remove all references to staticman endpoints that are affected by quotas

[onweru]

It is now obvious that the public instance of staticman is broken, and will probably stay that way indefinitely.

It would then make sense to remove any reference to that endpoint. Why? It's misleading as an option, since it doesn't work anyway. This is also true for all links that point to the dated staticman docs. It doesn't matter if some sections of those docs work if as a whole, they don't get the job done.

As a courtesy to users, it would be immensely helpful to only document options that actually work. Is it possible to document a concise guide so that one doesn't have to get lost in the weeds (issue threads and fractured documentation pages)? .

What we have at the moment reads like this is the default option but it doesn't work, but you can try it.

In hind sight, this 👇 too needs to be removed. Why? At this point, these information is another turn deeper into the annoying rabbit 🐇 hole.

ℹ️ This Binary Mist article could also be quite helpful :)
ℹ️ By default, this theme uses the official production instance at v3 instead of v2 due to a requests' quota issue reported in issue eduardoboucas/staticman#222.
⚠️ Since Staticman is an evolving project, things might work differently than they do at the moment of this writing.

We need to rethink the staticman section because as it is, it utterly frustrating to anyone looking to get up and running quickly.

Maybe we could even extract it to a dedicated documentation page(s), with procedures that work. All other references could then be linked as footnotes.

[VincentTam]

Move detailed instructions like the following to the Wiki page, and keep the essential parts.

---

👷‍♂️ Staticman Configurations for Beautiful Jekyll

📧 Please don't hesitate to ping me on GitHub/GitLab/leave a comment on my blog/send me an email to ask for help.

1. 🏁 Invite @staticmanlab as a collaborator to your repository (by going to your repository Settings page, navigate to the Collaborators tab).
   
   Image short link: https://git.io/fjZPx

2. ☑️ Help @staticmanlab accept the invitation by going to

    https://staticman3.herokuapp.com/v3/connect/github/<username>/<repo-name>
   

   In your case, that's

    https://staticman3.herokuapp.com/v3/connect/github/<username>/<repo-name>
   

   
   Figure: image showing setup for my fork of this theme
   Image short link: https://git.io/fjZXG

   
   Figure: @staticmanlab invited to GitHub repo
   Image short link: https://git.io/fjZXZ

3. ✍️ Fill in your repository and branch in the Staticman section of _config.yml.

    # Staticman support 
    staticman: 
      repository : "<username>/<repo-name>" # GitHub username/repository eg. "daattali/beautiful-jekyll" 
      branch     : "master" # eg. "master" If you're not using `master` branch, then you also need to update the `branch` parameter in `staticman.yml` 
      endpoint   : # URL of your own deployment (with trailing slash) (will fallback to a public GitLab instance) 
      reCaptcha: 
        # reCaptcha for Staticman (OPTIONAL) 
        # If you use reCaptcha, you must also set these parameters in staticman.yml 
        siteKey  : # Use your own site key, you need to apply for one on Google 
   secret   : # ENCRYPT your password by going to https://staticman3.herokuapp.com/v3/encrypt/<your-site-secret> 

4. 📝 Start commenting.

[onweru]

@VincentTam, yeah, something like 👆. staticmanlab doesn't have quotas, correct?

[VincentTam]

@VincentTam, yeah, something like . staticmanlab doesn't have quotas, correct?

It's running on a free Heroku instance, so there is quota, just like the official development instance. I won't regret my choice since it's supposed to be a test for Staticman's native GitLab integration. One might setup his/her own API endpoint or find other existing ones like @staticmanaima, @BloggerBust-bot, etc.

[onweru]

I think setting up their own instance is the safer option. It's not that straight forward though, is it? It's not to me 😄 . I think one concise guide on how to setup a heroku / docker instance would be more great.

This way, we don't even have to advise people to signup to specific bots. After all, those bots are bound to reach their respective quotas at some point 🤔 ... at which point they will find themselves where they began.

[VincentTam]

Staticman config in this project's Wiki page was published several minutes ago. It's copied from https://github.com/victoriadrake/hugo-theme-introduction/pull/119/files#diff-04c6e90faac2675aa89e2176d2eec7d8 with minor updated.

I think setting up their own instance is the safer option. It's not that straight forward though, is it? It's not to me . I think one concise guide on how to setup a heroku / docker instance would be more great.

• A anti-spam fight will never end, so it's too soon to say which one is safer.
• Without the help of the author of Staticman's GitLab integration, I wouldn't have set it up.
• Would you call Data Science blog's Staitcman self-hosting on Heroku guide concise?

This way, we don't even have to advise people to signup to specific bots. After all, those bots are bound to reach their respective quotas at some point ... at which point they will find themselves where they began.

I have another view of point. It would be hard to grab the message from a large chuck of text/code. Having some screenshots illustrating a concrete textbook examples helps interested users to understand what Staticman is actually. By using this example, I would be suggesting new comers to try my bot, but I think experienced users are wise enough to judge.

[onweru]

• A anti-spam fight will never end, so it's too soon to say which one is safer.

I meant, safer in terms of one's comment form being rendered useless after a quota has been reached. If one has an instance of their own 👇

If the only app that you are running is Staticman, however, it is unlikely that you will use up you free dyno hours because this would mean that you are receiving comments for more than 18 hours per day, which would be quite a lot.

• Would you call Data Science blog's Staitcman self-hosting on Heroku guide concise?

I have read it and I like it. I will try it and see how it goes. Thank you.

[VincentTam]

If the only app that you are running is Staticman, however, it is unlikely that you will use up you free dyno hours because this would mean that you are receiving comments for more than 18 hours per day, which would be quite a lot.

Not related to the title, just for responding to your comments about frequence of comments received.

You may see @zinefer's GitHub Page repo:
https://github.com/zinefer/zinefer.github.com/pulls?utf8=%E2%9C%93&q=is%3Apr+label%3ASpam
to see how the GitHub App Staticman spammed his repo. I believe (s)he isn't the only victim.

[onweru]

Looks pretty awful. Do you think it would be cool to add a disqus fallback which users can choose to activate?

[VincentTam]

Looks pretty awful.

Me2

Do you think it would be cool to add a disqus fallback which users can choose to activate?

I'm not keen on that, due to reasons found in the linked articles in Pelican-Elegant/elegant#288.

It's the user's choice to grant permission to a Staticman bot. What we, as a maintainer of a static theme, can do is to offer suggestions/guides for best practices, say, those mentioned in cc386c7#r34762244.

What I, as the maintainer of a Staticman instance, can do, is to fight spams.

[onweru]

What we, as a maintainer of a static theme, can do is to offer suggestions/guides for best practices, say, those mentioned in

Fair enough, let's keep things simple, as they're, and keep improving the docs wherever possible.

[zinefer]

Staticman did not necessarily spam my repository. I made a choice not to use recaptcha because I wanted my users to be able to add a comment with absolutely minimal effort. I am only using a BotPot field right now. It worked quite well for a while but it seems a few bots have figured it out. I have been considering changing up the field to try to trip them up again.

In the past I have written my own commenting system in PHP and that system had an order of magnitude more spam.