Web Editor Progress?


#21

I am gonna pop in and share some observations I’ve had here in the Hugo community.

When Bjørn implies things like, “the roadmap is in my head”, or “I work on things that interest me”, it isn’t a wall going up around development, it is modeling how one should approach the project. If you want to work on an API design proposal, that is in your head, and of course it is worth it.

But you gotta scale it to a community free software project. That means, iterate, discuss, iterate, reveal, that kinda thing. Check the #dev forum, or the issue tracker for a sense of how decisions get made.

The other thing is, no effort is punished. This depends on my prior statement about iteration. No one wants you to spend a huge amount of time working on something that runs counter to the direction of the project. But folks working to scratch their personal itches is a driver for Hugo development (and arguably any FOSS project).

Bjørn is the only devotee. In all likeliness e reads every post on these fora. That means if you want eir involvement, you need to get as many folks as possible excited about your idea. But a much better way is to take the initiative, do as much work as you can on your own, and then ask for assistance when you hit a wall, so you can continue on your own. Folks here will rally to help someone learn in order to improve themselves and the project.

If you are excited about a web editor that can plug into Hugo in a “native” way that assists in content creation, no doubt many folks will have input for that. My advice is to go for it, figure out an MVP, and post it for feedback. :slight_smile:


#22

But, re. the web editor/split view idea:

In my head, the hardest part is the GUI. I would get that right and wait with the rest of the Hugo integration. So:

  • A file browser (read from disk)
  • A Chrome window
  • A Markdown editor

None of the above needs an API – it needs to read from disk and a running Hugo server.

And @maiki is right, this is open source, so you cannot sit and wait for something. A good way is to start with a prototype, if that is really convincing you will get other people exited (me included) interested and involved.


#23

Thanks, y’all! Yeah, I’m totally cool with buckling down and getting stuff done. Had some bad experiences in the past doing work that a project didn’t want, so these days I always check in advance. :slight_smile: But seeing @maiki say this sets me at ease:

So all that being said, I can start to poke at a prototype of this along the lines of what @bep described. Like you said earlier in the thread, this is a bit more complex than other stuff, so it may be a little while before I surface again with the first prototype.

Thanks!


#24

I’ve been asked to add a blog to a site I’m building for a customer, one they can post to themselves, so quite interested in this. Needn’t be online though. I did come across Publii[1] recently which has an Electron[2] based GUI app.

I quite like the idea of something like that for hugo, I suspect anything more than basic features would require using compatible themes though.

[1] https://getpublii.com/
[2] https://electronjs.org/


#25

Looking into Publii app – they claim to be open source on the grounds that the Electron app contains the js/css that makes the app.

On OSX it seems to be in Publii.app > Contents > Resources > app .

Think I’ll have a play with Electron over the holidays and see if I can come up with something. If I do I’ll put the code up on Github.


#26

Thanks for this. I never really understood how ideas get moved into reality when it comes to development of app/platform design. Mostly because I develop websites and don’t spend a lot of time contributing to the platform I am using to develop with. But that was then, this is now. I’m transitioning my whole site from Drupal to Hugo. As one person said in this thread, I’ll never develop a website with a CMS again, it’s very difficult to migrate the data and its expensive to run as your site gets bigger as the database part of your site ends up requiring more and more resources.

I really like how Hugo is designed and how static site generators mix markdown with HTML design elements. It keeps content and design separate without having to rely on a engine to drive it the way it should be. I am just learning static site generation, it’s slow going but I am making progress. This is the first time I look forward to contributing to a tool I use to make my own sites. I want Hugo to succeed, it really has good design foundations, it makes me think of many possibilities to improve mine and others workflow in static content generation.


#27

Vim open on one side of my screen with Google Chrome open on the other side works for me. I save a file and the browser window updates with the changes. New extra software then what I already use.


#28

We have built a completely functional Hugo admin with Go, that runs locally or it could run over a network, if that was desired. Only been working on it for a short time. We think an easy to use editor would be able bridge the gap for many dissatisfied CMS users. Certainly our clients as well as ourselves will get great use out of it.


#29

@NexWeb - that looks really promising. Do you have a link you can share where we can experiment with it?


#30

Thank you for the feedback. At this point it works very well for us and our clients and we continue to add new features that we find of use. I posted it here to gauge the interest, which doesn’t seem to be overwhelming. Just as well.


#31

It’s because of the conservative UI. And that huge Gopher is a bit over the top.

Sorry. I didn’t mean to be harsh. But in 2018 people value design more than they did in 1998.


#32

this is what I’m looking for, I think it willl be great to have GUI and run offline for non-developer user.


#33

Getting ready to release some binaries soon. What is your favorite flavor, Linux, OSX or Windows?


#34

@NexWeb thats awesome, I’m using windows


#35

Hello there,

I’ve been developing my own Hugo CMS and I want to make it open source.

I’m using Electron, React JS and Material UI.

So far, it supports field types, fields nesting, options to expose only some configurations of a file (data or content), components customizations, one-click publishing.

Here are some screenshots I took today:

I the next few days I’ll release a video to measure the community interest.

Question to anyone who have an opinion: It’s OK to use the name Hugo CMS?

I’m eager to hear opinions.


#36

Looks promising. I am looking forward to the next steps.


#37

I think that you need permission for that.

CC @spf13 @bep


#38

I think @spf13 should answer that. I own the copyright to my own Hugo source code, but the logo is @ Steve Francia. Not sure about the Hugo name itself.

But what I would have done if I was making a CMS for Hugo (I don’t) was to find another cool and search friendly name totally unrelated to Hugo …

“wersdgsfdg - a CMS for Hugo”

wersdgsfdg should be search friendly, but probably not catchy enough :slight_smile:

It looks really nice, btw.


#39

Your screenshots do look cool. Is this CMS meant to be run online only or can it be used locally also?

Like @bep said above, you could give it a search friendly name and use a description line like:

A CMS for Hugo static sites.

Or something similar. That should be enough for search engines.


#40

Hello guys, thanks for the warm feedback.

Cool that you’ve liked the UI. I did spend some time on it.

Answering to @onedrawingperday’s question:

When I started to code, my initial goal was to make a web editor. But because a web editor requires more work, like designing the security rules for multi user access (user, groups, roles, token), and a server to host it, I decided to start with something simpler and let the web editor for the long run.

So, it’s a desktop app that runs on Windows, Linux and MacOS (I’m only running in Windows for now, but Electron support all those platforms). More about Electron.

Anyone who wants to use the CMS have to install the application and make a small setup to point to at least one site source (a Hugo site source). For now, the only available type of site source is the LocalFolderSource, but in the future we should see a GitSource, S3Source and so on. It will be very flexible.

If you want to develop a Hugo site and let a user with no code skills to edit it, you should give directions to the Hugo CMS (temporary name) installer and a small JSON configuration file.

Everything will become clearer when I release the first video.

About @bep’s name suggestion…

Nice name. Is Wersdgsfdg a mythological place or character? :sweat_smile:

I was avoiding the need to burn my brain mass thinking about a name - and a domain. It’s exhaustive. But yes, I’ll try to follow this painful path.

Maybe you or someone else can give me a clue about the meaning behind the name HUGO, then I can try to pick a related name.