Looking for your thoughts ...

Dear Reprappers,

first of all congratulations for your visionary project - that's really exciting!

At Hamburg University of Technology, where I currently do my doctorate, we do research on open design projects and the opportunities of transferring the open source approach from software to physical products. Right now, we are studying a number of examples and in this context your project is pretty interesting.

I've been contacting some of you directly and I hope to get a chance to talk to some people. To also get a broader view I would like to use this forum to ask some questions and collect your opinions.

One topic I would like to understand is how you coordinate all the different contributions: How do you ensure they can be integrated into one well-functioning product? How do you come to a decision, if you e.g. have several solutions to one issue?

I also wonder whether the management of the development process of your project needs to differ from OSS projects, simply due to the fact that you are developing a physical product, not

Hi Kerstin,

as now it's a bit chaotic in structure, organisation and such, but maybe this is a chance too?

When a project is developed to mature, it tends to 'freeze' in an optimized stage and new ideas have to fight their way through until the're common.

I think we actually are in a decreasing 'brainstorming' phase, where different concepts, solutions and options are tested for usability - look on the different 'RepStrap'-concepts and the talks about changing toolheads ...

Viktor

PS: check the German Usergroup ( [forums.reprap.org] ) too when you want to meet some local folk or chat in plain german.

>One topic I would like to understand is how you coordinate all the different contributions: How do you ensure they can be integrated into one well-functioning product?

There is a core team run by Dr. Bowyer.

[reprap.org]

All sorts of ideas are tried out by both the core team and the larger Reprap community. Really good ideas as judged by the core team are integrated into the current mainstream product, viz, the Darwin printer. That is not to say that other non-mainstream products don't don't get produced in the broader Reprap community parts of which are sometimes brought into the mainstream product. The switch from the PIC-based microcontroller boards to the ATMel-based Arduino approach that happened recently is a good example of that.

>How do you come to a decision, if you e.g. have several solutions to one issue?

Typically there is a lot of discussion over a period of time about a particular development stream which culminates in a vote of the core team members being called by Dr. Bowyer. A positive vote gets the new development included in the current mainstream product. If the new development has been created by a non-core team member more often than not that person is voted onto the core team.

>I also wonder whether the management of the development process of your project needs to differ from OSS projects, simply due to the fact that you are developing a physical product, not

Edited 4 time(s). Last edit at 06/03/2008 08:39AM by Forrest Higgs.

Being uninformed but bored, I'll wade in with my opinion... its more of a general open source rant, but it probably applies here too

Quote

One topic I would like to understand is how you coordinate all the different contributions: How do you ensure they can be integrated into one well-functioning product? How do you come to a decision, if you e.g. have several solutions to one issue?

The number of people with 'commit' access to SVN or the Wiki is minimal. This ultimately means that if they don't think something is a good idea, it doesn't go in. 'True openness', wikipedia style, where anyone can do anything requires an awful lot of housekeeping, and is of dubious worth in most circumstances.

As for several solutions to one issue, how does anyone choose between multiple options? Rarely are all equal. And if they are, why not supply both, and let the more popular one triumph in the long run thanks to greater support and interest? Sure, it spreads interest and resources thinner, but if no-one cares about a project you can just let it fall by the wayside. There's more than one way to do it, and all that.

It seems to me that open project evolution is somewhat Darwinian... many fall by the wayside and are abandoned, some fork into newer and better projects leaving the parent behind, some just keep on going successfully.

Quote

I also wonder whether the management of the development process of your project needs to differ from OSS projects, simply due to the fact that you are developing a physical product, not

Thanks a lot for your answers, this is already very helpful. Thinking about your answers some more questions came to my mind.

>>but maybe this is a chance too?

>I'd credit this to Dr. Bowyer's management style.

It seems to me that within this project a lot depends on Dr. Bowyer as a person. From your experience, do you think it is important to have one strong leader driving the project? If so, which activities should he perform and what decision authority must he have?

> We have had several team members whose level of participation has dropped off for a variety of reasons, usually personal. Some are still on the team list and a few have been dropped, if I recall correctly.

How does this change affect the organization of the project, e.g. concerning roles or task distribution? Or, more general, do team members choose the activities they want to perform or do (sometimes?) responsibilities need to be assigned?

Cheers
Kerstin

I've done a lot of thinking about meta-aspects of the RepRap project.

My thoughts:
I think RepRap is a little different than 'your average open source project'. There are various branches at various aspects of the system, for example, multiple people are developing a stepper driver controller based on different and sometimes the same chip.

While many software products branch off, we are still working towards the shared vision that is central to RepRap. Making an evolving, powerful technology a commodity for anyone to have. I think there is convergence after something is developed that works well. Building a RepRap allows for some choices, that you can combine in different ways. In that sense it's more a set of libraries that you 'assemble' into a whole, than an 'entire product' that you can install.

Anyone needs to assemble their system anyway, so not allowing choices based on materials available, price sensitivity or usage intentions would be a waste.

With an open source software package, you will normally not change bigger chunks.

I think our shared ambition and vision allows us to set our ego's aside and switch to someone elses design. On the other hand, if we have hardware that is assembled, you may decide not to change this completely but build the new thing as an addition. In software, trying a beta and reverting is much easier of course.

Because every bit (as in piece) is necessarily seen and assembled by someone that builds a RepRap, there's a very complete peer review going on. In software, you might just focus on interfaces and (re)program parts that fit in, and leave the rest as-is.

Software copies error-free. With the RepRap, deviations ("To err is human") are actually a lot of different things that are tried.

Another thing that is different: we communicate a lot in terms of narratives and with pictures. I don't see people talk about options that we have for RepRap like they do about software options. There are best-practices in both hardware and mechanical design.

The RepRap is both an idea, a physical thing and it can make 'all sorts of things'.

I also think that the team is more multi-disciplinary than most software projects.

>Did you know each other before?
I didn't know anyone before. I found the website looking for a electro scematic and the idea struck me as wonderful!

---

Regards,

Erik de Bruijn
[Ultimaker.com] - [blog.erikdebruijn.nl]

Kerba Wrote:
-------------------------------------------------------

> It seems to me that within this project a lot
> depends on Dr. Bowyer as a person.

I'd say that is very true.

> From your
> experience, do you think it is important to have
> one strong leader driving the project?

In an open source project, I'd say that a "strong" leader in the sense I think you mean would result in the team very quickly shrinking to a membership which includes only that leader and no one else. Mind, when I was much younger I ran an open source software project and very nearly had this happen to me.

I would characterise Dr Bowyer's particular genius in this regard to be more one of having boundless and infectious enthusiasm for the project and, and this is very important, a rather deft touch and considerable cleverness in the handling of a very diverse group of people whose only incentive for participating is a sharing of that enthusiasm.

> If so,
> which activities should he perform and what
> decision authority must he have?

I can't relate to generalities for this sort of thing in that I expect that there are a variety of ways to successfully run such a project. In terms of this particular project I can say that Dr. Bowyer leads from the front by being probably the hardest working and most inventive member of the team. Most of us, and I include myself, struggle to keep up. In terms of decision authority, decisions are usually taken by a poll of the core team members with Dr. Bowyer posing most of the questions being polled. As I said before, he is very deft in his handling of his team.


> > We have had several team members whose level of
> participation has dropped off for a variety of
> reasons, usually personal. Some are still on the
> team list and a few have been dropped, if I recall
> correctly.
>
> How does this change affect the organization of
> the project, e.g. concerning roles or task
> distribution?

We haven't had so many people drop out of the project who were doing really critical tasks that I could give you a good answer to that question.

> Or, more general, do team members
> choose the activities they want to perform or do
> (sometimes?) responsibilities need to be
> assigned?

From what I've seen responsibilities are "assigned" by mutual consent. In most cases, a particular task is taken up by somebody who is already interested in that aspect of the work.

> In terms of decision authority, decisions are usually taken by a poll of the core team members with Dr. Bowyer posing most of the questions being polled. As I said before, he is very deft in his handling of his team.
One more question concerning decision-making: are there any premises or norms on which the core team bases their decisions? e.g. related to the identities of the team members?

On "future plans" I've been reading about what you're planning to develop. Is there a pre-defined timeline, maybe only internally, stating until when which steps should be done?

Also I'm wondering, how do you finance this great project?

Regards and again thanks a lot for your answers
Kerstin

Kerba Wrote:
-------------------------------------------------------
> One more question concerning decision-making: are
> there any premises or norms on which the core team
> bases their decisions? e.g. related to the
> identities of the team members?

A major premise that I do not share is that open source development software ought to be used whenever possible in the project. My feeling, and it is very personal and not shared by the majority of the team, is that open source development software has been used in circumstances where it really wasn't good for the project. Again, that's my personal opinion only.

> On "future plans" I've been reading about what
> you're planning to develop. Is there a pre-defined
> timeline, maybe only internally, stating until
> when which steps should be done?

We haven't discussed any sorts of timeline for Mendel for quite a few months. The major focus has been to get Darwin out of the door. I have no doubt, however, that in a few months when things settle down a bit that a major Mendel discussion will begin.

> Also I'm wondering, how do you finance this great
> project?

Dr Bowyer has received a modest gov't seed grant to get Reprap going which provides for some materials purchases and keeps his graduate student Ed employed. In addition, the project has received, in the past, private donations from a variety of sources, both private and industrial.

Beyond that, ad revenue from the Reprap website has been very helpful. Other team members largely finance their own work, though some disbursements from the ad revenue account have been made in special circumstances to hurry along promising developments. I shudder to think how many thousands of dollars that I have spent on Reprap-related work. I am certain that I am not alone in that regard.


> Regards and again thanks a lot for your answers
> Kerstin

Edited 1 time(s). Last edit at 06/06/2008 09:56AM by Forrest Higgs.