Hi all,


I somehow woke up thinking about sustainability and growth in open
source projects, particularly of openACS, and I would like to share my
thoughs and start a non-flames thread on this.


      There are four phases in the our software value chain:


      1. Design

      2. Implementation

      3. Integration/customization

      4. Maintainance


      Besides these, consulting companies often do strategy and/or
operation.

      Open source companies like openforce, furfly, etc... make
revenues out of integration/customization and maintainance. aD
probably manages to make money out of the first  4. Consulting
companies like Scient and Accenture make their money on the 6 areas.


This means that  we are effectively doing the 4 phases but usually
only get money for the last 2. The same for Oracle and Vignette, they
do the 4 and they make the money out of the last 2 PLUS licensing
(Oracle probably makes almost as much out of consulting).


      In our case this happens because people in the open
      source community do the first 2 usually for free (as
      part of creating the toolkit/infrastrucure) or as paid
      by another project. The last one is the case for
university people who are paid for other deliverables anyway. Until
recently the openACS community seemed to be relaying on aD design, so
we did not have to worry about that. But this is obviously changing
since aD is not supporting tcl versions any longer.


      It might happen that the last 2 pieces of the pie (intregration
and mainatinance) are big enough, and we can all sustain our work with
that. My guess is that it is, but:


      1. The community would grow faster if we do all the 4
      more often.


2. It gives you a competitive advantage over Vignette
      kind of company if you add the clients requirements in
      the first 2 (it gets closer to doing strategy). Again
      this means growth for the community.


      In general the open source community doesn't worry much about
      revenues or marketing. I believe that this is because its
origins lie in the group of people who had another income to pay
      for the project (like in Universities) and probably
      because we are not as greedy . But for the sake of
      the community we should worry a bit about this things.


I am aware of openforce's network initiative, and I think this is a
very good approach. But I think it would be usefull to the community
to have an open discussion about this.


My first question would be: what are the best strategies to make the
community grow?

Malte,
I did not have the time to read all your documentation completely, but I definetely will.
I think is in Innovation Management, Strategies implementation and profits by Allan Afuah that they say that virtual organizations don't get along very well with R&D companies. But this (open source) might be the opposite! Anyway, I think that several of the things you mention are important.
Some of these things should be public and some should be shared by those that contribute or are members of a network (I am talking only about commercial information)
I believe that openforce network has this kind of thing in mind, although your idea of a virtual organization seems to go much further.
There seem to be two different areas of work: technical and commercial. How are they separated? or does people think they should go together?
I would add some "netiquette",some of which is already in place, some might need more definition:
1. Making all code GPL, unless the client explicitly requires us not to (Done)
2. Fair play when tendering for a project. Essential if you are going to work with your competitors in everyday community life. (Done?)
3. No price dumping. So if you are at Univ. you should try not to take on a project that someone else in the community might be getting paid for.
I am sure you have though about these things often, and you might have your own analysis/rules. I would like to hear them. I am new to the openACS project, I would particularly like to hear from the core team and openforce network people.

Thanks Adam, this is definetly something to write a paper about.
I think innovative ideas will not come from a a team of 100 people (o more than 2000 if you count all OpenACS members). The ideas will come from smaller teams were the conditions are met. People in Boston are lucky in this sense (but you don't have the beaches I have!) since there is a critical mass.
OpenACS provides the infrastructure, the innovation will come from its components and the architectural design. Each component is made by a small closely knit group of people, and the "global" architectural innovation comes from the core design team.
Brooks in The mythical man month talks about "conceptual integrity", that usually is set up by a lead architect and then made a product by the people implementing it. In fact he has a great metaphor about how great cathedrals were made over several generations by architects that followed an original vision.
Phillip probably defined ACS vision quite well in his book and around the ACS classic. Time will tell if this vision is followed or a new approach is taken by the future architects.

touche... you are right about ACS.

Anyway I would like to point out that innovation or conceptual integrity does not come from technology (or software engineering in this case).

The "single data model" in particular has huge implications on the "soft" side of things. What is the common thing between b2c, b2b, b2e? answer: the character "2". That character means communication  and collaboration. The single data model makes possible the management of the knowledge you gain from this collaboration.

Have you looked at birdnotes.net? that is a great example of innovative collaboration. Ok, we are not going to have many clients paying for a service that "Graph species in order of abundance by location". But you can show how innovation can be "empowered" by the right tools.

How innovative are going to be the businesses echanging excel files in .NET?

Don't forget the problem of information hoarding.  In many situations, it is not at all personally advantageous for someone to share their knowledge.  Resolving this issue often means changing a person's incentives, or rearranging workflow.  Both are painful.

The issue is that the Open Source business model is not a super effective one. aD seems to be building an open source model without a Noosphere-- odd concept.

It is. If you read the article by the CEO and Chairman of aD, you can see their perspective on the issue. Free software's primary benefit to business is business and technology risk reduction.

Free software reduces business risk by not chaining you to a vendor. If your collaboration-software vendor takes off in a different direction, you are not forced to go with them. You can keep the code, for it, select a different support vendor. You are not screwed.

It reduces technology risk simply because the code is available to you. I am living this right here, right now: I'm supposed to be installing and evaluating Sybase, Oracle, and PostgreSQL for a large client of my employer. Oracle has thrown me into DLL hell, since it's incompatible with recent releases of glibc which are used through Linux-land. It will only work with older versions of glibc (2.1; 2.2 is a big pain). Postgres can be rebuilt with no problem, and minimal risk. If there are problems, I have access to the source and can fix it myself. Again, a technology risk reduction.

aD is using the GPL to sell risk reduction; they aren't willing to open the software up to the community to the point that the community can change the direction of the software. The only choice is to fork the code.

enhydra.org where their is a strong community of both commercial and non-My first question would be: what are the best strategies to make the community grow?

Eric Raymond's list of preconditions seems right here.

Here's the thing:

There are a huge number of non-profits out there. Many of them
are very relationship-oriented like to deal with local companies.
There's no reason why the companies who are interested in this
piece of business couldn't do the same thing that the ACES folks
should do (but aren't really doing yet either):

- Work together to develop an ACES-style configuration of
OpenACS targeted at this audience

- Work together to develop a marketing message for that
configuration

- Have a conversation about coordinating their efforts to go after
different marketing segments (e.g., maybe one company goes
after museums while another goes after environmental groups)

- Have each company work hard to reach out to their own market
segment rather than having everybody pile up on the few
companies that come looking for an OpenACS solution.

From where I sit, I see OpenACS development shops
increasingly competing against each other for the same jobs.
The community as a whole is not being pro-active about
expanding the pool of potential customers rather than fighting
over the handful who find their way here on their own. Perhaps it
is difficult to do too much marketing before 4.x is finished, but at
the very least, it's not too soon to start talking about it.

Michael, it would be sad if you are right with your opinion. That's why I said the companies involved with OpenACS should start working together instead of seperatly, especially in the areas of marketing. To make just one example:
I will go to the german linux conference, talking about knowledge management solutions for the future. This is sponsored by AD. What will I do ?

• Talk about Knowledge Management in general.
• Promote the ACS as a plattform for knowledge management.
• Promote ArsDigita as a company (at least in Germany) which knows how to handle KM within large corporations.
• Talk about OpenACS as a community which supports ACS for lower budget companies and non profits (and universities).
My point is, I'm not going to put my own kind of business into the foreground. It's AD for large companies and OpenACS as a PRODUCT for smaller enterprises. So, if you go out and have the opportunity to talk to people, don't talk about your own business (yes, I know it is important and you poured your blood into it), talk about OpenACS and it's community. Get people interested in the Open Source idea. And if they check it out and think, gosh that's the way to go, they will first come back to you and ask you if you could help them with their web based solution (not someone else in the community).

Maybe I'm too idealistic, but I've seen enough money generated out of virtual networks, who promote their JOINED services. We don't need to convince venture capitalists that our service based approach is better than a product based approach. And last but not least, our competitors are not within the community. It's not even AD. It is everyone else, who is not using the (Open-)ACS as the plattform for building webbased services.

Know where your enemies are, and give us the Business BBoard.

These really are critical issues that the community is just starting to look at. I've spoken with Janine and Michael Feldstein privately about how the community should work together, and Malte and Rafael have done a very good job of talking about it on the bboards.

Part of the reason we are holding the OpenACS socials (and we're planning on two more in the near future, one in July and one in September) is to get people more involved, meeting one another and talking about this stuff in particular.

Not many people chimed in about the last one, but I think everybody that showed had a great time. At the very least, people got a chance to chat a little about the progress they're making in porting the system, how it compares to other systems and so on. I expect that as the socials gain more momentum, more people will start showing up and the community will grow stronger.

However, these clearly are not enough since some of the key people are spread around the world. Rafael in Austrailia, Malte in Germany, Don in Oregon, etc. So we do need to figure out ways to collaborate better.

Also, though, I think we need to figure out ways to advertise the system. Perhaps in the future we should consider setting up something like an OpenACS foundation so that we can do things like advertise in trade journals and magazines. I think it would be nice if we could have an add in Linux Magazine or something asking people to just come and check out the website to learn more.

Perhaps more importantly we need to work out a way to explain what the system is. Many times I want to tell newbies that post to the bboards to RTFM, but then I realize there's no good FM to direct them to. This is more than just tech docs, and less than P&A, but somewhere in between.

So I think my basic point is that there are a number of things that the community should clearly do. Since there is an increasing number of companies dedicating their business to the OpenACS, it makes sense that they would start the process of generating these materials. But there are also many independent consultants out there interested in using this stuff too.

So maybe we should talk not just about a business bboard, but perhaps a steering committee or working groups dedicated to building an infrastructure for these kinds of things. It might be nice to have a group whose purpose is to focus on OpenACES, another working on improving ecommerce, another on non-profits, another for system integration, etc. while being overseen by a core board.

I know that this idea is introducing a great deal of beuracracy, but that may be required in order to push OpenACS farther. We have a very good, very large system here that is competing with many other packages. If there are enough people dedicated to seeing it succeed, this may be a way of ensuring continued improvement and innovation.

talli

The main thing is to develop a unified marketing message about the platform. Once we have articulated that message, we should post web pages on this site that essentially act as marketing materials for anyone trying to sell OpenACS.

This reminds me of a point I've been meaning to make. Mike has been looking at Zope, to see what we're up against and also to have some idea of what it can do in case we're asked to use it some day. His main impression so far is that they do a *far* better job than we do of presenting it on their website. Of course there is a company behind Zope and we are all volunteers, but that doesn't mean we can't do a first rate job of putting together a site to evangelize OpenACS; it just may take us a bit longer to do it (any resemblence between this and furfly's perpetually "under construction" website are purely coincidental :).

Company Y should come along and say, "Hey, we're going after higher ed too. Tell you what; we'll stay away from the Northeast for now if you stay away from the Southeast for now.

Although I agree with the sentiment, Michael, I think you're breaking into a trot here. :) We already know that marketing a company like ours is difficult, or at least I will assert that I believe it to be so. So how are we going to segment ourselves, geographically or otherwise, when we don't know how to target the customers in the area each of us has chosen?

The hard part is that our businesses are not geographically based. furfly has clients all over the world now. How would we advertise to reach them? When I think small business marketing I think of things like ads in local papers, sponsoring the local Little League team, things that will get your name in front of the people who might be your next client. And that's fine if you're a local CPA or whatever. But all of us need to reach a national or even global audience, and the only medium which can do that cost-effectively is the Internet. That doesn't solve the fundamental problem, though, beause then we run smack into the problem that has plagued many a dot-com with far deeper pockets than ours - how do you get people to visit your site? Hence Talli's idea of an ad in Linux Magazine - I'm not too sure that people read those any more than they click on banner ads, but it might be worthwhile to try at some point.

Since none of us know the answer to The Big Question just yet, I think we need to set it aside for now. Let's work on evangelizing OpenACS, and see what happens. Sometimes, in situations like this, the right answer becomes obvious when it's the right time to implement it. Not to get too New Agey here or anything. :)

This reminds me of a point I've been meaning to make. Mike has been looking at Zope, to see what we're up against and also to have some idea of what it can do in case we're asked to use it some day. His main impression so far is that they do a *far* better job than we do of presenting it on their website. Of course there is a company behind Zope and we are all volunteers, but that doesn't mean we can't do a first rate job of putting together a site to evangelize OpenACS; it just may take us a bit longer to do it (any resemblence between this and furfly's perpetually "under construction" website are purely coincidental :).

Janine, I do not agree that our competitor is Zope. Most people has never heard of Zope or OpenACS, most have heard of Open Source and Vignette/Microsoft... I think that when a company choses a product the shortlist will be made between open source and not. So you go and compete with a guy selling Vignette, another selling Websphere, 2 selling Oracle,etc ... if there are 20 offers maybe you get another guy selling Zope.

As a consultant I think you have to sell solutions, "kill customers' pain". Some solutions require licenses, are less customizable, less robust,etc... other are royalty free, completely customizable, robust,...

Zope vs. OpenACS is a discussion for us geeks, is not the dilema for the CEO or CTO.

Lets say you have a market of A$/year for internet consulting (morgan Stanley Dean Witter has excelent research with these numbers). Of that B$ are for software and C$ are consulting, I believe that for commercial products such as Vignette C=2B.

With respect to software, if X% of the market is made of commercial solution and Y% is made of opensource solutions, I believe that X>>Y, so when a CTO has to make a decision, he will first decide about commercial vs. open source, if he ever gets to know about open source systems like OpenACS. Again he cares about the kind of solution and the business model behind it, then he might care about the programming language ...

Very interesting!

Do you have a proactive or a reactive approach?. I mean, could it be that your statistics are based on a sample that is already skewed? or is it that my sample is skewed by  companies that do not know of either (ACS/Zope)?

My approach has been proactive in the sense that I go to the paper and look for tenders or go to managers I know and show them what you can do with OpenACS (a screencam or live demo or both). It has to be proactiove since I am now in a University but doing consulting...

In one case the other options were Vignette and Interwoven.
Another I expect to have Webct and Blackboard.
In a couple were internal developments.

My biggest "sale" was an internal project against another internal system with less features but with a nice "wygiwys" inteface (we discussed this in another thread "wordprocessing in forms"). I implemented the intranet module there (using ACS/Oracle).

Being from the Midwest (Okay, Cleveland but we are usally lumped into that area) the perspective is quite different.  Since I have (and will continue) to do quite a bit for work for non-profits will "marketing" towards small and mid-size business in the area, I'm quite use to "in-expensive" solutions and utilizing software as much as possible. It was actually uptime that brought me towards ACS (since I was looking for something for an email customers that didn't want to move to me for hosting at the time).

Looking at the services offered for free by Arsdigita (and reading "the book") is really what got my looking at ACS and since I really don't like Oracle, looking at OpenACS.

From my perspective was that ACS/OpenACS (from now on refered to as OpenACS) offered a great platform that could be expanded and have the features that many people wanted, at a reasonable price.  I then setup 6 sites for mid-sized customers for their INTERNAL networks (so I can show anybody them *grr*).

I think OpenACS gives everybody at rather good chance to present a well supported product, based on strong ideas to their customers, but the continued support and the show of support from the OpenACS community would benefit both sides.  Since, I believe, we are all fairly "spread out" I don't think marketing by the OpenACS community will step on those involed toe's.  I know my customers are not going to goto Arsdigita because of the price, but I show them OpenACS and they want something that I can not do or I don't have the time to do, it would be nice to be able to pull from the community so that I can support my customers.

Since I have been involved with a few GPL'd projects over the past 1/2 dozen years, I've seen things that I, personally, think can be done better.  Yes, the code has to be there, but a strong commitment from the community is one of the best selling points for anybody using/selling/support the software.  You can look at Linux and how distributions hiring developers (and/or support development) has help many of them gain the all important "mind share".

I agree that staying away from the Zope vs ACS debate (or at least not focusing on it) will be helpful and important as there is a larger market out there.  I still find it hard to believe that I'm seeing more and more sites based on the slashdot type model of community software, when I've yet to find ONE that will send me an email when something new and/or interesting comes up for discussion!

It might be worth looking at creating and/or supplementing some of the free services (uptime, clickthroughs, bboards, etc.) that Arsdigita does as a demostration of what OpenACS can do.  It is a small part, but I think it does give a nice place to view the operation of the software and yet give potential "customers" the ability to use it in a way that they want and not simply to dicussion the package themselves. Plus, we should port that stuff to Postgres anyway :)

Okay....off my soap box there.

I don't think Janine was suggesting that we ought to start a p.r.
war against Zope. Rather, what I got from her post was that the
Zope community has set a strong positive example for us in
terms of marketing and documentation. While they may not have
a huge market share, it looks like the Zopatistas have grown
their share to be significantly larger than that of the ACS. We
ought to study what they are doing so that we can expand our
market share in a similar way.

Hello Talli,

<P>I like your idea about an operating foundation for the OpenACS, though I think this might take some time as it is probably to early now, but we should keep it in mind.

<P>As for the steering comitee, well yes, sure, definitively. What would be the main tasks:
<ul>
<li>Coordinate marketing activities of the community:<ul>
<li>Maintain a list of events, related to OpenACS
<li>Provide speakers with material for conferences
<li>Creation of an elevator pitch for OpenACS
<li>Creation of a two pager about OpenACS and its functionality
<li>Collection and distribution of case studies
<li>Collection and distribution of articles around OpenACS and other things (something like ASJ. And NO, this is not meant as an offense toward AD, but would open their base to a larger audience)
<li>Support with/for the press</ul>
<li>Coordination of the development<ul>
<li>Keeping track of who is working on what
<li>Keeping a list of suggested release dates {might come in handy for marketing reasons}
<li>Tracking a wish list
<li>Distribution of OpenACS and related packages (read: RPM for RH {e.g postgres},SuSE {e.g Oracle})</ul>
<li>Run a plattform for virtual corporations:
<ul><li>Ressource exchange (as in, I've got a client, but not enough manpower, anyone able to help)
<li>Collection of legal ressources (list of lawyers, contracts,...)
</ul>
<li>Setting up an foundation (in the future, I like Talli's idea)
<li>Keeping track of competitors and competitive technologies:<ul>
<li>List of companies people of the community have competed with and the outcome
<li>List of technologies OpenACS was compared to by potential clients. Writing of comparisons for these technologies for later "pitches"</ul>
<li>Running OpenACS website:<ul>
<li>Coordinate with the server provider (thanks AD, hopefully for a long time)
<li>Operate a demo server as David suggested. Might be the same as openacs.org.</ul>
<li>Structure the bboard information in e.g. articles, FAQ.</ul>
<p>
That are just the things which jumped into my mind, probably there are a lot more. The most important jobs (as in development) are already been done outstandingly well. I'd volunteer to keep a list of todo's and people willing to contribute for the non development related issues.

As an adjoiner, I would like to recognize Roberto's yeoman efforts to document OpenACS up till now. I don't want to suggest he hasn't been up to par, just that I think his work ought to be standardized so that he doesn't have to work as hard.

Also, I think one of the issues that has been raised is how poor documentation for the newbie or non-techie is. This is often called marketing, but really it's a way of describing what the system is in 500 words or less. A one-pager, if you will.

This stuff may be the most important result of our efforts. So far, the only way to learn about OpenACS is to slog through the bboards or download it and do the problem sets. While the latter may be the preferred response to a newbie's question, it's a real tough way to convince clients that they ought to use this stuff.

Allow me also to touch on the Zope issue really quickly, which can be called the Webobjects issue, or the Userland issue, or even the ACS5 issue. We need to have effective arguments for what sets us apart, not necessarily anti-Application X arguments. This means having competitive analysese (sp?) that provide current developers and potential clients lists of our advantages. Furthermore, they will be effective documents for knowing what we need to *improve* in our next release.

talli

A few points on this very interesting thread:

• competition: competition is good and it should be maintained. I don't see a need to segment the market through some sort of agreement between OpenACS vendors. First it would be bad for clients because they would lose choice, and the whole point of open-source is choice of vendor. Second, it would be horrible for all of us because it would make us look like there is a hidden conspiracy between OpenACS vendors. That is the last thing clients want to see. Sometimes we'll compete (Furfly, OpenForce, Ybos, MuseaTech have competed at various points). Sometimes we'll work together (MuseaTech and OpenForce have made one joint proposal, and Furfly and OpenForce will soon work together on a project.... stay tuned). This is healthy.
• documentation: lots of talk about this. Let's write it! You don't need to ask for permission or put forth grand plans of what needs to be done: go ahead and write it. Take example on Don and Roberto who have just gone ahead and written significant documentation because they knew they had specific knowledge to share. If you know you can do something that's not explained clearly on openacs.org, write it and send it in!
• marketing: our biggest mistake so far is that we pitch OpenACS as the product. ecommerce? use OpenACS. KM? Use OpenACS. eLearning? Use OpenACS. We know OpenACS can solve these problems, but it's bad marketing. Instead, we should have specific solutions for each vertical, solutions which can be footnoted as "powered by OpenACS." This will allow each vendor to focus on and develop marketing materials for each vertical that they are specifically interested in. Companies should develop marketing materials, not some volunteer organization.
• more frequent news updates: just submit news and I will approve news items that are relevant!

A very interesting thread that made my thoughts run berserk


I'm pretty convinced that the majority of the participants in these bboards have become members of this community after having read "the Book". Hence we have found salvation, or at least chosen to believe (brainwashed or not), in the supremacy of OpenACS... "Farkas's Three Conditions" above, explaining why ACS was innovative and thereby unique (it still is, to my knowledge), indicates that we have reason to believe. Personally, had I known of a better toolkit I wouldn't be on this bboard writing this.


I'd like to offer my customers the best solution I know and the solution I know best. In my mind there are no alternative solutions to OpenACS. So, if a potential client will not share that conviction I frankly don't need him. Saying no to a customer? This may seem harsh and stupid but IMO it's not. If I in any way so much as imply that other solutions (e.g., the Z-word) are comparable competitors to OpenACS I'm only shooting myself in the foot because that would be devaluating the market value of OpenACS which is the solution I want to sell. I think this devaluation will occur as a result of a re-active marketing approach since, by doing so, I'm accepting clients that seek *any* solution instead of actively picking the clients who need "my" solution in the pro-active way of marketing.


I agree with Ben when he says: "Companies should develop marketing materials, not some volunteer organization." That doesn't have to mean that we can't discuss ways of marketing on these bboards. But one thing we don't need IMHO is for the comunity to become a common voice speaking to customers (end-users who roll their own, sure, but not people seeking vendors). What we certainly do need is for the OpenACS community to develop the additional role as a "business association" of individual companies worldwide, as opposed to a centralized "company" turning to the customers and not to their different vendors.

The only way a centralized "Uber OpenACS TNC" might deal with global marketing would be to utilize the concept of "re-active" marketing, i.e., trying to please potential clients looking for *a* system. Using this strategy OpenACS would have to drag itself down into the mud and subject itself to the humiliation of being equalled with so called "competing systems"... Moreover, using this strategy, by necessity the Uber OpenACS would get pretty much involved with the "wrong" customers. If these were the only customers out there we would indeed have to make due with them... My point is that the re-active marketing strategy is responsible for the wrong clientele taking an interest in us in the first place.


The alternative strategy is the "pro-active" way of doing things. This way suggests that the OpenACS vendor actively approaches the targeted clientele that she supposes share the essentials of our philosophy. Several stories from the community as well as my own limited experience imply that non-profits constitute an ideal market.


I myself find the re-active marketing practically useless since I'm based in Stockholm (Sweden) and can only work with clients in that region. If OpenACS would do worldwide marketing on behalf of the community it would run the risk of creating an unsustainable situation where interested organizations around the world get the impression that OpenACS is a company that can meet the needs of a global market. Imagine how disappointed these people will get when they realize that OpenACS is no more (but to us no less) than an online club upheld by a few thousand members and a handful of companies supplying their local markets, and what's more, most of them situated in the US.


Remember that the fact that our software and ideas may reach a global audience doesn't make our market global since the task of providing a web solution to a paying customer is a physical task in the "real world", much like the work of your local plumber or carpenter...


FWIW,

Talli - With respect to search gurus, don't overlook the excellent work being done by Neophytos Demetriou. He has been working on the Russian solution mentioned by Don in his status report.

Although, it is less mature than swish, OpenFTS has the potential for much tighter integration with openacs. In alot of ways, it offers functionality equivalent to intermedia in Oracle.

Talli,

I agree with you that I am missing the point, or at least
misunderstanding it. You are proposing a fairly bureaucratic solution
to a problem that isn't clear to me. How is the situation untenable?
What have you (or anyone else) been unable to accomplish because of
this "over-centralized" authority in the community? I'm not talking
about what hasn't been done, because certainly much work lies ahead of
us. Rather, I'm not seeing what you and others have actually been
*prevented* from doing. Having a new committee won't help you actually
get the job done if you don't have the resources (financial and
otherwise) to take action. Solving the "invisibility" problem won't
happen without serious OpenACS live sites, participation in
conferences, and partnerships with bigger open-source players. These
things require significant funds much more than they do central
committees.

Don and I immediately noticed the over-centralization of the OpenACS
community as soon as 3.2.2 was released, and we saw the move to 4.x as
a perfect way to decentralize the process. The core, much like the
Linux kernel, would remain somewhat tightly controlled, but would
probably need no more than 2-3 yearly upgrades. Then, packages would
be created completely independently of the central organization. Any
organization can create an OpenACS package or combine a few of them in
the same tarball to create a "solution." That organization can and
should take responsibility for marketing that solution. That
organization should have no obligations to the central OpenACS team
other than begin decent open-source citizens.

Take a look at the other successful open-source projects. Is there a
Linux marketing committee? Is there an Apache marketing committee?
Not as far as I can tell. The Linux and Apache companies do their own
marketing, and in fact they often pitch the open-source products they
support in very different ways. It is already difficult to agree on
one marketing message within a single company. Coordinating marketing
between 4 or 5 companies with different financial goals and
responsibilities is extremely ambitious, and possibly impossible to do
effectively or in a way that adjusts quickly enough to market changes.

As for the state of the community, our OpenACS bboard traffic keeps
increasing, our number of users is growing polynomially, and progress
on the port is booming thanks to as many as 15-20 hard-core
contributors. All efforts, including the MuseaTech socials, are to be
credited for these awesome results. I'm not sure where you see the
community being unstable or having growth issues.

In short, I don't understand the problem you're trying to solve. What
precise structure are you proposing, and what specific issues will
this structure address? I would like to hear more details about it,
given that you clearly consider the situation to be quite serious. I'm
also interested in hearing how you think a committee will give you and
all community members more power and flexibility than we currently
have in addressing the problems at hand.

Boy, this is a hot topic!

A few responses to the messages of the day...

I have a feeling that some of the posters feel that reactive
marketing is bad;  there's almost an implied value judgment, like
if you're not being proactive you're doing something wrong.
IMHO, that's only true if what you're doing is not working!  Being
reactive (that is, taking the clients who find you instead of going
out in search of them) costs nothing, in marked contrast to the
alternative.  If it works for you, do it!

I respectfully disagree with Ola that we shouldn't offer solutions
other than OpenACS.  That's probably true for a very small
company, because you don't want to fall into the "jack of all
trades, master of none" trap that comes with trying to offer too
many things.  But for those who have the people to do it, I don't
see anything wrong with offering OpenACS, Zope, and whatever
else seems useful.  No one product will fit every client, and most
clients appreciate being offered choices when it makes sense to
do so.

I don't agree that building a web site is a physical task like
plumbing, either - we can and do build sites for clients far away
who we have never met.  However, in general it's probably true
that not too many small companies really have a need to market
themselves globally.  Nationally applies to almost everyone,
though, and IMHO is just as difficult to achieve.

I'm not sure I'd go so far as to form committees, but I do think
there are some areas where we can help each other out, most of
which have already been mentioned:  good documentation (both
technical and non-technical), a demo site, a forum for
discussing business issues, white papers comparing OpenACS
to other technologies people have run into.  How about a job
board, to connect companies and consultants?  What about an
area for specs, where people can post ideas for new modules
and have them debated with others who might even help build
them?  There are lots of things we can do which will improve the
OpenACS experience for everyone and will benefit our
companies at the same time.  There were others mentioned in
this thread already, but these are the ones which are the most
compelling for me personally.

When someone has already read Philip's book, it's easy to
discuss an ACS site with that person.  But what do you do when
a potential client says "I want a site which will encourage my
users to share with one another" and has never heard of the
ACS?  What do you show that person?  Where can they go to
read about the ACS (Open or otherwise) which will give them an
understanding of what it is, what it can do, and why it's so great?
Right, there is no such place right now.  That's what we're talking
about fixing here.

Ben, when we talk about community marketing we're talking
about marketing OpenACS, not marketing for our companies.  If
OpenACS becomes more well-known, the pie will grow larger
and we can start fighting over slices instead of crumbs.
Companies might use some of the Ope nACS marketing
material as their own and that would be fine too;  it would be put
out there for anyone to use in any way that would help them.

Of course, nearly everything we're discussing here requires
adding "stuff" to openacs.org, so I guess the first task we need to
set for ourselves is convincing Ben that we're not all crazy. :)

Janine,

I haven't vetoed any contributions to openacs.org to date. If
someone wants to contribute content, I'd love to put it up on
openacs.org! I continue to coordinate the site because I spend
most of my time evangelizing OpenACS and I believe I still have
quite a bit to contribute in that domain. However, I do *not* want
to control/edit all of the content.

I do not intend to be a bottleneck for this community, and I hope
most people realize that I've stepped out of the way of decisions
as soon as I needlessly slowed things down. Note the fact that
Don and Dan are managing the 4.x release much more than I.
Note that Roberto managed the 3.2.5 release. I'm in talks with
Mat Kovach who wants to take over uptime.openacs.org, and I'm
more than happy to hand that over to him. If someone wants to
set up an OpenACS demo site, I'd love to point
demo.openacs.org to it.

I don't think anyone in this community is crazy or out of line. I *
am* worried that, for some reason, there's a lot of talk going on
as if action weren't possible. Go ahead and act! Build
documentation. Put up a demo site. Do a Zope/OpenACS
comparison if you've experienced both products. I will happily
post them on openacs.org!

(some people have been asking me about the new forums. I'm
waiting for the polls to stabilize out to see what everyone thinks.
The non-tech forum is looking extremely likely, while the web/db
forum is significantly more controversial).

In various forums recently, both here and abroad, there have been some statements made to the effect that it's hard to participate and hard contribute to the OpenACS project.  Perhaps too, that it's hard to find information about the OpenACS project on this website.

Generally I agree with these sentiments.  I know that in the past I have wanted to comment on documentation, and been told the best way to do that is to bury the fixes in the SDM.  I know that I have asked for new forums, and that's waiting until the current response (70% for, 30% against settles out) (would that our presidential elections had such mandates).  I know that I've suggested that bboards are not the most useful form for storing/finding/structuring long term information.  I've suggested alternatives tcl wikis, zope wikis, etc., and not much has happened.  This is a bit more understandable to me, but not completely understandable.  (Must OpenACS.org only use openacs technologies, or for the purposes of moving the project forward, is it okay to use the technologies of some of our competition, until our own modules come online?)

I've wanted to propose new projects, new modules, that might be developed in parallel with both OpenACS 3.2.x and OpenACS 4 (ACS Wiki, ACS Search, ACS XML-RPC/Soap).  But I haven't.  It just doesn't seem that the response from the community will be anything but "start a new thread".  As I believe we need more than just forums and can do better than just forums, starting a new project supported by just a new thread seems a particularly poor use of my time.  It won't generate a solution quickly.  It will be hard for myself or others to find information later on.  And it's not the best way I might contribute to the community.

In response to these experiences, I've pretty much determined that new projects that I start will be worked on from my own sites, where I can organize them in the way I see fit, where I can devote necessary resources to them, and where they feed back into the marketing of my own skills.  To the extent that I can I will make these modules and their documentation open to the community.  And  should anyone want to throw in and help out, I will happily create forums, faqs, etc.  On my sites though, not here, because I can't.

Now maybe I'm a lightweight.  But perhaps my experiences are felt by others in the community.  As some evidence note some of the other openacs support sites around the world where one can download openacs patches, modules, or experiments.  Some of these sites are on sourceforge, some not.  I don't know why the creators of those sites chose to host them elsewhere, but I have no doubt they felt it was better for them to host these sites elsewhere than to host them at OpenACS.org.  I feel that's a shame and a real loss to our efforts.  It takes momentum away from this site.  It removes the networking effects of hosting those projects and providing the information at this site.  And I've joined them.

I would like to see it easier for new initiatives, new projects, new documentation to be formed here.  Not elsewhere.

I think the basic problem is, and this is not to get personal at all, but the basic problem is that, understandably, Ben has only so much time to administrate this site and no more, and that results in a very conservative approach to building this site, favoring using existing tools, and favoring not doing things that might compromise security.

My personal preference would be for a more radical/experimental/performant approaches.  I would even be okay utilizing less secure technologies even if that meant that on occasion we get 0wn3d and have to restore the server from backup.  Now that's easy for me to say of course, I'm sitting in California and it's not my time that would be involved in hoofing it over to Boston to restore any machine.  I'm just expressing my preference for using technologies that work today, as opposed to waiting for an OpenACS solution.  But I'm wandering now....

Should the community of developers be responsible for how the site looks and operates?  That's risky, but I say yes.

Do barriers to community participation exist?  I don't know.  It might just be my perception.  We might run a poll to find out.  We might just 0wn up that they might exist and stipulate that as a community we don't want them to exist.

I'll assume that such barriers do exist and that we would prefer to lower them.  How?  There are many alternatives:

1.  Implement a wiki at this site.  Let folks build their own site.  Advantages: little apparent effort required from webmaster.  Lets site get built using community efforts.  Lets projects proceed at the inherent energy of the project members.  Disadvantages: the tcl wiki has always felt slow to me.  The zope wiki has potential security problems and/or doesn't feel like the zero maintenance answer the webmasters here need.

2.  Use sourceforge for new efforts, and create some sort of project page announcement/signup site over here:  http://openacs.org/projects or http://openacs.org/freshmeat.  Advantages: good support from sourceforge and it lets developers concentrate on development.  PR available from freshmeat.  Having healthy support from many projects may increase OpenACS recognition.  Disadvantages: dilution of efforts, reduction in apparent momentum at OpenACS, reduction in network effects from registering, participating at OpenACS.

3.  Do nothing.  Advantages: secure, and known.  Disadvantages: if barriers exist, this does nothing to lower them.

4.  Create a sourceforge module for the OpenACS.  Advantages: wonderful.  Disadvantages.  May have a long wait ahead of us.

5.  Open up this site itself.  Create a committee of five to seven with a deadline and the goal of producing a charter and/or bylaws and/or procedural mechanisms for changing this site.  Make sure one thing they produce is a mechanism for replacing themselves.  Create a new forum for the committee to post potential charters/bylaws/etc. to, that will allow the large community to comment/foment.  At the deadline, let the community vote on the various proposals.  Turn the keys over at that time.

Is this neccessary?  Well /directory/browse and emacs tell me that 2581 users have given personal homepage addresses when they registered here.  No telling how many are duplicates nor how many folks signed up but didn't give a personal homepage address when they did.  It's not absurd to think that with a community of 2581, it might be time to have a more formal organization.

Advantages: a more formal organization may be taken more seriously by the press, by other developers, and not least, by potential users and buyers of our services.  A more formal organization may secure better funding, from the aDs, or even the IBMs.  Overall, better recognition will do wonders for all of us, from resume enhancement to greater cred from potential buyers.  A more formal organization may in fact be more responsive and sensitive to environmental changes than one or two individuals that may not have the time or resources to administer the site.  A more formal organization may be more democratic.

Disadvantages: it may be too much overhead at this time.  It might be more democratic.

Anyway, those are my thoughts.  I favor 5 & 1, then 5 & 4.

Um, I hate to say this Jerry, because it is exactly what  you don't
want to hear, but I think you need to start a new thread for this.

There are two separate issues here. One, which was the original
point of this thread, was how the community could work together
on business issues. The other, which seems to have been
pulled in by several posters, is how the community interacts in
general--but mainly around development processes.

Personally, I am much more inclined to believe that there's a
problem worth worrying about in the former issue than the latter.
The biggest bound on the growth of this community, IMHO, is not
Ben; nor is it the fact that we don't have a wikki board. It's the fact
that the development is being done by an all-volunteer group.
Given that fact, and given the fact that the 4.x port is a
monumental task, I think the community is doing rather well.

But that's all irrelevant, because it's not the issue at hand. I think
if we try to address both these problems in one thread, we will
end up addressing neither with any degree of satisfaction.

I just did here
http://openacs.org/bboard/q-and-a-fetch-msg.tcl?msg_id=00025v&topic_id=11&topic=OpenACS