply throw that investment away and move to the Linux kernel requires there to be a serious ROI to overcome the decades of sunk cost if that were to happen.

There is an issue of control and the fact that Microsoft likely finds it nice to have complete control over the operating environment. Control allows Microsoft to build complimentary software that is essentially guaranteed to work exactly they way they want it to when running on their OS. Moving to the Linux kernel along with an open source desktop environment on top of it means now Microsoft has to play in the sandbox with many others. Some changes may not be in their best interest as development on these projects continue. Since Microsoft would not have direct control of the project, it won’t be up to them whether or not changes are approved.

Lastly, I continue to take issue with this notion that Linux is somehow the magic security pill that all end-users and organizations need running on their desktops. The reality is that Linux is really no more secure than Windows and an article on Tech Radar by Darren Allan pointed out earlier in 2020. Linux is perceived to be more secure because it is still not widely used outside of IT departments, academic institutions, and software developers. Find me a significant number of people that work outside of a technical field that are running Linux on their desktop then maybe I will change my tune.

Imagine of an additional billion plus devices all started running the Linux kernel. Do you think malware and ransomware authors might take more interest in attacking the operating system? If Microsoft were to create “Microsoft Linux” this is exactly what would happen and the notion of Linux is so secure would start to fall out of favor. All it would prove is that the idea of Linux being more secure was born in a bygone era. An era before Windows had a fully developed file system permissions structure and the ability to restrict certain operations to privileged users.

So while I love the idea of Microsoft moving all of their software and servers to Linux, it just doesn’t seem likely. I think they are very happy developing and integrating the WSL into their current Windows software stack. This let’s them and their customers do whatever they would like in Linux while Microsoft retains control of the core components of the OS. Microsoft can say they love open source software and contribute to the projects they like while not ceding any control of their OS to other developers.

Maybe some day there we will complete harmony among all operating systems. So long as there is a financial incentive to maintain the separately it won’t happen. I also don’t think Microsoft is at a point where they are willing to give control of the kernel and desktop environment to anyone. At that point they might as well keep developing what they have now. Forking the projects means Microsoft will be doing just as much work to maintain the functionality as they are now.

23
Aug
2020

Blockchain: Still Vaporware for Most

Jesse Frederik wrote a nice article over at The Correspondent which sums up what most of us in the technology space have been thinking for a long time. That thought is that blockchain technology is one of the most over-hyped technologies of the past decade or so. While the article is a little light on the technical details of blockchain concepts, its point is valid. Point to a situation outside of a crypto currency where blockchain technology is being used where it could not have been just as easily done by an existing technology. Not only that, but the existing technology likely is higher performing and easier to maintain. Ultimately I think Ehud Gavron over on Slashdot sums up the challenge with blockchain not fitting into most applications well with his comment written in the style of a press release:

Available immediately:
– new database
– stores records forever
– no purging of old records, obsolete records
– guaranteed to grow in size forever
– can’t edit records
– sequential processing with complex calculations so it’s not Order(1) or O(n) or even O(n^x) but a complex polynomial that grows by yet another O(n^y) each time another entry is added
– guaranteed to always get slower over time — it’s the nature of cumulative calculations to verify the data each and every time it’s accessed

Ehud Gavron via Slashdot

Some of the facets of blockchain are quite handy, such as not being able to modify a record once it has been written. Immutable records are very handy when it comes to transactional ledgers or document custody chains. The issues really start to come in when you can’t prune records off the end of the chain, when you need to find more and more systems to be peers to verify the chain, and when the number of transactions being processed hits the millions or billions per day. It no longer makes sense to bother with blockchain, you may as well go back to a tried and true data storage methodology where you can set field level permissions on data, prune data when needed, and not require substantial processing power to verify every transaction.

Don’t get me wrong, I think blockchain has a role to play in the future of data transmission and the management of the chain of custody for electronic records. Being able to track a contract document from creation to full execution where all parties agree it is in the correct state is very valuable. However, many people think that something with “blockchain inside” must be better than something without it baked in. Others, like the town mentioned in Jesse’s article, go so far as to ignore it when the developers of an app try and tell them they are not using blockchain. After all, how could the tout how advanced they are if it is just some old fashioned database application?

In the end, the moral of the story is use the right technology for the problem, not try and make the problem fit the technology. Blockchain isn’t magic, it won’t solve all your problems, and when your technology staff tell you it isn’t needed to solve a business problem, listen to them. When blockchain is the right answer, they will let you know.

13
May
2020

The Work From Home Revolution – COVID Edition

The Verge (alternative source: Buzzfeed News) reports that Twitter is extending its work from home (WFH) allowance “forever” should staff choose to continue to do so. They are the latest technology firm that will transition to a culture that fully embraces working from home. Google also announced that they will allow work from home to continue through the end of 2020 at the very least.

Yes, technology firms are generally the tip of the spear when it comes to adopting forward thinking staffing policies. However, they are a good indicator that there will be a mounting push by staff in other companies and industries to allow for the same type of work location flexibility. What will be interesting to see is how organizations that have historically been resistant to remote work adapt to this new reality. Remote work is no longer seen as a perk and instead it is seen as an expectation by staff. Companies that adapt will attract top-tier talent and retain staff more effectively than those that don’t.

As leaders we must look at our company culture and policies and not dwell in the past. The time is now to change the norms of how and where we work. There has never been a better reason to do.

13
Nov
2019

The End is Nigh! Time to Ditch Windows 7 Now

ITWorld has a very interesting long running series of articles chronicling the slow but steady demise of Windows 7 and the slow but stead rise of Windows 10 in terms of market share. Come January 14th 2020, Windows 7 support will officially end (unless you want to keep paying Microsoft for security updates on a per PC basis) and you will no longer get all of those critical updates needed to keep your organization secure.

What amazes me about the whole process is the prediction by Net Applications that Windows 7 may retain 10+ percent market share well into 2022, long after support has ended and almost every known flaw will be easily exploitable. Don’t get me wrong, I know first hand how painful it can be to update and replace thousands of physical PCs to get rid of an old OS but as hard as that may be, it is well worth it. In my own experience, the reduction in vulnerabilities just from going to a fully patched version of Windows 7 to a fully patched version of Windows 10 will make a world of difference on your audit scorecards.

Please do you and your organization a favor and move to Windows 10 now. You will be happy you did and it will allow you to sleep better at night.

12
Nov
2019

Robotic Process Automation Goes Open Source

If you have had your eyes and ears open at all for the past year or so, you know the new hotness is Robotic Process Automation (RPA) in enterprise IT. Basically that is a really fancy name for a system that mimics a user’s actions on another system so that a person doesn’t have to do it. Truth be told, there have been scheduling and automation platforms around for a long time that have done a lot of what modern RPA solutions are doing. The biggest difference is that the focus is now more about interacting with a GUI versus just focusing on what could already be done through scripting like moving files around.

This week Robocorp and the Robot Framework have been starting to make a splash within the industry as the first organizations looking to take the RPA movement into the open source space and make it more accessible to organizations that don’t want to buy into a major commercial platform or that want to do something more custom with their current tool set.

As a user of commercial RPA technologies currently, the idea of an open source framework and a company looking to make that more accessible to the masses is very exciting. The cost of current RPA solutions is a significant barrier to entry for many smaller organizations and Robocorp has the chance to increase the user base for RPA significantly by making it more cost effective for these smaller organizations. Just knowing that this is coming in the future makes me want to spin up a virtual machine with the Robot Framework running to start playing around. Then when Robocorp has a product ready, I can be primed to pick up and start using their solution.

After all, as their site says:

If you can document it, you can automate it. Never send a human to do a machine’s job.

https://careers.robocorp.com/

That is music to my programming ears 🙂