OOP 2024: No-code does not mean no model

TL, DR; Embrace no-code in order to explore more models and throw most of those models away. You will quickly discover what works, and what matters, in the business process that you are automating. If it matters enough, you can extract it into a high-fidelity design in code.

No code has entered the chat...

For a while now, no-code vendors have been telling organizations that they shouldn't be limited by expensive software engineers to build systems that are useful. No-code aims to commoditize the software production process. Commodification of technology leads to value if it removes a limitation, but successful adoption only works if the rules and policies that initially helped us overcome the limitation are replaced as well. Practices such as DevOps have to be adopted in order to reap the benefits of the commodification of compute and storage in the cloud. In order to benefit from serverless, system components need to be decoupled through message-driven designs. In order to benefit from no-code, people have to organize around the software production process in a different way.

Within software engineering communities no-code has been dismissed as a fad, saying the need for writing code will never go away because the needs of most software systems are too complex to capture in a visual design environment. This viewpoint ignores the argument that software engineers act as a gatekeeper, a limitation for the stakeholder to get what they want. It is reductionist to say that no-code means no-code. No-code is as much about no-code, as wireless is about the absence of wires, or serverless is about the absence of servers. No-code means less boilerplate. And no-code does NOT mean no-model.

The inability to deliver meaningful results in a reasonable amount of time is never out of bad intent, it's the consequence of rigidity in the system of work. If there is no room for experiments, for error, for trying again, then we shouldn't be surprised if people attempt moonshots. But if we can reduce the cost of experiments, then we should be able to iterate more, learn faster, and as a consequence produce more value.

Let's explore how no-code is able to remove the time to market of our ideas to explore new models. Join this session to uncover which rules, policies, and practices around modeling and design need to be replaced in order to reap the benefits that no-code has to offer.

Transcript

1. huizendveld.com #OOPmuc No-code does not mean no model
2. Marijn Huizendveld I help leadership teams at fast growing organizations accomplish more with the people that produce their software. marijn@huizendveld.com marijn.huizendveld.com
3. huizendveld.com Why I’m interested in this topic 💸 Software production eats budgets 🦾 Potential to reduce mechanical aspects 💭 Additional focus on the intellectual, 
 and collaborative aspects of design #OOPmuc
4. huizendveld.com 🧪 #OOPmuc
5. huizendveld.com 👥 #OOPmuc
6. huizendveld.com 🎯 #OOPmuc
7. Intentionally left blank
8. huizendveld.com #OOPmuc If you want a tools comparison…
9. huizendveld.com #OOPmuc If you are looking for a one-size fi ts all solution…
10. huizendveld.com #OOPmuc If you have been seduced to chose between code or no-code…
11. huizendveld.com #OOPmuc If you expect the mother of all buzzwords: AI…
12. huizendveld.com My goal is to start a conversation on no-code and DDD Because… #OOPmuc
13. huizendveld.com #OOPmuc Let’s get started → The future isn’t here yet, and that allows us to shape it.
15. huizendveld.com #OOPmuc 🌎 Context, 💬 Example, and ➡ Implications Our journey Today 🌎 Context,
16. © Atlantic Records
17. © Roger Parkes
18. Wha t are t h e c o m mon al i ti e s b etwe en Ed S h e era n a nd N ei l G a i man ?
19. huizendveld.com So what..?! #OOPmuc
20. huizendveld.com #OOPmuc 👩🎓 Learn, ✂ Copy, and 🎛 Adapt Explore other domains
21. © Chris F
22. Intentionally left blank
23. © DDDEurope 🥳
24. © Ed Sheeran
25. huizendveld.com Barely any ideas get thrown away #OOPmuc
26. huizendveld.com #OOPmuc 💡 An i dea Ex pl a i n val u e 👥 🤝 Ag r e eme n t Awa it c a pa cit y ⏳ Sta rt p rodu ci ng 👩💻 Dep lo y s o l u t i o n 🚀 🗳 Gat her d at a We i g h o p ti o ns ⚖ 🔁 Ite rat e
27. huizendveld.com Most organizations I work with attempt overnight success #OOPmuc
28. N ≧ 30
29. huizendveld.com #OOPmuc In the eyes of a business owner the status quo time to market is bad…
30. Wh y i s t here a m arke t f o r n o -co de to ol s ?
31. huizendveld.com #OOPmuc 💡 An i dea Ex pl a i n val u e 👥 🤝 Ag r e eme n t Awa it c a pa cit y ⏳ Sta rt p rodu ci ng 👩💻 Dep lo y s o l u t i o n 🚀 🗳 Gat her d at a We i g h o p ti o ns ⚖ 🔁 Ite rat e
32. Whe n doe s E d v is i t t h e r e c o r din g st ud io ?
33. huizendveld.com #OOPmuc Un c e r t ai nt y Des ign l i f e c y cle
34. huizendveld.com #OOPmuc Overnight success is a myth
35. huizendveld.com Why we aim for overnight success? 💍 Assumption: Single model to rule them all ✨ Environment: Institutionalized perfectionism 💔 Personal: Di ffi culty killing our darlings #OOPmuc
36. Tak e a m o me n t an d a sk y o urs el f : Is th e o r gan iz ati o n I w o rk f o r at tem p tin g t o ac h i e ve ov er nig h t su c c e s s es ?
37. huizendveld.com ✅ Next Action #OOPmuc
38. Intentionally left blank
39. huizendveld.com #OOPmuc
40. © Domain Language, Inc.
41. huizendveld.com #OOPmuc Whirlpool fl ow ≠ Artifacts fl ow
42. huizendveld.com ♻ 🚰 #OOPmuc
43. huizendveld.com Dismiss ideas as fast as possible #OOPmuc
44. huizendveld.com Leverage the model, and the abstractions it provides #OOPmuc
45. huizendveld.com Naive modelling in code 1⃣ Single path (e.g happy path, speci fi c case) 👩⚖ Little to no restrictions enforced 👀 Look back and ask: “What have we learned?” #OOPmuc
46. UsageId = NewType('UsageId', str)
47. class UsageId: usage_date: DayOfUsage customer_id: CustomerId @classmethod def from_string(cls, as_text: str) -> UsageIdType: [provided_prefix, provided_customer_id, provided_usage_date] = as_text.split(':') if provided_prefix != 'usage': raise SorryInvalidUsageId.because_of_unsuported_prefix() if provided_customer_id == '': raise SorryInvalidUsageId.because_of_missing_customer_id() if provided_usage_date == '': raise SorryInvalidUsageId.because_of_missing_usage_date() return UsageId( CustomerId.from_string(f”customer:{provided_customer_id}"), DayOfUsage.from_string(provided_usage_date) )
48. huizendveld.com We will get to the no-code part later #OOPmuc
49. huizendveld.com What can we learn from Ed Sheeran? 🚰 Make more songs domain models 🗑 Throw most of them away 🔁 Discover what resonates #OOPmuc
51. huizendveld.com #OOPmuc 💬 Example, and ➡ Implications Our journey Today 🌎 Context, 💬 Example,
52. huizendveld.com Leasy: Easy Car Leasing 🏢 Successful fast growing company ⚠ Big deals involve lots of manual labor 🧟 Many abandoned purchases due to poor UX #OOPmuc
53. huizendveld.com #OOPmuc Con vin c e th e p ros p ect Mak e it h ap pen Mak e the de a l
54. huizendveld.com An opportunity to explore domain models using no-code to throw most of them away #OOPmuc
55. © Kelly
56. © BMW
57. © Fox
58. © Cottonbro studio
59. huizendveld.com Prototypes deliver value Often not for the end-user, but for the designer #OOPmuc
60. Wha t is t he c on c ep t t h a t i s mos t o ften mi s sp el led in D oma i n -Dri ven D e si gn ?
61. huizendveld.com Ambiguous language 😝 #OOPmuc
62. huizendveld.com Ubiquitous language #OOPmuc
63. huizendveld.com Prototype: “A fi rst, or preliminary version of a device or vehicle from which other forms are developed” - Oxford Dictionary #OOPmuc
64. Hav e yo u eve r a ske d yo u r s elf : Wha t is n o -co de ?
65. huizendveld.com What is no-code to me? 🖥 Produces working software 🔑 Turnkey runtime 🔷 Visual programming paradigm #OOPmuc
66. huizendveld.com ⚒ Form Builders #OOPmuc
67. huizendveld.com ⚒ Automation Platforms #OOPmuc
68. huizendveld.com ⚒ Visual Databases #OOPmuc
70. huizendveld.com 🤔 What about domain models? #OOPmuc
71. huizendveld.com #OOPmuc Domain modeling is not about modeling data
72. huizendveld.com No code environments help to model data rather than domains #OOPmuc
73. Do e s tha t m a t t er ..?
74. huizendveld.com BYOM #OOPmuc
75. huizendveld.com Bring Your Own Model #OOPmuc
76. huizendveld.com “No code is about the absence of code, as wireless is about the absence of wires, or serverless about the absence of servers” – Romeu Moura #OOPmuc
77. huizendveld.com Software production is more than writing code #OOPmuc
78. How do w e bi nd t he mo de l to t he i m p l eme n t a tio n if we d o n’t h a v e a n y c od e ?
80. huizendveld.com
81. huizendveld.com Challenges ∞ Many moving parts and many assumptions 🔄 One part in fl uences the other and vice versa 🔮 Human behavior is impossible to predict #OOPmuc
82. huizendveld.com #OOPmuc
83. huizendveld.com The proof of the pudding is in the eating #OOPmuc
84. huizendveld.com #OOPmuc Con vin c e th e p ros p ect Mak e it h ap pen Mak e the de a l
85. huizendveld.com #OOPmuc
86. huizendveld.com #OOPmuc The value of no-code
87. huizendveld.com ✅ Removes busy work #OOPmuc
88. huizendveld.com ✅ Shortens the feedback loop #OOPmuc
89. huizendveld.com ✅ Proves assumptions #OOPmuc
90. huizendveld.com Downsides 🏗 Encourages production grade solution 🖇 Lock in on platforms 📈 Scaling is hard #OOPmuc
91. huizendveld.com Bene fi ts 🗑 Less attachment 👥 Larger group of authors ♻ Cheap to replace #OOPmuc
93. huizendveld.com #OOPmuc and ➡ Implications Our journey Today 🌎 Context, 💬 Example, ➡ Implications
94. Wha t do y o u t hi n k : If o ur p r o t o t y pe is go o d , s h o ul d i t s ta y i n n o -co de ?
95. huizendveld.com Should it stay in no-code? ⏱ Rate of change 🚦 Service expectations 📑 Decoupled using interfaces #OOPmuc
97. © Spearhead
98. huizendveld.com “[…]Long before the availability of this technology we developed […] rules to help us accommodate the limitation” – Dr. Eliyahu M. Goldratt #OOPmuc
99. huizendveld.com Technology adoption according to Goldratt 1. What is the power of the technology? 2. What limitation does it diminish? 3. What were the rules to deal with the limitation? 4. Given the new technology, 
 what rules should we replace them with? #OOPmuc
100. huizendveld.com My recommendations #OOPmuc
101. © Marijn Huizendveld
102. huizendveld.com 1. What is the power of the technology? #OOPmuc
103. huizendveld.com There are less things that require deep technical knowledge #OOPmuc
104. huizendveld.com 2. What limitation does it diminish? #OOPmuc
105. huizendveld.com The need for a large engineering team to produce working software #OOPmuc
106. huizendveld.com 3. Until now, what were the rules to deal with the limitation? #OOPmuc
107. huizendveld.com Rules up until now 🛣Predictable > innovative 🌊Big design up front 🏔Do not discard 🏁Finishing > succeeding 👥Outsource to agencies ♻Rework the same artifact #OOPmuc
108. huizendveld.com 4. Given the new technology, what rules should we replace them with? #OOPmuc
109. huizendveld.com Optimize for throughput: 
 Reach a volume of bad ideas fast #OOPmuc
111. huizendveld.com #OOPmuc Looking back
112. huizendveld.com No code 🔁 No code helps shorten the feedback cycle ⚠ Bring your own domain model into the tools 🔚 “Begin with the end in mind” to discover parts #OOPmuc
113. huizendveld.com Artifacts ✅ Not all artifacts add value to users ❌ Useful if they disprove assumptions fast 🔍 Help identify the artifacts we care for #OOPmuc
114. huizendveld.com Implications ⌨ No only writing code to produce software 🙈 Software production is more than code alone 🚰 Optimize for the throughput of ideas #OOPmuc
115. huizendveld.com #OOPmuc Final Words
116. huizendveld.com #OOPmuc
117. huizendveld.com In short 🕵 Use no-code to explore more models 🗑 Throw most of those models away ✂ Extract to code if needed #OOPmuc
118. huizendveld.com #OOPmuc What is on your mind? → My goal is to start a conversation 
 on no-code and DDD…
119. huizendveld.com #OOPmuc Did you like this presentation? Tell your friends Tell your boss Tell the internet
120. Are you eager to learn more? Marijn Huizendveld B.V. Hillegomstraat 12-14 1058 LS Amsterdam The Netherlands If you take the time to ask a good question, then I’ll take the time to write a good answer. marijn@huizendveld.com marijn.huizendveld.com