I would like to open a thread to discuss the idea of open-sourcing and distributing softwares to the mass (I have mostly github in mind).
There are obviously strengh and weaknesses, let's discuss them here.

What do you think nixers, about sharing your code to the whole internet and letting everyone review/submit changes to it?

Your real aim is to discuss Github.

Open source used to be something esoteric, the guys knew each others or stumbled upon them.
With the advent of the internet we've seen a rise.

What used to be exotic is now common.

Search engines have pushed this to a new limit, we can now freely access anyone's open source software.
Another issue with open sourcing is that you have to find a way to distribute it, make it available online for others to download.
The only way available is by using a friend's or community server or your own.

Github is one of those place but unlike ftp servers or whatever, it's a closed source ruby based webserver.

The great thing with web ui is that they make it easy for non-very-technical people to download and check projects.
The annoying thing about a closed source Github is that it's completely paradoxical and an oxymoron of what their project actually stand for. It makes you genuinely question it.

Maybe it's related to how "corporate" they want to appeal to companies.

Back to today.
Open source has become a hipster thing, we see a lot of "unicorn ninja open source evangelist"-whatever which gets on everyone's nerves.
We are seeing a bunch of companies suddenly coming out and releasing their softwares as open source, mainly for branding and reputation.

Open source is usually coined with "stability" and "security" because it lets everyone asserts if the code is fine or not. However, if a company suddenly releases a code that has been worked on for more than 10 years and expects others to contribute to it this will rarely ever happen. Open source projects go through long periods of time and gradual improvements and everyone agrees to the changes.
The only thing the people can profit from when a company open sources their software out of the blue is to be able to fetch some important parts of the code they have or to install the software for free, if it wasn't already.

I think I've developed enough to get a conversation going.

As a general rule I think the sharing of information is always a good thing. Especially if the choice is analog. This is a science, and some people operate from that perspective, and there are others who understand the science and create/execute solutions with artistry. Information and discussion makes for better science, and art is pretty much someone/somethings expression and interpretation of information. I know I've benefited from having an easily accessible library of unique peer reviewed solutions where I can get to the heart of the thing and bypass all the white noise of a google search, or a forum (not this one). I think sometimes it can seem like that theory operates on a bell curve though..

(18-07-2016, 12:32 AM)venam Wrote: Your real aim is to discuss Github.

Definitely.

I think that releasing source code in the wild is a good thing. I don't like the way github do it.
First of all, it makes you rely on a private company to hold your software. If tomorrow github decides to shutdown, I've found no mention in their policy that they will let you retrieve your own code. That's actually not a problem if you are aware of it though, and take special care to have backups elsewhere. After all, it's no different than bitbucket or sourceforge.

The second point bothering me is how they put forward people instead of code. Code is organised around its author, not around a project. This means the only thing that can differ between two projects is the name of its author. That might only be me, but for an opensource code-sharing plateform, I think code is more important than who wrote it.
Just take a look at someone's page. There's a big picture with their avatar, graphs, stars, followers, etc.. but your need to click on a tab and scroll down if you wanna see the actual work of the author! For a website that is supposedly code-oriented, it shouldn't be this hard to see code.

Third, their "fork" policy. They make it so easy to steal code from other people that it bother me. The idea behind forking is to make it easier for people to contribute. In the end, it makes it easier for people to get credit for someone else's work.
It's easy to fork and not contribute, and you see that a lot. From personnal experience, a bunch of people forked some of my repos, changed a few things they didn't like, and advertised it as a "Fixed version with less crap" or whatever, without even noticing me about the original issue, or submitting the changes. That is a behavior I hate, and github just makes it too easy to perform.

Finally, I hate how they put "fame" forward. With the stars and all. Open-source should be about sharing code, not getting fame for it. I've seen a lot of people submitting shitty/useless pull-requests, simply to get their avatar shown in the "contributors" page.

I much prefer the "old-school" way to share code, using git-daemon and a website, on a custom server. This way, you're responsible for your code, and people refer to YOUR website to access/modify/fix the code. And when they send patches, they do so to get something fixed rather than having their name mentionned.

Just for sport:

If we're being honest, they allow a standard sized photo (of whatever you like) and some informational bits. The rest of the site is dedicated to the data. It's really not much more than a "Hi My Name Is" and theres no requirements you personalize. But sure, I understand your point. Code is king and more important than the person behind it.

(18-07-2016, 03:46 AM)z3bra Wrote: z3bra Wrote:
From personnal experience, a bunch of people forked some of my repos, changed a few things they didn't like, and advertised it as a "Fixed version with less crap" or whatever, without even noticing me about the original issue, or submitting the changes.

But I thought we established code takes precedent? Im sure this isnt the case, but if someone took your code and created something you couldn't imagine and the only credit/notice you got was being the crap they shined up are you content that the better code lives on?

Also I wish I would have picked up on the fact that this was more of bone to pick with github thread.

I think Github goal was to push forward the social aspect of coding, adding features such as commenting and who did what contribution.
It's more oriented towards making it easy to manage for companies.

That's the reason why:

(18-07-2016, 04:37 AM)oxa Wrote: If we're being honest, they allow a standard sized photo (of whatever you like) and some informational bits. The rest of the site is dedicated to the data. It's really not much more than a "Hi My Name Is" and theres no requirements you personalize. But sure, I understand your point. Code is king and more important than the person behind it.

Users need the minimal info there.

Unfortunately, z3bra is right, a lot of people are attracted by fame and will do things to get their name up there.

(18-07-2016, 03:46 AM)z3bra Wrote: Third, their "fork" policy. They make it so easy to steal code from other people that it bother me. The idea behind forking is to make it easier for people to contribute. In the end, it makes it easier for people to get credit for someone else's work.
It's easy to fork and not contribute, and you see that a lot. From personnal experience, a bunch of people forked some of my repos, changed a few things they didn't like, and advertised it as a "Fixed version with less crap" or whatever, without even noticing me about the original issue, or submitting the changes. That is a behavior I hate, and github just makes it too easy to perform.

I don't agree that this is Github specific.
Sure it has made it easier to fork/copy someone else's work but the "stealing" aspect or credits appropriation has been there before in a lot of other places and not only in the programming field.

The social aspect has probably amplified this a bit, especially the hype I discussed before.

On the other hand, this has brought a lot of good stuffs.
Let's not hide under a rock and deny what the popularization of open source has given us today.

(18-07-2016, 04:37 AM)oxa Wrote: But I thought we established code takes precedent? Im sure this isnt the case, but if someone took your code and created something you couldn't imagine and the only credit/notice you got was being the crap they shined up are you content that the better code lives on?

I'm certainly not perfect, and neither is my code, nor are my ideas. For each project I create, ithere are issues, bugs, and it lacks features. I open-source my code, because I want people to use it, and help me fixing issues. The problem with github "forking" is that it distributes the efforts between multiple persons/organisations. Take the "ed25519" library on github for example. Original implementation is by orlp. There are at least 2 significant forks: rdeker's and evilrix's. One making it compatible with other implementations, the other adding support for signing in chunks. Which one should you use? That's 3 implementations of the same software in the wild, each being "valid". If you look at the fork graph, you can see that some of these forks have been forked, and so on, resulting in even more "valid" implementations. This leads to a very confusing state, where if you want a feature, you start relying on a "not-so-unofficial" version, which might at some point disappear, breaking your software.

On the other end, take a software like [dwm](http://dwm.suckless.org/). There is a specific website for it, the code is hosted in the same domain, and to contribute, people send patches. There is only patches applying on the original code, and no full copies of it available for download. It means that there is only one code to rely on, which is better for the users.

I have seen a lot of effort duplication in the open-source world, and that's what bother me the most. I find it irritating that some people find bugs in your code, figure out a fix, and decide to shit on your head by simply saying "Hey, I fixed his code, use my version now", leaving the author with a buggy version.

---

(18-07-2016, 06:43 AM)venam Wrote: he "stealing" aspect or credits appropriation has been there before in a lot of other places

That's true, of course. But I personnally think that github makes it even easier to do so. I do agree though, that github helped a lot to promote open-sourcing, and I'm grateful to them for that. But hey, as the proverb says:

Quote:You either die a hero, or live long enough to see yourself become the villain.

To go back to the original topic, I think opensourcing to the mass is a good idea. It helps improving the software a lot, getting new/fresh ideas and so on. But special care needs to be taken to avoid losing control over the project, and that's a pretty hard thing to do.

(18-07-2016, 08:42 AM)z3bra Wrote: I have seen a lot of effort duplication in the open-source world, and that's what bother me the most. I find it irritating that some people find bugs in your code, figure out a fix, and decide to shit on your head by simply saying "Hey, I fixed his code, use my version now", leaving the author with a buggy version.

Copyleft licenses, anyone?

EDIT: Even our old UH license has it.

I don't think any licence forbids contributors to maintain their own copy of the software. I might be wrong though, I suck at licensing

This question binds together with the fame of a project; a small sub-10-stars project will not have as many "fans" as a big project such as neovim has got, and thus, an act of someone "taking over" a small project will in many cases go unnoticed by the public.

That being said, what power (if we exclude the public backlash that can occur) upholds the content of a project's license? Does Github do that?

i personally like github. and nothing is stopping you from having a mirror.

github as a "social network for programmers" is a cool idea in my opinion. having your public commits be published can give your small project a higher visibility. like, just searching new commits for a given language will show you all kinds of projects you may or may not have an interest in that you might have never seen. github has a team of people who keep the server running (they have a pretty decent uptime) and a whole team of security people working to try and keep your stuff secure, which you'd have to do on your own with self hosting.

i completely agree about have your code hosted by a company, but i dont see sites like savannah.gnu.org and repo.or.cz being much better. and github has a much better reputation for quality than sourceforge or bitbucket. gitlab being an open-source project that is self hosted, is the one i think might be the best alternative, but guess what? it's also run by a company gitlab LLC.

I must admit that the ability to "browse" random project is pretty good.

I'm not against using these platforms (github, gitlab, sourceforge, ...) to share your projects. They do help promoting. One just has to make sure they don't lose control over their project.

(18-07-2016, 03:46 AM)z3bra Wrote: Third, their "fork" policy. They make it so easy to steal code from other people that it bother me. The idea behind forking is to make it easier for people to contribute. In the end, it makes it easier for people to get credit for someone else's work.

I have thought about the question if there was a cryptographic method
of proving that some information was created by a person and only that person.
The only method I could think of is blockchain, which is quite complicated and
requires some motivation to verify the chain (some sort of money). This is too
hard.

(18-07-2016, 03:46 AM)z3bra Wrote: I much prefer the "old-school" way to share code, using git-daemon and a website, on a custom server. This way, you're responsible for your code, and people refer to YOUR website to access/modify/fix the code. And when they send patches, they do so to get something fixed rather than having their name mentionned.

I am also quite convinced that the old way with patches over e-mail or mailing lists for big projects.
But I think it is a problem to assume every programmer or open-source enthusiast is willing to set
up a git/web server. It is simply not convenient, and it's not easy.

I would like to see a decentralized way, such as with ipfs, simply cloning from ipfs with git,
writing your patches and sending them per email.

But centralization is of course always bad. Github is not beneficial, it's a company,
and I have a deep natural distrust for companies.

(22-07-2016, 07:03 AM)pranomostro Wrote: I would like to see a decentralized way, such as with ipfs, simply cloning from ipfs with git,
writing your patches and sending them per email.

Git itself is decentralized! You can easily setup a local 'git-daemon' server that people can use to clone/pull from your own code.

But I still need a domain name if I want other people to be able to clone, don't I?

And I perceive the action of 'getting a domain name' as difficult. Maybe that's because
I have to pay money (if I understand the system correctly) and interact with other
people.

Maybe I'm just lazy, but this is way harder than using github.

You don't have to interact with other people, and the money you have to spend isn't really a lot. But I agree, github is way easier.

An URL is nothing but a simpler way to remember an IP address. You can easily NAT git-daemon port on your home router, and get people to fetch your changes using the public IP give by your ISP

I disagree with Github's new management's "ethical reasoning" (all that feminism and social justice humbug), I publish all of my new stuff on Bitbucket, licensed under the WTFPL. Which leads me right to OP's question: No, I don't think "the mass" has any advantage of being able to read their kernel sources, because, seriously, who cares? The only reason I put my stuff into WTFPL licensing is that I really don't care what is done with it. I develop software, people may or may not like and/or use it. That's all that matters. I don't care if anyone takes it and adds his name. I don't lose profit anyway.

(25-07-2016, 03:46 PM)jkl Wrote: I don't care if anyone takes it and adds his name. I don't lose profit anyway.

I used to think the same. But it now bother me when people use my work to entertain their ego. You have all my respect though. I wish I could detach myself from the code.

My code does not disappear when people take and modify it, is's still mine, published in my repositories and saved somewhere in my backups. It's an honor when people fork and modify it. One of my old projects on GitHub has new forks every now and then. I randomly look into them and cherry-pick their changes if I feel like it. ;-) Of course, as my repository is the official one, the chance to drive away traffic from me is rather small.

Sure, I never said so! At this point, I think it's only a problem with me having issue dealing with assholes that think they can just go famous by selling my code as their own. I used to think I could go with it, until it happened. Maybe it just bother me when jerks can satisfy themselves!

So your problem is that it's not you who became famous?

(26-07-2016, 03:32 AM)z3bra Wrote: Sure, I never said so! At this point, I think it's only a problem with me having issue dealing with assholes that think they can just go famous by selling my code as their own. I used to think I could go with it, until it happened. Maybe it just bother me when jerks can satisfy themselves!

Again, licensing solves all this.

(26-07-2016, 05:12 AM)venam Wrote:

Quote:Sure, I never said so! At this point, I think it's only a problem with me having issue dealing with assholes that think they can just go famous by selling my code as their own. I used to think I could go with it, until it happened. Maybe it just bother me when jerks can satisfy themselves!

Again, licensing solves all this.

I know, and that's why I switched from WTFPL to MIT for my code.

The MIT license doesn't disallow anyone to take your code and make a shitload of money with it. It just forces them to place a note somewhere in the README (as in: WILLNEVERBEREAD) mentioning you.

(26-07-2016, 05:12 AM)venam Wrote: Again, licensing solves all this.

I haven't had any experience of somebody trying to steal my code (I'm not famous enough, I guess :P)
But I don't think anybody really cares about licenses if they are willing to market other people's code
as their own.