Thoughts on editors (Emacs in particular)(pinard.progiciels-bpi.ca)
pinard.progiciels-bpi.ca
Thoughts on editors (Emacs in particular)
http://pinard.progiciels-bpi.ca/opinions/editors.html
21 comments
The Hacker News community would do well to minimize discussions about optimal editors. It just isn't that interesting or important.
It's a bit shallow to sum a 5000 word essay which spans the history of the FSF, the rise of linux, and the cultural friction between the two, viewed (oddly, but intruigingly) through the lens of one's editor with 'you said editor, let's not talk about editors' .. well it's sort of missing the point of the essay, IMO.
I think you're right. I over-simplified things.
But that doesn't mean I find the discussion on this page interesting. Editors are very utilitarian, and for some reason lots of people talk about them. Feel free to have at it. I'll refrain from comment and up-votes.
But that doesn't mean I find the discussion on this page interesting. Editors are very utilitarian, and for some reason lots of people talk about them. Feel free to have at it. I'll refrain from comment and up-votes.
Why do you say that? It would seem to me that editor choice can be as important as language choice, and understanding the features of the editor are important to fast, effective development. One of the most interesting articles I read this year was about editors.
Neither choice matters much. Smart, flexible people tend to get the job done one way or another. If a bottleneck becomes apparent they work around it: revise, redesign, change languages, tools, editors as necessary.
Historically a lot of useless dogma and arguments have surrounded the choice of editor. There is a desire to avoid that sort of low-quality discussion.
Historically a lot of useless dogma and arguments have surrounded the choice of editor. There is a desire to avoid that sort of low-quality discussion.
Depends. Smart, flexible people can get the job done with any tool as long as that tool is "good enough". However, a bad tool will cripple even a good programmer.
The Emacs/vi/IntelliJ/Eclipse/Netbeans/Kate/jEdit/TextPad debate is meaningless, because all the above choices are "good enough". However, a decent programmer's productivity will be severely hampered by Notepad. Similarly, the Rails/Django/Pylons debate is pointless, but choosing JSF can easily kill your project.
The Emacs/vi/IntelliJ/Eclipse/Netbeans/Kate/jEdit/TextPad debate is meaningless, because all the above choices are "good enough". However, a decent programmer's productivity will be severely hampered by Notepad. Similarly, the Rails/Django/Pylons debate is pointless, but choosing JSF can easily kill your project.
The Emacs/vi/IntelliJ/Eclipse/Netbeans/Kate/jEdit/TextPad debate is meaningless
To the extent that it becomes a religious discussion about which one is the best, which happens all the time, I think you're right. However, the exchange of ideas that results can be meaningful. This is especially true of emacs, since there's so much room to use it creatively, but is applicable to all those other editors as well.
To the extent that it becomes a religious discussion about which one is the best, which happens all the time, I think you're right. However, the exchange of ideas that results can be meaningful. This is especially true of emacs, since there's so much room to use it creatively, but is applicable to all those other editors as well.
Agreed. Exchange of ideas is valuable, esp. w.r.t. emacs.
I understand what you're saying but I think you're missing my point. If any tool was empirically "not good enough" or "crippling" it would be abandoned and the show would go on. The focus is on the end result. The SFP gets there one way or another.
Heh, you'd be surprised at how many businesses cling to crippling tools because management has made their decision and refuses to abandon it. Moreso with languages than with editors, though I know of companies that standardize on a toolset (often one that's not top-notch, though I don't know of any that would mandate something as ridiculous as Notepad) and refuse to let you use anything other than that. Sometimes customers mandate certain technology choices too - I know of one government-sponsored project where they had to use Java/EJB even though virtually every engineer in the company thought it was a terrible choice.
Many of the projects I've worked on have succeeded in spite of lousy tools, insane deadlines and other various hardships. Those experiences were the genesis of my personal philosophy that good people get the job done one way or another. Luckily the people I've worked with have all been pretty smart!
Of course it is interesting - this is your major tool. Photographers are obsessed with cameras, programmers with editors...
I do quiet a bit of photography, imho photographers should not be obsessed with camera. and that further implies that this is not a good explanation for programmers to be obsessed with editors. The key word is "obsessed". You should definitely think about it. But save the obsession for other important things.
Such people are properly known as "camera fondlers". [1] Only some of them are actually photographers.
[1] http://nemeng.com/leica/002e.shtml
[1] http://nemeng.com/leica/002e.shtml
The constant complaints about inappropriate posts bother me more than the inappropriate posts themselves. Why is it even necessary to comment? If you a new item is not interesting, just don't vote it up?
Point taken. But I don't think voting will be enough. This topic isn't the worst, but some topics are really bad and have momentum. To fight the momentum, not up-voting doesn't seem enough.
When I consider the time spent in thinking about a line of code and typing it into my editor, I couldn't agree more.
Yea... and then when you're finally know what to do, you're held back by bottlenecks and deficits of your editor/IDE. Doing "context switch" on your brain, switching from the problem you're programming to "how the heck do I tell it to stop auto-indenting 8char tabs?!" is frustrating. It's slowing you down. Things get only worse when you're dealing with many files.
It's not the "typing time" that good editors save you, it's the "thinking time" they save since you're not forced to think much about them.
When I was forced to switch from Visual Studio 6 to VS2003, my productivity really suffered. Things did not work the way I wanted and I kept cold-rebooting my brain after every little interruption. Now I'm switching to vim and while it may not be as "fast" for certain tasks, its consistency and transparency makes me forget it's there. Besides, by not relying on arrow keys, pageup/down, del/ins and so on, vim allows me to be equally productive on unknown keyboards, especially on laptops, easily switching computers, which I happen to do quite often.
It's not the "typing time" that good editors save you, it's the "thinking time" they save since you're not forced to think much about them.
When I was forced to switch from Visual Studio 6 to VS2003, my productivity really suffered. Things did not work the way I wanted and I kept cold-rebooting my brain after every little interruption. Now I'm switching to vim and while it may not be as "fast" for certain tasks, its consistency and transparency makes me forget it's there. Besides, by not relying on arrow keys, pageup/down, del/ins and so on, vim allows me to be equally productive on unknown keyboards, especially on laptops, easily switching computers, which I happen to do quite often.
I agree with both of you, but I think it is wise to choose a good editor when you start. If the editor is any good, it can help you and reduce the time it takes to program, so more time is spent on thinking than typing. I'm sure you don't use Notepad or Nano.
But, yeah, dead topic to me, I already chose my editor and could care less what someone else says, only if it's about a new one, then I may care.
But, yeah, dead topic to me, I already chose my editor and could care less what someone else says, only if it's about a new one, then I may care.
The choice of editor matters in the sense that you would gain quite a lot by using an IDE with automated refactorings.
Yes! Or, an IDE which integrates a debugging tool, key bindings, API-sensitive auto-complete, run-time manipulation, and many other language-specific features geared toward easing development in that language.