Troubleshooting Brave mobile browser problem

When my main site got switched to https, I had some problems with sub-domain sites. One of them was my Old School Drummers river of news site. I cleaned up some http/https references, and got it to appear on desktop and mobile, but the mobile version did not show the text. I used Developer Tools in Brave and cleared up what I think is the remaining issues, but it still does not display in the Brave mobile browser for Android (that is, news items do not appear). It does work in Google Chrome for mobile. Lazyweb – any thoughts?

Trying to get my WordPress site back

Earlier this week, I wanted to start posting again on my website, AndySylvester.com. When I went to log in, all I saw was a white screen. After some searching, I found that this was a thing for WordPress sites (“white screen of death”). Ugh – just what I needed! I found a few articles (https://themeisle.com/blog/cant-access-wp-admin/, https://www.hostinger.com/tutorials/fix-wordpress-white-screen-of-death, https://www.hostinger.com/tutorials/fix-locked-out-of-wordpress-admin-issue)

and did the following steps:

• Added a line in wp-config.php to increase memory to 64 MB
• Checked my WordPress core files (they were up to date)
• Renamed my plugin and themes directories
• Added a DEBUG MODE statement to wp-config.php

At this point, I was able to see some error messages:

Warning: Cannot modify header information – headers already sent by (output started at /home3/andysylv/public_html/wp-config.php:1) in /home3/andysylv/public_html/wp-login.php on line 462

Warning: Cannot modify header information – headers already sent by (output started at /home3/andysylv/public_html/wp-config.php:1) in /home3/andysylv/public_html/wp-login.php on line 478

Warning: Cannot modify header information – headers already sent by (output started at /home3/andysylv/public_html/wp-config.php:1) in /home3/andysylv/public_html/wp-includes/functions.php on line 6828

However, I still could not log in. I spent some time with Bluehost tech support (both chat and phone support), but did not get much help there. As part of their troubleshooting, I turned SSL support on and off (that had a side effect, will discuss later in this post).

Per this post, I removed the closing PHP tag in wp-config.php.

Per this Stack Exchange post (https://stackoverflow.com/questions/42832860/wordpress-utf-8-without-bom), I resaved wp-config.php and wp-login.php to force UTF-8 encoding without BOM. At that point, the error messages were removed, and the login fields remained, but now I was getting messages related to cookies, and saw some redirection in the URL field in the browser. Read through these articles (https://wordpress.stackexchange.com/questions/215032/cookie-domain-setting-confusion, https://wp-staging.com/how-to-fix-the-error-cookies-are-blocked-or-not-supported-by-your-browser/).

I tried to get Bluehost to turn off SSL for my site (although I already de-selected that). They said they turned it off, but subdomains on andysylvester.com still showed up as secure sites, which messed up some of the display (had links to http which were interpreted as “unsecure” – arrrgh!).

I decided to change the URL of my site in the MySQL database from http://andysylvester.com to https://andysylvester.com.

Also, per this post (https://themeisle.com/blog/add-wordpress-https/) , I added the following line to wp-config.php:

define(‘FORCE_SSL_ADMIN’, true);

After I did this, I was FINALLY able to log into my site – whew!

I do not know how this whole thing started, but I am glad to get back to posting!

How to organize information for use

Ken Smith had two recent posts (“For the team” and “Or maybe not“), discussing the ideas of forming a community, collecting information on a topic, processing information as a group, and providing a summary of that information. I think the collection and processing tasks are attainable and occur on a regular basis. The task not being performed well is the curate/share/maintain steps.

As an example, I started collecting information on the topic of personal book lists in a Github repo. I provided a chronology of recent posts on the subject, collected tools and techniques, and also created some tools of my own. The recent Github repo for Drummer stuff would be another example, as well as other “awesome-fill-in-the-blank” sites on Github. A mega example of this is a tech guide put together by Stephen Downes on resources for creating an online community, class or conference. 

Perhaps the main thing is to start a “beach-head” of some sort where people can collaborate, then publicize and support that beach-head. Thoughts, anyone?

(cross-posted from my Drummer blog)

Today is the one year anniversary of the January 6th insurrection at the US Capitol. I created a liveblog of the events of that day, click this link to see the OPML in Little Outliner. The original post is here, and another rendering of the liveblog showing all entries expanded is here.

Getting started again with rssCloud

Andrew Shell contacted me recently to mention that he has his rssCloud server online – nice! rssCloud uses the <cloud> element in RSS 2.0 to connect a loosely-coupled Twitter-like network of people and 140-character status messages. I went through my archives (looks like I last looked at this in 2015 and 2016) and notes, and found that I had a copy of a script (Github Gist link) Dave Winer wrote as a test app in 2015. I uploaded it to my current server, modified it to point to my current domain, and found that it was able to register with the rssCloud server and get updates on the feed listed in the test app – excellent! I plan to review the rssCloud Walkthrough document to guide my next steps.

Resources:

https://gist.github.com/scripting/dbb07695736de85b3882 – Dave Winer test app for Andrew Shell server

https://blog.andrewshell.org/2020-02/18/updating-rsscloud-server/ – Update on rssCloud server

http://rpc.rsscloud.io:5337/ – Running instance of server

https://blog.josephscott.org/2009/09/07/rsscloud-for-wordpress/ – Info on rssCloud plugin for WordPress

https://wordpress.org/plugins/rsscloud/ – Current plugin page for rssCloud WordPress plugin

https://plugins.svn.wordpress.org/rsscloud/ – Current SVN repo for rssCloud WordPress plugin

http://home.rsscloud.co/ – New home page for rssCloud info

http://walkthrough.rsscloud.co/ – Implementers Guide to rssCloud

https://github.com/rsscloud/rsscloud-server – rssCloud server source code by Andrew Shell

https://andysylvester.com/2015/11/22/learning-something-new-can-be-hard/ – My 2015 post on rssCloud work

http://notes.andysylvester.com/2016/08/07/exploringRsscloudAndPubsubhubbub.html – My 2016 post on rssCloud work

I am a WordPress user, but not as prolific as I would like from this diagram. I do think the “static gen basin” is totally accurate! Had a hard time stopping laughing…

Source: Rakhim.org

nodeStorage: Feedback on new migration changes (affects 1999.io)

I wrote a post recently about a 1999.io server that I set up for a friend. That post documented a problem that happened when I initially set up the server. A few weeks ago, he wrote me to say that he was getting 502 Bad Gateway when he tried to access his site. After some review, it appeared that the 502 Bad Gateway was due to the fact that the server process had stopped for some reason. I went to restart the server, but got an error immediately:

Cannot find module ./main.js

After some more review, I saw that the nodeStorage repo had been updated in October (https://github.com/scripting/nodeStorage). This is the backend of the 1999.io blogging tool. I remembered that in the past, some of these apps would try to update themselves periodically. From looking at the installation directory, this appeared to be the case. The update was to create a NPM package for nodeStorage (it appears this was to help with making future updates easier). I agree that the manner of updating 1999.io in the past has been somewhat awkward, so I hope this will improve the environment for future updates. However, I wanted to get my friend’s server working again, so I started trying to figure out how to fix things as quickly as I could.

The description for the new update process (https://github.com/scripting/nodeStorage/blob/master/package.md) had the following steps:

  1. Download the package.
  2. The code you need is in the example folder. Copy the two files into your app’s folder.
  3. Edit app.js to the name of your app, and update package.json accordingly.
  4. You can delete the other files.
  5. At the command line, enter npm install.
  6. You still have to have a config.json file as before.

I began to work through these steps, here are my notes:

  1. Download the package.

Since I did not think my install was ready to do the ‘npm update’ type solution (future state), I instead downloaded a Zip file of the current nodeStorage repo (https://github.com/scripting/nodeStorage/archive/master.zip) and unzipped the file.

2. The code you need is in the example folder. Copy the two files into your app’s folder.

I reviewed the unzipped file and saw that there was a folder called “example” and it contained two files (app.js and package.json). I copied these files into my existing nodeStorage directory.

3. Edit app.js to the name of your app, and update package.json accordingly.

I decided that I wanted to keep the storage.js app the same, so I changed the app name in app.js from “nodestorage” to “storage” (which corresponds to storage.js). I decided to keep the current package.json for now.

4. You can delete the other files.

I decided to delete the npm_modules directory within the nodeStorage directory (due to step 5)

5. At the command line, enter npm install.

I executed this command from the nodeStorage directory

6. You still have to have a config.json file as before.

I had a config.json file, so left it as is.

I then started the server using the command “node storage.js”, and the server appeared to come up correctly. I then stopped it and started again using the command “forever start -a storage.js”

I then repeated this operation with a second 1999.io install, except that I left app.js and package.json from the example folder as-is. I then used the command “node app.js” and saw the server come up correctly. Finally, I stopped it and started again using the command “forever start -a app.js”, with everything working normally.

Conclusions/suggestions:

  1. Recommend changing step 1 to say “getting the nodeStorage repo Zip file” rather than “Download the package”, as this gives the impression that you are supposed to issue some NPM command right at the start.
  2. Be more specific about what files could/should be deleted in step 4. I deleted the node_modules directory, but perhaps other files could be deleted as well.

1999.io: How to make sure posts get into the RSS feed

Recently, I helped a friend (Ron Chester) set up a 1999.io instance for a blog on ham radio. In his initial posts, however, he did not see the posts appear in the RSS feed for the site. I did some debugging (will go through that in a post some other time) and found that the problem was due to a setting within 1999.io. There was a reference on the 1999 blog for Facebook Instant Article support, saying that the checkbox needed to be checked for IA support.

In the default install, this box was checked. As a result, if a user posted something without a title, it would not appear in the RSS feed (the 1999 blog post linked above states this). To make sure that all posts (with or without titles) appear in the RSS feed, this box should be unchecked, as shown in the above picture.