ERP systems are complex, but implementations are often large and complex mostly due to inadequate tools and practices, and generally poor organization, resulting in large volume of code that is highly redundant and inflexible. It doesn't HAVE to be this way, it just usually is.
If the discussion of Moqui in the context of OFBiz is the framework (toolset), then Moqui Framework is already far more advanced than the OFBiz framework. That has been true since late 2012, and Moqui Framework has still progressed quite a bit since then. The Moqui Framework code is also only about 15% of the size of the OFBiz Framework code. There are many reasons for this including a well defined and minimally redundant API (using DSL style classes in many places), implementation in Groovy instead of Java, and extensive use of existing open source libraries. On the business level the Moqui Ecosystem is structured different from OFBiz. Instead of a single project there is a framework project, a business artifacts project (with entities and services, no UI), and then an ecosystem of projects based on them. The Mantle Business Artifacts project, even just comparing to the services in OFBiz, does not have as much functionality as OFBiz. The existing applications based on Moqui are also not nearly as extensive as those in and based on OFBiz. For those who don't believe that this can be done quickly and efficiently, development in this area only really got started in late 2012 in the Moqui Ecosystem. The initial functionality was focused on basic order placement (with a simple UI in the POP Commerce application), and more extensive end-to-end project management and billing (in the HiveMind PM app). End-to-end order processing functionality including shipping, invoicing, payments, inventory, and accounting/GL were implemented in around 60 days when I finally had space to work on it nearly full-time as opposed to occasional part-time efforts before. This can be done quickly with good tools, good design, and well structured, flexible code. The current functionality in Mantle Business Artifacts (with various end-to-end automated tests that are also a good reference of some of the high-level services and the resulting database records) are summarized here: http://www.moqui.org/mantle.html Most of this was finished in late 2013, and there are already a couple of custom ERP systems for specific industries (that I'm aware of...) based on these that are nearly to the end of initial development and ready for deployment. On the open source front the most extensive application is the HiveMind Project Manager, which was an interesting initial focus as it is something never really done completely in OFBiz. It addresses general needs of services organizations and includes project/task, request, and content (wiki) management with support for multiple vendors, multiple clients, flexible billing rates, time recording, expenses, invoicing, payments, and general ledger. More details and releases are available here: https://github.com/jonesde/HiveMind This wasn't true a year ago, but right now I would recommend that clients use Moqui instead of OFBiz whenever extensive customizations are needed and the plan is to build a complete UI anyway. Most larger projects based on OFBiz are in that category, including nearly all of the OFBiz projects I have consulted on over the past 3 years. That said, I should also say that until this year most of my consulting work was based on OFBiz, including various multi-million dollar custom ecommerce and ERP projects. As far as general velocity and future potential, things are already moving quickly and just getting to a point where the ecosystem structure will become more important and allow things to grow very quickly. The general idea is instead of a single project with various groups compromising over functionality there is an ecosystem of complementary and competing open source projects and commercial products. This allows as many people to get involved as want to with no stepping on toes. With a common framework and foundational business artifacts these are implicitly integrated (as much as app developers follow existing patterns and use existing services), and the opportunity for various user interfaces focused on different industries and organization sizes. Being business process and target user specific these are also much more usable OOTB, and still customizable given the toolset and foundation business artifacts they are built on. Lost cause and not heading there? Hardly. The current market environment is FAR more difficult than when OFBiz started. There is much more competition and many more open source alternatives. When OFBiz was started just being an open source eCommerce and ERP system, even with pretty basic and poorly implemented functionality, it got huge interest. I spoke at all sorts of conferences and landed all sorts of contracts just because of the novelty of the thing. It's really quite a wild story, and I feel lucky to have been a part of it. That just isn't enough to be interesting any more. Even still, the Moqui Ecosystem projects are far further along that OFBiz was 3 years in, and in spite of few contributors and users, the software itself is advancing at a faster pace. Could OFBiz (both the software and ecosystem) be restructured and organized differently to leap forward even more? Probably so. I've already communicated the reasons that I chose to do these as separate projects with a clean rewrite of everything (no code at all comes from OFBiz), but others could choose a different path and perhaps do just as well... and the existing framework and functionality in OFBiz could perhaps be cleaned up enough to act as the same sort of foundation for growth. I am not writing this to try to recruit Moqui contributors and users from the OFBiz community. To date that hasn't happened anyway and I don't know that it will very much. OFBiz is still a good system and many contributors and users already have a lot of momentum with it. Currently the vast majority of Moqui contributors and users did not come from the OFBiz world, only a few have. Even investors I've spoken with in recent years consider my approach to be a bit crazy given the existing momentum that OFBiz and organizations around it already have. My desire is to do better. The development efficiency I had in mind in the early years of OFBiz has been dampened by a number of things, and it just doesn't have to be. My path is my personal decision, and I've enjoyed it very much so far, with both OFBiz and the Moqui/etc projects. Pursuing the best possible solutions in whatever way is necessary is deeply satisfying. The most difficult thing I face these days is that I still do a lot of consulting based on OFBiz, and it is hard to explain that there are good solutions to various problems and try to explain why those aren't done in OFBiz itself already. It doesn't have to be that way, even if I did personally choose another path for myself. Predicting the future is a tricky thing. The only guarantee is that the more detailed one gets the more one is incorrect. Anything could happen. For those who think things can be better, just do it. All it takes is what is these days a cheap computer, some time, and personal experience and knowledge. Competition isn't a bad thing, we can all learn from each other and choices in a "marketplace", even an open source one, are a good thing. I'd still love to see OFBiz and its various derivatives take off and go amazing places. In various cases I still recommend OFBiz and its derivatives, especially the ones that are more publicly available and that I'm aware of (like Big Fish) even though I know there are many others made by different companies that are just used internally or for their clients only. -David On Mar 9, 2014, at 9:25 AM, Mike <[email protected]> wrote: > I agree and disagree. If anyone could create a full blown ERP using Moqui, > David could do it. After all he has already managed to do it once, but > yes, it is (probably) years away. However, I have no idea of it's current > state. > > Regarding the current imperfections of OFBIZ: All large software projects > have hacks and shortcuts built-in. I bet SAP has thousands of hacks that > the original designers wish were not there, but are now so ingrained into > the framework they are impossible to change. For me the biggest question > is "does it work"? OFBIZ already seems to. Good enough. > > > On Sun, Mar 9, 2014 at 6:42 AM, Paul Piper <[email protected]> wrote: > >> I have to second Adrian's opinion on this: >> >> From my perspective Moqui is a rather lost cause. Not because I am not >> impressed by David's work, but rather because it is years away from >> becoming >> an actual thing. From an industry perspective it is difficult if not almost >> impossible to implement an ERP framework to begin with. OFBiz is a >> marvelous >> thing in that extend as that it managed to actually become feature >> complete. >> It is, however, an exception and not the rule. Any new "framework" would >> require a huge amount of either time or money, and frankly Moqui won't be >> heading there. >> >> Again: this doesn't take away anything from David's accomplishments, but if >> you want my personal two cents on this - ofbiz is what you are looking for. >> >> >> >> -- >> View this message in context: >> http://ofbiz.135035.n4.nabble.com/Reposting-from-David-Jones-on-Moqui-tp4648884p4648953.html >> Sent from the OFBiz - User mailing list archive at Nabble.com. >>
