Tech: Website certificate expired, etc...

Subject: RE: Tech: Website certificate expired, etc...
From: DaveRo
Date: 12 Oct 23 - 09:17 AM

Jon Freeman wrote: ...I type anything more than a couple of sentences into libreOffice firsrt so I’ll do it as a LibreOffice macro. I’ll have it scan the whole document for URLs and convert each occurrence to a mudcat link

What about link text? You could type it in like wiki:
[http://example.com|this is an example].

When I wrote my Simple Linkifier I tried to make it automatically fill in the link text from the <title> tag of the page linked to, to save having to copy it between tabs. But I couldn't: the browser Same Origin Policy I referred to doesn't let one page pull data from another, unrelated, page. For obvious reasons. But you could do it in a separate program.

Subject: RE: Website certificate expired...
From: DaveRo
Date: 10 Oct 23 - 12:16 PM

Jon Freeman wrote: By my current thinking, without the protocol, the link given is relative...

A link without a protocol is a Protocol-relative URL. A link without a server - first 2 slashes - is a relative link. A link starting with one slash is relative to the document base, one that has no slashes is relative to the page containing the link.

There's no good reason to generate relative links within mudcat.

I don't know why there are two subdomains either - www and none. A redirect from www.mudcat.org to mudcat.org or vice-versa would be simpler. Everything would then have the same origin: Same Origin Policy

Subject: RE: Website certificate expired...
From: DaveRo
Date: 07 Oct 23 - 03:01 AM

https still works:

https://www.mudcat.org

The "blue clicky" tool is still broken if you try to link to Mudcat itself, or if the URL is very long. Bookmark my Simple Linkifier

Subject: Website certificate expired...
From: Steve Shaw
Date: 03 Oct 23 - 02:16 PM

Er, are we still OK?

Subject: RE: Website certificate expired...
From: Stilly River Sage
Date: 07 Oct 23 - 10:39 AM

There is no www in Mudcat dot org, and if you enter it in any "blue clicky" it will break the link. That is typically what I have to fix when the links back to other Mudcat pages are shared in posts but don't work.

http://mudcat.org (My system adds the "s" and it goes through.)

Subject: RE: Website certificate expired...
From: Stilly River Sage
Date: 03 Oct 23 - 06:34 PM

Max was notified this morning.

Subject: RE: Website certificate expired...
From: Joe_F
Date: 06 Oct 23 - 05:41 PM

In my case, oddly, I had to remove the "www" from the URL to get in.

Subject: RE: Website certificate expired...
From: Dave the Gnome
Date: 03 Oct 23 - 03:19 PM

Aye, same here.

Subject: RE: Tech: Website certificate expired, etc...
From: Jon Freeman
Date: 13 Oct 23 - 03:00 AM

and after another very long sleep...
yes Dave. I will have link text.

I've been caught by same origin a few times with my plays with Javascript

Subject: RE: Website certificate expired...
From: Jon Freeman
Date: 12 Oct 23 - 06:07 AM

The link generator was never intended to to more than giving me something to think about but I’ve decided I will write something that makes links along those lines for my own use that would suit the way I post to Mudcat these days. That is I type anything more than a couple of sentences into libreOffice firsrt so I’ll do it as a LibreOffice macro. I’ll have it scan the whole document for URLs and convert each occurrence to a mudcat link.

I did have a quick look at this before drifting off into a very long sleep yeaterday. LibreOffice offers JavaScript as one choice of scripting language for its maros so I tried the code I already had for a starter but no go. I don’t think their JS supports URL (not essential but a nice convenience). I’m not certain but I think they are using Java Rhino. I used this years ago when or home automation used my Java server as the means of dynamically providing code for events. The only other time I wanted to do something like that I used Groovy but I can’t remember what remember why I dropped Java Rhino in favour of this.

Oh well, I’ll get something written and it doesn’t matter to me whether it winds up in their version of basic (never used by me but the best documented for LO), python or JS.

Subject: RE: Website certificate expired...
From: Jon Freeman
Date: 10 Oct 23 - 04:29 PM

Thanks Dave. To try to clariy I wasn’t thinking in terms of generating a relative link but questioning the correct interpretation for an absolute link from user input such mudcat.org/... and www.mudat.org/… Whatever, a server rewrite rule would be a good idea. Far from all links here are created using a generator plus there must be a lot of threads pointing to dead subdomains around.

I think when I started here, probably late 1999, Mudat was using www as its main website domain. Other sub domains came in as the result of crashes and us learning how to access individual servers.   We could for example find www inaccessible but loki up and running. I think using “backdoors” (if any exist) is frowned upon these days because it can cause complications with links.

Subject: RE: Website certificate expired...
From: Jon Freeman
Date: 10 Oct 23 - 04:43 AM

While wondering what to do to occupy some time, I decided to think about what I might do with a link maker. Here’s one attempt at a JavaScript possibility:

  function rep(url, remove=false){
          prefix = "https://mudcat.org";
          u = new URL(url, prefix);
          pathpart = u.pathname;
          if (u.hostname.search(/(mudcat.org)/i)>=0) {
              mc = u.pathname.search("mudcat.org");
              if (mc >= 0 && remove){
                  pathpart = pathpart.substr(mc+10);
              }
              hostpart = prefix;
          }else{
              hostpart = u.protocol  + "//" + u.hostname;
          }
          return '<a href="' + hostpart + pathpart + u.search + '">' + 
                  ' + hostpart + pathpart + u.search + '</a>(input url: ' + url + ')';
  }
  
  and some test urls.
  <a  href="https://www.google.co.uk/search?q=mudcat">https://www.google.co.uk/search?q=mudcat</a>(input url: https://www.google.co.uk/search?q=mudcat)
<a  href="https://mudcat.org/thread.cfm?threadid=172747">https://mudcat.org/thread.cfm?threadid=172747</a>(input url: /thread.cfm?threadid=172747)
<a  href="https://mudcat.org/thread.cfm?threadid=172747">https://mudcat.org/thread.cfm?threadid=172747</a>(input url: http://mudcat.org/thread.cfm?threadid=172747)
<a  href="https://mudcat.org/thread.cfm?threadid=172747">https://mudcat.org/thread.cfm?threadid=172747</a>(input url: http://www.mudcat.org/thread.cfm?threadid=172747)
<a  href="https://mudcat.org/thread.cfm?threadid=172747">https://mudcat.org/thread.cfm?threadid=172747</a>(input url: https://loki.mudcat.org/thread.cfm?threadid=172747)
<a  href="https://mudcat.org/thread.cfm?threadid=172747">https://mudcat.org/thread.cfm?threadid=172747</a>(input url: https://mudcat.org/thread.cfm?threadid=172747)
<a  href="https://mudcat.org/thread.cfm?threadid=172747">https://mudcat.org/thread.cfm?threadid=172747</a>(input url: https://www.mudcat.org/thread.cfm?threadid=172747)

  (remove=false)
  <a  href="https://mudcat.org/mudcat.org/thread.cfm?threadid=172747">https://mudcat.org/mudcat.org/thread.cfm?threadid=172747</a>(input url: mudcat.org/thread.cfm?threadid=172747)
<a  href="https://mudcat.org/www.mudcat.org/thread.cfm?threadid=172747">https://mudcat.org/www.mudcat.org/thread.cfm?threadid=172747</a>(input url: www.mudcat.org/thread.cfm?threadid=172747)

  (remove=true)
  <a  href="https://mudcat.org/thread.cfm?threadid=172747">https://mudcat.org/thread.cfm?threadid=172747</a>(input url: mudcat.org/thread.cfm?threadid=172747)
<a  href="https://mudcat.org/thread.cfm?threadid=172747">https://mudcat.org/thread.cfm?threadid=172747</a>(input url: www.mudcat.org/thread.cfm?threadid=172747)

As I guess it represents "mudcat best practises" these days, links to Mudcat pages are returned as starting "https://mudcat.org" regardless of whether http or https or subdomains (eg. www.) are given in input.

The last 4 examples show bit of a puzzle to me. That is when mudcat.org is supplied without http(s)://. By my current thinking, without the protocol, the link given is relative so, eg. www.mudcat.org is part of the path and a result like https://mudcat.org/www.mudcat.org/thread.cfm?threadid=172747 is correct. On the other hand, I think this behaviour has been considered a bug. I’ve left the question open in this example. Calling the function with remove set to true removes the apparent duplication.

Subject: RE: Website certificate expired...
From: Jon Freeman
Date: 07 Oct 23 - 06:20 AM

I see the certificate has been renewed. It is for both mudcat.org and www.mudcat.org.

Subject: RE: Website certificate expired...
From: Jon Freeman
Date: 03 Oct 23 - 02:34 PM

I have let the expired certificate pass but it is an error that needs fixing. I've allowed it as I think Max has simply forgotten to renew the certificate (or something similar) but it's up to others whether they want to take the chance.

Subject: RE: Tech: Website certificate expired, etc...
From: GUEST,.gargoyle
Date: 12 Oct 23 - 08:33 PM

Very, VERY, Strange Things have been "reconfiguring" on the entire internet ... over the past four weeks.

It reminds me of the original, late 90's internet browser wars.
HTML needed to be configured in three formats ... depending on the Browser the client might read.

Sincerely,
Gargoyle


In late 2023 it appears "FREE" serices ... are direct linking to their corporate fee services.

Frustrated ... with gmail new approach ... this last week ... I fell into ... a world I never imagined ...Within seconds...data was phoned in, posted to a spread sheet ... verified and posted.

Subject: RE: Website certificate expired...
From: GUEST,keberoxu
Date: 03 Oct 23 - 06:10 PM

The Mudcat just had a birthday (Sunday)
so that might have something to do with the interval of renewal.