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

I credit Solarized with setting the baseline allowing proper "dark" themes to be more commonplace. I'm going out on a limb, but perhaps macOS' Dark Mode, or VSCode's dark mode, wouldn't exist without Solarized blazing that path.

It's kinda funny how, before Solarized, I would see tons of gimmicky themes like some that were going for a Matrix or "classic terminal" feel, going all in on a certain theme but eschewing real usefulness. I want my basic terminal colors to remain adjacent to their intended hues! I want different keywords/tokens in my code to be distinguishable!

Ethan's application of color theory/science to the landscape was like a beacon in the darkness.

Before Solarized, I used Zenburn (I actually still do in some environments, I'm not too keen on the blueish hue of Solarized backround.) I also like Tango Light/Dark.




What exactly do you mean proper “dark” themes? Being able to swap to coding in light test on a dark or dark text on a light background goes back to at least windows 3.11.

Solarized dark theme to me looks just like several other similar themes that predate it ex: https://superuser.com/questions/156979/how-to-i-use-textmate...

But I assume you have something specific in mind here?


I would assume 'proper' means 'adequate in terms of function and aesthetics'. maybe something like 'themes that don't eschew real usefulness'. those themes did exist before solarized, but they were not the norm in many circles (mine included), as many authors did sacrifice function for the look&feel.

solarized demonstrated how to apply colour theory/science to the landscape.

I would add that it has been not just the theme that serves as an example, also the _how to make_ a theme following those principles.


> not the norm in many circles (mine included)

Fair enough, I have the opposite experience where most people using dark UI’s where doing so for utility reasons not aesthetics. Outside of a brief fad of Matrix themed terminals which quickly died out.


I feel like we've had a similar theme journey. These days I use darcula and its derivatives I some places, Solarized in others. Tango dark as well. And fond memories with Zenburn. Monokai gets an honorable mention.


If you used Borland C++, DBase, Lotus 123, WordStar and others... or used a monochromatic monitor (green, cyan or white on black), you used light text on a dark background.

By 2010, when Solarized appeared, the larger portion of the people that started coding on a light text on dark background had already retired.

Solarized did a good job in terms of distribution, I give you that 100%. But it is not the most important theme in history or anything of the sort.


Not sure why you're downvoted, but anyone that fondly remembers the "Borland Blue" days probably agrees with you.


Possibly the downvotes are because folks (like me) who started coding in light text on dark backgrounds in the 80s had emphatically not hit retirement age by 2010? They'd be between their late thirties and maybe late fifties, unless you want to push it all the way back to the 1970s or earlier -- the build-out of computing infrastructure over the past century has been exponential, so there is always a preponderance of young/new developers, but with retirement age nearing 70 folks who began programming in the early 1970s are in many cases still working.

(I went sideways and left software development entirely. But I'm 58 and very much still working with text on screens and routinely use dark mode.)


As a 48 year old who's feeling quite burnt out, I'm curious to learn more about this sideways move... I'm feeling a bit trapped right now. :(

EDIT: Ah, checked your profile and history and it seems like you've become a writer? Hmm... maybe I could do that...

(BTW, the SSL cert for your website has expired!)

EDIT2: Thanks for the advice on writing, all! Honestly, I was thinking of taking the technical writer route. I've always been pretty good at authoring things like technical specs, so it seemed like a potential "lateral move," albeit one that probably pays a hell of a lot less than software development. I'm suspect that technical writing has some of the same soul-crushing aspects of software development, so perhaps it wouldn't even be worth it.


If you want to become a writer, you should definitely have a financial cushion. Replicating Stross’s success will almost certainly take years of minimally paid hard work, plus a fair bit of luck. Writing is a hard, hard business.

Unless you can write proficiently in the romance genre. There’s gold in them hills.


I wrote my first novel-shaped object when I was 15.

Sold my first short story for actual money aged 21.

First novel sold when I was 36.

... Stopped doing other stuff (mainly freelance computer journalism) to sub the fiction when I was 42.

The sole consolation is that I'm doing okay now and my career didn't really get started until most successful pop stars are at the touring-with-the-greatest-hits-playlist stage (or have given up and gotten a job as an accountant).

On romance; yes, it outsells all other commercial fiction genres combined. But don't assume it's easy money: there's a lot of competition, and every genre turns out to be much harder than it looks once you dig your teeth into it.


Fair point, it’s all hard. I intended to call out romance as a (long shot) potential source of overnight success. Reading the various self-publishing forums, any post that starts with “I earned six figures on my first book” ends with “and it’s a romance.”

I didn’t mean to imply that those authors don’t work hard or that those stories aren’t crazy outliers. Just that there’s more money sloshing around from a lot of avid readers.

As an aside, please keep doing what you do. Accelerando hooked me deep when it came out, and I’ve had a ton of fun reading through The Laundry Files. Thanks for optimizing your career for something other than money.


Re: writing, there are also people making decent money with web serials and Patreon (then publishing to Kindle/Kindle Unlimited). I suspect the most successful ones are exceeding Stephen King levels of output.

You'll find some of them on Royal Road. Lots of fantasy, litrpg, xianxia - fluffy serializable things.


I ran a novel on Royal Road as a serial. Definitely a hard market to get traction in if you’re not doing litrpg or xanxia. I published a techno thriller at 2+ chapters a week for six months and ended up with ~20 readers for the conclusion.

My three reviews were solid (all five star), but if you’re not writing to the site genres you’ll struggle to find an audience. One of my reviewers even mentioned they’d never read a techno thriller before.

Also tried Wattpad and Inkitt - basically zero views on the former, hard to say on Inkitt. Their stat reporting is super bad. Got one review there, so presumably less readers than RoyalRoad.


I've read a few web serials this last year and change (Unsong, HPMOR, and Worm).

They're interesting from many angles, although I think they distinctly lack in mass appeal. Unsong and HPMOR in particular seem to me like the kind of thing you would either love deeply, if they're in your wheelhouse, or otherwise find positively insufferable. (I loved both).

Worm, on the other hand, while in need of some editing, seems like it would be absolutely killer as an animated series in the vein of Amazon's adaptation of Invincible.

One issue that seems to come up in this space is the authors having issues getting into publishing deals. For instance, the author Worm don't currently distribute ebooks because of nebulous concerns about the effect on future contracts. Unfortunately this seems to leave such works in a state of legal limbo.

Idk what my point is, just rambling at this point. I have very much enjoyed all the stuff I've read in this area so far though.


If you are 48 years old, you were about 6 years old in 1980.

I was certainly not referring to you, but people who were in their 20s during that time.


Certainly! I remember being quite annoyed when I started encountering editors/IDEs, that would not let me type light-on-dark!


> the larger portion of the people that started coding on a light text on dark background had already retired.

We're not that old!


Yeah! (shakes cane)

I began coding in the late 80s as a hobby. It led to a career in programming, beginning in the 90s, just out of high school. Turbo Pascal was my first "serious" language. I'm not retired yet.


Macintosh and Sun machines had black text on white backgrounds. Suns had incredibly large screens with lots of dots.

Commodore 64s defaulted to white on blue.

Apple II and IBM machines all started out with monochrome white (green, amber) on black.


Amiga had white on dark blue by default as well, until 1990 (release of AmigaOS 2.0), when they switched to black on grey by default.

There was a long and significant period of reaction to the light on dark UI's where a lot of applications arrived either with light on dark as the default or with no option to change it, so I can understand why someone would see solarized as that start of a change towards (renewed) support for light on dark interfaces, though I do think it's overselling its importance.


I don't remember C=64 colors (friends had one, we did not) but my PC had MSDOS. The text editor was blue background, grey or white foreground. But DOS itself was gray on black. And I used DOS as my OS because it was lightweight.


I'll add WordPerfect v5.1 for DOS.

White text on Blue background.


And Quattro Pro

[dot matrix printer sounds]


I used Word 1.0. Same.


Memories!… QBasic was white over blue, and Turbo Pascal yellow over blue.


My first love for this kind of colorscheme probable goes to KDE's CDE theme (CDE itself being before my time).


> I credit Solarized with setting the baseline allowing proper "dark" themes to be more commonplace.

Apart from the IBM 3279 color terminal, released in 1974, and literally hundreds of devices following (and maybe preceding) that.


keyword being "proper" as in "good", as opposed to "whatever the technical limitations allowed"




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

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

Search: