Show HN: Avoid editing while writing your first draft
rolandasb.github.io
rolandasb.github.io
I basically write a couple of sentences or paragraphs, stop, review, reword things if needed, change things around and move on. Then at the very end I'll give it all a final reordering / fixing until I'm happy with it.
I feel like I can't progress to the next chunk of an article until the prior section is 95% edited because what I write next depends on what was previously written.
Does anyone else work like that?
A book people might find interesting is Joe Moran’s "First You Write a Sentence: The Elements of Reading, Writing … and Life". He examines writing from the ground up. I found it a really wonderful book to read. Each sentence in it is great.
Creativity doesn't work that way. You'll literally inventing new things. It would be like saying, "let me plan out how I will innovate." Makes no sense.
I very rarely know what I'm going to write until I start writing. I don't even outline anything. I just start writing and relentlessly edit as I go.
I do the same thing for 1,000 word blog posts as I do with 400,000+ word course scripts.
If anything editing is more important for the larger pieces of work because in order to cleanly flow to the next lesson or section of a course the prior stuff has to be pretty much in its final form.
I'm not really a big outline person either for either writing or presentations. I almost always have some idea where I'm going though it's not unusual for that to diverge.
A good way to think of creative writing, if you're a technical writer, is to imagine attempting to write a textbook when you don't know the subject at all, and the "demand" from each new paragraph causes you to do the research necessary to acquire the knowledge necessary to write it.
How would you even structure a thing like that, in advance?
How would it ever attain a sensible shape if done a paragraph at a time, when something you learned while doing the research necessary to write paragraph 100 invalidates everything you wrote before it, or causes a complete restructuring of your mental model such that you realize the topic should be presented in an entirely different order? (And then you realize that again, and again, and again...)
A similar thing occurs in investigative journalism. How would you know how to present a story—know what story you're telling, really—before you know all the key facts? In est, before you've completed your investigation? You could certainly write notes about what you might write, but those have little to do with drafting the final story.
For many disciplines, the actual research involves field work, performing experiments, etc., and the thesis is primarily about reporting on that process and connecting it to the pre-existing body of research. For other disciplines (for example, philosophy), the text of the thesis is all there is to the research. Maybe the later is more like creative writing and the former more like technical writing??
(My uncle did a PhD in creative writing, his PhD thesis was a novel.)
The advice to "don't edit while you write" usually comes after the advice to "make an outline" before you write. You are supposed to figure out how the pieces fit. Then quickly write it all out. Then edit edit edit.
The general idea is that is easier to edit afterwards than to edit on the fly. Its true for me.
Same thing with writing. Yes, you are literally making things up and inventing new things, but that doesn't mean that one is writing freely and not needing to edit along the way. Outlines give framework, and editing gives time to reflect and plan things that might not have been planned in the beginning. I'm not saying that no one works this way - just sits and writes, or does freehand paintings (I do this last one, even with watercolor, and accept a certain failure rate) - but how one works doesn't really reflect on their creativity nor how their creativity works.
Not that I know how the mind of a great author works, but it seems that it would be difficult to write a great story without having some solid idea of what the final product should look like before you start. Like creating a product, maybe they pivot a lot after realizing that something doesn't work well (Margaret Atwood apparently did that for a book once, completely started from scratch after realizing that she had chosen the wrong point of view for the protagonist). BUT it's difficult for me to imagine they don't actually have a good solid idea of the story they want to tell before they start writing. If not book by book, then at least chapter by chapter. But hey, I'm not an author, maybe their minds really do work differently. It's possible.
I guess my overall point is that I don't think having a solid idea about what the final product should look like is the reason why authors have the "never edit while drafting" because I think they also pre-structure their ideas, and diverge on "editing while drafting" for some other reason. But of course I have no data or evidence to back this idea up.
Hamming agrees in his famous speech (see the Q&A section on this page - https://www.cs.virginia.edu/~robins/YouAndYourResearch.html) that 50% of the time is just spent writing/presenting research.
* Conventional development = writing, then editing
* Test-driven development = writing and editing simultaneously
The preferences are endless and neither removes the most difficult part which is the conceptual architecture.
Personally, I like to splatter a whole bunch of ideas down first and then later go back and commit.
For shorter academic and work pieces (and back when I blogged) I’ve done this as well, but I do an outline for longer papers first, and then edit as I go. Of course I wasn’t anywhere near nickjj’s level of output.
[1] https://www.panmacmillan.com/blogs/science-fiction-and-fanta...
For this reason I trained myself to make the first draft the final draft. Now with computers this is no longer relevant as I'm not hand writing multiple drafts on paper.
There are other factors, but I think this one is major for a lot of people.
I also start writing in a fairly basic text editor, not a word processor, which makes it easy for me to reorder that point form outline with my keyboard and get the flow of the major points down before I start the actual writing.
The text editor also prevents me from spending too much time on that 5% of improvements until the very end. But, I do edit major grammar issues and color as I go since, like you, I feel like I can't move on until it sounds right, and the way I say something often shapes what comes next. Still, having an outline helps me get into those weeds a little bit in the moment and not lose sight on where I'm going next.
I can't stress the outline enough. I wish I did it in the early days, but I wasn't able to "see" the whole story from the beginning back then. Now, I find it is the thing that helps me see what is important to the topic. Sometimes I feel like I need to talk about a point up front but after roughing it out I realize it's actually more interesting and easier to understand if I hold it back until the end. Conversely, there are some really interesting things that can't be explained until the end but I can tease them up front to build excitement and then explain them after the foundation is laid.
The words, style and delivery of my intentions through a pen/pencil/crayon differ from my spoken word and typed thoughts.
Trite "avoid ..." theses are often ... trite.
I've tried the 'never edit your first pass' approach before, and I found that my ideas flowed like sandpaper instead of silk. It was awful. I simply do not think in that manner.
It's not really real editing though. You don't catch everything because you're still in the writing mindset. Plenty of times i've just gone back and completely re-written whole sections during editing. I won't usually do that while writing something. Editing and writing are two somewhat different skillsets.
Editing requires a detachment you can't quite get while something's fresh. You need a bit of time away from something before you can go and do a full real edit. It helps you look at things more objectively. Re-reading a first draft of something I wrote after usually only a couple of hours kinda makes me cringe sometimes even with some inline editing while writing.
I look at it like coding. I'll write a section, test it, fix simple mistakes, keep going until I hit a bigger complete milestone, then go back and fix all the real mistakes while I'm debugging.
ETA: Sometimes I wish the edit windows were longer on hn. There's been a few comments on here I wish I could go back and change after i've had time to.look at them more objectively. Not always downvoted ones.
When writing my first draft, I often have very little idea how the whole thing is going to turn out. I simply don't have the big picture yet. If I spend time honing my sentences then, most of the editing time is wasted because I'd later rewrite them anyways.
I like the blur, but I find it hard to use. I make extremely many typos and my normal typing method involves simply hitting backspace a lot of times. Obviously I should address that instead of complaining about your blur feature on HN, but the gist is: if I can't even see the last few words I wrote, I'll end up writing gibberish :-)
If there'd be a "blur everything except the last 15 characters or so" kind of mode, or something where it gradually blurs characters as they're further away from the cursor (like a warcraft "fog of war" for typing), then that'd be ultra neat.
That said: WOW that app is nice. You got yourself a new user. I love how it's essentially device local and signup is optional. Super user friendly that way, in a literal sense (i.e. it's very kind & friendly to the user)
Just FYI, in "typewriter" mode, you can still delete if you use Ctrl + Delete or Option + Delete on a Mac. And then if you delete everything in "typewriter" mode using either of those key modifiers, using the other modifier on the empty text input crashes the app.
Does it support overstriking? Typing o<BS>/ is a legitimately good way to type ø, for example; it isn't a universal input method by any stretch but when it works, it's very mnemonic.
Blurring everything makes it impossible to fix (or see) typos which is annoying, knowing that the output of the thing is going to be full of typos. Hiding texts has another issue (even if not hiding the current line), that is you don't even have a rough idea of how much you've written and what the paragraph lengths look like so far.
If we do blur, but don't blur the current line, it solves both problems for me.
Works a treat, even if you liver may not fully appreciate it.
[0] Whiskey may be substituted for your preferred liquor, but it is a solid traditional writers' choice
[1] TL;DR 'Write drunk, edit sober'
I'm working on a very similar problem atm: https://ulysses.sonnet.io/
My approach is slightly different though:
- you can see the previous lines (the last 4 lines of text), i.e. they don't fade, so you can still keep some context
- you cannot select, edit (the UX makes it hard). I'm still working on making the UX around it more explicit, curious to hear feedback.
There are several reasons for that (or use cases):
- I use it to foster a better writing habit
- During the day, I keep it on split screen and write to organise my thoughts when I work on more complex problems
- I use it as a diary (and paste the text to the Notes app)
Do you use it in a similar manner? Did working on the app change your writing habits in any way? The reason I ask is because in my case writing this way feels somehow more natural. I ended up writing 800 words per day on average (with spikes to 2-3k).
Also, I'm following the same `div.fade` approach to mask text in my app. The main reason for that was incorrect colour correction/mixing on Firefox (banding or shifting the colour by 1 point).
PS. I'll send over some of the apps here to my partner (a writer) and see which one suits her needs better. I'll let you know if I get any feedback.
I hated writing then. It was just painful. I procrastinated any writing assignment till 2am the night before it was due.
Now I love writing. Writing is often something I do when I'm procrastinating something else. (ummm, that would include right now I guess)
I'm not a linear thinker and forcing myself to do it that way would just be a way to make me hate writing again.
Either way, there was no way I'd turn in my rough draft as a final, as it would either be a mess to look at, or terribly/confusingly written.
I just don't see the benefit of imposing a linear process. I prefer to start my writing with what comes easily, and immediately start tweaking.
It doesn’t work on mobile, but you should use it with a proper keyboard anyway. There’s nothing sent from the JavaScript to any server; it’s a static page hosted on Netlify with very simple code. It’s just an index.html with style and script tags, 39 lines total.
I use it in the evening when the baby is asleep and I just want to reflect and think for a while. Maybe you will like it too.
ctrl+d to exit.
You can edit the current line (to fix spelling misteaks) but not anything else. You can reference as much as your terminal shows.
This is a perfect example of the superiority of the Unix way. Simple programs that do one thing and work well together, combined into things their authors didn't think of. This is my idea of beauty.
cat > yourfile.txt
All perspectives have been successful tech book authors.
The more difficult thing is how you collaborate with coauthors with different writing style that might conflict with yours.
I think this works for writing as well.
Personally I also use this in programming. When I was young I directly dove into the details but that just doesn't work. Now I'm more like a painter. First big strokes and taking smaller brushes as I move along.
Many many many fiction writers have said they don't plan things out this way. A vast majority of writers I've read interviews of do not plan things out this way. Salman Rushdie, Zadie Smith, Jeffrey Eugenides, Stephen King, and George Saunders are a few I've seen talk about this in person.
Reported pieces are sometimes different, but even then in narrative non-fiction that narrative can evolve along the way. For technical pieces it could be the same.
The reason it works is that it is far easier to futz around with the master plan/outline than it is to get your ass in a chair and write. Also, in fiction certainly, the characters reveal themselves as they go. You can't build more than a pastiche of a human with an outline.
I'm not sure why this is a good thing though. Often times I find I need to go back and edit something for the purpose of continuity.
This doesn't prevent that, but it does hide the reason for it. But, hey, if it works for the user I find no reason to criticize those who'd use it.
The key, for me, is to retype into a computer and edit during that phase.
Acquiring and operating a typewriter is definitely not, in my view, a simple alternative.
Just writing with pen and paper seems simpler on all counts.
Typewriter acquisition is about an hour of time and $50.
Admittedly, I've owned a typewriter for longer than I've owned a conputer. I put it down in favor of WP5.1, but picked it back up a few years ago.
The advantage over pen and paper, for me, is speed and legibility.
I suppose with a typewriter today, you could still just OCR the result using an old scanner you can pick up at a garage sale or thrift store for nothing, if you didn't want to re-type everything by hand.
R.R. Martin has said he still uses and old word processor with Word Star on it, to have fewer distractions. He transfers things over via floppy disk.
Repetition is under-appreciated today. I think it's something we lost when we gave up typewriters. I'm not saying we should go back to ink and paper in all cases but I do think there is an aspect we're missing out on, like someone whose fitness deteriorates when they start driving everywhere, or who doesn't know where anything is now that they have a GPS-enabled smartphone.
After that, you write.
The virtue of editing on a computer is that you can easily correct as you go. If you don't think that's a good idea, you might as well go back to a typewriter or handwriting.
Generally by that point I find the post is ready to go, but sometimes I go through three or four edits on pieces I'm not sure about. Feedback is also good- I typically ask a friend or my brother to see what they think.
Planes and trains are great for this; I find that I get a lot of writing done while traveling. Even on the subway, sometimes I just type up thoughts into the Notes app, and then pull it up on my computer later.
Now editing, that’s a whole different ballgame. I find editing very hard to do on the phone.
[0]: https://www.newyorker.com/magazine/2013/04/29/draft-no-4
This gives a word sprint of 15 minutes with a visible goal of 1000 words. Hitting ESC leaves it ready to paste elsewhere.
No backspace, no delete, only enough feedback to know your typing is captured, and whether you are on pace.
https://github.com/SteveJSteiner/BantamScribe
It is also entirely local.
Spelling and grammar errors take something like 60 seconds to fix in any reasonable editor that has suggested spelling fixes.
The goal is to focus on the idea and dispense with what should be automated. This is for exploring ideas not refining wording.
For a novel, the choice to edit as you go is orthogonal to avoiding editing during a word sprint. The real issue is during revision you’ll find you corrected a bunch of stuff you are cutting. If that editing was cheap who cares? If you sweated over a sentence for 15 minutes you might do the wrong thing and not cut it even though that is better for the story.
Why not just make it uneditable?
FWIW I learned to write longhand and then using an early electronic typewriter with only three lines of text retained in memory. Modern word processors have been great. Sometimes, though, when planning something out, I'll use a paper notebook.
Feature idea: prevent you from leaving the page or doing anything else on your laptop for the time period, that way you can’t research anything.
More tips: https://dtrejo.com/how-to-write-consistently-painlessly-and-...
But I'm not sure what the point is. I mean, searching yields "We couldn’t find any repositories matching 'rolandasb'". And I'm certainly not going to write anything substantive on this site. So how would I do this locally?
It only disappears when I'm done writing the para, i.e. when I move to next para. Which renders the feature obsolete when I'm writing big paras.
Could this be done to letters, after x secs they disappear.
1. Make It Work
2. Make It Right
3. Make It Fast (maybe)
2. Ah it doesn't work
3. Make It Right
4. Big refactor, better rewrite it. Ah no time.
5. Make it Work, but without it being Right, what is the bug level we can get away with shipping?
6. Make it Fast, scale up the VMs!
Okay. Look: here's an actual snippet from the first chapter of a novel I wrote a few years ago. It's not unusual for dialogue to be written like this: short paragraphs, one or two sentences each, right?
---
If he’s based on Panorica or one of the half-dozen platforms that have a compact with them, he could lose his license for that. Or worse. But a lot of the private yachts berth at places where space law is more space suggestion. “Who else have you called?”
“Just you. I’m pretty sure the ship’s completely dead. The crew either got out already or didn’t make it.”
“Jesus, Randall.” She runs a hand through her hair, stopping outside the bar. Could she lose her license for following up on this? “If it’s dead, how’d you find it?”
“I picked up an emergency beacon. It stopped before we made visual contact.”
“What kind of ship is it?”
“The beacon data was for a Horizon class freighter.”
“A Horizon went missing and nobody noticed? When was this?”
“Yesterday. Maybe fifteen hours ago. I’ll send you the telemetry data.”
---
Now: imagine trying to write that when every paragraph immediately disappears after you type it.
I'm sorry, but that's just not going to be helpful. Writers don't need to have their past paragraphs hidden from them; we may, in fact, need to see those lines for context. If you want to grey them out the way iA Writer does in "focus mode," fine, although I confess I remain skeptical about how much benefit that truly brings. If I'm actually typing, like I am at this moment, my eye is following the cursor and I'm focused without the extra benefit of disappearing text; if I stop typing, it's because I need to think about what I'm going to write next, and I may need to, you know, read what I've just written in order to do that. If it so happens that I notice a typo in the previous sentence -- or even, heaven forbid, the previous paragraph -- I don't actually feel like my editor is helping me if it prevents me from doing that.
And this is the problem I have with an awful lot of these "let me help you write by being super super minimal and throwing in one neat trick you haven't seen before that makes this even more minimal than those other minimal" editors. I sincerely appreciate the ideas and the work and, yes, the aesthetics. This particular one is really elegant and I don't want to take away from that. But it's part of a whole class of editors that feel, at least to me, like they're kind of solutions in search of problems.