January 2020

Refactoring continues with light at the end of the tunnel :-)

We are now in our fifth month of our refactoring to Svelte. Occasionally we question our sanity in making the decision to switch from Vue, but it is easy to forget the reasons why when a project seems to be burning time. So let's re-cap.

Reason 1 - Vue

Vue was great to us and gave us a head start, but due to the architecture of an SPA the Vue library was getting bigger and bigger as we added functionality and that caused page speed to suffer. 

We weren't having that. Jolojo has always had speed as a foundation requirement.

Reason 2 - KoP & Templating

KoP? What the heck? 

Ok, in plain English 'Kit of Parts'. 

Still no clearer... Another foundation requirement was to allow front end developers to create sections of website (Headers, Content Areas, Footers) as KoP's that could be re-used across the site (and potentially other sites). These would then be included in a library that could be dragged on to pages. So the aim was to make things easier for FE developers who in turn would be making this easier for Content Publishers (or webmasters) of various skill levels. 

But we built in a bug in the development of the Vue version, this meant that we could only have a single KoP on a page, if multiple KoP's were added to a page (of the same 'type') they couldn't create individual IDs so became the same. 

Templating for websites is hard. Coming up with a framework that is both quick and powerful without overloading the page load with unnecessary CSS and Javascript is really hard. Just look at some of the site builders out there and see how efficient they are. Some are OK, most are rubbish.

Although we are happy with the existing engine and can create really good, responsive templates pretty quickly. We decided that this could also do with a few tweaks. So in the refactoring pot it went.

Reason 3 - OpenSource / Licencing Jolojo

In a nut shell, we weren't 100% happy with the quality of the documentation from within the code. We are not sure that 100% happy is achievable, but as and when we release the platform with an OpenSource licence, we want to be confident that it can be used quickly as opposed to developers having to reverse engineer the code to work out what the heck is going on...

We know we could do better, and time is being spent to make sure we are much more confident that other developers can get going much quicker.


We are getting there.... We haven't bogged ourselves down with re-inventing the wheel. Jolojo Vue was damn good, Jolojo Svelte will ROCK!