I had to replace my 2015 MBP last week, because the kb started failing (it wasn't one of the ones eligible for a free repair, sadly, and I can't live without a machine for however long it takes them to fix it anyway). I was really torn, because although I could reluctantly live without the function keys, the escape key is a complete dealbreaker.
After much debate I decided to take a chance on a surface book. I figured I'd try it, try and get my dev environment in a decent place, and see how I felt about windows after 30 days, with the idea that I might have to return it if things didn't work out. I'm 6 days in, and I can pretty safely say there's no way I'm going back to apple. Windows has managed not to suck at all, my dev environment is arguably better than what I had before, and I'm absolutely loving the machine itself. Most joyful new laptop experience in many years.
Not my experience. I have to use windows at work. Got ubuntu running in a vm. Buggy as shit.
- sharing screen in hangouts puts chrome on every virtual desktop so now you can't click on chrome to navigate to that desktop
- slack crashes silently every 24 hours
- virtual desktops is a mess, if one window overflows into another desktop, then when you click on that app ubuntu will take you to the desktop where you see the least amount of your application, usually just a few pixels on the top of the screen
- copy-paste sucks balls, in some applications it's ctrl-c in others its shift-ctrl-c and some applications will copy to clipboard if you select a string
- keyboard support is terrible, in intellij i can't type `or ´ or ~ but I can type all those in gedit and then i have to copy paste them over
- ctrl-a and ctrl-e don't work in most applications, even though these are old school shortcuts that work nearly everywhere on a mac
- installing applications is always terrible on ubuntu. You can install from their app-store, if that works (50/50 odds) but you can't get intellij there, or sublime. To install sublime I had to find a blog post with 4-5 steps I needed to do. For intellij I also found a blogpost, but when I need to upgrade intellij I couldn't find it anymore and had to delete everything inside my intellij folder and copy a new intellij in.
> installing applications is always terrible on ubuntu.
there are three main was of "installing applications on ubuntu":
1. main software repositories; this is essentially 1-click install (if you go through the software management application)
2. installing from a PPA (e.g. Sublime); this is 1-line copy/execute (add-apt-repository XYZ), then update (apt update) and 1-click install
3. rarely, you install binaries directly; this is a double click.
none of them is terrible. the vast majority of the Ubuntu/Debian linux software is provided and installed via 1. (update obsessed users use 2.), which is a much faster workflow than windows, and immensely safer.
If a software is fidgety in a version 3 installer, it has nothing to do with Linux, as it's almost exactly the counterpart of the typical windows installation, it's just badly packaged by the developer; binary installers are a small minority anyway.
> copy-paste sucks balls, in some applications it's ctrl-c in others its shift-ctrl-c and some applications will copy to clipboard if you select a string
this is not correct. Ctrl+C/V are the standard, the exception (not "some applications") being the terminal, because Ctrl+C has different semantics.
> keyboard support is terrible, in intellij i can't type `or ´ or ~ but I can type all those in gedit and then i have to copy paste them over
you may have a misconfigured keyboard. run the live version of ubuntu, test the keyboard with the installer, then configure the same on your installed system.
>> copy-paste sucks balls, in some applications it's ctrl-c in others its shift-ctrl-c and some applications will copy to clipboard if you select a string
> this is not correct. Ctrl+C/V are the standard, the exception (not "some applications") being the terminal, because Ctrl+C has different semantics.
I agree that it sucks (in a CMD shell). It's the reason I switched to a Mac when I became a developer. Maybe it's come a long way. I've heard good things about PowerShell, too, but honestly haven't played with it much.
As far as the Linux subsystem, I wouldn't use it for anything serious. It choked the first time I tried installing Elixir.
I'd probably rely on a Vagrant VM running in VirtualBox (or whatever VM host makes the most sense on Windows).
When using the "Ubuntu Software" application I am shown results for snaps, as well as results from the apt repos. Creates a bit of confusion as to which one should be installed, even more so when version numbers vary between the two.
I use snaps for isolated software that I want more or less continuous stable updates for, apt for system components that'll stay on the same major version until I upgrade the OS.
I managed to install docker according to docker's own tutorial. Then 4 months later I get an update from ubuntu to upgrade my docker. I thought, that sounds about right....why not??? So after the "upgrade" I now had two different versions of docker installed at the same time in two different locations and neither of them wanted to start. That was a great start to my day and cost me an hour or so to fix.
I was a windows guy for many years before switching to OSX in 2006 (I actually worked for microsoft in the early 90's, but we all hated windows as much as the rest of the world, maybe moreso, because we had to develop on weekly builds). Therefore I have seen windows suck in a wide variety of ways, hence my desire to try it for 30 days and return it if needs be. In all my years in the windows world. I've never seen an app use anything other than ctrl-c/ctrl-v for copy paste (ok, I think you used to be able to do ctrl-ins, shift-ins, but that was optional), but the rest of your experiences sound quite plausible. I haven't run into any of them.
Also, I don't have to restart my wifi 5 times a day anymore, which is kind of a nice win.
EDIT: Oops, I thought these were complaints about windows (highly plausible!). actually it's unix under windows. Apologies.
I know classic Linux defense bullshit but these all aren't typical 'Linux issues'.
I can only guess the desktop you used but I assume we talk about Unity/Wayland. Ubuntu removed Unity meanwhile for the Way cleaner Gnome Shell (however I don't know if they have some custom bullshit in there again)
Shortcuts are often miss configured on system level. I noticed nonsense issues therelike as well on Ubuntu.
There are highly different software repository concepts like pacman and the AUR on arch where you have a package selection right there you could never experience anywhere else.
What I am trying to say is that there is no one Linux, I don't like vanilla Ubuntu. However having used any OS in the recent years I love going back to MY Linux.
Edit:// oh the shortcuts are also likely due to the VM I just realized
He also complains about a Chrome bug, a Slack bug, and an IntilliJ bug. 3rd parties doing no QA and not respecting UX conventions on the most popular distro is a problem, but not really with Linux.
2 out of three are proprietary software that you pay for and all 3 have free alternatives that are better.
He's also running unknown version of Ubuntu, probably old, that he doesn't know how to use properly. Odds are really great he's one of those that pastes unknown terminal commands from blog posts to achieve goals he doesn't quite understand and his install is screwed up in some way that would require hours of troubleshooting or 30 minutes to install over top of.
Its like reading an analysis of Chinese quisine written by a guy who only buys past date 99c frozen dinners.
That CTRL+SHIFT+C thing in Gnome Terminal/Ubuntu (and some others) has driven me nuts for years!!! yes, I get that CTRL+c means something different in the context of shell execution. But it's still maddening.
Middle click to paste is an unrelated thing that drives me up the wall in Linux. Many of the programs I use on a daily basis have specific actions triggered by a middle click that I use often, but if I miss my target and hit a text area instead, it injects whatever I had in my clipboard into the text, often without me realizing it until later. (In one instance, my clipboard contents made it into an email I sent out. Thankfully, it was minor enough that it just looked like a typo, but it could have been much worse.)
Unfortunately, I'm not aware of any way to turn off this behavior.
I have a similar setup at work, albeit Ubuntu VM in a Mac OS host (I don't like Mac OS). A lot of these sound like VM issues. For example - the copy to clipboard if you select a string. That's how your VM is "sharing" the clipboard between the VM and the host. Try highlighting a string in the VM and then hitting paste in a text editor in the host and you'll see what I mean.
> - ctrl-a and ctrl-e don't work in most applications, even though these are old school shortcuts that work nearly everywhere on a mac
The shortcuts for text fields on Windows are not consistent either, are they?
I am very used to the organization of shortcuts on macOS, especially those for text editing. Strangely, that's what I miss on Linux the most now. However, it's probably easier to provide them to users when there is only one mainstream GUI toolkit (and, for instance, in Office for mac some of those shortcuts don't work).
I have 3 different distros running on as many machines all work fine and one is based on ubuntu LTS. Probably what you should be running but whatever.
"- keyboard support is terrible, in intellij i can't type `or ´ or ~ but I can type all those in gedit and then i have to copy paste them over"
Do you really believe people have this issue? Do you think we all have emailed ourselves documents full of all the characters are OS are too broken to recognize? Something is screwed up about how your vm or your vm software is configured that you are blaming on the OS. I have used over a dozen distros in a multitude of versions and none had any problem with hitting a key and having a character appear on the screen.
"virtual desktops is a mess, if one window overflows into another desktop, then when you click on that app ubuntu will take you to the desktop where you see the least amount of your application, usually just a few pixels on the top of the screen"
This sounds sub-optimal but may I suggest you keep windows on one screen. Most environments provide a hotkey to move a window to a certain screen and many provide hotkeys to resize it to either fill the entire screen or some portion thereof like the left/right/top/bottom half or just drag the window all the way onto the screen you intend to use it on.
Personally I prefer the way i3wm handles virtual desktops. They are PER monitor. Optimally you just assign each workspace to a particular output. Example 1-5 on left monitor 6-0 on right and switch a given output and to a given workspace with one click of super or cmd + number.
Want to view your chat app on the right monitor while leaving your browser alone on the left cmd + 7. Its also substantially lightweight which means it would work better on a vm.
The great thing about it is that there isn't ONE way to handle virtual desktops there are a bunch of different environments available one of which probably works optimally for you.
"- ctrl-a and ctrl-e don't work in most applications, even though these are old school shortcuts that work nearly everywhere on a mac"
Many applications are configurable but honestly different environments different conventions this isn't a bug its just a difference.
"- installing applications is always terrible on ubuntu. You can install from their app-store, if that works (50/50 odds) but you can't get intellij there, or sublime. To install sublime I had to find a blog post with 4-5 steps I needed to do. For intellij I also found a blogpost, but when I need to upgrade intellij I couldn't find it anymore and had to delete everything inside my intellij folder and copy a new intellij in."
Installing applications is fantastically easy and works every single time. Most common stuff is available in the repo and you can apt install foo or use the gui if you like.
Most less common stuff is available in a 3rd party repo. If you need such the workflow is google ppa somesoftware note the ppa involved and run
if you get bored and want to automate that in fish
function ppa
sudo add-apt-repository ppa:$argv
sudo apt-get update
if [ $argv[2..-1] ]
sudo apt install $argv[2..-1]
end
end
ppa someaddress somesoftware bam installed
example want newer gimp
ppa otto-kesselgulasch/gimp gimp
When you say it works 50/50 you don't mean that it fails half the time you mean that it installs almost all things just fine save for a few quite frankly mediocre poorly chosen apps that you simultaneously actually pay money for but don't get good support for your environment.
This is probably because your environment is some 7 year old version of ubuntu that is similar to what your servers are running. Seriously the last time I installed intelij I just had to unzip an archive, add to path, and tell it where to get java. This "Install" in /opt actually survived multiple distros and worked without complaint.
Shockingly the instructions you found on the internet probably didn't cover the case of running new sofware on an obsolete base.
Ubuntu 16.04 LTS is what I'm running. Set it up about 9 months ago. Nothing special, haven't configured much expect hosts and the number of virtual desktops (which requires googling and special terminal commands to do, why is there no setting for this in the UI when it's a UI feature???)
Then I guess it runs Unity and when it comes to Unity I'll personally agree with some of the criticism:
Ubuntu Unity was always too opinionated and lacked the necessary UX testing IMO. (Yep, they seemed to have tested it, but not on enough power users; some power users (a lot of them around here it seems) love Unity. Others like me find it a producrivity disaster, probably for some of the same reasons as you found.
That said, a lot of the rest seems like a mix of misunderstandings and broken config either in the host or in the VM.
the only viable linux platform remaining is the one loosely organised and licensed around the world called "Android". with recent advances in ARM64 chips and extremely fast processors some OEM's are releasing desktop modes for android. Its truly the final frontier of linux, hopefully this idea can aleviate those issues you've had with linux. Applications work consistently on Android although there is still the app gap of stuff not yet ported onto google play store.
I was extremely unsatisfied with the promise and then the delivery of my surface pro, ended up selling it to a friend with lighter computing needs. The touch bar macbook seems pretty useful itself considering how insanely good its IO is (4 thunderbolt ports!!!). I haven't been able to use it as much and im waiting for better eGPU options to come out.
in the mean time I'm stuck just like you in the world of linux. My primary desktop pc for the past 9 months was my galaxy note 8 with a usb c to hdmi adapter. it has an incredible chromebook like UI when docked and has all the linux tools i need! I've even been able to write an app and submit it to the app store all from there. You really cant go back after you've tried one device life and it makes filming and editing content for my youtube channel seamless since video never leaves the 256gb SD card inside my phone.
Huawei has a similar mode and i cant wait for google to support it. they do natively support windowed mode and moving apps freely but i mean just add a chromebook ui to the screen when u dock a pixel 3 and you will be golden :)
> The touch bar macbook seems pretty useful itself considering how insanely good its IO is (4 thunderbolt ports!!!)
Can a machine have "insanely good IO" if dongles are required for everything from iPhones to hard drives to memory cards? I could agree that a machine with 4 thunderbolt ports on top of the regular complement of power, USB-A, and SD slots would be insanely great. But I'm living the dongle life and not loving it.
A machine can have "insanely good IO" if you're not judging it in the context of older hardware, yes.
Four thunderbolt ports is great. One adapter to plug in all of the things you mentioned is also great. iPhones and hard drives don't need dongles, they just need cables. A USB-C to USB-B 3.1 cable is $1.33 from Monoprice; USB 2.0 version is just $1.12. Not too shabby.
For iPhones: the only times I've had to plug my iPhone into my laptop is doing development, but a cable for that only costs $20.
But let's compare it to the previous (2015) MacBook Pro: two USB ports and two Thunderbolt-and-Displayport ports, and HDMI. Five ports total, but no flexibility. Our office was full of DP-to-DVI dongles already, so swapping one dongle for another seems like a no-op. Difference now is that we can buy USB-C monitors which also serve as a power source and a USB hub, meaning one cable to plug in my entire desk. That's something you can't really get on the old MacBooks (at least, not without spending a lot more than it costs to do it on the new MacBooks.
So yeah, the machine has "insanely good IO", and once the rest of the world catches up it'll be even better.
I feel that these discussions are always going back and forth because TB3 is a fantastic docking station port, and a niche port for everything else, so both sides are right. Comparing the 2016+ port selection to the 2015 one is a false dichotomy created by Apple. The straightforward upgrade path would have been to replace TB2 by TB3, like they did on the iMacs.
Exactly. There's nothing wrong with TB3 as a connectivity option, but it shouldn't be the only option.
I understand the desire to make the machine thin, which means the old USB had to go. But taking out the SD card slot makes it seem like they don't want you to have an easy way to add more onboard (-ish) storage space.
>So yeah, the machine has "insanely good IO", and once the rest of the world catches up it'll be even better.
I mean, that's the thing. Apple jumped ahead, and USB-C just isn't common. (At least, for me. I can't name any device I've used or even heard of that uses USB-C.) So, it's a less than practical option if you're in the market for a laptop, not so much insanely good.
I wouldn't buy a computer with only serial and parallel ports because none of my devices use them, and I wouldn't buy a computer with only USB-C ports, because none of my devices use them.
> if dongles are required for everything from iPhones
I never understood this complaint. AFAIK, I've never seen a macbook with a lightning port or the older 13-pin iphone port. If someone had an iphone, they had a USB-A to lighnting cable. Now they need a USB-C to lightning cable. Then you can use the macbook to charge your phone, or the usb-c charger that charges your laptop to charge your phone. Or yes, you can get a usb-c to usb-a dongle - but you can just as easily get the appropriate cable. I'm also seeing that Apple is going to start producing usb-c chargers for iphones and ipads (with a usb-c to lightning cable) in the near future.
I'm a macbook + android user, so I've had USB-C phones for a while, so now I can use the same cable and charger for my laptop or my phone.
> If someone had an iphone, they had a USB-A to lighnting cable. Now they need a USB-C to lightning cable.
iPhones don't come with these, so it basically means you have to go out and buy another cable, which won't work with any of the wall charger power bricks that came with your previous iPhones. I understand a change needs to be made at some point, but it's strange for Apple to ship computers with no ports that are compatible with their iPhones.
iPhone syncs on WiFi to Mac. I do not charge my phone with my laptop. I have an inductive charger on my desk at home, desk at work, table next to bed. I just set the phone on it.
I guess that works great for people who spend hundreds of dollars on inductive chargers!
I work in many environments and just bring a charge cable with me.
Of course, the iphone was just one of many things that I was noting isn’t compatible. External hard drives? Thumb drives? SD cards? All easily plugged into my previous Apple laptop, but require a dongle here. I have a mini-hub, but even that isn’t great because it can’t charge anything.
Glad they're cheap now! I stand corrected on that point. Curious to know your thoughts on the question of hard drives etc. Seems to me like Apple jumped the gun here.
Eh, is what it is. I recently swapped my MacBook Pro with TB for a MacBook. The HD is plugged into the Apple (LG) 4K. It runs at USB2 in this setup. I duplicated the same keyboard, mouse, monitor, HD at work and at home. I like 1 cable and the smaller size of the monitor given the resolution works fine. For me, the size and weight of the laptop was key due to 150K miles a year on a plane. Does the rsync of the HD to the home NAS take longer, yup. Does it matter, no. It’s a script that runs when the device is plugged in, in the background. If I need to do heavy lifting, CPU, compile, etc, I ssh to a rack of build servers in a colo. 3 racks of 48 cores a blade with a huge SSD NAS and 10G or 25G NICs sort everything. It is really about YOUR workflow and what works for you. My laptop is pretty much a window to google docs and for everything technical it is pretty much the same as a dumb X terminal. It works for me as my goal workflow is always the same no matter where I am.
> If someone had an iphone, they had a USB-A to lighnting cable. Now they need a USB-C to lightning cable.
The problem is that when they purchased the iPhone, it came with said cable.
I purchased an iPhoneX this year and it comes with, you guessed it, a USB-A cable. So in order to connect it to my new Apple laptop I need to buy either a $20 cable or a special dongle? If Apple wants to go all in on USB-C then they shouldn’t still be shipping their top of the line “futuristic” phone with USB-A cables.
But obviously they already know that they would have gotten many dissatisfied customers complaining about having to buy a cable just to ‘connect their phone to their iTunes’ if they did that.
Termux emulates Zsh for me and gives access to apt-get. so I installed nodejs through apt and made the project with react native and expo. compiling the binary for uploading to Google Play Store was a tricky step and I hope to improve the workflow in the future but for now had to cheat. Expo.io provides a build service for free where they will compile and sign your app for you.
ideally i need to get a copy of OpenJDK installed and build with gradle but this was much easier.
Expo also allows you to build an iOS blob from Android entirely assuming you have a way to get it onto an iphone through testflight.
My Windows machine at home is awful. Every time I go to boot it up, I have to wait another 5 minutes for it to… whatever it does, finish windows updates I assume.
I shut it down before a trip to Montreal; two weeks later I came back and it wouldn't boot. No idea why. Had to restore from a previous system snapshot and re-install a bunch of drivers and windows updates.
It's a lot better than it was, but there's still strange, ridiculous stuff that crops up seemingly out of nowhere.
My windows machine at home is always on, it's 6 years old and it's great. The only thing that has gone wrong is the motherboard failed, I replaced that and it's still chugging along quite nicely!
I cut my teeth on Windows with Microsoft Visual Studio Basic in the 2000s, then Windows/Dev-Cpp because of poor MSVC++ support for some C++ features I thought I needed at the time, switched to Mac/Vim, then had a 6ish year stent with Ubuntu and Vim as an IDE. But then a botched upgrade from 14.04 LTS to 18.04 LTS lead to a few frustrating hours of my time to get X to even load. I'll never see that time again. Now I'm back on Windows, and somehow have drudged along without installing Cygwin yet, although that's coming. Developing Java has been a pleasant experience with IntelliJ IDEA. The GUI use is taking getting used too-- but configuring IDE things like language servers is mostly automagic and have very reasonable defaults. The biggest issue I've had is Windows upgrades requiring random restarts completely interrupting my workflow. But, overall, rediscovering Windows has been a pleasant experience for me too.
Is there some missing repository you need to add to the Windows Ubuntu? Every time I've tried to install some one-off tool Stackoverflow has suggested (yesterday it was ncftp, the other day it was some ssh configuration, ifconfig isn't there) apt can't find it.
it's the same userland as "real" Ubuntu, so apt will give you the same results on both.
Perhaps do a "sudo apt-get update" (updates the local resources of apt, not the packages themselves) so that when you do a "apt search" or "show" in the future it has something to show?
Filesystem performance is a bit crap and you can't run Docker in Windows Home, but otherwise I found WSL pretty great. The Docker thing finally turned into a dealbreaker for me though, so after about two years of WSL I'm back to linux.
The lack of Docker support with WSL stopped me in my tracks too, or I'd have probably moved back to windows as my daily driver. I don't enjoy dual-boot for gaming as I only game occasionally and have to deal with a backlog of windows updates when I do.
> but configuring IDE things like language servers is mostly automagic and have very reasonable defaults.
What does configuring IDE things have to do with Windows? Wouldn't IntelliJ have the same reasonable defaults on Linux and macOS? Or is it a different product somehow (haven't used it myself.)
I just moved from an XPS to a surface book and I'd say if you can afford it go for the surface book, it's everything I liked about the XPS but with a detachable keyboard and a pen extra.
How does it compare to the XPS 2-in-1 models? The 3:2 aspect ratio seems tempting (even if slightly lower resolution) as does the battery life. Pen support should be quite comparable (wacom AES vs N-Trig), I guess?
Reviews said batterylife with the screen detached is very short?
The big drawback for me for the surface book would seem to be the subpar support for other OSs.
Pen support is better than the original surface pro's, which I believe were wacom. Hard to pin that on pen tech vs OS improvements, though. The pen needing a Quad-A battery is a bit annoying, but they're not too hard to find, and just one lasts a long time.
Battery life with a detached screen is short, but you'll only really need it detached if you're walking around with it. If you're sitting or lying down, you can just attach it backwards and fold it back onto the keyboard. Heavier, but not to the point of annoyance, and it gives it a nice tilt for reading things if it's on a desk.
Battery life with the keyboard attached is incredible, especially if you can dim the monitor a bit and run in battery saving mode. (The battery saver mode does have a pretty significant perf hit, though). Just don't try to use it at full speed, at least with an i7 and discrete GPU - I can kill my battery in under 2 hours if I try to run something like a graphics-intensive game. For dev stuff I've never had a problem.
Multi-OS support is probably an issue, though WSL does seem to do a pretty good job of most things.
Windows because I use it primarily for gaming/light browsing when away from my desktop, VM for other things. I bought it when I injured my leg and couldn't easily get up to my office.
Source? I can only find articles of people complaining that they can not get the switchable graphics to work or can get it to work and have horrible battery life. This guy (https://medium.com/@kemra102/linux-on-the-dell-xps-15-919e6d...) tried 11 linux distributions and describes his experience as "horrible".
I can't speak to the 15" versions, but my XPS13 (onboard graphics only) runs KDE Neon beautifully with no noise or build quality issues.
It's probably my favorite work laptop to date. The only minor gripe is that I haven't been able to find a good combination of settings to prevent accidental touchpad activation while typing.
People also often mention coil whine which I haven't noticed on this model, the 15" might be different.
What are you using for the terminal? I just started a new job with a Windows desktop provided and I can't get much done on it because every terminal emulator I can find sucks compared to iTerm 2 on my MBP, so much that I'm going in to IT to beg them to let me use it, or at least let me dual-boot Ubuntu. Bash on Windows has no way to add solarized light/dark or groovebox themes, much less switch between them daily like I like to do when I get stuck on something and need to see it differently. Also the Windows bash home is a no-mans-land once you want to open something up with Atom on Windows. Directories I make in Bash don't show up in File Explorer once I spend 5 minutes navigating to the right place. Font rasterizing and color-coding is awful too, on the native Bash, CMDR, and Hyper.JS. It's so frustrating.
Yet, the customizability rocks, and since xterm.js now renders using Canvas, performance is decent, and with my incoming MBP 32GB, I can afford the memory hit even more!
Although installing the plugins is asking for your fans to spin up.
You shouldn't use windows to access the WSL filesystem, there are several issues with that (which I believe are being worked on). If you want files accessible from both put them in the Windows filesystem.
Regarding themes I can't help that much, I have successfully solarized vim though. I needed:
let g:solarized_termcolors=256
set term=screen-256color
set t_ut=
If you're like me you're stuck in an older version of Windows 10 where WSL has serious bug the most recent version are very stable but if your work has you stuck on an old build it might not work well.
I opted for a Xiaomi laptop. Best purchase I ever made. Amazing build quality, out of the box compatibility with Linux, great hardware specs. I can't believe that near MBP clone only set me back 550€.
How does the trackpad on the Xiaomi compare with the MBP? I'm considering buying a cheaper MBP alternative and running Linux on it but I don't want to be carrying an external mouse around with me.
Hey, what was wrong with your keyboard? Was the trackpad also not working when you had keyboard problems?
You may have had a common issue with the 2015 Macbook Pros, literally tens of thousands of people have Googled that problem on Youtube. Apple won’t admit that it’s a problem with hose models but if you take it to an Apple Store they’ll conveniently take $500 to replace the keyboard.
Anyways it’s an introduced bug due to a design flaw in the models released in 2015. They placed the ribbon cable flat on top of the battery on those models, and when the battery heats up and expands, as it does, it damages the ribbon cable. Prior models had the ribbon cable avoiding the battery by going around it. You can fix it yourself by replacing the ribbon cable. The ribbon cable costs $10 online and it’s the first thing you see when you open up the macbook, you don’t have to dig to get to it. No special equipment needed besides the screws to get the Macbook open.
For awhile the 'e' key was only working about 90% of the time. Then it began working about 120% of the time, so I had to remap it to nothing, move the 'e' to the backslash, and plug in an external keyboard. The trackpad and all the rest of the keys were working just fine. I usually replace machines every three years or so anyway, so the key issue was just the trigger. Thank you for your advice. Do you think that might be the issue, given that the trackpad worked fine?
Intermittent keyboard problems are the early signs. But I haven’t heard of it affecting just a single key.
The trackpad and keyboard both use the ribbon cable, so when it fails completely, both the trackpad and keyboard will stop responding.. With the exception of the power button.
It’s a toss-up if it’s the same issue with only one key being unresponsive. Could just be a bad key. But I’d give it a shot since it’s only a $10 fix with a t6 screwdriver. Or if you want, mail it to me and I’ll fix it and use the laptop as a secondary computer to run my data visualization computations =D
Heh... the current plan is to hold onto it as a backup (in case windows decides to rise up and ruin me in a couple months) and just use it with an external kb.
Trading a touchbar escape key for Windows? As a Mac user that uses Vim, that touchbar escape seems easy to adapt to compared to adapting to an entirely different OS (and Windows at that!)
I recall when touchbar first launched, someone on HN pointed out that Esc is a poor way to use Vim anyway and there are far better mappings. I swapped to Ctrl-[ before I even had a new MBP and much prefer it. The “Esc problem” is a drastic reason to throw your whole hardware/OS stack our the window IMO.
As a long time emacs user, I've always had cap-locks mapped as my control key (plus the emacs key-bindings work in html forms and in anything using gnu deadline) which is a bonus. So on my MBP with touch-bar, I have had this mapping so I can have a physical ESC key:
CAPSLOCK -> CTRL
LEFT CTRL -> ESC
It's not great, but it works. The last time I had to deal with a company who put the ESC key in a weird place was back in college where we have several of these HP Apollo/720 keyboards:
My sense of the comment was not that he was trying to change your mind, but rather saying it seems unusual to be so fixated on the escape key that nothing else matters.
Having said that, yes, Apple should allow individuals to, for example, always require the touch bar to provide an escape key (or any other key) on the far left (or anywhere else on the touch bar). The user should get to control what the touch bar does through an easy and powerful interface and API (so third parties can extend the capabilities). Can we even make touch bar keys act as new meta keys? That would rule! The benefit of the Apple approach is that even if this has not been done yet, it's all software: it could be done.
As far as Windows, I guess it must come down to which features are important to you, and personal taste. I use Windows a lot, and have owned a Surface Pro, and boy I hated it. The very, very best thing about Windows is the Linux compatibility layer, and even that's not as good as if it were a real Unix-like platform all the way down.
I am indeed glad you're happy with your laptop (seriously; not sarcasm), but I've been there, done that, and hated it.
I wouldn't say I'm so fixated on the escape key that nothing else matters. The base OS matters a great deal! I pursued three lines at this particular fork in the road, I considered buying a 2016 MBP, a refurb or whatever, I considered switching to unix on whatever hardware (leant toward dell XPS), and I considered a windows box, with unix living underneath, somehow. The first option just seemed crazy to me, perhaps irrationally, but making my new machine a two year old machine just felt wrong, the second option could have mostly worked, except that I really need excel, and I really want ableton. I decided to give the third option a try, thanks to the money back guarantee, and I'm amazed at the improvement.
I suffered the temporary pain of switching to OSX in '06 because I'm not religious about these things, and at that time macbooks running OSX were providing an exceptional environment for devs like me. It was a hard transition, but after a couple of months I was (mostly) fine. A fair amount has changed in the world in 12 years, and I'm no longer convinced that OSX inherently provides the best environment for a dev. I figured give windows a shot and see how it goes. It took me 24 hours to get the machine in a state where I could work as before (this would have been impossible on a windows box in 2006), and the question now becomes, why would I go back? It's only been a week, so perhaps windows will show me why very soon, but my gut feeling is it won't.
Also worth considering, is that from about 2000 onwards, apple was catering to open source focussed devs, at the exact moment that microsoft was making their lives impossible. I think that trend may have been reversed. Every upgrade to OSX for the past six years or so has represented more pain than pleasure for me. YMMV, but microsoft's offerings, even when flawed seem to be headed in the right direction.
As far as the comment, yeah, I was a little snarky I suppose, but a common response to complaints about the lack of an escape key is "you don't actually need one". I'm sorry, but I do. Ok, I don't need one. but I really want one, and this is a laptop we're talking about, and I use it as a laptop most of the time. I also want ports, a headphone jack, etc. But the escape key is for me the last straw, and enough of an incentive to actually look on the other side of the fence, which I've avoided doing for 12 years. I like what I see there.
it goes beyond the displayed icon certainly, but definitely not to the physical extent of the whole bar, for at least mine and another that I've tried.
There's at least 2-3mm of dead space on the LHS of the bar.
Sadly I already remap caps-lock to ctrl. Don't worry though, this SB2 is so nice I'm no longer feeling any pain. Literally my only complaint about it so far is that the headphone jack is in the wrong place (it's on the edge of the screen in the upper right, which is sort of an odd place for a wire to be coming out. It should be on the lower edge either left or right.)
I left Mac for a Surface Book a couple of years ago. It was a challenge at first. Largely because of WSL speed, but that has been mostly fixed now (test suite may be a few seconds longer, but it isn't 10x longer)
I am browsing the thread out of FOMO and don't feel any urge to switch back.
I hear you. The little annoyances are far outweighed by the positives. Even the display seems crisper which I didn't expect over a retina display. WSL is not amazing performance, but it's probably about the same (the machine itself is faster, phoenix under WSL is not. Perhaps that's the perf issue at play)
So my main concern was being able to easily run elixir/erlang, with some flavor of unix as a production target. Therefore running natively on windows was, at best, a nice to have. On OSX I could just work in the native environment, since it's unix, albeit a bit of a peculiar unix with the occasion random odd thing installed to trick me. On windows I've found two possible solutions so far. The first is Ubuntu installed under WSL. I've heard people complain that WSL doesn't always work, but so far I've had no issues. easily installed elixir, erlang, npm and vim. Phoenix works no problem. I installed elixir natively on windows as well, just to see if I could, and again, no problems at all, and it works fine. I'm using cmder, which gives you a much more unix-like command environment. Chocolatey is the equivalent of homebrew, and actually seems to be superior, or at least more complete. ("choco install chrome", "choco install spotify", kinda handy). I'm mostly a vim guy, but I'm not religious about that. Neovim worked fine in WSL, so that's all good for now. Might give vscode a shot, as it has some nice elixir plugins.
Finally, I installed vagrant under windows, with virtualbox. I haven't used that before, but I think it could provide another alternative dev environment in case WSL turns out to be unreliable. My ultimate fallback plan was always to use vmware or similar and just run unix for dev, windows for things like excel, but my feeling is I'm not going to have to do that.
We've banned this account for violating the site guidelines. If you don't want to be banned, please read them (https://news.ycombinator.com/newsguidelines.html) and then email us at email hn@ycombinator.com and give us reason to believe that you'll follow the rules in the future.
After much debate I decided to take a chance on a surface book. I figured I'd try it, try and get my dev environment in a decent place, and see how I felt about windows after 30 days, with the idea that I might have to return it if things didn't work out. I'm 6 days in, and I can pretty safely say there's no way I'm going back to apple. Windows has managed not to suck at all, my dev environment is arguably better than what I had before, and I'm absolutely loving the machine itself. Most joyful new laptop experience in many years.