it's kind of ironic that the main value proposition of static site generators like Gatsby is that it's just a file so it's portable and you can take them anywhere, instead of getting locked into server/client based database systems, yet here they are, trying to monetize by building the "the best way to build and maintain Gatsby sites".
From my point of view, if they succeed in achieving that goal, that means Gatsby the open source project has failed, because it contradicts its purpose.
I can understand them raising a few million dollars, but $28M? Not so sure, I actually think the VC money will do more harm than good.
it's different from wordpress, for the reason I mentioned above.
The main value proposition of static site generators COMPARED TO 1st generation CMS like Wordpress is that they don't require a database and are lightweight and portable.
It is common sense that not everyone will want to run their own database and a server, so a wordpress.com makes sense.
But static sites can be deployed anywhere and there already is a huge ecosystem of vendor independent and free static site hosting sites like Github (which is subsidized by Microsoft), Gitlab, etc.
For Gatsby to survive and thrive in this ecosystem that is completely different from the wordpress ecosystem, they will probably need to do a lot of things to create a lockin, in which case there is no point in using Gatsby. After all, all Gatsby does is just compile some files into static site. Nothing magic in there.
The kicker with Gatsby is it’s actually not compatible with GitHub or any hosting where you can’t control caching headers. Gatsby relies on page-data.json files to not be cached[1] in order to work correctly. This is not possible with GitHub pages. After deploying a new version of the site, some users will use a cached version of page-data.json which can cause them to see a flash of the new data that quickly gets replaced with the old data. It can also cause the site to throw a JS exception in some cases.
Or only partially replaced, such that some React components get updated and some don't so if you have i.e. a list of stuff with text and pictures, the text will get updated but the pictures will not. It's a mess
Sounds like their offering will basically consist of some fairly simple CI and some static hosting, right?
Those things (particularly the CI) are incredibly trivial to setup for someone who is technically proficient. It might be a hurdle for someone who is primarily a designer with a bit of HTML/CSS knowledge and a "copy and paste heavy" understanding of JS.
I think a large proportion of static sites are built by people in the second category, so the offering makes sense to me. If my reasoning is sound, that just leaves the major issue that this offering is incredibly easy to duplicate, and everyone from Github to CircleCI to Netlify could put in a week of dev time to roll out their own matching offering (if they don't already have this).
From what i know, their CI offering tries to aim for very quick build times. Quick as in quick enough to actually allow dynamic looking website to be built with gatsby statically (eg. site rebuilt on every user post). Even their pricing reflects this, they charge by how many times the site is updated, regardless of build time.
I think that looks like a pretty hard problem, i wouldn’t call their CI simple.
> For Bob the marketing guy, it's just another tool to create a website.
To me this is the "emperor has no clothes" about the whole ecosystem. The actual users don't care if it's a static file, a database CMS, or a herd of cats running around frantically typing up HTTP responses. So it's odd to me that the big selling point is "we did something neat in how we made our product!". Right, but what does it do to make your users life better than using ghost or wordpress, or whatever?
I know that some devs see it as a rails like thing to combine with graphql, but if I was going that route I personally wouldn't start with something that's ostensibly a static site generator.
Bob the marketing guy couldn't care less where he hosts his Gatsby generated HTML website (at the end of the day it's just an HTML/JS/CSS bundle). He can just take that HTML and host it somewhere where it's FREE or provides much better quality.
Basically, the hosting can be unbundled from generating the site. And this is the territory which big companies like Amazon, Google, and Microsoft are already good at.
of course, Gatsby cloud won't want that, and the only way to solve that problem is to create artificial lockins. And when they do so (because they need to justify their valuation), people will leave.
I think we need to take over evaluations as a serious signal of danger in the economy. They are a signal of the advancing size of the financial class - which I think will have massively negative consequences.
My thinking here is heavily influenced by this documentary I just watched:
From my point of view, if they succeed in achieving that goal, that means Gatsby the open source project has failed, because it contradicts its purpose.
I can understand them raising a few million dollars, but $28M? Not so sure, I actually think the VC money will do more harm than good.