My Thumbnail Topic Image Link Collection
-
Issues that I am aware of that we face with internal image hosting:
- How do we handle uploading to multiple nodes? This is a NodeBB hurtle, is anyone aware of how this would be handled practically?
- Is there a way to make NodeBB upload to a media node (without using NFS to mount a directory of course) so that we could have images "hotlinked" from a separate service like media.mangolassi.it? Same as local uploading but not assuming a local machine?
-
Okay, main icon has been moved to local hosting. Image uploads seem to be working. That took a few tries, not sure why it was failing, but seems to be fine now.
-
@scottalanmiller said:
Just checked, all other services have been removed from NodeBB, even ones that are not "official." There used to be plugins for several services, like S3. But when I search, it looks like Imgur is the only thing that is supported. @JaredBusch do you see any other options? Imgur is a plugin, but it is an official one from the core team like the markdown.
You did not check real hard then as NodeBB does not remove anything. More specifically, you most likely went to the control panel and did not see it in the plugin list. This does not mean it was removed from NodeBB. This means the developer has not updated the compatibility string in their plugin to note that it is compatible with the version of NodeBB you are running.
-
@scottalanmiller said:
When Imgur was set up, we went through their policies. It may be that the ToS is different depending on the type of account that you have set up. They have CDN as a specific service offering. But we can look at moving to S3 if people are really disliking Imgur.
I have no idea what the Imgur ToS says in detail, I never looked. I said as much earlier.
-
@JaredBusch said:
@scottalanmiller said:
Just checked, all other services have been removed from NodeBB, even ones that are not "official." There used to be plugins for several services, like S3. But when I search, it looks like Imgur is the only thing that is supported. @JaredBusch do you see any other options? Imgur is a plugin, but it is an official one from the core team like the markdown.
You did not check real hard then as NodeBB does not remove anything. More specifically, you most likely went to the control panel and did not see it in the plugin list. This does not mean it was removed from NodeBB. This means the developer has not updated the compatibility string in their plugin to note that it is compatible with the version of NodeBB you are running.
Do you know of one that works with NodeBB currently then? I've tried going outside of the plugin system and grabbing the Rackspace Cloud Files one via GIT and it would not even show up to activate.
Where are you seeing them?
-
@scottalanmiller said:
@anonymous said:
Also, don't use Imgur to host image libraries you link to from elsewhere, content for your website, advertising, avatars, or anything else that turns us into your content delivery network.
It's specifically what the NodeBB platform does. It's built in with one of the original plugins.
That NodeBB wrote a plugin or core functionality for it has nothing to do with the legality of it.
-
@JaredBusch said:
@scottalanmiller said:
When Imgur was set up, we went through their policies. It may be that the ToS is different depending on the type of account that you have set up. They have CDN as a specific service offering. But we can look at moving to S3 if people are really disliking Imgur.
I have no idea what the Imgur ToS says in detail, I never looked. I said as much earlier.
They've been contacted directly to query about the use case to get their opinion. In the meantime we went to local uploads and are trying to get Cloud Files working, but that looks to be unlikely any time soon as the plugin looks to be deprecated.
-
@JaredBusch said:
That NodeBB wrote a plugin or core functionality for it has nothing to do with the legality of it.
I understand that. It is, however, suggestive and that Imgur has built, supplied and documented how to specifically do this as well is odd if they do not intend it to be used.
-
At least when uploads to Imgur are done through the site it is done openly to Imgur. They are totally aware that it is being used. ML is a registered application account with Imgur. People uploading directly to Imgur would be different. Might be the same rules, but it is two different systems. There has been nothing hidden from Imgur, this was all set up through them. It's not like an anonymous posting system that is just leveraging them without their knowledge.
All uploads and views of the content here are registered with the site.
-
@scottalanmiller said:
@JaredBusch said:
http://definitions.uslegal.com/t/theft-of-services/
https://en.wikipedia.org/wiki/Theft_of_servicesDid you read those links? Not remotely related.
Of course I did. And they do apply. I pay my colo for a set amount of bandwidth. When you hot link to an image I have uploaded locally, from a site with millions of views in a day, that is millions of loads of my image and jacking my costs up because you just caused me to blow past my bandwidth allocation. You are stealing my service.
-
@JaredBusch said:
@scottalanmiller said:
@JaredBusch said:
http://definitions.uslegal.com/t/theft-of-services/
https://en.wikipedia.org/wiki/Theft_of_servicesDid you read those links? Not remotely related.
Of course I did. And they do apply. I pay my colo for a set amount of bandwidth. When you hot link to an image I have uploaded locally, from a site with millions of views in a day, that is millions of loads of my image and jacking my costs up because you just caused me to blow past my bandwidth allocation. You are stealing my service.
No, I'm not pulling anything from your site, I'm referencing it. Very different things. Same as a link. You linked to uslegal.com. How is that different? If my browser or cache pulls links automatically it is identical in that you provided a link and it was pulled. Linking is not hijacking and uses no bandwidth.
-
If ML itself was pulling data from those sites, that would be totally different. It is not. Nor is ML being provided through those sites. It is just providing a document with links. It's how the web was designed to be used.
Like I said, create a thread to discuss how links work and why some people feel hot linking (which is just another term for linking) is an issue and why some do not.
As someone who has had for years systems that pull all anchor links automatically (to a depth, of course) I can assure you, even the links you used are a form of hotlinking.
-
@scottalanmiller said:
@JaredBusch said:
@scottalanmiller said:
@JaredBusch said:
http://definitions.uslegal.com/t/theft-of-services/
https://en.wikipedia.org/wiki/Theft_of_servicesDid you read those links? Not remotely related.
Of course I did. And they do apply. I pay my colo for a set amount of bandwidth. When you hot link to an image I have uploaded locally, from a site with millions of views in a day, that is millions of loads of my image and jacking my costs up because you just caused me to blow past my bandwidth allocation. You are stealing my service.
No, I'm not pulling anything from your site, I'm referencing it. Very different things. Same as a link. You linked to uslegal.com. How is that different? If my browser or cache pulls links automatically it is identical in that you provided a link and it was pulled. Linking is not hijacking and uses no bandwidth.
You are trying to rationalize theft. Just stop.
Just because you are only providing the link and not actually using the bandwidth does not take you out of the loop. Of course the actual bandwidth is getting used by all the people that view ML and not by ML itself. But ML is the legal entity that embedded the link to enable the theft of bandwidth.
A link to my site is much different because it takes the user entirely to my site. It makes it known that it is a different site (well unless I was maliciously trying to clone ML or something) and as the owner of the domain I can know that the sudden spike of a million users are real people on my site viewing the entirety of my content and not some asshat taking advantage of the fact that I did not disable hotlinks.
-
@JaredBusch said:
You are trying to rationalize theft. Just stop.
There is no question that it is not theft. You are trying to rationalize calling it theft. It isn't. Not in the least.
Not generally accepted? Okay, I'll take that. And I'm working on changing how the site works. But calling it theft is beyond ridiculous. You can't look at an HTML page with links and call that theft. Referencing is not theft.
You linked to sites, what makes that different? Nothing, of course.
-
@JaredBusch said:
Just because you are only providing the link and not actually using the bandwidth does not take you out of the loop. Of course the actual bandwidth is getting used by all the people that view ML and not by ML itself. But ML is the legal entity that embedded the link to enable the theft of bandwidth.
Using the term "embed" is misleading. It is a reference. Yes the browser will embed it into the final, displayed product, but the link itself is just a reference - the same as any link like the references.
If you link to an image in text, you could make the same claims. But would you?
-
@scottalanmiller said:
@JaredBusch said:
Just because you are only providing the link and not actually using the bandwidth does not take you out of the loop. Of course the actual bandwidth is getting used by all the people that view ML and not by ML itself. But ML is the legal entity that embedded the link to enable the theft of bandwidth.
Using the term "embed" is misleading. It is a reference. Yes the browser will embed it into the final, displayed product, but the link itself is just a reference - the same as any link like the references.
If you link to an image in text, you could make the same claims. But would you?
Not, it is not the same. There is no bandwidth being used on the other site by just providing a link.
-
@JaredBusch said:
A link to my site is much different because it takes the user entirely to my site. It makes it known that it is a different site ...
That's by convention only. Both take you to the full referenced resource. The idea of going to a "whole site" doesn't actually exist. And even a reference link only goes to a page, to a text document. It's still a singular resource.
From a technical perspective, there is no difference. it's only a human perception thing that makes one seem different from the other.
In both cases you load the "entire resource". In both cases it is "just a file." In both cases it is "publicly serviced willing by the provider." In both cases it is exposed to the end user if they choose to look and hidden if they don't.
-
@JaredBusch said:
Not, it is not the same. There is no bandwidth being used on the other site by just providing a link.
I explained earlier how that isn't true. By convention, today most people do not automatically download links, but some do and always have. By convention, most people automatically download images but some do not and always have not.
Both pull automatically at the discretion of the end user, not determined by the site itself.
-
You are using convention of client default settings, that are only recent conventions for use, to determine what is theft and what is not. That's not a good way to determine theft. You can't apply it like that.
-
@scottalanmiller said:
@JaredBusch said:
Not, it is not the same. There is no bandwidth being used on the other site by just providing a link.
I explained earlier how that isn't true. By convention, today most people do not automatically download links, but some do and always have. By convention, most people automatically download images but some do not and always have not.
Both pull automatically at the discretion of the end user, not determined by the site itself.
What an absolute bunch of shit.
How in the hell can you rationalize that most people do not download links when the entire point of this is the fact that you are providing a service to let people see images without paying to host the images yourself.
