Theia developers: The Eclipse brand in the IDE/code editor world is toxic to many developers; You may want to disassociate Theia from Eclipse if you’re hoping for wide usage or pulling devs away from VS Code.
Eclipse is known for being clunky, but at least it is _known_.
It is also trusted. When I wrote my first Android app, it was the only practical way to write and test. Not sure if that has since changed (am no longer doing apps), but my guess is that a lot of people have Eclipse on their system.
There is a saying that even negative publicity is still publicity.
The most important thing that Theia needs to convince would-be contributors, is that they will still be around tomorrow. Eclipse has been around forever, and most devs assume it will stay around. That has a value that should not be underestimated.
> Eclipse has been around forever, and is going nowhere.
That's true of everything, right up until the day it's gone. I heard the same thing about Borland. And Perl, Fortran, 1-2-3, Netware, Lycos, Blackberry, ... The history of computers is littered with brands which were once impossibly dominant, and now just a footnote.
In my research field, about 80% of the software is Fortran. Actively developed Fortran. I wouldn't really compare it with 1-2-3/other DOS-apps or blackberry which are basically the software equivalent of consumables.
While I understand your point about “Eclipse”, as a long time open source web developer, macOS user who has had to work at times with Microsoft servers, I did not think the best of “Visual Studio”
I worked professionally with VS doing Windows app development in C++ and I found it super powerful, especially for handling our 20 years old gigantic codebase. I don't get the hate, and I spend most of my days in Emacs nowadays (doing more research-y coding). Yes, it's huge, but it's huge because it is full of powerful tools for developers.
I think hate for giant IDEs & their corresponding resource consumption will generally come from devs who don't need all the features, or who have the usage of the features forced on them by their org when they may not be needed or there are more lightweight alternatives outside the IDE.
It's kind of an even-more-resource-intensive Microsoft Excel - almost no one needs all the features, but they all need a different subset.
It is just that I have always done a different style of coding for over 20 years. More focused on text editors (BBEdit, Sublime, VSCode, vi, etc) and supporting command line tools.
I have seen the power that my IDE colleagues have for their use cases.
But honestly, I think many of us text editor focused developers where very skeptical of Visual Studio Code due to its VS branding. For myself, I don’t see Eclipse’s brand any better or worse than Visual Studios.
I was under the impression that VS is amazing and the best-of-the-best IDE as long as you're entirely working within the MS ecosystem and your employer is paying for all of it. I never heard anyone hating VS itself, just MS or perhaps M$.
But the Eclipse Foundation is a pretty good place to develop vendor-neutral open-source software, kind of like the Linux Foundation. There are many great open-source projects under the Eclipse Foundation umbrella, and anyone is free to use them; contribute to them; build businesses on them; etc.
IntelliJ is prettier, but just as bloated. Buch of that bloat is just features. I've used both for years and thought both were good with different strengths/weaknesses.
The last time I checked (long ago, when no IntelliJ frorks like PyCharm etc existed) Idea' code autocompletion was a way more smart than that of Eclipse. Also the screen space is less bloated with IntelliJ.
How is the screen-space less bloated? I thought pretty much all IDEs are the same: navigation tree panel, code panel, and optional extra side/bottom panel for log/test/debug/docs/etc. Also they're all super configurable.
That's fine. They want people who appreciate Eclipse during these early stages, and young people who haven't used Eclipse before don't have preconceived notions.
If the IDE broadly reduces the efficiency of developers, and causes developers to waste time contorting their workflow to the IDEs poorly conceived approach to doing things - especially in relation to other IDEs... I'd say that's toxic for the Eclipse brand.
Replying to my own post but I also did this with VS Code. Brand has a lot to do with first impression. When someone I knew told me to try out VS Code my first impression was I have spent too many years in Visual Studio already why would I want to go to what is obviously the lite version. It was only after seeing it, and seeing the support it had in the community did it overcome that initial bias. I hold that same bias towards Eclipse and my initial reaction was great web enabled Eclipse, no thanks.
Out of curiosity, what sold you on VS code over VS?
I use Sublime to edit smaller code bases, and web technologies because it gets out of my way, and obfuscates nothing from me. I use VS if I need to work with larger codebases in compiled languages with static typing, complex build processes, or multiple dependencies.
I tried it about a year after it first launched, but only for a couple days. VS code strikes me as a hybrid between the two, but from what little I've tried of it, my experience has been that it's less snappy than Sublime, while attempting to obfuscate the things the way VS proper does, which I find distracting when working with js/Python/html. When it came to the large CPP code-base I was toying with, its felt slower to work in than VS, and I couldn't find any compelling reasons (killer features) to learn its conventions.
I know quite a few people using code for web development, but they haven't really told me why ("it's just better").
I get the impression I'm missing out on something wonderful here, and I don't want to be an anachronism. But I just don't know why I'd want to switch unless I needed a heavier IDE on my Linux box.
Two things, integrated and well put together debugger for node and chrome debugging in the IDE would be the first feature for me that stuck out. I am a debugger or REPL forward developer, I like to digest code in small chunks. VS Code has some fairly slick debugging features. Second it's a text editor that acts like and IDE or it's an IDE that acts like an editor, I cannot really decide which it is and I think that is the sweet spot. It also has a huge base of quality plugins for a host of languages.
I am not sold on strong typing for the web and SPA's there is just so much tear down rebuild and remodel that I don't know if the gains are there. Given that I take that view I tend to shun a full and robust IDE in favor of a text editor with language support. VS Code is something a little bit more than that, and I it hits a sweet spot for me. When I have to support our legacy Java apps I jump back to Intellj Idea but it's not my favorite and goes a little too far towards the IDE side.
When I do embedded C, I tend to prefer a full blown IDE as it's just too easy to get lost in the weeds of a large C code base. Which is usually CLion.
That is just my personal opinion, and I recognize that tastes differ, but due to years of (bad) experience across multiple domains for a plethora of reasons, I won’t allow and will actively root out any whiff of Eclipse near my codebase.
Yes, others have different options and IDE selection is only slightly less polarizing than politics or religion but... that’s the branding quagmire right there.
Maybe they are not so much trying to convince end users about the project, but other companies who might be willing to incorporate this into their products and contribute back to the project.
For that kind of use case having an organization like Eclipse Foundation, Apache Foundation or Linux Foundation on the background adds credibility.
Agree. When I read Eclipse, I thought "icky" and scrolled down to see if maybe JetBrains will turn it into something usable. Since they're not in the list, I'll pass on this.
I'd rather rent a Windows Server so that I can remote desktop into IntelliJ than suffer through another IDE.
The productivity boost that you can get from using a great tool that you are intimately familiar with is just way too much to sacrifice in exchange for what boils down to less setup work for getting a new laptop ready for duty.
It is also trusted. When I wrote my first Android app, it was the only practical way to write and test. Not sure if that has since changed (am no longer doing apps), but my guess is that a lot of people have Eclipse on their system.
There is a saying that even negative publicity is still publicity.
The most important thing that Theia needs to convince would-be contributors, is that they will still be around tomorrow. Eclipse has been around forever, and most devs assume it will stay around. That has a value that should not be underestimated.