I started at CompuServe in 1990. I met Steve a couple years later when I interviewed to join his team.
I didn’t make the cut and I thought he was a bit of a jerk during the interview. Turns out I was the jerk and Steve was actually being really kind to a kid who wasn’t even close to being ready for that level.
Anyway, what I can say is that Steve was a lot more than GIF. In fact, back then, no one really thought too much about it, but they thought a lot about Steve.
TAR files from 1970s? ROFF from the 1960s is still used to create and maintain Unix man pages for all flavors of Unix. MIDI from 1983 is older than GIF....
Maybe mbox (https://en.wikipedia.org/wiki/Mbox), the Unix email message data file format? It hasn't changed much in a really long time, but it's also extremely simple.
Also the diff/patch format (https://en.wikipedia.org/wiki/Diff) is almost as old as GIF, but Wikipedia says unified diff support was added in 1990-1991.
GIF89a left some things undefined (larger than 256 color palettes, timing 0 frames, partial frames, etc), so while the text of 89a is fixed, modern GIF decoders handle s superset of GIF files, just like modern CSV files.
Also GIF89a supported embedded text to be drawn, which is not implemented in any modern decoder I am aware of. Web browsers certainly don't support it. So GIF89a is not technically still in use - we're using a more recent, less standardized flavor.
If I recall, there are other things like this related to modern GIF versus old GIF89a gifs - I ran into a lot of this working on hiding information in GIF file encoding choices (not in the images or file, but simply in choices made by the encoder). Then I scanned zillions of online GIFs to see if someone else had done this, and kept finding weird edge cases in GIF89a.
I think there are also many currently used GIF extension blocks that are in use that were not in the GIF89a spec. I'll see if I can find a list and dates they appeared (remember, GIF89a allowed anyone to add extensions to the format, and many companies did). MIDI is one type of extension block that was once embedded in GIFs by some vendors. PCM audio is another block that appears in some GIFs (so would these not count as GIF format being modified later?)
So even GIF89a is not exactly the same :)
MIDI, v1.0, still works and reads just fine. And it's still widely used (likely the most common flavor of MIDI in use).
Current readers read ancient CVS files just fine, so there are modern behaviors, but the original file format is still also in use.
The tar format in use was made a POSIX standard in 1988, older than GIF89a. The extended formats do not have .tar extensions.
TROFF seems to use the exact same file format it did in 1973: https://troff.org/history.html. The extensions you mention are making it have different output formats, ore making a new version to process the same file format but with code under different legal provenance.
Thanks for the detailed reply, I didn't know about the unofficial changes to GIF89a itself. So, yes, in that case, my point is refuted.
But the examples you gave ("an older format still works today") weren't my point. My point was that GIF89a didn't get any updates to its spec since 1989, yet it remained extremely popular as it is. That's different than a format getting occasional updates to its spec and that stays backwards compatible. If that were the case, even BMP/ICO would predate GIF.
There is no CSV standard. The ancient ones are still completely valid. Some groups have tried to define some regularity, but there is no spec. So by all angles, this predates gif by a long shot.
ROFF and TAR are the same. No group owns some standard for them. And akaik ROFF is still the 1970s format.
Yes, that is one of many, many places trying to be the standard. But that is not a universal, used by everyone, used everywhere standard. CSV is simply not standardized, as any data scientist working with CSV files from all over can quickly tell you. Or a simple google search will tell you. For example, Excel, likely the biggest consumer and producer of CSV files, does not implement RFC 4180. So that is the end of that.
From Wikipedia:
"The CSV file format is not fully standardized. Separating fields with commas is the foundation, but commas in the data or embedded line breaks have to be handled specially. Some implementations disallow such content while others surround the field with quotation marks, which yet again creates the need for escaping if quotation marks are present in the data.
The term "CSV" also denotes several closely-related delimiter-separated formats that use other field delimiters such as semicolons.[2] These include tab-separated values and space-separated values. A delimiter guaranteed not to be part of the data greatly simplifies parsing."
From the Library of Congress database on formats: A CSV file is "A simple de facto format, for which no single, official specification exists."
Just because one guy calls his flavor a standard does not make it a widespread standard, any more than one company adding MIDI to GIFs makes that the newest, universal standard.
I get that you must really, really want GIF89a to be the end all of unchanged golden formats. But the lengths you go to to ignore all this other stuff, without putting in equivalent effort to understand them, is not making the case. Simple spend a minute on Google before posting stuff that is simply untrue. It's easy to find all this stuff.
Since you don't seem to care to read about each format, and understand the nuances and that this question is not the black and white case you want, I'm done providing evidence or cases. Good luck.
It remained unchanged for over 30 years, but there are actually two versions of the GIF file format. The second revision, known as 89a, added a few minor refinements like a frame rate for animations and transparency (but only binary, e.g. fully opaque or fully transparent).
PNG could arguably be seen as GIF 2.0, since it addressed the major shortcomings of the format. But since PNG had no animation capability, it could never fully replace GIF which held on for that one niche use.
Every word (except gin) that starts with G, then a vowel, then an F, is pronounced with a hard G. For example:
Gaffe. Gift. Guff. Guffaw.
Most one-syllable words that start with G have a hard G (not an exhaustive list):
Gab. Gad. Gag. Gal. Gam. Gap. Gas. Gay. Get. Gig. Gill. Gimp. Gird. Girl. Git! Give. Go. Goal. Gob. God. Gone. Gore. Got. Guide. Guild. Guilt. Gull. Gulp. Gum. Gun. Gust. Gut. Guy.
The word "gift" is the closest word to GIF, and it has a hard G. To pronounce GIF, just say "gift" without the "t".
There has been a huge debate about the real pronunciation of GIF over the years and many users have continued to offer their own arguments.
But in 2013, Steve settled the debate once and for all during his acceptance speech of his lifetime achievement Webby Award, revealing that GIF is pronounced with a soft “g”.
“It’s pronounced JIF, not GIF,” the beloved creator revealed to the crowd.
Etymology wise it inherits that from the sound of "gen" from "geneva." Since GIF is a new word it can't inherit a pronunciation from a former spelling.
To reason in those terms, you should first define such «natural pronunciation» principle and rules. 'PGN', if made an acronym out of the initalism it is, could reflect both 'pigeon' and 'pagan'; it's "Paris German News", so the direction of 'pigeon' is the one. An example of an acronym: the last in 'OPEC' - which per se could be soft or hard - has to reflect that the C is the initial of 'Countries'.
Nah, once it's out there people can do whatever they want with it, in the same way that once "xerox" became a common word, it wouldn't matter if Xerox Corp. came and said "hey btw it's pronounced as ZEROZ".
Except that GIF is not a personal name, and all those meme GIFs won't be mad if we call them "jifs" or "gifs".
Names for things get lives of their own.
Similarly, I don't think it merits looking at similar words for the "proper" pronounciation. If anything, it makes sense to look at expanded acronym (Graphics Interchange Format) which still points at the hard G (unless it's Jraphics Interchange Format :)). But hearing it from whoever coined the term is definitely similarly applicable.
Still, I would reiterate, names for things that get wide adoption get their own lives, and we'd face an uphill battle to stop people from using them how they want. As long as we can understand each other, we are golden. And when we can't, it's not that hard to get back on the same page.
> If anything, it makes sense to look at expanded acronym (Graphics Interchange Format) which still points at the hard G (unless it's Jraphics Interchange Format :)). But hearing it from whoever coined the term is definitely similarly applicable.
This. I first learned what the acronym came from and assumed the G was a hard G from Graphics. I just assume anyone who pronounces it with a hard G learned it on their own.
Can't you make this argument to just bully people you don't like?
If you get a majority of people to say "zzleeper" is actually pronounced "crap bag", then majority rule wins, right, and you must acknowledge people when they call out "crap bag"?
I just don't really see how majority rule should follow. I'm also curious to what extent we allow this to function. Is it just limited to names?
You make it sound like word pronunciations are logical, follow rules, behave similarly to other words that are close in origin or in writing.
That’s not how it works. That’s like trying to convince someone who names his son “Richard” how the last letters are supposed to be pronounced. There’s no amount of debate you could have that actually makes any progress one way or another.
Would you pronounce it “Grin-drrrrr” because you just want to normalize to a hooked-on-phonics view of pronunciation while using a language that is a crash-merge of proto-Germanic and French with a sprinkling of centuries of colonial cultural appropriation?
“GIF” is like a brand. It’s a spec; a proper noun. As such, it has a clearly intended pronunciation that was unambiguously provided by the creator. That guy is dead now, so he may not notice, but try honoring his intentions for a day.
Everyone who gets dogmatic about “I don’t care what the spec is like” just comes off seeming to me like old Englishmen who quite clearly intentionally mispronounce non-English words (except French, which they relish…).
It used to be that the only folks I knew who pronounced this out of spec were my parents and grandparents (because the didn’t know better and were computer illiterate). Now, somehow, we have rafts of people who have adopted the posture of my parents in the 90s with pride.
I imagine that the next 24 hours will see a lot of these sorts of discussions…
the word gin is closest to grin, maybe we should say gin without the "r"....
these arguments are really bad, there is no real rule to follow, especially for an acronym. Anyways...
this gent was a gem, a real gentle giant and genius.... these debates are like germs and generate generic retorts, lets be more genuine.... I hope you are getting the gist of this comment.
Methinks HN should always include a brief who/why summary in their death-notice headlines - as much as I'd like to think we all already know who early leaders in tech are by-name, I'll admit the low-information "Stephen E. Wilhite has died" headline passed me by and while skimming HN I misread it as "Stephen Wile E. Coyote has died" and didn't care to click further.
...it wasn't until hours later when I was bored and idly reading everything that I saw that I earlier misread the title and learned the CompuServ/GIF connection in the HN comments.
It isn't uncommon for letters in an acronym to be pronounced differently from they are in the words they originated in. eg: laser, scuba, and NASA aren't pronounced "lah-seer", "scuh-bah", "nay-sah".
Unfortunately I know people that have died fully dosed, from COVID. The mortality rate has always been "low" but it's higher among very elderly, and especially among men.
I think it's bad taste to start speculating if someone dies from COVID whether they were part of some kind of ideological group that put themselves more at risk, he's actually part of a group that is just inherently more at risk.
I didn’t make the cut and I thought he was a bit of a jerk during the interview. Turns out I was the jerk and Steve was actually being really kind to a kid who wasn’t even close to being ready for that level.
Anyway, what I can say is that Steve was a lot more than GIF. In fact, back then, no one really thought too much about it, but they thought a lot about Steve.
George Jones, another CompuServe alum, penned this a few years back. I think it’s a nice summary. https://web.archive.org/web/20220129204133/http://curious.ga...