Tag Archive for linux

Bashing Bash

Ubuntu on Windows
Coder Radio listeners who already caught Monday’s show will know that I have been playing around with Bash on Windows. To be specific I was using the Ubuntu “app” available on the Windows Store and made not modifications to it or my system.

The experience of just getting it installed is pretty terrible. You have to join the Windows Insider program and after you’ve done that restart your computer a number of times until Windows Update notices that you’re in the program and picks up the required update. There’s no direct command line script you can for this you just restart, restart, and restart… until… eventually… it works. Having spent a lot of time in UNIX-like environments, I’ve gotten used to things be a little more precise than “restart and pray.”

Once it got the Windows Insider issue sorted out and installed the Ubuntu “app” from the Windows Store, I started messing around with the BASH interface. On the whole, it functions like a real BASH interface and the commands you’ve been using on Linux / macOS will work just fine. However, that was pretty much the end of my enjoying the experience. Like many things in technology my primary issue is one of expectations. I expected to simply be able to open a BASH window that would start in my home directory and be able to interoperate easily with the local Windows file system. This is not how BASH on Windows works. The Linux file system is separated into an obscure section of the hard disk, making even the simplest tasks of working between the Linux systems and Windows needlessly challenging. From a practical perspective what I wanted to be able to do was edit some code using Visual Studio Code somewhere in my ‘Documents’ directory and run BASH commands against that code easily.

Because I am doing a lot of work that crosses into the legacy Windows world, I really wanted this to work and find it particularly disappointing because other Microsoft tools (Azure, VS Code, and Typescript) have proven to be surprisingly useful for my latest efforts. It would have saved me a ton of headaches. However, it’s just not ready for prime time yet. Ultimately, if you’re working in a Windows environment you’re still better off with just learning Powershell or using something like Cygwin. Let me know what you think on Twitter or in the comments. Also, heard of Docker Compose but not sure what it’s all about? Checkout my quick explainer here.

Pop!_OS First!_LOOK

Pop!_OS
After about three weeks traveling technologically marooned on macOS island, I made a decision to try a new operating system – Pop!_OS. Yes, I’ve been using Ubuntu Gnome for the majority of my home office computing but due to some odd proprietary VPN requirements that only work on macOS and Windows, I’ve defaulted to traveling with my MacBook Pro. No more! From now on my System 76 Lemur will be dual-booting Windows 10 and Pop!_OS with Pop as the primary operating system from most of my development work. It’s been a few days and I have some initial impressions of Pop.

At first glance, Pop looks bright and modern, blending the simplicity of WebOS with the Material Design aesthetic of Android. Long time Ubuntu users will notice some similarities to the soon to be discontinued Unity user interface from Canonical. On the whole, I like the bright aesthetic. For the most part System 76 has done a good job of creating icons for common applications that it’s target market of “makers”, however, it is very easy to find icons that do not match the visual design language and they stick out like a sore thumb.
enter image description here
I particularly appreciated that my editor of choice, Visual Studio Code, and some of the more common text editors used by developers have on-brand icons that fit into the overall system well.
enter image description here

From a practical perspective Pop is little more than a flavor of Ubuntu. That may sounds like a dig, but it’s actually one of the best advantages of Pop, since that means it has access to all the Ubuntu repositories and is compatible with all Ubuntu software packages. This was actually very helpful to me, since it allowed all of my system bootstrap scripts to run unmodified to setup my new install.

So far, my usage of Pop has been pretty smooth. It seems to be fast and snappy. The one point of UX annoyance is the lock screen; it’s just the Gnome one and the requirement to slide up to login is annoying. I’d like to see a customized lock screen that was more on brand with the design and didn’t require that extra step.

It’s important to note that Pop is not yet a fully released product, so a lot could change / mature in the coming months. I definitely like the direction it’s going in but I do think that if it’s going to be branded a OS for “makers” there should be something like per-configuration profiles / setups similar to what Dell does on the Sputnik laptops; the idea there is that you’d have profiles for different types of “makers” that automatically installs industry standard FOSS tools for them. Maybe that’s not something advanced users would use much, but it would make provisioning a small shop with Pop much easier. Also, they should just brand it “Pop”, not Pop!_OS as the current name is confusing.

Want to learn more about what I’m doing in the AI / bot space? Follow me on Twitter.

Why I’m Selling My MacBook Pro – Focus

In a word – “focus.” There are a lot of cool technologies available to developers to today and the truth is that, I’ve been spending a lot of time chasing a lot of different albeit very interesting technologies and trying to figure out what makes sense for myself and for Buccaneer. Here’s just a brief list of the things that I’ve found incredibly interesting technology-wise over the last eighteen months:  Docker, Swift, Linux, iOS development, Android development, Arduino, 3D Printing, DevOps, Angular, React, and about a thousand other things. All of them are very cool, but there’s not a lot of depth one developer can get in any given technology if he or she is focusing on more than one or two of them at a time.

So what does this have to do with selling a Mac. Well, I spent years writing iOS apps in Objective-C and a significantly smaller amount of time writing them in Swift and that was fun for a long time, but now Buccaneer and I have moved on to the exciting world of Containerization via Docker and the wider DevOps space.

For a time, I was trying to juggle these two priorities but what I found was that if you have two messages, then all you’re really messaging is a cacophony of white noise. Another aspect of this is looking into the future based on the current tech trends. While there still are some advantages to native apps over hybrid apps, most enterprise customers are correctly focusing on hybrid or web-powered apps or those who have a little more tolerance for a performance hit are skating to where the puck is likely going in the form of progress web apps. Enterprises are really deciding between hybrid apps with tools like React or Ionic or full on progressive web apps using something like Angular or Google’s Polymer. The reasons for this tend to be the usual development and maintenance costs arguments but also that most of the complexity in enterprise systems tends to be on the back-end rather than the client-side.

My solution is extremely simple — to just skate to where the puck is going and that’s toward thinner clients with complicated back-end systems. These back-ends will be hard to maintain and that’s where containerization provides value and I find that sort of work is best done on a Linux workstation, so I’m going Linux 100%.

Mac Exodus Over?

Many commentators myself included have been making some hay out of the trend of developers and other pros moving away from Apple’s macOS in favor of various (usually Ubuntu) distributions of Linux. Vendors like Dell and System76 have seen gains in the professional workstation market against the less then well received MacBook Pro, but Apple is waking up and smelling the professional angst. Apple’s pronouncements in favor of professional computing on macOS and the promise of a revised MacBook Pro as well as a re-designed Mac Pro with a more “modular” design. We’re already seeing the so called Mac Exodus being blunted by Apple’s announcement. The questions becomes less a contest of Linux vs macOS quality and more a race against the inevitable tide of macOS’s professional resurgence. The overall goal for Dell and System76 should be to gain as much market share in the professional workstation space before Apple actually launches new hardware for that market. To that end, I’m going to play “CEO for a day” of Dell and System76 and game out a strategy for both of them respectively. I’m picking on these two firms, because I like them and also feel like they have the best shot of actually being successful.

Dell has money. Lots and lots of money. That’s great but also can lead to conservatism. Their success with the Sputnik project was one of the early and most successful ventures of a major desktop manufacturer into the Linux space. The product it produced – the XPS 13 Developer Edition – is still one of the most compelling Ubuntu laptops available. Dell needs to widen their Ubuntu product-line to include larger higher power models as well as something more akin to the MacBook Air. There will be an R&D / product development cost to this, but it’s going to be worth spending. The other key here is that Dell has a huge asset that System76 won’t – it controls its own production pipeline and has the manufacture of PCs down to a science. That should lead to better yield over competitors which at any reasonable volume means there are some margins to play with there. Dell should cut these margins on select base models of Ubuntu Linux workstations to the bone, nearly selling them at cost. This will make a dramatic cost comparison against Apple, given their already high prices and should also make Dell a very attractive supplier to creative agencies and the like as they look to cut costs in an increasingly competitive environment. Remember, the goal here is to gain market share fast and hopefully create career spanning Linux customers who otherwise would have gone to Apple.

System76 doesn’t have Dell money but it has something else focus. In many ways, they’re already taking the right steps to up their hardware game by moving away from Clevo and Sager hardware and toward producing their own, but more can be done. My expectation is that within the next eighteen months we are going to see more Apple quality hardware from them once their production lines and processes are fully up and running. Sadly, some of it is going to come at a greater cost than money. System76 has good relations with the Linux community and in particular the Ubuntu community. Canonical (the company behind Ubuntu), in what can be described charitably as a pivot toward reality, is dropping its Unity desktop user interface in favor of GNOME and seems to be more focused on IOT and “cloud computing” than the desktop. This makes sense, given that Canonical has limited resources and needs to make real money somehow, someday, someway. The folks at System76 who I’ve met and like very much need to find a way to show leadership in the community by guiding it into a direction that strengthens the Ubuntu desktop as the leading choice for professional workstations. The key here is to lead the community in the right direction but resist the temptation to commit too much of their own limited development resources to the effort. I know what I am suggesting is less being a good community citizen and more leveraging the community, but the reality is that the Linux community has been wasting development resources on alternatives to alternatives for things like package management and window managers — strong leadership could finally close some of these questions and focus the communities efforts.

This is a race against the clock and make no mistake, the window is closing quickly. If Linux workstation vendors such as Dell and System76 can’t make significant gains in market share quickly, then this whole “Mac Exodus” will be little more than a blip in the history of Apple’s domination of the modern professional workstation market. If you have any questions or comments, Tweet me and please checkout my Youtube channel where I offer Docker and DevOps tips.

Lemur Review

Coder Radio listeners will know that I have been agonizing over replacing my three-old MacBook Pro and with the recent Apple announcement of their new MacBook, I ended up ordering a System 76 Lemur laptop with Ubuntu 16.10 Linux pre-installed. If you’re interested in the specific specs of the system, you can see them here. My thoughts in the future of computing (or the next “big thing” if you prefer) being AI powered by Linux on the cloud. My crazy predictions about the future of AI and the stages it will go in are a post for another day – for today, let’s go over the Lemur itself as it compares to the MacBook Pro it’s replacing.

The general hardware is good but a little more plastic feeling than I am familiar with. I like the inclusion of a USB-C port but do wonder who is still using VGA and if that space on the body would not have been better used for another USB port. I was pleased that my high-res Dell monitor and peripherals all worked out of the box with the Lemur – that’s a big win for System 76, as device compatibility concerns (real or imagined) tends to be a main thing that keeps would-be-switchers from going Linux. The general “just works” quality of the Lemur with Ubuntu 16.10 is by far the most impressive aspect of my experience with it.

On average, I am getting about 4-5 hours of battery life with the screen just about all the way turned up and playing / streaming music or podcasts basically constantly. That’s less than ideal, but I am confident, I could bleed an hour so more out of it if I did less streaming and dimmed the screen some.

The matte screen on the device is surprisingly good and has helped to bring me around on matte screens in general. I might have liked a more “retina” screen but the 1080 resolution is more than fine for my needs.

The keyboard is a bit of step down from the MacBook Pro keyboard that I am used to. There is far less key travel and at times it feels a bit too insubstantial to type on, however, it’s about on par with most non-Apple keyboards on windows laptops with the possible exception of the current Dell XPS 13 and is by no means the worst laptop keyboard that I’ve used.

The track-pad is just bad. Even compared to other non-Apple laptops, the track-pad doesn’t cut the mustard. It at times feels “jumpy” and at other times feels slow and unresponsive. It’s also awkward feeling on my fingers but that’s probably a bit too subjective to worry about for most. I’ve taken to keeping a USB mouse in my bag with the Lemur.

The sound from the on-board speakers is adequate but leaves a bit to be desired when compared to the MacBook Pro speakers. Neither is very good, so there’s not a lot to worry about here – in general, if you care about audio like I do you’re probably always disappointed with laptop speakers. The most glaring issue here is that at high volumes the sound becomes tinny.

All in all, I think the Lemur is a good choice for someone looking for a Linux laptop that could be great if just a little more care was taken with the track-pad and keyboard, as the other minor points I’ve made here are probably specific to my tastes / expectations.

Comments? Questions? Forest burning FOSS rage? Hit me up on Twitter.

Linux Adventure Pt 3 Skylake Hell

My adventure in the world of Linux continues and I’ve come to the point where I have purchased a Dell XPS 13 non touch and went ahead and installed Ubuntu 16.04 on it. It’s important to note that I did not spring for the Dell Developer Edition, since there was a sale on the regular (read Windows 10) version of the machine and since my intention was to immediately upgrade to 16.04 once I got the machine.

My first day working with Mate, things were great, but there was a lurking problem hidden by the fact that I was running my XPS to an external monitor that entire day. You see the XPS model that I have has the newest Skylake hotness in it. Unfortunately, Skylake has been problematic in terms of screen flickering issues on Windows 10 due to driver issues – the Linux situation is just as bad if not work. The screen flickers every few seconds and is basically not usable unless connected to an external monitor.

Because of this issue, I was forced to use the recovery tool provided by Dell to reset my machine back to Windows 10 and am using it with Windows without issue. This is a pretty disappointing problem but is not necessarily the end of my adventure. My plan is to simply wait until the community or Intel provides a driver that resolves the issue. In the meantime, I’ll be checking out Windows 10 as a development environment and try out that BASH on Windows stuff.

Let me know what you think on Twitter.

Linux Adventure Pt 2: Ubuntu Apps

UbuntuMy Linux adventure continues on my modest Dell workstation. I’m pleased to say that so far things are going very well and Ubuntu continues to bring new life to my otherwise underpowered workstation. After getting over a few hurdles, what’s really impressive about my experience working on Ubuntu daily is how uneventful it is. Still, there’s always some room for improvement and the most glaring pain point is the lack of decent apps available for the operating system. Ubuntu just doesn’t have a good app ecosystem compared to MacOS and the Software Center is little more than an embarrassment.

Developer Interest: The simple and most basic cause of this is that there just aren’t many apps available, since developers don’t see Ubuntu as a platform worth developing for. Unfortunately, that’s probably true to a point. A simple Google search for developers considering moving their app project over from MacOS or Windows to Ubuntu, doesn’t yielding very heartening results. There also is something of (what I believe to be a misconception) among some developers where they believe that Ubuntu users are unlikely to purchase software.

App Distribution: Canonical, the developer of Ubuntu, released the Software Center several years ago with the hopes that it would become the equivalent of the App Store on MacOS. Unfortunately, the Software Center was poorly implemented and little to no effort was made to draw developers to the platform. Failing the Software Center, developers are left to their own devices for delivering their apps and there’s little standardization on Ubuntu or Linux as a whole for that matter when it comes to the easy distribution and installation of GUI apps.

The advantage of Ubuntu and Linux operating systems in general is that there are steps that the community can take to resolve issues on the platform. For instance, the community could develop an open-source alternative to the Software Center and encourage its adoption. Of course, Canonical could accelerate the process by throwing their development and financial weight behind such an effort and making a clearer statement about where the platform is headed.

Let me know what you think? Do you see Ubuntu as a viable development platform? Reach out to me in the comments below or on Twitter.

 

UPDATE: I have been made aware that the Software Center launched before the Mac App Store. I appreciate the correction. This only makes Canonical’s failure deeper, since they’ve had more time to work this out. Maybe the GNOME store will be better but I don’t think being first is in any way valuable in terms of being a developer and considering developing commercial software on the platform.  

Linux Adventure pt 1: First Look

Coder Radio listeners will know that I started using an Ubuntu workstation a few weeks ago for over about half of my general development work. While it’s true that I can’t do any native iOS or MacOS development on my workstation the majority of my current work tends to fall into one of the following technologies: native Android (Java), Ruby on Rails development (Ruby), Ionic development (JavaScript). Due to the death of my Macbook last week, I’ve spent about a week 100% on Linux at work and it’s really given me some perspective into how much things have changed on desktop Linux since I last used it seriously sometime around 2009.

The machine I’m running is Ubuntu 15.10 on a slim Dell tower with no graphics card and 8GB of Ram – it should be noted that it started with 4GB but I found that I had a spare stick lying around, so I went ahead and pushed it to eight. Overall, the performance has been phenomenal and there have been no obvious speed or animation issues in the Unity desktop.

However, the general UX has been a little more problematic. Many applications have small visual bugs in Unity under the default theme and rounded corners in particular proved problematic, leaving a dark triangle near the edges of the application’s window. I ended up solving this problem by using the GTK Arc-Dark theme. There was also an issue where the cursor kept getting stuck on the loading icon that I had to solve using the GTK Tweak tool.

In terms of actually getting work done, things are going fairly well. Installing my Android and Rails development tool-chains was a breeze with the exception of Postgres, but Postgres tends to be terrible on MacOS as well, so I don’t hold that against Ubuntu. Once I really got down to coding, my JetBrains tools were just like their MacOS versions and Git was well Git.

On the whole, I like the workflow that I’m developing on Ubuntu and am going to stick with it for a while. Another striking aspect of the changeover was how much of my workflow is Chrome dependent rather than being operating system dependent – ie I find myself using a wide variety of web (and particularly Chrome) based software. I expect that my usage of it will evolve as time goes on and as I find more sophisticated Ubuntu applications that fit and expand my workflow.

If you’d like to follow my continued Linux usage and get other development / tech insights, please follow on Twitter.

2014 Predictions

I hope that you have all had an awesome New Year’s Eve! I’ve decided that I’ll start off 2014 by making a fool of myself by trying to make some predictions for the tech industry for the new year. This is not a what you will see list, but nor is it a what you won’t see one; in reality, I am trying to focus more on trends than anything else. Overall, 2014 is looking to be a transition year rather than a real game-changer. This is in no way a bad thing and makes sense for where we are in what is usually a twenty year tech cycle; it is important to remember that the mobile revolution is not even half way done and there are still a lot of incremental advancements that need to be made in that area before it can be considered complete. Still, this article will not focus on mobile exclusively but will rather jump around with no other aim than what I find interesting.

Docker: Docker is the darling of the developer community right now and for good reason — it solves a problem that (outside of the BSD community) hasn’t really been properly addressed. There is however a risk of certain segments of the community drinking a little too much of the Koop-Aide and using Docker in ways that it wasn’t intended to be used; just think about what we saw with Rails a few years ago and the hype surrounding that and you will have a good idea of what I am concerned with. Still, at the end of the day Dockery is going to be a major tool in a lot developers’ (including this one’s) toolboxes for 2014 and probably beyond.

Windows 8: In the consumer market, Windows 8’s RT offering is in a lot of trouble — that is undeniable at this point. Windows Phone suffers much the same fate as Windows RT, though Windows Phone does enjoy a good holding in South America and some other parts of the world. Microsoft has already hinted strongly that they plan to merge WinRT and Windows Phone into one mobile operating system alla iOS. This is a great Idea but is several years too late. It also undermines the “one Windows” pitch that Microsoft has been making for Windows 8 over the last few years. In 2014, Microsoft will still be dealing with the fallout of their bumbling launch and marketing of Windows 8.

Azure: Azure has grown far beyond Windows in the cloud and at the close of 2013 is a rival to Amazon Web Services and pretty much every other cloud offering. I’ve used Azure several times myself over the year and am pretty impressed on the whole; there were some bumps in the beginning and middle of the year, but these have largely been addressed and it looks like Microsoft has some ambitious plans for Azure in 2014. Over the last month or so something called Midori has been leaking out of Microsoft and, though the pundits seem to think it is something to do with Windows on the client-side, my bet is that this is some sort of evolution of Microsoft’s cloud offerings. Either way, 2014 is going to be a good year for Microsoft in the cloud and for Azure.

Mac OS X: Twelve months ago it looked like OS X was veering dangerously toward an iOSification that would have proven intolerable for professional users. With Mavericks however Apple has found a good balance between their desire for control and the reality that the pro-market has been driven to OS X due to its being a UNIX system that has a late vendor for support and an attractive user interface. Despite the apparent back peddling, it is important to note that Apple has gotten one major change in the OS and managed to implement it in a way that is both useful to the average consumer and acceptable to the pro users — this feature is called Gatekeeper. In Mavericks, Gatekeeper does not allowed unsigned applications (to sign an application one needs to be an approved Apple developer) to be installed on the system by default. The key is that this is a default that any sophisticated user can change. However, I must admit that I have kept this default. Going forward, Apple’s belief in signed applications (or perhaps some slightly watered-down version of it) makes a lot for sense for the future of computing and I actively support refusing to install unsigned applications from untrusted sources. If the current path holds, Apple will be balancing making OS X simpler for average users and new users who came over due to the halo effect of iOS while balancing the needs of the professional user market.

Ubuntu / Linux: Canonical has done an amazing job of sullying the Ubuntu name over the 2013 and has done little more than make a fool of themselves with their naïve attempt at breaking into the mobile space. Ubuntu will still be a very popular desktop Linux operating system among new Linux users and will continue to be a major player on the server-side. Canonical the company however will fail to monetize their offerings in any significant way. The only ray of hope would be some sort of re-focusing of the company to be an enterprise focused organization much like Red Hat. Even in that case, Canonical will not be able to be a true challenger to Red Hat in 2014 and it is unlikely that they will even decide to try. The continued flailing of Canonical will contribute to a “brain drain” of passionate and talented Linux enthusiasts out of the Ubuntu community and into other Linux communities. Another side of effect this is that desktop Linux will continue to be the proverbial tempest in a teapot that it has always been. This internal discord will guarantee that 2014 will not be the year of the Linux desktop in any significant way.

That’s it! Those are my foolish predictions for 2014 — foolish as they may be, I am pretty confident that most of them will be, if not correct, then on the right track. I know they are not earth shattering and basically boil down to 2014 will more or less maintain the status quo. In a way, that’s a good thing. If we are constantly reinventing new technologies and never refining the technologies we are already have, then we will always be using unstable and half baked first generation technology. Have a happy new year and feel free to comment on Twitter.

Fingertip Tech, INC Goes Linux with Spindl

spindl_256x256x32Yesterday my company, Fingertip Tech, INC, acquired Zane Swafford’s Linux application Spindle. Spindl is a great tool for managing and tracking your time. I couldn’t be happier with the application and look forward to an awesome 2.0.

I hope that this acquisition communicates my and my companies commitment to Linux as a platform for desktop software development. I’ll be making more announcements before the end of this year and do follow Fingertip Tech, INC on Twitter to stay in the loop.