Hacker News Reader

Comment by scarmig | Hacker News Reader

Enterprise likes it, because it gives them more visibility and control over what's going on in the dev box. That gives more angles to control against e.g. exfiltration threats. It also makes dev environments much more homogeneous: you worry less about client machines. In principle you could have a fleet of stateless Chromebooks whose main function is ssh and http, all connecting to identical virtualized desktops that are easily provisioned and deprovisioned.

View on HN

I agree that enterprise likes it. Developers hate it, though. We are a lot that are very picky about our tools.

Virtualization means images means standardization means everyone is using the IDE and tools that IT and "Enterprise" decides.

It has it's pros and cons. But one of those cons is that it can make for an intolerable work environment. At least for me. I've gotten to the point where I now ask about dev environment during the interview process. If they mandate Mac or Visual Studio I'm not interested (not trying to start a flame war, those are my personal preferences). I know that certain organizations need their control for auditing, security and remote wipe reasons, and that is understandable. I'll just go work elsewhere. It's fine.

You don't speak for all developers, at least not me. In particular, having a working dev environment that accurately represents prod just handed to you, instead of trying to rig up something locally is actually kinda nice. It matters greatly what you're working on. If you're building a local C++ application, a devbox is going to be wildly inappropriate, but for a SaaS company where prod is a untangleable mess; databases and queues and microservices, it's infinitely more efficient to gave me a box already spun up that's a good enough faxscimile of prod. Bonus points for seeing the DB with test info. In IT, if a user comes to you with a broken mouse, do you try and fix it, or do you just give them a new one and get them back to work asap, and take the mouse in the back and fix it later? Same principle. If the dev box isn't working, it's far easier to spin up a new one rather than debugging on someone's laptop.

Dev boxes don't have to mean standardizing on One True Editor (although everyone knows that's vim), as long as you can get to the source, typically via ssh, and edit it, whatever you want to use is just fine.

If only there were tools to spin up repeatable environments based on a given config file locally, using either VMs or containers, reducing the setup process to a single command.

What a world that would be.

If your stack is trivially setup via a docker-compose script, great! there are more complex systems out there for which that's insufficient.

I think they meant Nix.

I didn't mean one specific thing. There are numerous tools in this space that approach the problem from different angles using different technologies.

My point really, was if your local dev environment requires you to jump through a bunch of hoops doing manual setup to get it working you're doing it wrong.

I think this hits the nail on the head. Plenty of companies have software stacks so large / complex to configure that there's no reasonable way to run the entire system on a single laptop.

As much as I dislike the experience of remote dev environments in general, there are certainly times where it may be worth the pain.

> have software stacks so large / complex to configure

I think that’s your problem right there

Sometimes thing are just comped complex. I know we want to see epicycles and simplify and be able to reduce the system, and sometimes that's even possible. (Don't get me started on being all in on all of AWS.) But sometimes it isn't and, well, that's where we are as an industry.

I haven't worked at many places like this, but the last time I did they had services that managed and operated on petabytes worth of data. The full suite of front end applications, rest services, ingestion services, maintenance, queues, billing, communication, scheduling and other systems is large enough that I probably only saw the source code for maybe 1/3 at most, let alone modified.

I do like local development better, every time it is feasible, but there's not much point in doing so if it can't reasonably reproduce a production system.

This is really not true - a poor implementation shouldn't dictate the direction of the "remote development" space. Sure, IT and security have their requirements, but primary requirement is to make the developers happy and provide them with reproducible dev environments.

I used to work at Uber and what we ended up doing with devpod (https://www.uber.com/blog/devpod-improving-developer-product...) was to enable the popular IDEs to connect to these remote environments - all the dotfiles etc etc were persisted so it literally felt like the local IDE, just way faster. Admittedly, it costs a bunch of money to build internally, but there's a path to having people be happy with dev environments.

(we collected data on what IDEs to prioritize based on surveys)

Why use a survey and not just ask the endpoints directly? Presumably the laptops are managed and are running something like Santa on them. Would remove bias to get the data this way.

yea - we had that too (good for understanding how laptop tooling worked, and what areas were starting to show latencies and therefore, needed to be worked on)

surveys were anonymized

When I discover a company prefers Windows desktops, Microsoft 365, and Sharepoint, it's often a red flag for me. I've worked with guys who didn't know how to use git, so they copy all their code into Sharepoint. Then there are others who didn't know how to merge branches, so they copy from one folder to another using the windows desktop. It's even worse when the people doing this are supposedly "senior." In a Linux or Mac environment, I never encounter these sorts of WTFs

Just like this company that deployed by mounting the production server on "the special machine" through SMB and dragged and dropped the files across.

You couldn’t drag and drop the whole folder, else you’d overwrite the numerous configuration files that were all over the place.

I really don’t miss it.

Yep. I ran into the “don’t overwrite these files” issue on a branch at a new company. It took months to get a sane branching/merging process in place. The “lead” loved copying crap with his windows desktop, did not even script his half baked processes. You should’ve seen what the actual code looked like.

> Virtualization means images

No. Some prefer image based provisioning, but it doesn't mean you can't have your 'own' machine.

I like it. I used gitpod.io before. Super nice to just be able to switch between projects in an easy way. Stopped when they changed pricing though from unlimited time to credits.

As someone who migrated our devs from local VMs to cloud VMs, I like it because despite the promise that every VM is the same, every dev laptop is not. There's so many ways for Vagrant and VirtualBox to screw up. Sure, the cloud VM is a little slower than local, but at least it fundamentally _works_, unlike VirtualBox where on any given day you might be suffering from any combination of 1) VirtualBox kexts not working on the latest version of MacOS, 2) file sharing silently breaking yet again with no logs, 3) docker containers inside the VM losing network connectivity for the third time this week because the bridge decided it didn't want to do anything, or 4) having VirtualBox installed just completely breaking MacOS and causing a crash during boot even after completely wiping, reloading, and trying to install VirtualBox again. Those are all real issues I spent weeks tearing my hair out over across dozens of devs. Now that we've moved to EC2-based VMs the only problems I have to deal with are minikube problems (which I had to deal with under VirtualBox as well) and when devs forgot they shut down their VM before going on vacation. The devs sometimes don't like that the default EBS volume is a little small, but every one that used the local VM knows it's a small price to pay

> VirtualBox

Even HashiCorp know vbox is scraping the bottom of the barrel for hypervisors:

> if you are using Vagrant for any real work, VMware providers are recommended since they're well supported and generally more stable and performant than VirtualBox.

Don't use VirtualBox. It's literally the shittiest VM you can choose.

What other option do you have if you want to support developers across Linux/Mac/Windows?

VMWare player? Besides it not being OSS, I didn't notice any performance benefit.

The bottleneck (for my web development needs), is around the spottiness of shared folders and JS tooling that needs to keep a watch for file changes.

If you're using vagrant to manage the VM and either use widely-supported or self built base boxes, each developer can use whichever works best on their platform.

So you might have a windows dev using hyperv, another usimg vmware workstation, a mac user with parallels, a linux user with lxc and another with kvm.

Parallels always took that crown in my book, though I don't have that much experience with VirtualBox.

Reply on news.ycombinator.com