Hacker News Reader

Comment by llamataboot | Hacker News Reader

I think there at least 3 things at play here:

#1) Software used to live on a disk you buy. That software doesn't change, until buy a new disk. Then it lived in apps, where you can update or not. Now often, it lives in the mythical cloud where changes can happen all the time.

Now, even on the disk front, things were always changing. Often for the better, sometimes for the worse. In the OS world Dos->Windows->Windows 95 were big changes, and the OS9->OSX change also huge! But now the changes are always.

#2) The entire software world is built on VC money. VC money is not looking for slow and sustainable growth. Or a happy userbase of 10k people. The VC world doesn't mind if 99 companies crash and burn trying to harvest the wind while building their sails if one takes off.

#3) Out of the VC/startup world, large companies must justify their existence and every team and every programmer on that team must justify theirs. No one ever stuck around by saying "everything is good, we literally don't need to do anything or acquire another customer, let's all cut out hours to 2 days a week, keep patching bugs and making security updates, be happy with our current level of subscriptions, and spend a bit of time in R&D to make sure we have other bait in the water too if for some reason our users stop liking this.

View on HN

> No one ever stuck around by saying "everything is good, [...] spend a bit of time in R&D to make sure we have other bait"

That's more or less what FogCreek did with Trello and CoPilot. While FogBugz was "good enough" to more or less run on rails, they made other bets to expand or productize some of their internal toolkit. It worked so well that these side-bets eventually got much bigger than FogBugz.

I wish more people tried that, instead of messing with their cash-cows. Yes, sometimes you have to keep up with competition or "innovate" in your own space, but if you got big already, chances are that you've done your best work already - leave it as it is and move on.

This has long confused me about energy companies. You're big in coal, or gas, or whatever, and see green energy as a competitive threat. Why not take some of those massive profits and treat it as lengthy and luxurious runway to leapfrog into those new markets instead? There's a defeatism there that I wouldn't have anticipated based on how confident and aggressive those companies have historically been.

It is really a shame that there is not a carbon dioxide tax and have it pay out for sequestering carbon dioxide (negative emissions). Sequestering CO2 in geologic formations would fit right in with oil and gas company's core competencies. Many of those companies supported a carbon dioxide tax and they could have transitioned to companies just sequestering carbon dioxide sometime in the future. This could still be done right now, but current politics is more interested in punishing the other side than finding win/win solutions.

You say 'this could still be done right now', but in many cases it is being done already. For example the company I work for [0] spent many years databasing where oil reservoirs were globally. Now we sell the same information for sequestering CO2 in the empty reservoirs :-)

[0] https://www.cgg.com/industry-applications/energy-transition/...

That's cool. Looks like quite a big international operation. Who are some of your customers and places where they are sequestering CO2? I had not heard it being done at commercial scale yet and don't see any examples on the website.

Some are, but it may not be obvious because instead of doing this directly within their company they are becoming venture capital funds for different energy markets.

For example Shell Ventures [0] funds companies like d.light (solar home energy) and Husk (off grid renewable energy)

[0] https://www.shell.com/energy-and-innovation/new-energies/she...

That's basically Shell's strategy.

So now you're Exxon, and you see that Shell has significant investments into oil AND Green energy (specifically Wind right now for Shell). How do you plan to beat Shell at their dual-strategy game?

Exxon double-down on oil. You can't just do the same strategy as someone else and expect to get ahead. You gotta place your bets on a different strategy.

> You can't just do the same strategy as someone else and expect to get ahead. You gotta place your bets on a different strategy.

No, you don't. Betting all on black is just a different gamble, and in my opinion a more risky one. But investors are ok with it because many prefer pure play gambles.

There's no law of nature saying that you have to zig when others zag. A big opportunity like clean energy has room for many winners.

It's a bit different in that space because the "new markets" are seen as a thread that will cannibalize the core business. It becomes a major political issue within the organization.

Not to mention the reputation of the organization might make it difficult to attract and retain the types of innovators and professionals it needs to succeed.

Whereas with FogBugz, they were investing in complementary products, not destroying their own market.

There's a popular book called The Innovator's Dilemma that assesses the phenomenon of incumbents being out-competed by upstarts.

https://en.wikipedia.org/wiki/The_Innovator's_Dilemma

Every energy company that I know of makes substantial R&D investments in an effort to patent the most likely technology paths forward. They all hope to develop a portfolio of patents that will allow them to sue any new competitor that becomes a threat.

Solo devs who want to make a few bucks should look into developing desktop applications. VC companies can't compete in that space, it's not profitable enough.

Only problem is we don't have any good UI toolkits anymore, all are second rate compared to the browser. Qt might be the best, but their licensing scares solo devs.

I want to make a living making desktop applications. However, as far as I can tell, unless you hit the lottery, it's not doable in general, because of a growing contingent of users who won't use your tool unless it's free or open-source (there's a bunch of people running around the internet, many on HN, who think that the existence of proprietary software is intrinsically evil), and an increasing number of devs who will make an open-source clone of your tool simply because yours isn't open-source.

With that, if the only thing that's going to happen is my ideas get taken and used in an open-source project, why not just keep my tools to myself and continue to work for a big corporation?

Three years ago I started a proof of concept electron app to design and print labels on thermal printers. That app has grown into mid-6-figures revenue from mostly “one-time license” sales. I do offer monthly subscription pricing with an expiring license and after a year my MRR has hit $3k. My competition is windows only and closed-source, largely backed by private equity.

I live in a small town in Minnesota so I don’t need to make much money to be comfortable, raising a family. I’m in my late 30s and have done mostly iOS and Node IIOT consulting the last 10 years. That’s what was paying the bills until my electron app started making “real” money.

On the technical side. I get emails every day saying how much a user loves my app. It’s really rewarding. Every year I get one angry email or App Store review saying that there are much better design/desktop-publishing apps out there. These reviews are usually written by technically savvy people that don’t quite get how stupid-complicated label printing can get. Like, don’t expect to design your company logo in my app. But surely, you can drag in your company logo and my app will dither your image for a 200 DPI zebra printer.

My app is basically a no-code image processing pipeline that effortlessly hooks up spreadsheet data, generates barcodes, resizes and fences text, dithers images, dynamically changes colors, and then talks directly to a thermal printer without using a driver.

And you can use it for free for 14 days to see if it works. It’s a single download. Yeah, it’s electron, yeah it uses some memory. But you can share the design files between Mac and windows and it doesn’t require internet for use.

I don’t know. I think it’s a pretty good solution.

Read more at https://label.live

Very interesting! I'm always amazed at kind of creative work I come across on here. Did not expect to read a success story about generating imagery for thermal printers this morning, but I'm glad I did.

Btw, what do you mean "without a driver". Do you bundle code that talks to thermal printers in you app? Is there a standardized communication protocol in this space or do you maintain a breadth of custom implementations?

Also worth noting they in my journey I’ve added the ability to submit the label jobs as PDFs to existing printer drivers. This allows for targeting full color label printers like epson and Primera, but also inkjet and laser printers that use sheets of labels. This is accomplished by rendering the labels on a larger page and then sending the resulting PDF to the system printer. There’s also the option of sending the images to a directory on the file system and naming the files using a column from the spreadsheet. Quite versatile … but difficult to market.

Yes, the app uses node-usb to read and write directly with printers matching vids/pids. Each vendor is a custom implementation with some similarity and lots of differences. It usually involves parsing status to get DPI, busy states, then sending raster image data, and other commands to place the image, feed speed, heat settings, cutter, etc.

>it's not doable in general, because of a growing contingent of users who won't use your tool unless it's free or open-source

Oh, don't worry about that. Trading money for things is an ages-old thing and will remain for centuries no matter what. Just make something useful and put the right price on it. People will buy it if it solves their problem. If someone won't use it because it's not free - they don't really have the need that your product solves. Just ignore them.

This is not as bad on the Mac platform as it is on Windows.

Anyway if a few devs can quickly slap together a functional open-source clone of your apps you might want to ask yourself if your app really provides enough worth to your users to pay for it.

I've been making desktop apps for two decades (proprietary closed source software).

My target audience is businesses, not HN users so I haven't ran into these open source zealots.

Some businesses do ask for open source (politely declined)

Some have also asked to place the source code into escrow for continuity in case the business folds (also politely declined but for some companies this a reasonable ask)

The thing is that people will ask for literally anything and the requests are asinine:

* Rewrite in java\python\ruby\c++\whatever language the IT guy read about last week (nah)

* We need a mac version for the one guy who somehow uses a mac (nope)

* Support oracle (fuck no)

* Support novel networking (lol)

* Make a webapp (lmao)

I'm also not afraid the "hacker" who is going to re-write my app. If theirs is truly better, I will just fork it and crush them :)

Don't let the bobble heads dissuade you. If you want to make a desktop app and if you think the desktop is the correct medium, do it.

Username checks out.

Loved VB6. I have a friend who is still writing and maintaining software he wrote in VB6 for an industrial plant to this day.

Curious, how would this solo dev convince anyone to _download and install_ something on their machine?

The fact that you are even asking this makes me feel so old :-)

Anyway, you'd be surprised. Once you venture outside the HN echo chamber there are large numbers of people who really prefer that their data stay on their own PC. Also, many are concerned about the same problem that this topic is about: having interfaces change without warning is a huge problem for most people.

Convincing them is largely about providing an upgrade path if they need it and not having forced upgrades.

> Once you venture outside the HN echo chamber there are large numbers of people who really prefer that their data stay on their own PC.

AFAICT, that's backward: people with that preference are over-, not under-, represented on HN.

No, I think parent is right. One gets the impression that HN only meets two kinds of computer user: their parents who just use a web browser[0] and tech company or large enterprise employees. There are a lot of people out there relying on highly specialized VB 6 applications to run their business or engage with their entirely-too-serious hobbies.

[0] anyone who just uses a web browser arguably doesn't even count as a 'desktop computer user'.

You're right. I really didn't say that well. What I was trying to get at is that there's a preference for webapps vs. PC apps on HN. Which is strangely at odds with the privacy obsession you find here.

Not sure why you got the impression that I haven't used software in the 2000s.

My understanding right now is that if you have a software that people need to install and your competitors have an in browser offering, your customers overwhelmingly choose the in browser offering.

This has been my experience even though I hate the subscription model as much as everybody else in this thread

Indie game studios have managed it. They're a response to the current era of bloated AAA titles that discourage single-player modes and nickel and dime the user with frequent microtransactions to keep the recurring revenue coming.

Speaking for myself, I have replaced Adobe CC with Affinity Publisher, Designer and Photo. A one-time fee for software I can use forever, as opposed to paying forever regardless of whether I use it or not.

I also have Dreamweaver and Flash-type animation tools (Bootstrap Studio and Tumult Hype) that are also one-time fee apps.

You aren't the only one. A lot of people and smaller shops have abandoned Adobe for this reason. A single, upfront cost is just a more justifiable expense than an upfront cost with a monthly subscription, not to mention, losing access to your work if you miss a payment.

Mark it as "pay once use forever" instead of the infinite cost sink of a subscription model.

Change the model to "pay the subscription fee otherwise we will alter it"

And pray we don't alter [the deal] any further

And more and more, "it will never change, no new features" can be a selling point.

Sure, unfortunately as a solo developer unless your app is a unicorn or you make several of these apps, pay once use forever doesn't pay your bills a year from now.

> or you make several of these apps

Yes, this.

Investing all your energy into features none of your users want for an existing app is a waste of energy for you and an annoyance for the users.

And maintaining the backend of a SaaS can mean more operations busywork for you, compared to a standalone application.

Build something new people actually want.

I'm afraid desktop apps mean a lot of annoying bug/crash reports from users with idiosyncratic local configs (e.g. antivirus software, or they modified some windows registry key a long time ago, because a YouTube video told them to do so, and now don't even remember it etc.).

They would publish it in a trusted App Store.

Do desktop users use these app stores? I'm yet to install a single app from the Mac app Store

My anecdotal evidence says yes. My wife was confused on how to "install an app" for something that wasn't in the Mac app store. She had been used to the iPhone/iPad model for so long that she had forgotten how it was on her Windows PC 10 years ago: if not in app store -> google company/go to company website, download the installer, install program.

I feel like general computer skills have decayed quite a bit in the general public, especially with people who don't work with computers 8 hours/day. See this recent article: https://www.theverge.com/22684730/students-file-folder-direc... The advent of the smartphone/tablet overtaking PC use for general/casual use, e.g. emails, photos, social media, online banking/billpay has really reduced the general public's skill level with desktop operating systems. Probably similar overlap with the subset of people who always dumped every file to the desktop directory instead of organizing files in specific directories.

Give them a choice between a web app, a desktop app (which can be on Steam or the Windows Store too), or mobile apps.

Do you not do this?

No unless you're counting browser, ide and everything CLI I install using brew

Interesting. I have the Adobe Suite, Ableton, a zillion audio plug-ins, ProTools, and the Unreal editor, not to mention DropBox, Outlook, Teams, Slack, Bear, excel … This is just from memory without looking at my apps folder.

Even for IDE I have XCode, Webstorm, Sublime …

In general, native apps are still far superior for most tasks I do.

What if this something did not need installation? Just download and run.

If it fulfills a business need it's a lot easier.

Unless you’re doing Linux applications you have all the toolkits you need. WinUI for Windows and Cocoa for the Mac, done. They’re not cross platform, they might not be all they could be, but they are the default and they work.

I believe you’re right in that there are money to be made in developing desktop app, but I also believe that the customer base expect thing to be native, so don’t try to do Windows application using GTK or macOS applications using wxWidgets.

>They’re not cross platform

Which is why everyone and their dog is going Electron. Linux compatibility is one thing (which no one really cares about, except for software in the public services sector), but users generally expect that their software is available on a Mac - not to mention that people using a Mac tend to be on the better-earning/willing to pay actual money for software side.

I disagree. I think the troubles of cross-platform GUI are overstated. People used to port software to a dozen different 8-bit micros and 5 game consoles using assembly, writing for Win32 and Cocoa can't be as difficult as that. Besides, it will incentivise you to make your interfaces simpler, while still providing you the ability to leverage platform specific features.

I think everyone is moving to electron because web developers are a dime a dozen.

True, but beyond VS Code name one Electron application that’s actually good.

Companies want their application on two platform, use Electron and now their product suck on both.

Electron is actually a reasonable solution for internal and bespoke software. It just not at a level where I can see it being something I’d want to pay for as a consumer. Again VS Code is pretty good, but it’s still clearly not a mac app.

Discord and OBS come to mind.

Discord certainly works, but when I see a chat box lag on an i7 8700k with 32GB of memory I die a little inside.

And that happens occasionally with Discord, without running make -j42.

Discord is awful to use a desktop application. It's sluggish and cumbersome, not snappy and lightweight like a native application would be.

I'm fairly sure that OBS is writing in C++ using QT.

Ah, I was thinking of this: https://www.electronjs.org/apps/streamlabs-obs

My dog is actually doing Mac only, since that’s where the money is.

I’m really tempted to learn Swift and the macOS UI toolkits just because I’m 100% on macOS these days and there’s a few things in my life that I think could be better as a stand-alone desktop app rather than a web app or Electron app.

SwiftUI is really nice but I've found performance to be questionable (they seem to be improving it over time) and there are some things that are impossible to do in it that are absolute must haves for some application types. For example, there is no way to change the tab order of which text entries get focus.

unrelated, but nice username!

Cheers, lots of things from San Francisco get featured on HN but the Grateful Dead have to be my favourite!

> all are second rate compared to the browser

As a non-web developer I'm genuinely curious, what is a good UI toolkit targeting web browsers/Electron that rivals Qt?

With my next business I’m aiming to do the opposite to number two, I want to address issues in a very old, traditional market and find a few thousand customers delighted with a reliable slow moving product that solves a big issue for them, and if the product were to ever get “feature complete” that would be great afaic.

If I ever get around to making a SAAS or whatever, all new features will have a toggle with a prompt letting existing users choose if they want to enable it or not, including interface changes. (And new users can disable features they don't want to keep things simple. KISS seems to be completely ignored in big tech.) If I can't program cross-compatibility (keep things modular), then I suck as a programmer. If Word Press can have themes, extensions, or whatever, so can I.

Sounds good, but impractical from many standpoints the least of which is the programming aspect. Options for users are the the opposite of KISS (but very nice for them).

If the SaaS is complex, this becomes impractical if toggles for "all new features" is the goal. Maybe a subset is doable.

If it's a simple SaaS by all means go for it. I really like the idea of being able to toggle things I like/don't like. But most of the times I become a "second-rate" citizen in the eyes of devs when I'm on a lesser-used subset of features.

Example is this service I'm using to manage my small company's finance; because I refuse to toggle on a feature that makes life easier for most people (but not mine), getting support for bugs has been very difficult.

Good luck hiring enough people to maintain a billionty different versions of your service

Websites don't run multiple versions of Wikipedia do they? Everyone would be on the same version, they'll just see different things.

If there are 30 feature toggles, there would be 2^30 ≈ 1 billion possible combinations.

Finding the combinations that result in something broken (through some unexpected interaction of features) would be impossible via QA. I guess you could keep a list of the more popular combination of settings and test those. More likely it means you'll let the customers do the QA for you. The customers will not like that.

Not sure if you've ever worked on complex systems before, but features simply don't exist in isolation. Features interact - sometimes in complex ways, sometimes in unexpected ways. The cost of supporting user-facing toggles for every feature in your product grows exponentially.

If you've ever maintained a system that uses developer feature flagging, the cost is large and typically this is for feature flags that should only exist for a matter of weeks or less.

This is versioning for TeX. It is the value of pi with added significant digits for each new release.

And when the author passes away, the final version will become exactly π, and all the remaining bugs (if there are any) will become features.

This sounds delightful. Best of luck with this!!

Well come on, what business is it (if you're happy to share at this point)? Some people will want to prioritize it just because of this perspective.

It’ll be in the manufacturing space building on an app I wrote for my own factory a few years ago

I think it's also the cadence. You alluded to it in #1. I used to work for a traditional software firm years ago. We actually cut CDs and shipped them. We would do major releases twice a year and patch releases twice a year, yielding 4 releases a year. That was considered a lot. Now it's releases every 2 weeks or so.

The nice thing about installing your own software is you controlled the updates. If there was no compelling reason to upgrade and you were stable, you didn't. Most people today could probably break out a copy of Office 97 and be totally fine with it.

#1 is important since you can't generally opt out of the change but I think it is largely #3 which speaks to the rate of the change. Everyone has to justify their existence. Everyone has to justify their bonus. The C suite has to present something to the board each quarter. Making great money is not enough, we need infinitely more. Lots of times the revenues go up but so do the costs and the profit goes down. Then there is even more pressure to add more crap to get the profits back up.

> 2) The entire software world is built on VC money.

I don't think that is the case (depending of course on how we define "the entire software world") I've been a developer for 20 years and never worked for a VC funded company.

There is a tonne of software development outside the realm of VC's.

#4) Attention is the new money.

For a lot of software, like social networks, getting new users and new downloads is the main goal.

When a company is not charging [most] users money, the second best thing is publicity or grabbing and selling user's data. Both options require user churn. Keeping users happy for a long time is not their concern.

Unfortunately this is happening in the FOSS world as well. Constant churn of software products, features and bugfixes to keep users addicted.

> Unfortunately this is happening in the FOSS world as well. Constant churn of software products, features and bugfixes to keep users addicted.

I suspect there's so much churn in FOSS, because developers love rewrites and starting from scratch. In the world of commercial software, written by companies, the bosses/owners reign them in. FOSS, on the other hand, does not have bosses or owners...

No, most churn comes from company-driven FOSS, especially around devops stuff.

Among other reasons, they use churn as a way to control the userbase and prevent successful fork.

> FOSS, on the other hand, does not have bosses or owners...

Corporate-driven OSS has plenty.

An extremely exhausting world, with no end in sight. I just wish all the money would run out or something, give (most of us) a break.

Maybe we need a name for the opposite of software that is exposed to such a pressure to change: Stableware.

#4) The dream of reusable components is here. Every major dev system from Emacs through Rust has a vast library of component dependencies. These things are a swirling unstable mass, leading to spacetime fluctuations up at the human level such as supply chain attacks, mandatory upgrade cycles, and orphaned deps. Then the complexity leads to wrapping standard deps to standardize them, leading to multiple standards...

You know it's bad if there's multiple XKCD's about it.

https://xkcd.com/2347

https://xkcd.com/1987

https://xkcd.com/927/

As a Python developer who just bought the wrong USB to USB-something (whatever IPhone SE uses) cable just last week the second and third links you provided hit very close to home. I don't know anyone from Nebraska just yet, though.

#4) We aren't innovating in the same way that we used to so many of us are doing rather pointless tasks to justify our existence.

I secretly think every major change in every software is somebody's L6 promotion

This is bizarre analysis given that one tool being talked about here is Microsoft produced, which has nothing to do with VC money.

Yes, the focus on features, and ongoing lack of bug fixing, is maddening. This applies to big companies as well as small.

VC = Venture Cancer?

Reply on news.ycombinator.com