Home / GitHub Page

Questioning about Joplin sent

I know that. I wanted to know, whether the OP recognized that domain…

I don’t know if that explain all. But yes I have tested that in a note. And when I click inside I have : tcp://151.101.120.133 : 443 (https)

Click inside what? Do you mean selecting a note or clicking on something within a note?

Just selecting the note. (When I click on it)

I didn’t activate sync yet. To start it was targeted to Dropbox default. And now file system.

It’s possible to block all of this exit demand, or make a request to the user before an IP call ?

No.

But you haven’t answered a previous question. Does your note include an external image? If you reference an external image, of course it is going to download that image every time you open that note.

If by “tested that” you mean you used my example link, that IP would make sense as it is Fastly which is a content delivery network (CDN) used by GitHub. One use is for serving up GitHub “raw” files (as in raw.githubusercontent.com) and that is where the Joplin PNG file is hosted.

You previously mentioned Evernote Clipper as the original source of the notes. I have not used Evernote for some time now so I do not know how it handles embedded content in a clipped note, whether that content is “simplified” or how that Evernote Clipper created code fares when subsequently imported into Joplin. Depending how Evernote clipped a page I suppose you could possibly have links in the note that call out for pictures, stylesheets, webfonts, video, anything.

It seems you may have sent @laurent examples of two of the suspect notes. However one thing to try is to view the notes in the markdown editor and look for any content that is accessed directly using http and / or https rather than using the Joplin (:/a6c18dab21de47829f6883aade793872) format. If there are any then that is likely what is causing this traffic rather than anything in Joplin itself, just as @tessus says (post above this).

Yes I looked at the note and it’s indeed simply an image so it’s normal that it’s making a request.

Notes clipped by Joplin normally would import all the resources locally so there won’t be external requests, but it seems notes clipped by Evernote can have such external resources.

3 Likes

Yes I confirm, external calls are made by notes who they have links with “.png” at the end.
I suppose there is no solution for that than delete notes (or part), one by one.
However in web mail it’s possible to block images before to have the user permission. Or an option to disable markdown links is it feasible ?

Well, if you have notes that request external images there’s nothing that Joplin can do. That’s the same as complaining that a web browser accesses the Internet.

You don’t have to delete the notes, but only comment the markdown for the images.

Maybe someone wrote a script that downloads all external images and attaches those instead. You can check out the #apps category.

Another way is to use a personal FW. e.g. I use LittleSnitch.

1 Like

Yes FW it is a good thing, but that need to make a rule for every note with an image link. I suppose it’s the same behavior in mobile apps. It’s more difficult to have a control in a smart phone. And search for this kind of link is not really efficient to put comment.

However Joplin is not a browser, when I put the same link in others programs, they can make hypertext link but they don’t try to connect to the target.

#Apps I checked but I didn’t find the script. Maybe is something in Electron with srcURL, native-image or from Mozilla with Data URLs, or not…

When an image is stored by content delivery network (CDN), Wikipedia give some information about security and privacy annoyance.

If an image link looks like this

[image](https://raw.githubusercontent.com/laurent22/joplin/master/Assets/LinuxIcons/256x256.png)

It will behave just like any “normal” hyperlink in that nothing gets automatically fetched.

However if Evernote clipper has prefixed the link with an exclamation mark

![image](https://raw.githubusercontent.com/laurent22/joplin/master/Assets/LinuxIcons/256x256.png)

the renderer is being instructed to fetch the image and display it. The Joplin clipper creates a local copy of images so that no further calls to the Internet are required. Joplin cannot be blamed for the failings of the Evernote clipper.

If these direct links to images in your imported notes cause you concern, edit them out.

  1. This is not relevant to your issue
  2. CDNs are widely used so that high volume sites can operate. If you feel that GitHub (or any other company) should not use a CDN you could raise the issue with them

If you are concerned about your system making calls to the Internet just to pull down an image I would seriously suggest that you re-evaluate your decision to use Windows 10. If left unchecked Win10 can tell Microsoft about pretty much everything you do on your computer.

4 Likes

Yes my links are prefixed with an exclamation mark.
I tested again in LibreOffice and in other program that, and the instruction is not followed.

I don’t want to blame Joplin, there have already been a great work inside.
I’m just a little surprised by the behavior link.
Yes I will probably edit notes out, just I keep unsure adding future notes with links or making copy/paste.

Yes I know but for now I prefer to use W10 because I can understand it better than Linux distribution.

LibreOffice is not a markdown editor/previewer. Somehow I have the feeling you do not know what markdown is.

That’s because you do not understand what Markdown is and how it is rendered. Download Typora for Windows ans enter a link with a !.

Once again, don’t use image links that tell Joplin to download the image.

1 Like

I have tested the Joplin web clipper in a webpage (clip complete page), and it looks working as Evernote web clipper for images.

Joplin clipper can’t extract resources directly from the browser ?

I have just clipped the page you linked to twice.

When using “Clip complete page” images appear in Joplin in the format:

![Sinclair delays interview containing Fauci Covid-19 conspiracy theory](:/7f622814299f45538b6be51f8b37d431)
    
...
    
![Israelis continue protests against PM's handling of pandemic](:/86588b8fea1a4c4fbbf1ba6e2859d3b1)
    
...
    
![Amid virus, uncertainty, parents decide how to school kids](:/870c5242ea9c42868266612bd267b9fa)
    
...
    
![German cruise ship sets sail, hopes short trip thwarts virus](:/279149dcc3de450ba4d5e16ed900e3f7)

These links show that Joplin has imported the images into the resources folder and has set up internal links to display them.

Only if I use “Clip complete page(HTML)(Beta)” are the image links actually URLs to the JPG files on the Yahoo site.

That occurs because that is what the clipper has been told to do.

What option did you choose?

I used “Clip complete page” in firefox, I think the clipper take firstly the text and links in the web page, and secondly Joplin app go to take images from web and transform it in local resources.
You probably can check it, if you download the page in the browser, cut internet connection and finally ask the web page to the clipper :

  • Sinclair delays interview containing Fauci Covid-19 conspiracy theory

- ![Sinclair delays interview containing Fauci Covid-19 conspiracy theory](https://s.yimg.com/uu/api/res/1.2/aAvkPppuD3orpS1FTZlubg--~B/Zmk9c3RyaW07aD0xMjM7cT04MDt3PTIyMDthcHBpZD15dGFjaHlvbg--/https://media.zenfs.com/en-GB/the_guardian_765/32faf4244f8c963ca53f03880f53ce58.cf.webp)

Hi Dino,
Did you find a better way to clip webpages ?
For me Joplin Web Clipper was supposed to take the image from the browser. Why Joplin desktop is pushed to do it ?
I don’t like to use the clipper in that way…

Jerry = Dino?

mal

@Jerry No, you can do PDF (print option) but is not the same result.
@malagrabla Thanks to try to increase the topic attention.
@tessus It’s difficult to correct ? :wrench: :nut_and_bolt: :gear: :clamp: