6 ms·
A great relative comparison as far as the keyboard goes is something like vi/vim or even QWERTY. I'm sure you could have better text editing shortcuts (power S
by urs2102 10y ago
A great relative comparison as far as the keyboard goes is something like vi/vim or even QWERTY.
I'm sure you could have better text editing shortcuts (power Sublime/Atom users maybe), or a better keyboard (Dvorak for example), but once people do things a certain way and develop proficiency, it's often tough to change.
- heisenbit 10y agoThat is a powerful moat. Vi users or Emacs users may die out at one point in time but they are unlikely to switch in larger numbers.
- kkkrist 10y agoI have gone from Textmate to Sublime to Atom and finally took the time to properly learn vi. vi has made my typing so vastly more efficient that I can't imagine to ever use anything else again. I guess you have never really used vi to make such a comparison, same as op has probably never used the Bloomberg Terminal.
- JBReefer 10y agoDo you feel that way about generic text or code in particular? I haven't seen Intellisense or project templating for, say, F# at the same level as a VS Code. Does this exist? I'd love to move more to Vim!
- nickpeterson 10y agoWell, except vscode is behind vs for F#. Try using a type provider on vscode, it runs like crap for anything non trivial.
- JBReefer 10y agoOh, maybe something like Js then? I just pulled the first language I thought of!
- woah 10y agoDo you think that using a vim key binding extension like those that are available in most modern editors would have the same effect? Or is part of the magic of vim being forced to use the key bindings?
- caconym_ 10y agoIt's definitely important to force yourself to use the bindings when you're trying to learn them. Also, the Vim bindings in a lot of other editors have subtle issues and/or missing features that make them a pain to use if you're used to using the full range of commands in real Vim.
- Qwertious 10y agoVim is an amazingly good concept, but the actual implementations completely lack a decent user experience. Like, when you do clearly-suboptimal key combinations (like going past a 4-line paragraph by hitting j 4 times instead of just pressing } once),there should be some side-window suggesting better practices. To be fair, Kakoune does that, but there's still a lot of low-hanging fruit like that.
- ptaipale 10y agoThis particular example makes assumptions: that hitting j four times is worse than entering } once. This depends on typist and keyboard; I use Swedish/Finnish keyboard and touch-type, and hitting j four times happens without looking or thinking. Hitting } requires me to lift my right hand from typing position, hold Alt-Gr, press 0, and then go back to typing position.
- skykooler 10y agoDoes Vim not have alternate mappings for other keyboard layouts?
- eru 10y agoThere's at least one remapping to make it dvorak friendly.
- jan888 10y agoI've used vim for years with Dvorak and no remapping. Works great, IMO.
- eru 10y agoOh, I'm using it that way, too. (Just don't tell anyone I'm not using hjkl for navigating; my Kinesis Advantage keyboard has cursor keys in sensible locations.) But there's at least one attempt at a nice remapping for dvorak.
- heisenbit 10y agoIt is worth noting that while the terminal (for layout see: https://en.wikipedia.org/wiki/Bloomberg_Terminal https://en.wikipedia.org/wiki/Bloomberg_Terminal) is a moat there are other reasons it is hard to attack them (Fortune: "Can the Bloomberg terminal be toppled" http://fortune.com/2014/03/20/can-the-bloomberg-terminal-be-toppled/ http://fortune.com/2014/03/20/can-the-bloomberg-terminal-be-... . While a moat makes it hard to attack it is not always the case that a moat is worth attacking. Just because the target seems weak makes it not valuable. Bloomberg is a private company so numbers are hard to get. What is obvious is that financial services consolidation, the movement to index investment and automated trading makes terminals not a growth market. https://www.nytimes.com/2015/09/10/business/dealbook/the-bloomberg-terminal-a-wall-street-fixture-faces-upstarts.html https://www.nytimes.com/2015/09/10/business/dealbook/the-blo... > All service providers for Wall Street, not just Bloomberg, are unusually vulnerable at the moment. The financial industry is in the middle of an aggressive run of cost-cutting as it grapples with new regulations and changes in the markets. A Bloomberg contract, which can be upward of $100 million at larger institutions, is a tempting target to whittle down. > The number of Bloomberg terminals grew only 1.9 percent, to 325,000, last year. In the 10 years before the financial crisis, the number of terminals grew at an average rate of 12 percent each year, with most companies signing on for multiyear contracts. There may be some demand for cheaper terminals but whether this amounts to enough to build out the expensive data collection organization is far from clear. As may have pointed out quality matters. Opportunities may become larger iff Bloomberg has to cut cost and quality decreases.
- hashkb 10y agoWe keep creating more vimpires every day. We will outlive GUI editor users by centuries.
- azernik 10y agoWhat gives me real hope is the insertion of vim modes into GUI editors. It lets me use languages that need IDE assistance to be practical (I'm looking at you, Scala), while still holding on to the Old Ways.
- actuallyalys 10y agoAlong those lines, I'm really excited about Neovim's work to make the editor embeddable. Rather than reimplenting a fraction of Vim/Neovim, you can embed all of it.
- pareidolia 10y agoHave you looked at https://github.com/ensime/ensime-vim https://github.com/ensime/ensime-vim ?
- flukus 10y agoNeither the github page or the website say what it is.
- SZJX 10y agoThis makes no sense. Vi emulators get into all kinds of text editors and IDEs for a reason, not the other way round. I haven't used a Dvorak keyboard so I don't know anything about that (though many sources seem to suggest its effect is debatable), but on Vi/Vim your comment couldn't be more wrong.