Who is the Jamstack for in your opinion? I'm a web developer that makes a lot of apps and marketing sites for small to medium businesses and I find it very hard to justify the added complexity of a Jamstack architecture - there's just too many moving parts.
It seems to me you need a few things:
- A dedicated development team to manage the stack
- Enough traffic to justify the complexity
- Regular content publishing to make use of the build pipeline
- A need for speed beyond what's regularly acceptable
Great work anyway on the book, it's good to see people charging for curated knowledge.
Added complexity? At its simplest, Jamstack is just a static site generator. A service like Netlify will do the entire build pipeline for you (for free) just by pointing it at the git repo. Or if you want to DIY the hosting, then all you have to do is find somewhere that can host static files.
At its simplest yes. But if you are proposing this as a solution to a client, they will likely want to edit content. That will have to happen on a headless CMS instance. That's whole new stack to maintain alongside. You then need to set that up and configure the SPA to talk to it (GraphQL/Apollo?). When I come back to client project in a month, is all this still going to be working ...?
Maybe, but it's certainly not less complex then a traditional Wordpress site. So I'm wondering what are the criteria for using Jamstack successfully?
As you said it could be that it's good for simple static sites but not great for more complex content-drive sites - I don't know, that's what I'm asking.
Obviously it depends a lot on which CMS you're using and I can imagine there being a lot of complexity involved.
That said, setting up something like Contentful and integrating it with NextJS is roughly a day or two of work if you're doing it for the first time. And it all works over REST so you don't have to worry about GraphQL.
I think it's good for simple cases where a technically minded person will be editing the content. And I think it's good for complex cases where performance is important and the organisation is large enough to support at least one employee to manage the system. I think it's (not currently) good for the in between cases.
Other than the simplest of static/marketing sites, jamstack is just taking standard server-side rendered frameworks and splitting it up into several layers: CMS and other content sources, external APIs, build process, backend "serverless" functions, etc.
This is more moving pieces and complexity. Yes it's easier to build the frontend using modern component models but I find the tradeoff is rarely worth it.
I think you might be underestimating just how important speed is becoming.
Not only is it now a very important ranking factor for SEO reasons. There's also going to be a point for most websites where the complexity (I'd argue that the standard WordPress install is a lot more complex than most JamStack setups, if you're going to be digging into the code) will eventually need to be offloaded to a build process that spits out a static site, just to meet the 'acceptable' threshold.
i think there's varying levels of complexity to the different solutions available jamstack or not
Jamstack, the complexity might be maintaining different services, such as a headless cms and the front end. the output though is static files that you can dump into static hosting pretty easily
on the other hand, you might have a serverful solution thats all in one stack. you run into other complexities with that like having to worry moreso about scaling and managing that server for traffic
there are tradeoffs between the 2. jamstack isn't a perfect solution but it has a lot of benefits. it's also relatively young in it's architectural lifecycle, so i would imagine a lot of these pain points to be worked out as it matures
the book tries to get into both the good and the challenging
I am a solo developer. I have built 6-7 projects which fall under Jamstack. I find it that it lessens the complexity, and as a bonus I can sleep much more easily at night knowing that my sites are hosted as static websites.
i think there's a lot of truth to that - but they do is so fantastically well
there are a LOT of pain points to deal with rolling out all that is included with the base netlify offering
hosting a static site is pretty easy - but configuring routing between cloudfront and s3 can be complicated so that the reequests don't always reference the root index.html
creating an autodeploy infrastructure can be challenging as well, tools like github actions might make that a little easier, but there are a lot of considerations there
it also makes these kinds of solutions more accessible to developers who might not have the understanding or interest to set up that kind of infrastructure. front end devs can build sites, connect, and go with little fuss
Agreed. As much as I would hate it as someone who utilizes their services, Netlify should look for an exit. AWS is probably the best bet. Between services such as Github Pages, Vercel, and Firebase there’s a lot of competition.
It's possible but so far Microsoft and AWS have not developed anything near as comprehensive as Netlify's solution. For example, Microsoft's static web apps solution feels very much like early days Netlify and I am not entirely sure how committed MS is to this product.
Honestly, I suspect that if any of the big players wanted to get into this space for real, they'd just acquire Netlify rather than try to recreate everything Netlify already has.
I like the idea of JAMstack and will likely use it for simple toy projects. But I do not like the idea of promoting paid products on HN. this link is basically just an ad telling users to "buy the book".
There is nothing for me to read, learn or discuss on the landing page unless I buy the book.
thanks - thats a good point - i plan on releasing the first tutorial free actually - i just haven't had the time to get that out yet, that'll be hopefully posted on freecodecamp.org in the next couple of days
Can anyone point me to a good headless CMS solution that doesn't introduce vendor lock in? I've tried Contenful and I only got nothing but good things to say about it. Probably something open source and self hosted.
Any git-based CMS would not have vendor locking (i.e. Forestry or Netlify CMS as two examples) since they are only an editing layer on top of content and data files in your git repo. Netlify CMS is also open source. There are limitations to using a git-based solution (reusability of content within a site or reusing content on a mobile app for example) but no vendor lockin.
For the book - I don't think there really is a high expectation of knowledge, except maybe just general web development from a high level. The hope for the informational part is to get anyone acquainted with it