Over the last decade I've built a number of different digital asset managers (mostly media files) that met the needs of my companies at the time. It is an area I enjoy working in. A month ago, when asked what was next for me, I jokingly said I would build another DAM.<p>Then on Saturday, two weeks ago, I learnt that Imgur was going to delete all anonymous & NSFW files on the 15th of May. It was pointed out that this would mean broken links in communities that had relied on Imgur. By the Sunday I had decided that I would build another DAM, initially with the intent of avoiding Imgur link rot.<p>It was challenging to find time to spend on this, the project was put together over about 8 evenings. It still has rough patches, this is an early MVP (a Michael Seibel "brick").<p>I have many ideas of where to take this project, but for now it only does one thing: backup Imgur files and produce new links that are easy to swap out for old soon-to-be-deleted Imgur links.
> Well done, you’ve defeated link rot<p>Until this site succumbs to the challenges of being a free image host?<p>FWIW ArchiveTeam is grabbing as much as possible so it could eventually show up in the wayback machine. Though what they have now, 47TB, is a drop to the petabytes imgur probably has now. There was an estimated 376 TiB in 2015.<p>- <a href="https://wiki.archiveteam.org/index.php/Imgur" rel="nofollow">https://wiki.archiveteam.org/index.php/Imgur</a><p>- <a href="https://tracker.archiveteam.org/imgur/" rel="nofollow">https://tracker.archiveteam.org/imgur/</a>
Could it be configured so if I go to, for example, <a href="https://i.haasie.com/something.png" rel="nofollow">https://i.haasie.com/something.png</a> and you haven't already backed up from imgur, you back it up when that URL is hit?
Interesting service, but an earnest question about your Terms and Conditions[0].<p>How closely did you read this? Since I see you used an TOC generator (which was nice for me learn about), I'm just curious how closely you read some of the conditions.<p>I just may be reading the Prohibited[1] section too closely as well as the User Generated Content [2] section, but it seems like your stated goal about protecting the NSFW content from Imgur, as I guess such content is potentially obscene, lewd, lascivious, filthy, violent, harassing, libellous, slanderous, or otherwise objectionable (as determined by [you and your team]).<p>I'm not here to trash your service or trying to "gotcha", just since you're responding in the thread, I'm only curious how closely you've read the TOC and if you stand by all the elements of it.<p>0 - <a href="https://haasie.com/terms_and_conditions" rel="nofollow">https://haasie.com/terms_and_conditions</a><p>1 - <a href="https://haasie.com/terms_and_conditions#prohibited" rel="nofollow">https://haasie.com/terms_and_conditions#prohibited</a><p>2- <a href="https://haasie.com/terms_and_conditions#ugc" rel="nofollow">https://haasie.com/terms_and_conditions#ugc</a><p>edit: fixed link formatting
First off: great work!<p>The age old question is how you afford to run this, and how if you get popular do you not become an imgur yourself and have their same pressures to delete, moderate and make money.
Could you export a list of all imgur URLs saved, so that ArchiveTeam can help save those to archive.org too?<p><a href="https://wiki.archiveteam.org/index.php/Imgur" rel="nofollow">https://wiki.archiveteam.org/index.php/Imgur</a>
> Well done, you’ve defeated link rot!<p>Except you haven't, because anyone that wants to view these images where they are linked needs to know to change the URL to Haasie instead of Imgur. So all the links are still broken.<p>And on top of that, even if the users know about Haasie, it only works if someone went to this website before the original image was deleted, and enter the URL on a form. Non-popular images will be long gone before people do this manually. Perhaps accepting any URL and transparently fetching it from Imgur would improve this a little. From my manual testing, this doesn't seem to be the case, just gives an S3 error from bucket ID haasie01, and nothing shows up there even if I wait and refresh.<p>All of this is assuming this website won't go down or go bad. Probably the Archive Team, and Archive.org can handle this task better.
Interesting work! Thanks for putting this together.<p>One potentially problematic thing about the new Imgur policy is that it will break machine learning projects that depend on Imgur for NSFW data. For example, here's a project relying on Imgur links in order to train a classifier that detects adult content:<p><a href="https://github.com/alex000kim/nsfw_data_scraper">https://github.com/alex000kim/nsfw_data_scraper</a><p><a href="https://github.com/EBazarov/nsfw_data_source_urls">https://github.com/EBazarov/nsfw_data_source_urls</a>
Thanks for all the questions. It has really gotten me to thinking about how to navigate the future of the project.<p>I have a question for all of you. My approach is that we are all understand that there is no such thing as a free service, and that for something like this to be long term sustainable donations (or some source of revenue) is crucial. If people get value from the service then it only makes sense to give back to the service to sustain it.<p>With this in mind, I'm toying with an idea that I would like to get a temperature check from HN on.<p>How would you react if, when donations are too low, Haasie serves a single image that requests donations to sustain the service in place of the original anonymously uploaded content, until the required donation level is reached? Once there are enough donations normal service is resumed.
That makes me think something like this could be a nice phpbb plugin.<p>Let forum admins compile a list of linked imgur posts that should be archived. Storage and bandwidth are cheap compared to what they used to be, so some may be able to self-host images now.