Hacker News new | past | comments | ask | show | jobs | submit login
Proposed new HTML tag: IMG (1993) (webhistory.org)
169 points by wamatt on Sept 26, 2012 | hide | past | web | favorite | 62 comments

I think the crux of that conversion is:

«Marc Andreessen (marca@ncsa.uiuc.edu)

Fri, 26 Feb 93 13:32:09 -0800

Tim Berners-Lee writes:

> I don't want to change HTML now if I can help it, until it has gone to RFC track

I absolutely agree in all cases -- my purpose in suggesting IMG is that things are reaching the point where some browsers are going to be implementing this feature somehow, even if it's not standard, just because it's the logical next step, and it would be great to have consistency from the beginning -- so that when HTML2 comes along, we're all still in lockstep.....

Cheers, Marc»


The HTML 5 designers/implementers fight has always been there, with implementers usually having the last word.

Mark Pilgrim on how much the WHATWG resembles the HTML 2 working group activity: http://web.archive.org/web/20110710091457/http://diveintomar...

And because of this, people were already predicting what would end up happening as a result of implementers rushing out features


Wait a minute -- let's temporarily forget about MIME, if it clouds the issue. My objection was to the discussion of "how are we going to support embedded images" rather than "how are we going to support embedded objections in various media". Otherwise, next week someone is going to suggest 'lets put in a new tag <AUD SRC="file://foobar.com/foo/bar/blargh.snd">' for audio. There shouldn't be much cost in going with something that generalizes.

That thread also predicted the weird and unrealistic "let's-stay-inside-the-existing-boundaries" responses:

> SGML does provide an official way of doing this folks and even if Dan C ain't here to round us up we maybe ought to stick to the track. > > <!ENTITY ICON6 SYSTEM "http://blah..>&ICON6;


Yeah, that looks good, but it would be nice if it had a way to specify different resolution copies of the same image for the future when we have varying screen sizes and varying ppi.

Good proposal though, and hope that this takes off.

You know, there used to be a 'lowsrc' attribute for the image tag. (A lower resolution image, which would be downloaded and displayed before the download of the 'src'-image file.)

It apparently is ignored by most browsers now and no longer allowed in HTML5.


Could do it like HTML5 video, and use the figure tag.

        <img src=blah.png dpi=72/>
        <img src=blah_hq.jpg dpi=300/>
        <caption>The fabled Blah stalks it's prey.</caption>
        <copyright year=1997>John Doe</copyright>
        <license href="http://cc.org/sharealike">CC Sharealike</license>
And then your browser could choose the DPI most like your screen.

Personally, I'm skeptical.

I got a bad feeling about this img tag - what if everyone just uses it to post pictures of cats and people making duck faces?! the internet might be flooded with useless content!

You missed the flippant humour in the parent.

Why does that is proposal have informal slangy cruft like "poly fill" and exclamation points? That is improper for a stabdard.

Can't you just scale down to half-size an interlaced gif/jpeg and have the browser intelligently decide to stop downloading once it is at sufficient quality?

"Good proposal though, and hope that this takes off." it was a proposal made 19 years back?

I can feel the breeze of it going over your head from my desk.


does not compute

"Otherwise, next week someone is going to suggest 'lets put in a new tag <AUD SRC="file://foobar.com/foo/bar/blargh.snd">' for audio. There shouldn't be much cost in going with something that generalizes."


Well, maybe not next week, but certainly 15+ years later.

Both <embed> and <bgsound> (the latter not standard, but certainly "proposed") have been around for a long time.

Jim Davis: 1) why have SRC instead of HREF?

Marc Andreessen: Because I wanted to avoid overloading HREF -- doesn't really make sense in this context, I don't think.


So... that's the reason!

Well, you want to embed the image from a source identified by URL, not refer to it. In fact, I wonder why we need to use href when invoking stylesheets.

Mark Pilgrim's Dive Into HTML5 has a chapter about this very conversation, providing some usful context: http://diveintohtml5.info/past.html

Great idea Marc. Just wondering what should happen with images stored on different web servers should someone in the future ever come up with a way for servers to identify individual clients.

Best sort something out now, before people get used to things working in a certain way.

Thanks for the feedback -- I'll see if I can work that in before I release the first implementation!

Was this the slippery slope that lead to the <blink> tag? (just to be fair, Microsoft introduced the equally horrendous <marquee>)

No way! It's that Marc Andreessen. What great humble beginnings!


Click next in Thread. Then look at the "Next in Thread" - it's that Time Berners Lee http://1997.webhistory.org/www.lists/www-talk.1993q1/0186.ht... And also look at the "Next Message". It's that Guido Van Rossum http://1997.webhistory.org/www.lists/www-talk.1993q1/0184.ht...

This was the same year that NCSA Mosaic, the first ever web browser, was co-created by Marc Andreessen. He was probably working on Mosaic when he wrote this email.

> the first ever web browser

Fact check: http://en.wikipedia.org/wiki/Mosaic_(web_browser)

> While often described as the first graphical web browser, Mosaic was preceded by WorldWideWeb and the lesser-known Erwise[5] and ViolaWWW.

Probably. If he wasn't, this bit from the first post becomes very weird:

> This is required functionality for X Mosaic; we have this working, and we'll at least be using it internally.

Guido Van Rossum posts to the same group a few messages later.

He's endorsing Romney? Did not know that.

@pmarca, why? If you don't mind : ) Just curious.

He says he believes that a Republican administration would be "considerably more open on regulatory issues, and the markets would be more competitive under a Romney administration."

Not to get all political here, but if I could somehow set aside social issues, endorsing Romney (and most republicans) would be a "slam dunk"...but alas,those pesky social issues persist.

Very strange, I tried searching http://blog.pmarca.com for obama and romney and there were no results.

But I swear that back in 2007-2008 I remember him having blog posts about meeting Obama?

edit: found it at the archive http://pmarchive.com/an_hour_and_a_half_with_barack_obama

I guess it was just deleted over time as the blog.pmarca.com was made more professional.

The reply was especially amusing:

"I was proposing to use the file extension (.xbm above) to tag what format the image was in, but with the intention that in future, when HTTP2 comes along, the same format negotiation technique would be used to access images."

Almost twenty years later and still no HTTP2!


Yes we do, it's just called HTTP 1.1 instead - and that exact functionality luckily made it in.


I'm old enough to remember reading this reposted on usenet (or was it gopher) at the time. Cue discussion about how gopher was much better as it already did images.

Little did we know back then...

There was also serious backlash from certain people (cough TIM BERNERS LEE cough) that adding images to the web would degrade the user experience, encourage bad use cases, and encourage the riffraff to use it.

Why yes, I still remember that, why do you ask?

But it did degrade the user experience, encourage bad use cases, and encourage the riffraff to use it. The easiest way to eliminate those problems is to not have users.

Why'd you have to go making things popular? I could still be using Lynx on a 1200 baud modem!

I'd rather be using Lynx on a 1200 baud modem. I was far more productive back then without the distractions of hacker news etc.

Tim, is that you?

See, if you hadn't let the riffraff in, you'd know for sure...

WorldWideWeb had image support well before this (in a separate window?), and certain included inline images c1993-1994. Seems like an odd thing to do if Tim didn't like them.

And here you were, an upstart kid, creating this newfangled "browser" so that anybody (anybody?!) could see the Web! The temerity of it! ;-)

I was proposing to use the file extension (.xbm above) to tag what format the image was in, but with the intention that in future, when HTTP2 comes along, the same format negotiation technique would be used to access images.

Nineteen years later and we still don't have HTTP2.

In the link provided by sbierwagen, it states:


“HTTP2” is a reference to Basic HTTP as defined in 1992. At this point, in early 1993, it was still largely unimplemented. The draft known as “HTTP2” evolved and was eventually standardized as “HTTP 1.0” (albeit not for another three years). HTTP 1.0 did include request headers for content negotiation, a.k.a. “MIME, someday, maybe.”

So, at least in this context, yes, we do have "HTTP2".

Oops, I was calling HTTP/2.0 HTTP2 as a shorthand seems that I was so wrong to do that :-)


Ha, so that's what rel was supposed to mean.

Wow, I have been meaning to look what rel means for years and I always forget.

Am i glad I saw this.

Mark Pilgrim had an insightful take on the development of the IMG tag in his book Dive Into HTML5:


It's too bad they went with `img` and `src` instead of listening to Tony Johnson and using `image` and `source`. Saved a few characters for needless abbreviation.

But how many more bytes would've been sent since 1993 were they image and source? :)

I'm actually glad they did.

HTML is already very verbose. Although 'header' and 'footer' appear (supposedly) once per page, styling inline elements requires to use 'strong' or the generic 'span' tag. And it's one situation when the tag is numerously repeated, which ends up in a very low signal/noise ratio.

My position may come from the fact that I type every line of code (I have no IDE, just Notepad++). But I think that there aren't enough HTML tags for us to have to use longer names in order to avoid confusion.

Do you also think <a href=...> should have been <anchor hypertextreference=...>?

TBL wanted to use the anchor tag for it: http://1997.webhistory.org/www.lists/www-talk.1993q1/0186.ht...

"I had imagined that figues would be reprented as <a name=fig1 href="fghjkdfghj" REL="EMBED, PRESENT">Figure </a>"

That would have really changed how things evolved. Probably then anything that was specifying an external resource would have all been anchors.

I feel sorry for Tony Johnson that had "something very similar in Midas 2", as it's the Mosaic implementation that "won" :)

It also shows here in his comment how 'small' the internet was: maybe everyone having fibre to their home was some far-off imaginings of crazy people.

"I somewhat prefer ICON since it implies that the IMAGE should be smallish"

I think the best part of this entire exchange is that it provides proof that shipping code wins.

Sounds like a good idea

We've landed on the moon!

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact