Independent dev. Craftsman. Passionate about code and the pursuit of excellency. Get training or coaching at https://xitter.cf/t.co/2JioYZ8xbz

Grenoble
Joined December 2009
Tests are like a crying baby. They don't tell you why, but you'd better listen.
2
29
1
33
Johan Martinsson 🇺🇦 getting rid of ⛽-addiction retweeted
Nous prévoyons une session de #formation #Tests Design et #TDD du 6-8 novembre à #Paris : formation.hackyourjob.com/ca… Rejoignez @johan_alps pour (re)trouver le plaisir du développement grâce au Test Driven Developpement Détails et leviers de financement : linkedin.com/posts/hackyourj…
5
1
7
282
Johan Martinsson 🇺🇦 getting rid of ⛽-addiction retweeted
En mars, nous vous proposons en collaboration avec @axon_iq un atelier (à partir de 25€ la place) sur Paris et Lyon d'une 1/2 journée sur le thème de la Conception de microservices avec @axonframework et les approches DDD, CQRS/ES, EventDriven. Cet atelier se déroulera en français et sera animé par @allardbz Il vise à introduire les bases du Domain Driven Design, de Command and Query Responsibility Segregation (CQRS) et Event Sourcing (ES), ainsi que des approches orientées évènements, puis de mettre tout cela en oeuvre sur un cas réel en s'appuyant sur l'Event Modeling et la boîte à outils Axon. Il vous est proposé : - le lundi 25 mars après-midi à Paris : billetweb.fr/concevoir-des-m… - le mardi 26 mars après-midi à Lyon : billetweb.fr/concevoir-des-m… Possibilité de financement au titre de la formation professionnelle en vous inscrivant directement depuis formation.hackyourjob.com/ca… #Formation #Paris #Lyon #Axon #DDD #CQRS #EventSourcing #EventModeling #Microservices #EventDrivenArchitecture
8
4
837
I had the chance to participate in a doc-as-test workshop at #breizcamp by @sebfauvel The idea is to use the documentation as a primary source for our tests. I absolutely recommend it sfauvel.github.io/documentat…
1
3
3
1,627
It uses approval tests by @LlewellynFalco in new way And it somehow reminds me about TextTest that @emilybache introduced me to.
1
1
135
It creates bits of asciidoc that can then be assembled, possibly in various ways.
1
137
Mais quel accueil pour le sujet #3P ! Ca fait plaisir de voir autant d'engouement pour le legacy à @breizhcamp
10
1,192
Johan Martinsson 🇺🇦 getting rid of ⛽-addiction retweeted
Je commence le @breizhcamp 🦖 avec "3 P pour sortir du legacy" par @johan_alps. Dans lequel il aborde les notions de refactoring, de tests, de coût, d'effort, de temporalité. Article relié : martinsson-johan.blogspot.co… #breizhCamp
1
4
2
6
827
Why small steps? I was discussing with colleague what would be the interest of coding a story with several mergeable steps (i.e. non breaking changes) One reason is Options ⏬ I made this viz while discussing, though I'd share it
1
3
475
One counter-point that was advanced is that, in particular, without tests it's too costly to stabilise every small step. Indeed that is true, but just because you have the option of stabilising with some manual tests doesn't mean you have to. That's what an Option is about.
1
94
This is a benefit that is valuable even when working alone on something When working in a team on the other hand there are a load of other benefits, like no merge hell, more opportunistic refactoring, etc. But that'd be a subject of some future thread
91
Excited to be on my way to @breizhcamp! I'll be promoting crappy 💩 tests as a part of 3P (Protect - Prepare - Produce), my favorite legacy mitigation technique.
1
10
875
Kent Beck on cost of software: Cost(software) ~= Cost(change) Cost(change)=cost(coupling) Reducing that also requires decoupling Cost(software) = cost(coupling) + cost(decoupling) #DDDEU
1
32
4
91
10,494
Decoupling ~= increasing cohesion Cohesion = Grouping of coupled elements i.e. Cohesion = Copling(e1, e2, ..) Meaning that we optimise the span of coupling so that things are closer together All from the great goodreads.com/book/show/5784…
1
7
593
Finally adding that when doing a behavioural change. We must consider separating into a stuctural change (refactoring) and behavioural change. Iteratively, of course.
3
7
665