Get ready for #Shopware 6.6 RC coming at the end of January. In this blog post, we cover you with everything you need to know.
Do you have questions or feedback? Let's chat on Slack in the channel #feedback_RC_6_6 💬
https://t.co/UmKl5itBN1 #ecommerce
My blog post on mocks and their influence on software design is now also published on https://t.co/VlydD0KPGP #DEVCommunity@ThePracticalDev
https://t.co/c8z6ZmUTr3
Interesting write-up by @KentBeck & @GergelyOrosz on how to measure developer productivity and especially the importance of differentiating between Effort, Output, Outcome and Impact.
https://t.co/9maZ8uP8rk
@mhgontijo@Sylius At least that's what we are striving for in terms of extendability.
/end
(I hope twitter/x does not completely mess up the answers, first time I'm trying to write a thread)
@mhgontijo@Sylius While I understand where you are coming from and seeing why it seems like to be a good idea from the outside, that's a direction we won't take. As IMHO that practice is simply not sustainable, because it creates such a high maintenance burden. Let me explain why: 1/n
@mhgontijo@Sylius 2. Faster adoption of fixes/improvements/new features, as they don't break public APIs that often
3. clearer ways of extending specific features that are use-case driven and not byproducts of the implementation
8/n
@mhgontijo@Sylius That means that we want to design specific interfaces that can be used as extension points, but protect our internal implementation from becoming part of the public API. That has the following benefits:
1. Less breaks in the extension points
7/n
@mhgontijo@Sylius What we want to do is not expose so many implementation details (in the form of interfaces for every class), but instead make behaviour extendable/swappable with the least amount of coupling possible. 6/n
@mhgontijo@Sylius Which additionally means the community can't adopt the latest features and improvements constantly as we finish them, but is hold back by the major release cycle, again bad for the community/ecosystem. So our approach is a little different 5/n
@mhgontijo@Sylius The other option would be that we slow down the rate of change, meaning that we will ship improvements and new features less frequently, again this would lead to more breaks in major versions as there is more stuff in them 4/n
@mhgontijo@Sylius Thus, a lot of changes will require a major version. That means that major versions probably are coming more often and with more breaks, as simply more changes are breaking, this will have a negative impact on the community 3/n
@mhgontijo@Sylius Interfaces introduce the highest level of coupling that you can have, even adding new optional parameters or broadening argument types is a breaking change, so as a system provider you are really limited in what changes you can make in a minor version 2/n
I've published my first blog post on https://t.co/VlydD0KPGP on why you should separate core code from infrastructure code #DEVCommunity@ShopwareDevs https://t.co/3R233ReQAp
Had a blast yesterday evening presenting shopware 6.5 from a dev perspective at the vienna shopware meetup. Thanks for having me!
You can find the slides for my talk here:
https://t.co/MBNWc76wdV
I'll be speaking at the @ShopwareDevs
meetup organised by moonshiners this Thursday in Vienna about "Shopware 6.5 - The Dev perspective".
https://t.co/QimmVmzZsu