Decoding of an animated WebP file is not supported.

kazitor

This error is immensely valuable and immediately actionable to me.

TwelveBaud

@kazitor I thought it was immediately actionable: Upload something other than an animated WebP file instead.

boomzilla

E_IMAGE_MAGICK_SAYS_NO

Gern_Blaanston

TIL: Animated WebP file is a thing.

But ...

cvi

@Gern_Blaanston Presumably for the same reasons people did Motion JPEG, APNG, HEIF with image sequences plus its bastard offspring (e.g. AVIF), GIF and stuff like animated SVGs.

Of course, the amusing thing about animated webp is that webp is derived from VP8 ("webm") and uses a subset to compress still images. VP8, of course, was designed for videos in the first place.

Tsaukpaetra

@cvi said in Decoding of an animated WebP file is not supported.:

Of course, the amusing thing about animated webp is that webp is derived from VP8 ("webm") and uses a subset to compress still images. VP8, of course, was designed for videos in the first place.

Lesser words, "Why not just webm?"

cvi

@Tsaukpaetra Yea.

loopback0

Is webm pronounced with a soft w or a hard w?

topspin

@Tsaukpaetra said in Decoding of an animated WebP file is not supported.:

@cvi said in Decoding of an animated WebP file is not supported.:

Of course, the amusing thing about animated webp is that webp is derived from VP8 ("webm") and uses a subset to compress still images. VP8, of course, was designed for videos in the first place.

Lesser words, "Why not just webm?"

Tsaukpaetra

@loopback0 said in Decoding of an animated WebP file is not supported.:

Is webm pronounced with a soft w or a hard w?

It's pronounced "Wemon".

Zecc

@loopback0 said in Decoding of an animated WebP file is not supported.:

Is webm pronounced with a soft w or a hard w?

Web' em, Danno.

Bulb

@kazitor said in Decoding of an animated WebP file is not supported.:

This error is immensely valuable and immediately actionable to me.

From the dwebp tool it's a good error. If you were running it directly, it would be actionable quite well. From imagemagick wrapper it still good enough because if you were running that directly, you could call the other tools directly too. From a web app … the text becomes a bit inappropriate. The web app should provide its own message.

… and dwebp and imagemagick could help with that by providing a catalog of error messages so the web app calling them can do that. But there is a whole herd of s that way so few applications ever do.

remi

@Bulb said in Decoding of an animated WebP file is not supported.:

… and dwebp and imagemagick could help with that by providing a catalog of error messages so the web app calling them can do that.

Or, rather than providing a catalogue of exact messages, maybe it could use some sort of simple code. Let's say an int. We could call that, I don't know, an "error code." Or maybe, just maybe, we could say that the process must always return such a code before finishing, with a special value to indicate "there was no error." We could call that a "return code" or an "exit code" and the OS or shell could provide a simple way for a process that called another process to get the exit code of the inferior.

Nah, that's too wild an idea, it'll never work. :sarkmark:

Applied Mediocrity

@remi said in Decoding of an animated WebP file is not supported.:

a special value to indicate "there was no error."

Like 502

TwelveBaud

@Tsaukpaetra said in Decoding of an animated WebP file is not supported.:

Lesser words, "Why not just webm?"

WebM implies some sort of medium-to-long form movie, as opposed to a mildly animated picture.

Put another way, you don't put smilies up full-screen on your home theater system.

Tsaukpaetra

@TwelveBaud said in Decoding of an animated WebP file is not supported.:

Put another way, you don't put smilies up full-screen on your home theater system.

Maybe you don't!

Filed under: DVD Logo.apng

Bulb

@remi said in Decoding of an animated WebP file is not supported.:

@Bulb said in Decoding of an animated WebP file is not supported.:

… and dwebp and imagemagick could help with that by providing a catalog of error messages so the web app calling them can do that.

Or, rather than providing a catalogue of exact messages, maybe it could use some sort of simple code. Let's say an int. We could call that, I don't know, an "error code." Or maybe, just maybe, we could say that the process must always return such a code before finishing, with a special value to indicate "there was no error." We could call that a "return code" or an "exit code" and the OS or shell could provide a simple way for a process that called another process to get the exit code of the inferior.

Note that there was a return code. 4. With explanation: unsupported feature. There is a use for these coarser-grained errors too.

Also, exit code is not an int, it's just a byte, and then the shell robs it of the highest bit.

Nah, that's too wild an idea, it'll never work. :sarkmark:

Some compilers have them, e.g. all Microsoft build tools do and Rust does. Which isn't exactly the use case, because toolchain errors generally make it straight to the user. I think in that case there are two reasons: different localizations, and referencing a more detailed explanation in the documentation.

But that's … about the only two cases I've seen lately.

dkf

@Bulb said in Decoding of an animated WebP file is not supported.:

Some compilers have them, e.g. all Microsoft build tools do and Rust does. Which isn't exactly the use case, because toolchain errors generally make it straight to the user. I think in that case there are two reasons: different localizations, and referencing a more detailed explanation in the documentation.

But that's … about the only two cases I've seen lately.

The exit code from pylint is a bitmask describing what categories of problems the linter found. You can use that to do things like making the CI build only stop if serious issues are found (useful because some style things it finds are very fussy and opinionated).

I've never seen any other tool that does that.

Bulb

@dkf That's not the same thing. It's just categories, not individual errors. Remember, it's just 7 bits that are available.

dkf

@Bulb said in Decoding of an animated WebP file is not supported.:

It's just categories, not individual errors. Remember, it's just 7 bits that are available.

Even if they had 32 bits, that wouldn't be enough. You can't just use a code per issue; there will often be many different problems found at once.

Bulb

@dkf You can use a code per issue. All Microsoft compilers do it. But you can't return them via exit code, because it's too small. Which is not a problem. In the above case the web app is reading the standard error of imagemagick, so it could pick a code from there and translate it.

dkf

@Bulb In short, you have to have custom exit code handling for near every damn program because so many of them have ideas about what to do with them.

I remember dealing with an IBM mainframe (back as an undergraduate) where errors were just codes. The codes gave the location in the wall of shelves of binders where the description of what the error was. (That machine was phased out toward the end of my second year.)

boomzilla

PRs might be accepted here:

GitHub - NodeBB/nodebb-plugin-imagemagick: ImageMagick parsing of avatars and other uploaded image assets

ImageMagick parsing of avatars and other uploaded image assets - NodeBB/nodebb-plugin-imagemagick

Bulb

@dkf Well, if you want to produce more helpful error messages for the error codes, you need custom handling for each of those codes anyway.

topspin

@boomzilla said in Decoding of an animated WebP file is not supported.:

PRs might be accepted here:

So other sites using nodeBB can enjoy new features?

Gribnit

@dkf said in Decoding of an animated WebP file is not supported.:

@Bulb In short, you have to have custom exit code handling for near every damn program because so many of them have ideas about what to do with them.

I remember dealing with an IBM mainframe (back as an undergraduate) where errors were just codes. The codes gave the location in the wall of shelves of binders where the description of what the error was. (That machine was phased out toward the end of my second year.)

I've taken to mapping legacy errors onto either the sysexits.h pseudostandard or the HTTP error codes just to make a sane surface.