The Linux Labor Force (editorial)

Stan Seibert has written a piece addressing a very key issue in the Linux community. In the beginning there was Linus. Back then, 100% of the users contributed to Linux. As it grew, the percentage of programmers dropped as the number of 'end users' grew. This will only continue as Linux grows. This is great because it means that proves that Linux is more than just a hacker's OS. But the growing number of users unable to actively contribute puts more pressure on developers (email pleas for help, bugs discovered, feature requests etc). The Linux community needs to adapt to this in order to succeed. Hit the link below and read what Stan has to say about it.

The following is an Editorial by Slashdot reader Stan Seibert

Harnessing the Linux Labor Force

By Stan Seibert

The Problem

Recently the author of Kirc expressed his frustration with answering all of the bug reports, installation questions, and patch submissions. Under the all of these pressures, he has now decided that he can no longer continue the development of Kirc. His frustrations are probably not unique; most developers are feeling the crunch of increasing demands on their time.

Ironically enough, this probably results from the increasing popularity of Linux. As Linux receives more press, more users try it out and experiment with it. Unfortunately, this creates a culture clash in the Linux community. Windows users (as most of these new converts are) pay for software and expect tech support; there is no other interaction with the developer. You give them money, they give you help. The protocol is part of the DOS/Windows culture. Even in shareware, most authors will only provide tech support to registered, paying users. The Linux community does not work that way (see Eric Raymond's Homesteading the Noosphere).

Linux (and all associated software) originally was written by programmers for other programmers. Since nearly all of the users were also developers in some way or another, the community was ideal for the free (think beer and speech) exchange of ideas. Programmers are used to digging through sparse documentation and cryptic source code files to find the answers to their questions. When completely stumped, they can at least phrase their question precisely in terms meaningful to other programmers, making it much easier to answer. Thus, everyone was able to use each other's tools with little or no burden on the developer, and in turn, made their own programs available.

The shifting profile of the Linux user now upsets the balance. No longer a Unix programmer, the new Linux user has a pretty good sense of how computers work, learned about computers on a Windows machine, and does not know how to program fluently. Such a user is probably baffled by the confusion and chaos that surrounds Linux development. Relying on their prior understanding of user-developer relations, they assume that software authors will have the time and inclination to provide them with complete, beginning-to-end tech support. As the number of users increases, this becomes harder and harder for developers to do. Paradoxically, the popularity of Linux could actually halt development.

Yet, if the proliferation of projects is any indicator, we have not reached that point yet. Here on Slashdot, we hear about new projects everyday to fill some software niche. Though some will succeed, many will fail, not for lack of ideas or worth, but for lack of manpower. It is probably safe to say that every live project in existence could use more help. I have lost track of the number of pieces of software still stuck in the "0.1 pre-alpha" stage of development. These projects could use an infusion of talent to help them get where they need to be.

The Solution

I do think that the changing situation can benefit Linux greatly. My solution to the problem relies on three assumptions:

1. There are lots of Linux users.
2. There are many development tasks that don't need programming skills.
3. Most Linux users who would be willing help do these development tasks.

The first assumption has been well established by surveys and polls. For evidence of the second, we simply need to look at the frustration of people like the Kirc developer. These projects need tech support, documentation, and targeted beta testing as well as coding. The third is just a guess. I think most users would not mind donating a little time and talent to helping the Linux community. There are probably numerous users who would like to help but do not know how to get involved. We need to change that if the Linux community is to survive. Since not every Linux developer can be lucky enough to get paid full time for what they do, we have to distribute the work of development among more people.

I want to create a web site that will act as a sort of project clearinghouse. Project developers could register and periodically update their projects with a database, including the skills they currently need. Users would have a place to learn how to get involved in development. It would have resources to help users educate themselves and learn new skills. They could also register themselves as potential volunteers, also listing They could also register themselves as potential volunteers, also listing their skills and desired type of work. At this point, some nifty Perl scripts would crunch and munch the two databases together, producing a match between talent and project. This would be a powerful way of leveraging the Linux Labor Force (as the title of my article calls it) to improve the Linux community. A developer that needs people with network programming skills, documentation ability, and Ensoniq soundcards to test their network sound system could automatically be informed of people who could do these things. With a better host than my ISP can provide, we could even expand the services to include roulette-style mailing lists (as I saw suggested in a comment here several weeks ago) where a single e-mail alias randomly distributes questions on a given project to one of several qualified people. All registered projects could provide a list of tech support people and an alias (ex: [e-mail]) would send general project questions (installation, etc.) to one of the people on the list. We could even have the mother of all such mailing lists, a [e-mail] (these are only example addresses) that would instantly forward a newbie question to one of a thousand Linux gurus around the world.

I don't have time to start this for another 2 weeks (until after finals), so I'm looking for someone willing to help me jump start this. I have 50 meg of web space, though I don't know if I can execute scripts, and I am certain I cannot set up any mailing lists without paying them some silly amount. So I'm looking for someone to help me who can set up a discussion list and a preliminary web site to get this thing started (See, I already need this site. :) ). So far as I know, other people have tried parts of this idea, but none have integrated the parts together or implemented the script idea.

Your input

• What do you think of this?
• Has it been done?
• What should be included/excluded?
• Are some of you out there uncertain of how to help linux development and would like this sort of think?

I am interested to know what the Slashdot readers think. Please email me if you want to help or have any longer ideas that wouldn't fit in a Slashdot comment.