I think there are amazon affiliate banner image links from the 90s that still display images, such has the 'buy a book' icon. How about google image logo links from 1998. I wonder if those still work.
As far as images that a modern browser can display, JPEG was created in 1992 and GIF in 1987, so there isn't going to be much prior to that. HTTP and Gopher date back to 1991, so the oldest image online that still resides at its original location would be on an FTP server.
Here's one of the oldest I can find, with timestamp 1988-07-06 00:00:
Most of the other old FTP servers I remember (sunsite.unc.edu, ftp.simtel.net, ftp.uu.net) have long since shut down. The ftp server at ftp.sunet.se is one of the oldest still running. Perhaps someone else knows of an older one?
I tried to find a copy of lena.bmp/lena.tga/lena.pcx since those file formats pre-date GIF, but I came up empty.
Lena (or Lenna, spelling varies) is not anime - it's a crop of a Playboy centerfold that has been a standard bitmapped reference image for decades. It's been used in literally hundreds, if not thousands of academic papers on image processing, as well as a variety of benchmarks, etc.: https://en.wikipedia.org/wiki/Lenna
Are we looking at the same image? The one I clicked on in GP's message is an anime-looking person holding a sword. It doesn't look much like the picture on the Lenna wikipedia article you posted.
It's from Urusei Yatsura[1]. It indeed looks like a bunch of Anime stuff. Scans of animation cells and other ephemera. It would have taken awhile to download on a modem.
These conversations happen so often on HN I wonder if there is some sort of working memory limit on the participants that rapidly removes context from the comment they're replying to:
> > > > > Here are some numbers: [6, 2, 34]
> > > > That's neat, one of them is prime
> > > 6 isn't prime, its factors are 2 and 3
> > Are we reading the same comment? There is a prime there
I'd like to propose that we generalize this particular part of the conversation, and coin the term "XY Argument" (unless a name for it already exists?)
> > > > That's neat, one of them is prime
> > > 6 isn't prime, its factors are 2 and 3
I believe it's the equivalent of the [XY Problem], except in the context of an (imagined) disagreement. It happens because the second comment is misinterpreted to be arguing against the first when really it may be in addition. It can be fixed by readers taking a charitable interpretation of the second comment, or by the second commenter's author making it clear that they're not disagreeing, as in,
> > > Yes, and 6 isn't prime, its factors are 2 and 3
I'm pulling out these two comments because it's a much more common pattern than just this comment chain; I think it happens all the time with text in general. Actually, I'm commenting here because it just [happened to me] in a different context. Someone I thought I was arguing with turned out to mostly agree with me — but because we failed to re-affirm the parts of each other's arguments that we agreed with, it seemed like we disagreed more than we did.
These conversations happen so often about anime I wonder if there is some sort of working memory limit on the participants that rapidly removes context from the comment they're replying to:
> > > > > Here are some waifus: [lenna, 6rem, natsuki]
> > > > That's neet, one of them is best girl
> > > lenna isn't from an anime, she exists irl
> > Are we reading the same comment? That's just your opinion, man
> 6rem is best spacing and material design is justice
I've gotten enough upvotes here (4) that I feel I need to ~~ruin~~ explain the joke, lest HN get the idea that this is upvote-worthy in general.
This is low-effort humor. I took the comment above and replaced words with memes from anime, tech, and earlier in this conversation thread. There's no (intentionally) meaningful relation between rem and Rem, or the number 6, I was just pattern matching things that sounded similar. Basically the equivalent of an "I hardly know her" joke. Funny only for the meme value, and not what I come to HN for.
In this particular thread, ^^that^^ is the joke — that we mistake unrelated comments as meaningful replies.
I personally found the metajoke amusing, whether it was intended or not. The comment, low effort as it is is no different from the comment chain it describes. Great stuff.
Netscape Navigator supported PBM (at least in some versions you could even write directly into a frame/window via JS in order to compose an image on the fly).
I completely forgot about that! I very vaguely remember experimenting with it, and I think I never got it to work. I can't find anything on it now, though.
The list of all webservers used to fit on a single page, which might be a good starting point. However, at this point I believe anything still functioning might fit your definition of a "mirror" as few things still speak HTTP/0.9.
The original question was ambiguous; it said "online" but seemed to imply "on the world wide web". I'm not going to judge whether the links I provided count; I leave that decision to the OP.
I would agree that an archive ala archive.org should be disqualified, because
few people would go to that site as the original. To me, Sunet is a little more grey, because while it is an archive/mirror site, many people used it and linked to it as if it were the primary source and not just an archive site. Before the web was indexed and search engines became popular, large mirror sites were the easiest way to access a variety of content, all in one place.
This is sort of a special case. It's a matter of republishing
using the original file metadata and file contents, using the same domain but under a different path.
(Meaning the the current maintainers could fix this by doing a single "ln -s" command, ehrm. @ acc.umu.se.)
If your sole criterion is "file in original directory on hard drive still functioning after X years" then I doubt we're going to go back much further than the late 2000s
Are we restricting format? Because queries like '"index of" pcx 1991' will reveal quite a bit (or TGA, TIFF, and PICT). Whether they've been up since the early 90s is another question.
I created a tool that will show you the images in the apache directory listings. Here's an example
I really appreciate the small details in that page... Such as using Date.getYear() instead of Date.getFullYear() to render the year in the guestbook, so the years are all 19121. Ha!
it's cool. I leave machines completely open for people to create free accounts on.
So it's not something I actually care about, at all. I've actually used that tool to point to various system files doing exactly that strategy.
So feel free to create an account on the pentium 90, find my printers ip address and make it print me a message. I'll find it exciting and adorable.
edit: I looked at the log, looks like you guys are having fun poking around my system. I love the creativity. enjoy. Be sure to reply with any epic exploits. I'm genuinely excited to see what you find.
I just plugged in a webcam. See if you can snap a picture! I'll give 0.1LTC to the first paparazzi.
I guess we need to keep in mind that index of is going to report file creation date, and it's hard to know when the file actually entered the web vs. "Was created, someone left it on disk, and then started a web server."
This is local copy of Anthony's WWW Images 1.8 released April 1996 by Anthony Thyssen. It was meant as a common set of common navigation icons for the web.
k12s.phast.umass.edu has been running continuously since late 1995. K12S was a spin off k12.oit.umass.edu (a free menu-based multi-user public access system for K-12 teachers and students). The S stood for SLiRP -- an open source SLIP/PPP emulator -- so they could offer emulated SLIP connections via the UMass modem-banks at no cost (local commercial SLIP accounts were $5.95/hr IIRC).
I actually owned that book circa 1996 and used Kermit in a production system transferring files via dialup between a manufacturing facility in Mexico and their US-based headquarters. It was incredibly fragile. No idea now why Kermit was chosen, but it was a horrible tool for the job.
I have a page, http://don.dream-in-color.net/books/ which has been continually updated since 1995. It's had some URL changes (it was originally served via FTP under a domain that I lost to missing a registration renewal back in 2004) and back-end technology changes (it was originally manually-updated HTML), but I was able to find a 1997 version of the page in the wayback archive: https://web.archive.org/web/19980121235911/http://quixote.co...
It definitely doesn't fit the bill of this thread (no images, not original links), but it's an interesting site nonetheless. I'm amazed at the number of books you've read over the years. I'm shooting for 3 books this year, and that's a 200% increase over last year!
Do you have good memory for the content of the books you've read? Or do some specific books stand out even over several years?
> I'm shooting for 3 books this year, and that's a 200% increase over last year!
I mean... there's probably no point forcing yourself if you don't want to. People who read a lot of books, myself included, don't do it because we're diligent, we do it because we enjoy it.
There's nothing special about books. You might as well say "I am going to finish 3 video games this year" or "I am going to go fishing three times this year".
If the achieving of goals itself is what brings you enjoyment, then by all means, go for it.
I suppose there's also a case to be made for people who tend to not take enough time off and who need to set themselves goals for leisure time for that reason.
In the past, I have found it very difficult to get through books, but I’ve never had a problem with just putting it down when I’m not interested. I think that’s a big part of my situation: I have had trouble finding things that really hold my interest. And even for some books that I did find interesting, my low reading speed made the process drag on for too long for my interest to hold up. But no, I’m not tortured. I do appreciate your concern, however.
>I'm shooting for 3 books this year, and that's a 200% increase over last year!
I'm gonna countersignal the other guy who replied to you and tell you that this is great news. I have had years in my life when I have read 40+ books, and years where I have read none at all. Setting a goal to carve off time for reading is huge. There is something special about books, something no other form of media does in terms of engaging your brain as an active participant. It's good for you.
I like reading books though I don't read nearly as many as I used to. It's one reason that I'm in the camp of going meh when people go on about needing free WiFi on flights. It's one of the few times I really unplug. OK, I watch videos and play games too but I don't actually miss being offline for 6-12 hours.
Having this list really helps to refresh my memory, but I do have an uncanny knack for remembering things I read. Perhaps the most notable case of this was that while taking a theology class in 1999, I was able to remember a passage in a book I'd read in 1993, but where in the book the passage was and could go to my bookshelves, pull down the volume and find the passage in about ten minutes. It's really handy when it comes to creative writing workshops in that I can often pull up a useful novel or story for a writer from my memory.
As for volume, a lot comes from what I prioritize. I don't watch much television (especially by American standards), and back when commuting was a thing, that was 60–90 minutes a day of reading time.
And, alas, I still fall victim to the buying more books than I can read trap (for a brief while I had the unread books down to a single IKEA Billy bookcase, but I fell off the wagon and now there are piles of books on most flat surfaces stacked with books that I intend to read.
This was posted to the WWW by Tim Berners-Lee in 1992 and I think it's the original URL, though maybe not. Surely there must be an anonymous FTP site that has images that predates HTTP...
"Berners-Lee made a webpage to advertise events taking place at CERN and uploaded Gennaro’s picture of the band. The image was tiny—only around 120 pixels by 50 pixels—because the early web would have struggled with larger images. It was about the size of a stamp".
They must be continuing to update this as in 1996 it wouldn't have had links to Privacy Policy pages on WB's current website. I think that makes it even cooler as they are actively maintaining it in its historic form.
In the earliest (December) 1996 version that Archive.org has, the site has an initial landing page you have to click through before you arrive at something similar to the homepage today:
I was wondering this, too, recently. spacejam2.com is kind of interesting in that they sell jams and also offer to sell the domain for a million dollars.
"You are welcome to buy our jam or you can also buy our website"
The answer is most likely an image hosted on an academic's personal site on a university server somewhere. Although how you would ever figure which is the actual oldest, I have no i idea.
As a college student, I helped create this site in, I believe, 1994: http://pynchon.pomona.edu/ I don't know that I really count as an academic, but I think the site has survived because it's sitting on a university server.
I like the idea of some extremely tired 486 hooked up to a thinnet cable in some storage closet still serving up pages off of their Slackware Linux 1.1 install using CERN httpd. Nobody wants to turn it off because they know the hard drive will never spin up again if they do.
Fun fact: The amusement park from his 3rd most recent tweet burned to the ground a few days ago, suspiciously shortly after the company declared bankruptcy.
Do I have to be the oldest fart in the room to just naturally assume that since Mozilla was the first popular browser to show inline images, that its homepage is the first place to look?
NCSA Mosaic was the predecessor to Netscape and Mozilla, and I think was the first browser to offer inline images. I remember going to Interop in 1993 and everyone was blown away by the demo of Mosaic and its inline images showing O'Reilly's GNN (Global Network Navigator) site. Somewhere, I might still have an advertising card the O'Reilly folks were handing out about it. You can see more here (and yes, bizarrely, wood-cut-look icons were a big design fashion trend for a period in the early '90s. Sun's TOPS used them extensively, too...) https://www.oreilly.com/gnn/
That said, there were several earlier browsers that could opened image tags in another window or app - Viola, Cello, etc. (Even Lynx, which still exists for some bizarre reason...)
As hard as it will be for the younger crowd to believe, at that time every new web site on the Internet got posted to the NCSA "What's New" page, so people would know what all's out there!
It’s remotely possible oreilly.com has one of the oldest continually existing images... they already had a shelf or two of books when Navigator launched.
> image maps from 1994 still perform like absolute crap in 2021
I don't follow. Are you referring to the progressive GIF rendering? It seems to be working beautifully and as intended. It's a 40K GIF that downloads at 5K/s, presumably because of limited server-side throughput (artificial or real). If you just open it locally or host it somewhere else it renders instantaneously, without any hint of its progressive encoding.
Or maybe that's what you were referring to--the possibly deliberate uplink throttling--and we just have different opinions about what's "beautiful".
I'm curious if anyone has written a JavaScript library that emulates the browser rendering experience of 1990s modem links. That'd be cool, both from an implementation standpoint as well as viewing experience.
Netscape Navigator was the second popular web browser to show images.
Problem is that the first one changed their domain name since then, from ncsa.uiuc.edu to ncsa.illinois.edu, so unless there’s some obscure little server still running, those are gone.
You’re going to want to look for one of the pilot sites of the internet, but find one that’s been continuously operational. The original CERN site went away.
Cello added support for the img tag in 1993, before Netscape was ever released. At that point in time, I recall it was the second most popular GUI browser, behind NCSA Mosaic.
Cello lost to Mosaic, Mosaic lost to Navigator, humanity lost to IE, then IE lost to Firefox, Firefox lost to webkit, and now on HN we hand-wring about whether humanity is losing to Google or not.
I have had many, many conversations about the "Browser Wars" over the years and it's like talking about cross-border rivalries in the Midwest. Nobody can quite agree on who is the best or worst. Most people can't even agree about who was actually involved (is Ohio in the Midwest? People in Ohio say yes. People in the Mississippi River Valley say no).
Everybody who is anybody agrees that they all hate IE/Iowa.
I was trying to put together some timelines at one point and was reading biographies about Spyglass. Do you remember Spyglass? According to the founder, the Browser Wars were Spyglass versus IE (or was it Navigator?), and Spyglass won.
Why? Because he was only thinking about corporate computers. That was the universe to him. Places that we all know were barely on the Internet, he felt was "The Browser War".
I thought the first browser war was Mosaic vs Netscape, but as many think the IMG tag is the first real salvo in the war, Cello vs Mosaic is probably the first war, IE vs Netscape being #3.
Hello fellow old farts. I rememer browsing on Mosaic, Cello, Slipknot, Netscape, Arachne, Phoenix/Firefox. There were doubtless others I've forgotten.
Anybody remember the various 3D browsers that would visualize site relationships? Apple's Hotsauce was the first one I remember, though it was more of a text based meta browser. Then there were a few graphical windowing attmepts.
The first images I remember seeing on the internet were the Clementine lunar photos, slowly rendered line by line:
but I think everything on the site was actually posted sometime in 1994.
Unfortunately, the University of Utah recently deleted all alumni home directories and websites (I managed to save what I could). There were probably some fairly early pages up there.
One image that certainly isn't the oldest but might be one of the oldest important images that is still online at its original location is the Blue Ribbin Online Free Speech Campaign animated logo (https://www.eff.org/files/br.gif).
This image dates back to 1996 when the EFF launched a campaign against the Communication Decency Act (CDA). Many of the pages across the internet turned their background images black in protest of the CDA. What remains of the CDA is today known as "Section 230".
"In the same month, the group purchased alien abduction insurance that would cover up to fifty members and would pay out $1 million per person (the policy covered abduction, impregnation, or death by aliens)"
I had never heard of Heaven's Gate. I just read (mostly skimmed through) some of the things in that website. It was quite horrible and dripping with morbidity (is using morbidity this way a correct nominalization?).
I'm very confused about the "NeoCities" image next to the IE and Netscape images. NeoCities is from 2013... Was there a different NeoCities that I'm not aware of?
The page actually contains modern things like references to Google Analytics and Apple meta tags, but that could be because the archival tool keeps modernizing the site.
Edit: ok, looking for "Maria Korhosen katoamistapau" gives results from Amazon music and YouTube... looking for the person's name got me nowhere.
The first thing this made me think of is the Astronomy Picture of the Day blog. Their first entry is on June 16, 1995 and they've been updating it ever since:
https://apod.nasa.gov/apod/ap950616.html
Michigan State University just finally (last year, maybe the year before) killed my personal site I had set up when I was a student in 1994. It had one picture on it.
There’s Usenet archives going back to at least 1990. I’d search those for image urls, then find ones that still work.
I just checked and sumex-aim.stanford.edu isn’t a host anymore :(
Here's one of the oldest I can find, with timestamp 1988-07-06 00:00:
ftp://ftp.sunet.se/mirror/archive/ftp.sunet.se/pub/pictures/anime-manga/UruseiYatsura/Images/mendo.gif
or via http:
http://ftp.sunet.se/mirror/archive/ftp.sunet.se/pub/pictures...
This is the oldest one I can find on nic.funet.fi, from 1991:
ftp://nic.funet.fi/pub/pics/comp/net/misc/dnesu.gif
They have mirrors of many of the old ftp servers of yore, so if there's an older image out there, that's the place to look:
http://www.nic.funet.fi/pub/mirrors/
Most of the other old FTP servers I remember (sunsite.unc.edu, ftp.simtel.net, ftp.uu.net) have long since shut down. The ftp server at ftp.sunet.se is one of the oldest still running. Perhaps someone else knows of an older one?
I tried to find a copy of lena.bmp/lena.tga/lena.pcx since those file formats pre-date GIF, but I came up empty.