Members for Democracy Archive

Open Source Unionism

I think this deserves its own thread. I'd like to focus both on open source unionism in the analogical sense that the authors of the essay intend (a method of creating networked unions), and in the literal sense of open source technology (a method of developing and licensing software) being a valuable tool for unions.

First off, in the other, rude thread, the creation of a union-friendly variety of Linux was proposed. Let's withdraw that. It's probable that among union locals you'll find a wide variety of legacy operating systems, probably going back to Windows 3.1, if not further (shudder). So rather than propose the development of a new OS and the subsequent upgrade hell, let's turn the focus to the development of open-source, platform-independent applications.

Currently, some rather expensive, proprietary solutions are available: check out UnionWare, a software company that builds tools that track union membership, dues, grievances, events, and registration; also check out CTSG, the AFL-CIO's choice of technology solutions, and rumor has it, the company that will be remaking the UFCW International Web site.

CTSG products include:

• Web Action Center -allows clients to easily create web pages, allows visitors to easily email or fax congress members, and contains a directory of Federal and State Officials, searchable by zip code, with key contact information, committee memberships, biographical data, staff information, and other data
• Activist Relationship Manager -Clients can send out targeted emails based on issue preferences, demographics, electoral districts, counties, zip codes, previous action taken, or desired frequency of receiving alerts. These emails can be personalized via email-merge to include information such as Congressional District, previous activist activity, etc. It also has an email to fax feature.
• Content Management
• ePetitions and ePostcards
• Interactive Dynamic Content -Localized Content, Polls, Online Stores.

The price for a UnionWare solution starts at base-fee of $9,500 with monthly billings of $450. I would assume CTSG's services are staggering higher.

Much of the functionality of the software these companies provide can also be found in free, open source software. The software that runs RetailWorker.com and BordersUnion.com has modules avaiable that let you send faxes for free internationally, send ePostcards, conduct polls, build mailing lists, manage expenses, track bugs (aka grievances), build job databases, manage contacts, build FAQs and knowledge bases.

The problem is, these modules weren't created with union-specific tasks in mind, and the functionality of the modules isn't integrated - they don't talk to one another like the software components offered by the above companies do.
Developing union-specific functionality for these exisiting applications, and integrating them, would require some sort of organization. That's why I proposed that the AFL-CIO announce its support for the open source concept and create a team of developers to create tese tools.

If you don't like the AFL-CIO, maybe it could be some other organization, but that organization would have to be big, because it has big enemies - media companies like Disney and trade associations like the RIAA who are engaged in an assault against the free flow of information via intellectual property legislation. The Digital Millenium Copyright Act was just the start. Senator Fritz Hollings' Consumer Broadband and Digital Television Promotion Act, in its present form could make the creation of free software legally and economically prohibitive in the US. Open source proponents need their lobbyists in Congress too, and since big labor stands to benefit from the availability of open source technology, why not add its voice to the fray?

By doing so, the labor movement would not only earn the respect and services of free software creators, but would also lay the technological foundation for the vision of "open source unionism" expressed by the authors of the piece.

quote:

---

Modules? .. union specific tasks? I don't get it.

---

Beats me. Maybe some of our techie friends can jump in here. It seems to me, though, that even within the Open Source Unionism discussion topic there's a technical thread and a social thread. Then again, the two depend on each other, too, because the social aspect of the union can only go as far online as technology allows, and technology developers need to ask the social questions so they know what sort of new technology is most needed.

I'm going to repost my questions from the Revolutionary Unionism thread, although I still think they're mostly rhetorical for now. We don't have any Open Source unions right now to look at for examples. The most we have now are a few stirrings that there's got to be a better way of doing things.

As an online union community emerges:

1. What will be the relationship between the membership of this online "virtual" community and real shop-floor worker communities, including conventional union-represented bargaining units? How much do or will they overlap?

2. Are there union services that can and should be provided specifically to an online community membership? What kind? What would be the relationship between these services and the traditional services provided to bargaining unit members? How should unions pay for all of this? Will they charge some of their online members dues? Or perhaps fee-for-service? How much should "virtual" members pay compared to shop floor bargaining unit members?

3. When these projects are undertaken by some entity other than an international union, to what extent might this project encroach upon their traditional jurisdiction? Does anyone here care if that happens? If and when conflicts arise, how will they be resolved?

I'd add as well, particular to the second and third point, that there's no reason why open source unions have to evolve from traditional unions. Traditional unions have the resources to make things happen more quickly, but they also have much greater organizational inertia to overcome.

Maybe the first real open source unions will evolve from purely virtual communities like MFD or retailworker. This is thinking really big, but what if some of our communities matured enough that we felt we wanted to do more than just get together and talk - that we wanted to undertake and coordinate sustained workplace actions, or that we wanted to start providing some conventional or unconventional union services through our web portals. The same questions would apply to us. Would we want to set up some kind of workplace or geographically situated branches? How would we deal with other unions and community organizations that felt we were encroaching on their "turf". How would we pay for all this?

Let me end by apologizing to Siggy. I couldn't answer your question, so I just fired off a range of unrelated questions of my own. Is there a way we can talk about both the technical and the social, bringing the two together in one thread?

quote:

---

As I asked in the other ("rude") thread; What are union-specific tasks?

What technology does a union require to obtain open-source-unionism functionality that doesn't already exist or isn't already in development by OSS developers?

---

Nothing very mysterious. For example, rewriting a help desk request tracker as a grievance tracker, etc...

A module is a software component (e.g. a chat room app) that can be easily installed or uninstalled without altering the main code of the core application.

quote:

---

Nothing very mysterious. For example, rewriting a help desk request tracker as a grievance tracker, etc...

---

Given a sufficient spec and a coffee-maker, you, me, and a boatload of O'Reilly books could do that in a weekend!

quote:

---

This, and the nature of OSS, IMO makes a centralized dev' team redundant

---

Who's advocating a "centralized" Dev team? You are. You're advocating each local as a center of development. I'm advocating a traditional open source project: the team is spread around the world. It just happens to be sponsored by a large organization.

The software would be scalable and modular, easily customized to fit the needs of any size local. Something like a cross between ZOPE (an application server platform) and postnuke (your basic Web site management software).

Actually, Postnuke utilizes a database abstraction layer, so it can connect to PostGre, oracle, and the rest.

But the team would be comprised of developers from locals - and whoever else volunteers.

Postnuke Core project has 121 registered developers and a team of volunteer managers.

Its modules projects has another 81 developers

Its theme project has 15 developers.

And then there all the unofficial developers who take the code and hack it for their own purposes.

That's the kind of project I'm talking about.

I disagree. I think your unthinking rejection of any centralized approach is kneejerk anarchism.
You won't even consider it.

Open source projects do occur within centralized organizations. The Zope project is run by the Zope Corporation, for instance. I think some labor organization should hire a team of project managers to get the ball rolling.

quote:

---

posted by weiser:
[QB]I see where you're going, JD. When you look at today's unions, you see poorly-run businesses. That's where unions are contributing to their own rapid demise.

You seem to be saying that with good "administration" everything will be all right with unions. If they are run to best business practices, they will flourish.

---

That has absolutely nothing to do with what I said. I was talking about the organization of a software project. Big projects need project managers. Good project managers, often but not always, like to be paid for their work. Once again, it should be clear that nobody owns an open source project, even if someone pays for its development. Having an organization manage the project will not give that organization ultimate control, it gives it provisional control, which if the managers screw up, will most likely result in a development fork - i.e., a group of developers with a different vision will take the code that has already been developed and mold it to their own specs. My contention is that the best way to get this project initiated is to have some kind of official sponsorship which would create publicity and the opportunity to influence law regarding the intellectual property aspects of open source software development.

Now, on the other tack, the structure of unions, the essay on revolutionary unionism was long on hope and short on specifics.

Here's a piece that argues against rev unionism and the idea of a decentralized approach:

http://www.law.nyu.edu/laborcenter/IUESSAY.2.html

It ahs mainly to do wiht reform of American labor law, so it may not mean much to most of you.

quote:

---

posted by slek:
If you can explain to me how a centalized team of project managers, sponsered by an org' like the AFL-CIO, can develop and produce customized solutions for the thousands of locals of different sizes, needs, and cultures, in North America - I might change my mind.
[/QB]

---

Sure. That's easy. The software is scalable and modular and open source.

oh but it also needs a core, an api, etc. Some rules.

Plus I've already given my reasons as to the greater social impact.

quote:

---

What are union-specific tasks?

---

A database of collective agreements with an elaborate set of meta data that allow a researcher or analyst to categorize contract language on a large (but not infinite) number of issues according to how good it is, and able to measure progress over time and index the full text of said same agreements.

An application that will take user input to generate a PDF of standard contract language or constitutional language to help a new local create by-laws or contract proposals for a first contract,

A flow-chart-like application to explain rules of procedure for union meetings (One for Bourinot's, one for Roberts' one for Code CSN or whatever else gets used out there)

A transaction application that allows locals to order supplies or paraphenalia that interacts with another system that tracks a local's membership and their payment of per capitas.

An interviewing application that leads an activist or shop steward through the process of filing a grievance or a health and safety complaint (imagine Quick tax but for your collective agreement).

That's off the top of my head. What are you waiting for? Get coding.

quote:

---

A database of collective agreements with an elaborate set of meta data that allow a researcher or analyst to categorize contract language on a large (but not infinite) number of issues according to how good it is, and able to measure progress over time and index the full text of said same agreements.

---

Database of collective agreements? No problem.

Tell me how you'd measure "good" and "progress".

quote:

---

An application that will take user input to generate a PDF of standard contract language or constitutional language to help a new local create by-laws or contract proposals for a first contract,

---

What type of user input would be required?

How would the input influence the output.

Give me some examples.

quote:

---

A flow-chart-like application to explain rules of procedure for union meetings (One for Bourinot's, one for Roberts' one for Code CSN or whatever else gets used out there)

---

Don't really need an app' for that. A web page/wap page would do fine. These probably already exist, although not in a comprehensive, member-friendly form.

quote:

---

A transaction application that allows locals to order supplies or paraphenalia that interacts with another system that tracks a local's membership and their payment of per capitas.

---

More database stuff.
The biggest problem I see with this is that the original order information has to move in two separate directions simultaneously: 1) into the DB for querying later 2) into whatever system the supplier uses for receiving orders. And any changes, omissions, and errors would have to be dealt with the same way.

I'm no DB designer but it doesn't sound terribly difficult.

Imagine the implications if this DB was available to the membership!

That's be some serious transparency!

quote:

---

An interviewing application that leads an activist or shop steward through the process of filing a grievance or a health and safety complaint (imagine Quick tax but for your collective agreement).

---

This sounds like a job for ASP.

I've never used QuickTax but I imagine this could be done with javascript and python and/or perl.

quote:

---

That's off the top of my head. What are you waiting for? Get coding.

---

Me?
You're the one on the payroll!

quote:

---

There ya go. Sounds a bit ambitious for your average local tech guy.

---

There's nothing new here, just custom databases. Getting DB's to talk to each other can't be terribly difficult - all that would be required are a set of standards - and standards are already built into whatever language you choose for the job.

Now I'm going to have to take my SQL books to work tomorrow.

quote:

---

quote:
--------------------------------------------------------------------------------
Database of collective agreements? No problem. Tell me how you'd measure "good" and "progress".

--------------------------------------------------------------------------------

Well, for example, annual leave provisions that gave someone three weeks holiday after five years on the job would be "better" than provisions that gave someone three weeks after seven years.

A contract that had gone from offering three weeks after seven years to three weeks after five years would be considered to be making "progress".

---

I meant, how are you going to teach an app' to measure and categorize "good" and "progress".

Contextual algorithms aren't quite what they're cracked up to be... yet.

quote:

---

While you can develop open source code for closed systems, I think it would be a mistake to develop any of this stuff on any proprietary OS. Especially if it happens to be Microsoft.

---

Ditto! Hence my embarassment at the mention of such blasphemy.

quote:

---

I have one or ten things on my plate right now already I'm afraid.

---

Join the club!
I'm in the midst of writing a CMS (in perl) for another labour related project. I may even distribute it if it's decent. This will be my first foray into databaseland - - - wish me luck

quote:

---

And we had/have people whose job it is to review all the thousands (literally) of collective agreements we have and to codify them so that we can see how badly or how well our people were doing and to see who has the best language etc etc.

---

Let's up the ante and let the users of the DB rate the language.

quote:

---

And this is an app that would be best developed and administered centrally

---

Not necessarily. It'd be more efficient to have each local update and maintain the contract entries they're responsible for themselves. This could be as simple as logging into a secure web page, cutting & pasting the appropriate text, then hit a submit button. The more I think about it, the simpler it appears.

One major problem I can think of is the lack of a standard form, or formula, or format of CA's. How could thousands of contracts be broken down into sections or subsections that could be easily compared & contrasted with others?

quote:

---

That's interesting. Would you be worried about locals or staff "modding up" contract clauses that were actually crappy so that they look better than they ought to?

---

Not at all. There's no motivation to cheat. Having the language rated by the users (assuming this would public/on the 'net) would render any "ballot-stuffing" useless with simple security measures anyways.

quote:

---

How do you deal with the "well, it's a lousy deal but it's the best we get so I'm gonna give it five stars?"

---

What would be the point of that?
How would anyone benefit?

quote:

---

I see merit in what you're saying but I do think that there has to be a material and practical scale for measuring this stuff.

---

There isn't. It's impossible. You can not measure the quality of language without introducing bias (and other assorted human faults). This is why you leave the measuring to the users, the public, where any bias is likely to be minimized by sheer numbers.

quote:

---

Five weeks after ten years is just objectively better than five weeks after twelve years, even if six people say it isn't if you win it after a six week strike.

---

Ya, you say so. I say you let the members decide for themselves what is objectively better.

quote:

---

Asking them to slice and dice their collective agreement into several chunks per article and paste it into form after form and then codify it into a series of relatively abstruse metadata is probably more than most are willing to do.

---

What? Can't be any more difficult than QuickTax.
This is why you'd put a tech-head in each local anyways.

Anyhoo, who's to say it can't be a single form, cut & paste, and have some exquisitely designed queries do all the grunt-work. Due to the varied nature of contract formats you would be limited to basic keyword searches anyways. There is no formula in CA design (that I'm aware of). There's too many variables to design it any other way.

quote:

---

From our perspective, we have data entry clerks who are paid a decent wage with secure employment to do this work. They're trained to do it and they do it systematically and consistently.

---

Trained to input data. Not trained to develop subjective metadata on which the queries would be based on.

quote:

---

Sending in news updates slashdot style is one thing. Creating a useful database is another.

---

Slash style news updates require a waaaay more complex interaction of separate components than a database does. A database just keeps information in cold storage until the right app comes along and takes it out for dinner.

quote:

---

..especially if you had 2200 different contracts being entered and interpreted by 2200 different brains.

---

It's more democratic than having it entered and interpreted by a team of ten. The entry and interpretation would have to be two separate processes anyways. As I stated earlier; your data entry team isn't trained to interpret language.