If you’d like, sure — and I say this not because I’m a diehard AI advocate (honestly, I miss the pre-AI days), but because I think it can be a good workflow for solving your particular problems here, under the right setup.
We have several general AI threads already, including but not limited to:
But feel free to start your own, if you want!
Back when I was working in software, I was something of a toolsmith myself. I would often create a tool for a specific purpose.. And make it available to my colleagues with the caveat: I built this to do XXX. I’m not interested in extending it, it has done what I needed. Play with it if you want…
It was obvious from this opening sentence of your OP that you are far too knowledgeable and skilled. LOL
For example, as a humble user with limited knowledge, I downloaded Linux Ubuntu, followed the simple installation directions, and then followed the simple directions for using it for the one reason I downloaded it in the first place. I have no frustrating conflicts because I know far too little to have any!
Heh, as horrible as I imagine Linux UI work to be… it’s just as bad, or maybe worse, in the Windows world: Microsoft Hasn’t Had a Coherent GUI Strategy Since Petzold | Jeffrey Snover's blog
Huh, well, doesn’t sound like that software was doing its job 
Ha Ha! The point is my tool did the job I wanted it to do. I didn’t write it for yours: but feel free to use it as a basis…
As working software developers, of course, we are accustomed to having to write our own tools sometimes.
In other fields… not so much…
It’s a reason the Hacker crowd has a couple of specific words for people who lack the expertise to tailor Open Source to get what they need out of software:
http://www.catb.org/jargon/html/M/mundane.html
Do “hackers” want good software that actually works and is powerful, or crap?
A bunch of FOSS is supported by mega (and also non-mega) corporations, anyway.
Eh, I have seen open-source projects with a big NO AI CODE rule, and others without. Certainly, it is a tool that competent professional programmers can leverage, while less than useless in the hands of others— and possibly onerous for the maintainers if their project gets spammed by crap submissions. None of which has anything to do with licensing.
It’s because the people writing open source code are doing it because they love the challenge and coding, not because they’re motivated to provide a good, solid, working software application with a well thought out user experience.
That’s largely my gripe with open source stuff. Sure, it’s equally capable as its for-pay competitors, but it’s usually fiddly and annoying to install, and fiddly and annoying to actually use. I’m at a point where I’d generally sooner just pay for software that I’m likely to use often, than mess with open source stuff, because I’d rather pay to not be so annoyed.
I think usability isn’t opposed to power, but tangential. You can have both, but it’s very difficult, and usually not a priority in many open source projects.
True, and sometimes those can become some of the more polished ones (like Firefox, which is basically a Google product now, by funding).
Sorry, I didn’t mean “AI can/should replace human coders in open source”. Just that LLMs can pretty readily solve the OP’s specific problem (of formatting plaintext into a specific format for submission).
That’s the nub, isn’t it?
Of course, the word ‘hackers’ covers a multitude of mindsets. I guess some of them do it because it’s intellectually fun and challenging. Others perhaps to try to show how much smarter they are than the muggles. (And perhaps they are no good at anything else…?)
In my case I ended up as a software engineer because it was a way to earn a decent living doing work I am good at and find interesting… though it was never a consuming passion for me… I have many other interests…
My advice would be to pay for it. This is true even for open source. FOSS projects are often open for donations and suggestions. If you want help with use or a tool better designed for your needs, requesting this from the devs (or commissioning their labour on this) is the best way to go.
Well despite Windows not yet unifying their settings menus under a single graphical scheme it still can boast having a native UI framework (WinUI) the same can be said of Apple (UIKit), Linux devs have two competing frameworks that do not really mesh well with each other. Not the end of the world, but it does add a little bit of jankyness to things.
ETA: Its interesting how Windows keeps all those depreciated native frameworks. For the points they lose in bloat, you can give them points on backwards compatibility. I guess.
You have the option to buy one or write one. As far as writing them, with AI coding tools, there’s never been a better time to write your own tool.
Open source doesn’t exist to provide you with free software. If you happen to get something useful for free, great, but there’s an implicit contract that you will “pay” for it by contributing your own improvements to it, new features or new patches. If you’re not willing or able to do that, then as a free rider you really have no legitimate basis for complaint. If someone shared something because they felt like sharing, and you didn’t like it, that’s a you problem.
And as I already said, open source doesn’t foreclose the option to buy or write your own wares. If you’re unable to do that, this is your own issue, not something that the open source community did to you.
Absolutely. As a music writer and composer, I use Reaper, for example. I am totally happy to pay their very reasonable licence to keep them in business, with new features and bugfixes ongoing. Good workmen are worthy of their fee. You can’t eat code…
Both sides have this same exact sentiment, and are 180 degrees apart on what constitutes powerful and crap. Understanding this difference in definitions is necessary when evaluating lots of software (open and closed source).
For example, most users of ffmpeg want a versatile library that they can integrate into their own project. Even if they have to read lots of documentation, to them, ffmpeg is taking care of the actual hard part of manipulating videos, they have the easy job of figuring out how to call it to get what they want.
Someone else just wants to turn a directory of still images into a video slideshow with some music, something command line ffmpeg is completely capable of. That person is likely to come away thinking ffmpeg is complete crap because there are multiple ways to do it, none of them are obvious, and now it’s just showing the last slide until the music ends.
They’re not really even deprecated per se, more just stuck in a “don’t ask, don’t tell” kinda semi-abandoned-but-still-official state… I wish they would just officially deprecate some so the devs can focus one official framework! It’s because of this mess (and mobile) that React Native and webviews won out over ANY of those frameworks.
Backward compatibility is Microsoft (and Wintel at large)'s double-edged sword. On one hand it gets and keeps them all the big enterprise-y customers, and on the other hand, casual devs and home users are stuck with a 30-year-old Frankenstein of legacy crap mixed in with half-baked portfolio pieces that some Microsoft PM threw in there with zero context. Ugh.
Hardware too; Apple Silicon was able to so thoroughly leapfrog them because they didn’t have to carry all that x86 baggage into the new generation, and even the newer Rosetta will be sunset soon too. Apple doesn’t fuck around when it comes to abandoning old hardware and software, while Microsoft supports them for decades.
And speaking of Apple, even UIKit has similar issues… there was AppKit and others before it, SwiftUI after it, and it’s a mess there too… Daring Fireball: SwiftUI Only Makes It Easy to Develop Bad Apps and Using SwiftUI to Build a Mac-assed App in 2026 - Paulo's Blog
Same in the Web world, which went through PHP → Rails → jQuery → Angular → React → Vue → Svelte → Hotwire → React in like less than a decade.
Ironic that it was Facebook who won the UI wars in the end…
But anyway, I guess UI work is never really “standard” anywhere, and frameworks rise and fall every year… it’s ugly. The intersection of people who are 1) good at design 2) also good at frontend development 3) willing to deal with framework churn across platforms and 4) on a volunteer basis… is quite tiny.
I’m not suggesting using AI to write Python scripts to solve a problem. I’m suggesting to use AI to walk through the steps to build an open-source project and answer questions when there are errors.
@Joey_P was saying that the steps aren’t always clear. Or if there is an error, it’s not clear what to do to fix it. Or explain how you use GitHub, etc.
Not at all. I just started writing fiction a couple of years ago and have only now reached the stage where I am starting to think about submitting some of it for possible publication.
So I am reading up on the rules for this: it appears to be a quite formalised process…
I am sure they want your manuscript to be formatted in a particular standard way, but don’t they want a PDF?? In any case, there are probably infinitely many ways to get there using open-source utilities.
FWIW, I am told that editors and agents are not very keen on .pdf format, and usually prefer a .docx in what is called Shunn modern format.
That is actually understandable because PDF ought to be thought of as “electronic paper”, whereas they might need more direct access to your original text.
I know that in some (many?) circles DOCX has been phased out in favor of formats like ODF, but perhaps editors and agents are not one of them.