Nice one that could serve as inspiration; they look professional, and it inspired confidence: https://chain.com/, https://onyx.org/. Those are the kind of websites you compete with, let’s not forget.
Jekyll is an interesting option. I’m not sure what the conversion effort would be for our sites not yet using it. When reviewing @Plats 's PR, I discovered the Tari.com website is running off NextJS, and I’m not entirely sure what it’s doing which is dynamic and requires that kind of horsepower. If we audit it and find it’s not doing anything that actually needs that kind of dynamic work (or said dynamism can be moved out) this sounds like an improvement.
RFC.tari.com was built on mdbook because that’s a standard Rust documentation tool and the site was originally built for the purpose of documenting the protocol and publishing the particulars of it by the Rust developers who were building it. Most of its contents might be portable to Jekyll, as both use Markdown, but care would need to be taken to ensure the URLs still worked, as they’re not intended to be moved.
The wiki doesn’t use static site generation because it’s intended to be WYSIWYG edited by anyone in the community with enough trust to write an article (a much lower bar than other tasks, and one that’s not very technical.)
I can’t speak for the others-- I don’t know their reasoning.
If Jekyll is too complicated for some users who don’t know how to code, wouldn’t the Tari developer be able to create an application with a user interface to help them?
I’d rather use a stable, maintained application rather than having a Tari dev go and create a one-off application that we’d then have to maintain in perpetuity.
I don’t know any application that works with Jekyll. Maybe you know one; I think of it more as a tool for the marketing team. They can use it for posting new posts and maintaining the website, but it could also be used to post on ALL your social media from one app, making it easy to track your virtual presence and results.
I thought you were referring specifically to the wiki based on the previous context.
I feel like there are multiple things being discussed here in parallel and it’s hard to determine which you’re referring to. In short I can see us moving the main website to Jekyll (or even WordPress) long-term. It would be good, if and when using a static site generator, to standardize on one. Whether that’s Jekyll or Astro or whatever doesn’t much matter to me so long as we do it. For things like the wiki I’m less jazzed about moving them to a static site generator.
Right now, though, all of these have working sites so it’s hard to see spending too much time on this when we can make improvements to things like SEO in the current repositories. It could be someone’s project to convert them, once we spend some time working on a consistent layout, perhaps.
Wordpress will be a better choice: easy for the team to work on it, multiple plugins for add-ons, a good team behind the project, free, a new AI tool, WordPress Studio – Agentic Local Development
Try it; download the Studio app and my backup for the Tari WordPress.
Most of the text is already done. Ootle is incorporated. I removed the update page and the store one; I updated the info in your user terms and privacy policy; now it says Tari instead of Tari Labs, but I did not touch the address at the bottom. I changed some pics in Tokenomic, but I don’t like them; they’re not professional.
I have never heard of Jekyll and I’m not entirely sure why we need to move to it?
Nextjs is fine. I’d argue given there may be more people contributing to the website using coding agents I did, nextjs is probably a more standardised framework to use not to mention its native to vercel if we wanted to use that for deployment.
Based on previous conversations, my understanding is the motive of this is seo-centric?
If that is the case, the website itself doesn’t need to use Jekyll.
I think if we want improve the seo friendliness of the website and our ability to contribute to it, I recommend we explore something like sanity.io for cms.
However like @Fox has said before, given that even my simple navigation changes haven’t been made live yet - there is some work to do before we can get here.
It’s not just SEO centric, it’s only part of a more global problem that you address one of it here, Make Tari Channels Great AGAIN! , Take a look at your YouTube channel, https://www.youtube.com/@taricommunity/videos, none of your videos have a description. Why don’t you put the summarizer into the description? How do you expect Google to reference your project if there’s no text? There’s no marketing team with experience, and your information is scattered across 4/5 different platforms instead of being grouped in one place; the same with your social media: how many channels do you need? A big clean-up is needed. You only need two channels, one slow and one fast. Your website is your slow channel where ALL your info/doc/university should exist, and your fast channel should be 1 chat room, either Discord or this forum, not both. All your social media should only be used for promotion. That’s only the start.
With regards to having the forum on Discord - I believe this has already been discussed and given the lack of persistence and continuity on discord, it’s not an effective tool.
On the Youtube description side of things I actually agree with you here. This is where I think it would be great if we had some sort of process around how we could get things added into the Tari Agent, as in principle this entire workflow is a community content workflow and with today’s tools, this can be entirely automated.
The issue with YouTube descriptions is that summaries cannot be generated until after the Tari Show has concluded, so I cannot do these ahead of time (we often deviate from the “agenda”), and then I forget later.
I fixed it last time I was reminded, and I can fix the latest one today.
It’d be great if we could just get the TariAgent to ping me on Discord or post in the writer’s room with a reminder to do these weekly tasks as they occur, because many of these things (like the summaries) are collaborative processes with multiple people involved.
We don’t need to. It would potentially make our various sites/subdomains more consistent since they could share code more easily.
Jekyll is very possible for static sites. I’d expect there to be plenty of training data available for it.
Jekyll would probably be a better system to build the website on, if we really don’t need its dynamic functionality. We haven’t determined that yet since no one’s audited the code to see what we actually use Next.js’s features for. However, the switching costs are an issue. As you note, we can address the majority of the SEO issues without spending the effort to rip out and replace the platform.
I think I know what caused the problem here, since the GitHub actions never seemed to run. I’ve opened a PR with instructions for @possum to give them a try.
