Hacker News new | past | comments | ask | show | jobs | submit login

If it's ok to tack on here, how do you convince developers to move to Cloud IDEs? I've been trying to convince my team, but they are glued to PHPStorm and thus my attempts to move them meet steady resistance. The value-prop for me makes sense though, get out of the business of trying to maintain a vm on everyone's machine. Granted maybe we could solve it in other ways with containers or such but Cloud9, Codeanywhere, and some of these other tools seem like the right thing to move to.



I'm going to get flack for this, certainly in this discussion, but... seriously? What is this obsession with cloud IDEs?

The reason you can't convince me is that nobody has ever been able to explain in clear terms the benefit or the point of using a cloud IDE.

A cloud IDE presupposes excellent uninterrupted connectivity. Guess what? I often work on the move and often don't have that.

I'll admit that keeping Visual Studio service packed can be tiresome but this irritation is vastly outweighed by the fact that I can use it on the move. Likewise Sublime Text, or WebStorm, or whatever. And it's not as if software updates can't be solved in a different way (Chrome, Firefox).

For me the combination of a locally installed IDE with a distributed version control system still offers the best value prop for working portably.

I'm not saying a cloud IDE is always a terrible idea but, after 4+ years of these things, I still don't really get the appeal.


As a former volunteer CS instructor (for high school students), the cloud IDEs were great for use on school owned hardware that got wiped randomly, often had boot issues, etc.

As long as the students had a working machine, they'd be able to get to their projects regardless of whether it was on the machine they used the previous day (without have to deal with git/etc).

In my professional life I would never use these tools.


I use them the same way. I find git as a single developer to be also a much better fit for me. I work in three different locations on three different machines and a laptop. Git works for me, but I have taught some and the cloud IDE is great for teaching.


That actually makes some sense, and I suppose it's potentially quite a big market (if one that's strapped for cash).

Which did you use, and how did the students find them in terms of complexity?


I think it is better to use the term "IDE on someone else's computer" and the whole "magic" disappears.


To be honest I think we need to stop thinking that "cloud development" needs to be in a browser. People like to do all sorts of things with their IDE that I certainly hope browsers never let pages do, like intercepting certain command keys, controlling the layout of multiple windows interacting with and knowing more about the user's display dimensions and multi-monitor, etc.

Not to mention things like vertical-selection, or multiple cursor modes for column editing, which can't be done at all in any of the current web-based editors which AFAIK are all based on contentEditable rather than a custom renderer like the original Bespin (I think, that thing changed codewords a lot).

Now some of these can be addressed with Electron or similar, but there's a lot of things that are nice to do in a code editor that just aren't compatible with the DOM itself, and need a deeper customisation of the rendering and input management.

TL/DR - it's mostly UX :D


Have a look at: https://github.com/Microsoft/monaco-editor . It's what powers visual studio code and has all of those things


Most of online ides use Ace or Codemirror both of which support sublime style multiple cursors and are not based on contentEditable


I didn't know this, cheers!


> how do you convince developers to move to Cloud IDEs

Well... to be honest that's a pretty tough sell. Especially in light of this (reasonably) popular one shutting down with hardly any notice.

As a former IT Exec I don't know of any development Directors who want to trust their critical path development to someone else's cloud based IDE.


On-premise installations of some open-source software seems to be the path.


Well, I can't speak for the ones here, but the three times I tried to demo c9 to my team (at the time) c9 was down, partially up and/or really buggy. Given this was a few years ago now, but that really soured me on the idea.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: