An inquiry on OFS vs OFN from Roy Guisinger

I’m not sure if Roy Guisinger is on here but I’ll invite him. I think @MikeiLL can answer a few of these and @Angie-US might have things to add, but this seemed like a great comparison and line of questioning for the broader community! I don’t have a lot of answers for this but here goes!

For reference: Roy has been leading efforts on Open Food Source, a piece of software that I believe was a significant inspiration for Kirsten and Serenity and has been around…10 years?

Open Food Source (OFS) compared to Open Food Network (OFN):

  • OFN is a centralized system where a single server hosts multiple organizations. OFS is a single-server system. Advantage of OFN is that it does not require an organization to own and administer their own server and requires significantly lower technical knowledge. Advantage to OFS is that each organization owns its own server and has relative autonomy in modification and customization of their own user experience.
  • Related to the first point, I do not know how OFN manages financial relationships with respect to the OFN server. Is this a software-as-a-service (SAAS) situation, or what? While OFN and OFS are both open source, is there any potential risk of unscrupulous organizations taking control and manipulating or exploiting the market. I think probably not, but I would want to understand the organizational structure better before I would be fully assured.
  • From a software development perspective, I am certain OFN is built to higher industry standards because it is relatively new and because OFS has evolved from a less-robust original implementation. This is partly the advantage of hind-sight, partly the advantage of being about ten years newer, and partly the advantage of having a supporting organization with capital to spend on development. On the other hand, OFS has evolved to fill a need, so (maybe?) it has features that OFN does not and may never choose to implement. I don’t know.
  • In terms of extensibility, OFN seems designed to expand to a wide foodshed with many autonomous elements playing cooperative roles. In contrast, OFS is built for umbrella cooperatives to host producers and consumers in a single collective environment. My plans (if they ever come to fruition) are to implement server-server relationships for OFS so that various organizations can connect together to form a more extensible network. I don’t know exactly how OFN handles these sorts of relationships with so many players (see next point).
  • Being designed as a full-service solution, OFS incorporates elements for producers, consumers, and shipping/scheduling. I believe OFN does something similar, but it seems that there is no coordinated scheduling. What I’m trying to say is that OFN seems to allow each of the producers and consumers to set their own schedule (which is nice) but lacks any advantages of coordinated pickup and delivery for those producers and consumers – instead making the connections a possibly less-efficient one-to-one model. In contrast, OFS coordinates all the producers and consumers in a lock-step cycling schedule, so the transport can happen in combination with other groups, rather than separately. Because of the set schedule, OFS is less flexible, but accommodates by allowing a certain amount of opt-in/opt-out for producers and consumers in each cycle. I imagine OFN could (or does?) accommodate the opposite by allowing external collaboration between producers and consumers. This may be the largest philosophical difference (and it may also be very different from my understanding).
  • OFS incorporates methods for payment directly within the system. I don’t know whether OFN does that or not. One nice thing with OFS (particularly for wholesale consumers) is the ability to buy from multiple producers and pay with a single invoice. I don’t know whether OFN can provide that option. Likewise, producers are compensated with a single check/payment rather than multiple streams.
  • The infrastructure to run a system like OFS (and maybe/probably OFN as well?) is necessarily complex, requiring quite a lot of human attention. This seems to be a large problem with the system. With OFS, many producers individually wrap and label products for individual customers, which are then handled and routed uniquely. Some users of OFS have modified their implementation to allow shipping of bulk products (i.e. eggs, potatoes, etc) to locations where they are then allocated to the consumers. This takes some of the workload off the producers, but places a greater workload upon people at the customer locations. Certain products (e.g. those needing weights) can not be effectively shipped in bulk.
  • In most instances I know, OFS is used for small farm producers to individual family consumers. Some amount of wholesale sales have been added into the system, but it evolved from a retail-centric model. It may be that OFN (maybe?) is developed to focus more on the wholesale market consumers. I honestly believe OFS does a good job for the market niche it serves, but I am beginning to question whether that market segment is sustainable, regardless of the system that serves it.

Thanks, @eric, for the invitation to this discussion. Just for the record, the bullet list which arose from an email was thrown together fairly rapidly. I hope it is useful in starting a conversation, but it may be imprecise on certain points. Recognizing that this is an OFN-sponsored forum, I am happy to have this discussion, but I don’t want to be seen as trolling to promote OFS over OFN.

Ultimately, I don’t know OFN, and I have not had time to investigate it beyond a cursory level, but I’m sure it is very good at what it does.


Maybe a good next step would be for one of us to extract a list of OFN questions from Roys line items above, which could then be addressed individually in the forum: possibly be added the agenda of (or simply mentioned at) the next international discussion?

For example: Is coordinate scheduling a feature?

Actually at the moment I’m wondering what coordinated scheduling means in this case. Coordinating timing of product delivery with single pick-up day? Coordinating product availability schedules so that the products from multiple producers are available through the web interface during the same period?

Yes. There could be any number of OFN instances in a given region, which could theoretically be configured differently in terms of install environment, code base, etc. Within a single instance, hubs are limited to whatever functionality the OFN instance offers. Ideally, additional functionality can be integrated into the main repository/code base, otherwise we are dealing with forks and all of the divisions, if not divisiveness that often accompany forking. From my perspective OFN is also an exercise in (international) consensus building.

I don’t see anything to prevent ethically disabled organizations from using the code base in endeavors to exploit and profiteer. But there are OFN values being developed that would prevent such people from being considered part of the global OFN community. Again, in many ways we are using this fundamental aspect of life, our food system, as the foundation for building an international community where support and transparency are honored and cultivated. On a practical level, we have learned that each instance requires a significant amount of consistent technical support and it’s likely that the percentages of transaction amounts that can be added to each transaction can fund this. That fact means that there is still a “middle man” so to speak, but if I understand correctly, the financial aspects of the system are transparent to where each user is able to see what money is going where and why. @Selmo is this correct?

1 Like

This is very interesting - please could you clarify what the purpose of the comparison is? @MikeiLL @guising

It is interesting @pmackay, and I had been hoping that some of the US group would have delved further into it by now. The purpose of the comparison is partially that as per the recommendation of @NickWeir, we are hoping that @guising might become involved more with the US group, but in a larger sense, as OFS has been active and in use and development here in the US, it makes a good measure for some of the various features and use cases that might be in consideration as OFN is developed and evaluated. I hope this explanation is accurate and makes sense. @eric, @Selmo?