Does the .png image format have a text metadata field?

nospam wrote:
In article , mechanic
wrote:

One day file extensions will be outdated and machines will work out
how to display the data from a file without the preconception
inherent in assigning a file extension.
that day was back in 1984 with the original macintosh and classic mac
os, which did not use file extensions.

mac os x, being based on unix, does use extensions.
No we don't need extensions in UNIX,

yes we do. change the extension and things break.

for example, rename a .tar.gz to .jpg, a .html to .png., a .pdf to .cc,
or remove the extension entirely and see how well it works out for you.


You have to know what you're doing, for that to work.

Rename a .tar.gz to .jpg.

Open your File Manager (nemo, thunar, nautilus, ...).

Right click the file, and the item showing will
be two instances of "Archive Manager". Archive Manager
will "mount" the mis-named file as a .tar.gz in the
same way as Windows cabview.dll will show the contents
of a cab file. Once you've navigated to the bottom of
the Archive Manager mount, you can interact with
whatever file is in there.

If you force the issue, like doing something stupid
(in other words, you intentionally avoid the
automation protecting you)

vlc some.tar.gz

then VLC will likely tell you that it has nothing
to parse or view that with. You do not expect VLC
to have an entire File Manager inside and a running
copy of the "file" command, to out-think you. VLC is not
going to say "I sniffed that, and the output of /bin/file
tells me you would have better luck with a gzip pipe".

If you force-feed VLC like that, it's just going
to yap back at you that it cannot read the file.

File extensions remain in either system, as a
"suggested serving". You can invent additional
steps in the computing process, to consider that
style of metadata extension, but it is not trusted,
and eventually the parser on some application
ultimately gets to decide whether anything is
going to happen.

If you do

vlc some.tar.gz

VLC does not go "OK, OK, already, I'm eating this,
here's your video", while on the screen you
see random colored dots caused by trying to
display a ZIPped file on the screen. Software
used to do stupid **** like that long long ago,
but we've long since surpassed the "naive loading"
approach. Programs usually need to parse a bit of the
file (say, a length field), before they can be coerced
into stupid stuff.

The closest you'll get to "loading garbage" these
days, is in cryptography. And there are probably sufficient
protections after an attempted decryption, to tell that
the file was not a candidate in the first place.

Paul

In article , Paul
wrote:

No we don't need extensions in UNIX,

yes we do. change the extension and things break.

for example, rename a .tar.gz to .jpg, a .html to .png., a .pdf to .cc,
or remove the extension entirely and see how well it works out for you.


You have to know what you're doing, for that to work.

changing an extension does not change the underlying format.

a tar.gz that has been renamed to jpg is still a tar.gz. it is not a
jpeg. renaming a .jpg to a .gif is still a jpeg.

you can sometimes force apps to open it anyway, but not always and
certainly not by double-clicking.

it *will* cause problems.

nospam wrote:
In article , Paul
wrote:

No we don't need extensions in UNIX,
yes we do. change the extension and things break.

for example, rename a .tar.gz to .jpg, a .html to .png., a .pdf to .cc,
or remove the extension entirely and see how well it works out for you.

You have to know what you're doing, for that to work.

changing an extension does not change the underlying format.

a tar.gz that has been renamed to jpg is still a tar.gz. it is not a
jpeg. renaming a .jpg to a .gif is still a jpeg.

you can sometimes force apps to open it anyway, but not always and
certainly not by double-clicking.

it *will* cause problems.

Not in File Manager. And in the GUI cosmos, that's
what you're supposed to be using. File Manager knows
what that is, and will present the correct application
(Archive Manager).

There's a reason you can't (easily) find Terminal
in Linux distros. That's the reason. To wean newcomers
off command line invocations.

Paul

In article , Paul
wrote:

No we don't need extensions in UNIX,
yes we do. change the extension and things break.

for example, rename a .tar.gz to .jpg, a .html to .png., a .pdf to .cc,
or remove the extension entirely and see how well it works out for you.

You have to know what you're doing, for that to work.

changing an extension does not change the underlying format.

a tar.gz that has been renamed to jpg is still a tar.gz. it is not a
jpeg. renaming a .jpg to a .gif is still a jpeg.

you can sometimes force apps to open it anyway, but not always and
certainly not by double-clicking.

it *will* cause problems.

Not in File Manager.

that doesn't interpret the contents of the file.

And in the GUI cosmos, that's
what you're supposed to be using. File Manager knows
what that is, and will present the correct application
(Archive Manager).

not always.

it can 'work' if the underlying format is the same. you could, for
example, write raw html into a .txt file, then rename it as .html and
continue editing it or put it on a server, but that's only because both
are plain text.

There's a reason you can't (easily) find Terminal
in Linux distros. That's the reason. To wean newcomers
off command line invocations.

separate issue.

"J. P. Gilliver (John)" on Thu, 20 Feb 2020
13:52:44 +0000 typed in alt.windows7.general the following:

Another odd thing about RTF: I don't know what
it is, but the display always looks cheap to me. Fonts
are not crisp looking. Color options are limited unless
you customize the code. Not as polished as other file
formats.

I know what you mean! It used to have the advantage that at least anyone
with Windows could edit it, as (I think) it's the native format of
Wordpad, which comes with Windows (and for a "free" WP is a lot better
than it's given credit for - in fact, if it wasn't for compatibility
issues, I'd say is sufficient for many users' needs). And there are
people who _don't_ have Office (or one of the free alternatives - people
who aren't computerate enough to get one). But Word, in at least some
cut-down form, is now tending to be on all (Windows anyway) machines, so
that's less relevant than it used to be.

I like RTF and the old Wordpad. Quick, "easy" cut and paste and
save most of the formatting.

But feeping creaturism means that the simple program of old, gets
more and more features, until it is no longer quick to load, use or
.... some days's I'd like to have the source code for some of the 16
& 32 bit programs and just compile it for a 64 bit OS. "Not like the
old days!" if 640KB was good enough for Bill Gates --- what the hell
am I saying?
--
pyotr filipivich
Next month's Panel: Graft - Boon or blessing?

"Mayayana" on Wed, 19 Feb 2020 22:55:36
-0500 typed in alt.windows7.general the following:
"J. P. Gilliver (John)" wrote

| I use the free plugins Microsoft issued to allow Office 2003 to read *x
| files. (I agree that there's little _need_for .docx to exist - but sadly
| it's becoming increasingly "supported", in that I am receiving more and
| more of them.)

I don't see any point to .doc, either, in most cases.
Word processors are for writing business letters that
will be printed. I actually keep a script on my desktop
to convert doc to txt.

Cool. "Back in the day" I had a small program which stripped out
the unprintable characters. Where it is, I've no idea.

--
pyotr filipivich
Next month's Panel: Graft - Boon or blessing?

nospam on Thu, 20 Feb 2020 08:12:39 -0500
typed in alt.windows7.general the following:
In article , mechanic
wrote:


One day file extensions will be outdated and machines will work out
how to display the data from a file without the preconception
inherent in assigning a file extension.

that day was back in 1984 with the original macintosh and classic mac
os, which did not use file extensions.

mac os x, being based on unix, does use extensions.

As I recall from the late 80's, the "advantage" of Unix is it
treats everything as a file: keyboard, monitor, files, output. A file
might have a extension to let humans know what type of contents to
expect, but the system was perfectly "happy" to open a (text) file
with a (music program) file. Not going to say the result would be
good, mind you, but the system did no hand holding. type
rm -rf *
at the prompt and there's no "Do you mean to do that? Y/N" - your
files are gone.
OTOH, C let you add letters, so that was useful.
--
pyotr filipivich
Next month's Panel: Graft - Boon or blessing?

pyotr filipivich wrote:
"Mayayana" on Wed, 19 Feb 2020 22:55:36
-0500 typed in alt.windows7.general the following:
"J. P. Gilliver (John)" wrote

| I use the free plugins Microsoft issued to allow Office 2003 to read *x
| files. (I agree that there's little _need_for .docx to exist - but sadly
| it's becoming increasingly "supported", in that I am receiving more and
| more of them.)

I don't see any point to .doc, either, in most cases.
Word processors are for writing business letters that
will be printed. I actually keep a script on my desktop
to convert doc to txt.

Cool. "Back in the day" I had a small program which stripped out
the unprintable characters. Where it is, I've no idea.

https://stackoverflow.com/questions/...om-a-unix-file

I think Windows has a strings program too :-)

When using this to scan EXE files, set the -n to a larger value like
maybe 10 or 12 or something. There's a lot of 4 character junk in files which
the "default" value will pick up. Cranking the value of N a
tiny bit, really helps performance. I see this also has a Unicode
option (for Unicode-only I would guess).

https://docs.microsoft.com/en-us/sys...nloads/strings

Paul

"J. P. Gilliver (John)" wrote

| Interesting. (My old Turnpike can handle truly embedded-in-the-text
| images in email and news postings, but hardly any [I don't know of any]
| other mail/news clients can, so I've learnt to attach any images - or
| other attachments - at the end when sending. [Other clients _appear_ to
| embed images in text, but they actually put them at the end, putting a
| _link_ - such as "cid:" - in the text where the image is to go. Or, of
| course, these days, don't include the image at all, but a link to its
| location online.]) I hadn't known about the ability of HTML to embed
| images: is that HTML 6 or something?

It's been around for a long time, but it's not in big demand.
However, I think IE has only handled something like 32 KB.
Though the latest version might work better.

It goes like this:

IMG WIDTH=800 HEIGHT=64 SRC="data:image/gif;base64,xxxxxx" /

The xxxxxx is the base64 encoding. It can also be used in
CSS.

"J. P. Gilliver (John)" wrote

| I think you're going a bit too far in totally dismissing word processing
| (or dismissing it for private use).

I don't distinguish "private" use. I use it for my
own business receipts and contracts. I use it when
I need to print a page with formatting, logo, etc.
But for most things I don't need that. I have lots
of articles I've saved from online as TXT. I don't need
images. I don't need formatting. Just text, the same
it would be if I were reading in a newspaper or magazine.

And have you noticed the images with webpages
these days? They generally use free, stock images
that add nothing to the article.

| Presumably your script fails for some characters - I don't mean just
| accents, but for example where someone's used Wingdings - "J" for smiley
| (you probably hate those too), and ")" instead of "Tel:", for example.

What kind of nut would use wingdings?

In message , Mayayana
writes:

"J. P. Gilliver (John)" wrote
[]
| location online.]) I hadn't known about the ability of HTML to embed
| images: is that HTML 6 or something?

It's been around for a long time, but it's not in big demand.
However, I think IE has only handled something like 32 KB.
Though the latest version might work better.

It goes like this:

IMG WIDTH=800 HEIGHT=64 SRC="data:image/gif;base64,xxxxxx" /

The xxxxxx is the base64 encoding. It can also be used in
CSS.


So you mean the encoded image follows the comma (does it need a newline
after the comma?), and is followed by '" /' (without the '')?
--
J. P. Gilliver. UMRA: 1960/1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

Time is an illusion - lunchtime doubly so. (First series, fit the first.)

In message , Mayayana
writes:
[]
I don't distinguish "private" use. I use it for my
own business receipts and contracts. I use it when
I need to print a page with formatting, logo, etc.
But for most things I don't need that. I have lots
of articles I've saved from online as TXT. I don't need
images. I don't need formatting. Just text, the same
it would be if I were reading in a newspaper or magazine.

Newspapers and magazines use italic, fonts - even images, for that
matter! Even bold and underline, though granted not as much as WP
documents.

And have you noticed the images with webpages
these days? They generally use free, stock images
that add nothing to the article.

I don't disagree with that.

| Presumably your script fails for some characters - I don't mean just
| accents, but for example where someone's used Wingdings - "J" for smiley
| (you probably hate those too), and ")" instead of "Tel:", for example.

What kind of nut would use wingdings?

It's pretty universal these days - been with Windows since at least '95,
I think - so why not use it? It contains a lot of genuinely useful
symbols, and being TrueType they scale nicely. The example above is a
trivial one: using ")" in Wingdings will give a little telephone symbol,
which some people like to use as just a little light relief before their
telephone number.

--
J. P. Gilliver. UMRA: 1960/1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

Time is an illusion - lunchtime doubly so. (First series, fit the first.)

"J. P. Gilliver (John)" wrote

| IMG WIDTH=800 HEIGHT=64 SRC="data:image/gif;base64,xxxxxx" /
|
| The xxxxxx is the base64 encoding. It can also be used in
| CSS.
| So you mean the encoded image follows the comma (does it need a newline
| after the comma?), and is followed by '" /' (without the '')?

Exactly like above. No spaces within SRC. The 6 x's
represent the base-64 code. The CSS version is a bit
different, but not so relevant for most cases.

Here's the entire drag-drop .vbs script, in case anyone
wants it. It will handle JPG, GIF, PNG, putting the entire
IMG tag on the clipboard for pasting. The usual caveats
apply:

* Watch out for wordwrap. Unlike JS and HTML, VBS
recognizes line returns. They mark the end of a code
statement. So, for example, the long list of numbers
in the array for generating base-64 need to be all on
one line. Ditto for the long lines after MsgBox and any
comments that wrap.

* Drag-drop may run into restrictions on Vista+.

* God knows what Win10 will allow. For all I know
it may just pop up a message box that says,
"Buy this script at the Windows Store! And click
here to see a free preview of Frozen2, also available
at the Windows Store!! Need more toilet paper? We've
got it!!! At the Windows Store!!!"

'-- begin script -------------------

Dim ImInf, Arg, Ret


Arg = WScript.Arguments(0)
If Len(Arg) = 0 Then
MsgBox "Drop a JPG, GIF, or PNG onto script to get a data URI IMG tag.
The pre-prepared tag, with image converted to base64 encoding, will be put
on the Clipboard for pasting.", 64
WScript.Quit
End If


Set ImInf = New ImageInfo
Ret = ImInf.LoadImage(Arg)
If Ret 0 Then
MsgBox "IMG tag on clipboard. Total size of Base64 image is " &
ImInf.Size & " bytes. IE8 will clip an image at 32 bytes. Other browsers
have no realistic limit.", 64
Else
MsgBox "Invalid extension: " & ImInf.Extension
End If

Set ImInf = Nothing

'-------------------------------------------------------------------------------

Class ImageInfo

Private IE, TA, iWidth8, iHeight8, sExt8, Qt8, PPI, LenPic, FSO1

Public Property Get Width()
Width = iWidth8
End Property

Public Property Get Height()
Height = iHeight8
End Property

Public Property Get Extension()
Extension = sExt8
End Property

Public Property Get Size()
Size = LenPic
End Property


Public Function LoadImage(sImgPath)
Dim OPic, LRet
LoadImage = 0
iWidth8 = 0
iHeight8 = 0
On Error Resume Next
Pt8 = InStrRev(sImgPath, ".")
sExt8 = LCase(Right(sImgPath, len(sImgPath) - Pt8))
Select Case sExt8
Case "gif", "jpg", "jpeg"
Set OPic = LoadPicture(sImgPath)
iWidth8 = CInt((PPI * OPic.width) / 2540)
iHeight8 = CInt((PPI * OPic.height) / 2540)
Set OPic = Nothing
Case "png"
LRet = GetPNGSpecs(sImgPath)
If LRet = 0 Then
LoadImage = 0
Exit Function
End If
Case Else
LoadImage = 0 '-- not a web image.
Exit Function
End Select

SetIMGTag sImgPath
If iWidth8 0 Then LoadImage = 1
End Function

'-- LoadPicture doesn't handle PNG. fortunately, W/H are easy to get from
PNG.
Private Function GetPNGSpecs(ImgPath)
Dim TS1, s1, A2(23), i2
On Error Resume Next
Set TS1 = FSO1.OpenTextFile(ImgPath)
s1 = TS1.Read(24)
TS1.Close
Set TS1 = Nothing
For i2 = 1 to 24
A2(i2 - 1) = Asc(Mid(s1, i2, 1))
Next

If A2(0) 137 Then Exit Function
If A2(1) 80 Then Exit Function 'P
If A2(2) 78 Then Exit Function 'N
If A2(3) 71 Then Exit Function 'G

iWidth8 = (256 * A2(18)) + A2(19)
iHeight8 = (256 * A2(22)) + A2(23)
If iWidth8 0 Then GetPNGSpecs = 1

End Function

'------------------------------------------------------------------------------
Private Sub SetIMGTag(sImg)
Dim sTag, TS1, OFil1, LSz, sIn, sOut

If iWidth8 = 0 Then Exit Sub

sTag = "IMG WIDTH=" & CStr(iWidth8) & " HEIGHT=" & CStr(iHeight8)
sTag = sTag & " SRC=" & Qt8 & "data:image/" & sExt8 & ";base64,"

Set OFil1 = FSO1.GetFile(sImg)
LSz = OFil1.Size
Set OFil1 = Nothing

Set TS1 = FSO1.OpenTextFile(sImg)
sIn = TS1.Read(LSz)
TS1.Close
Set TS1 = Nothing

sOut = ConvertToBase64(sIn)
LenPic = Len(sOut)

sTag = sTag & sOut & Qt8 & " /"

'--put the IMG tag on the clipboard.
Set TA = IE.document.getElementById("T1")
TA.value = sTag
TA.Select
Set TA = Nothing
IE.ExecWB 12, 0
End Sub

Private Function ConvertToBase64(sBytes)
Dim B2(), B76(), ABytes(), ANums
Dim i1, i2, i3, LenA, NumReturns, sRet
On Error Resume Next
ANums = Array(65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78,
79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 97, 98, 99, 100, 101, 102,
103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117,
118, 119, 120, 121, 122, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 43, 47)

LenA = Len(sBytes)
'-- convert each string character to ASCII value.
ReDim ABytes(LenA - 1)
For i1 = 1 to LenA
ABytes(i1 - 1) = Asc(Mid(sBytes, i1, 1))
Next
'-- generate base 64 equivalent in array B2.
ReDim Preserve ABytes(((LenA - 1) \ 3) * 3 + 2)
ReDim Preserve B2((UBound(ABytes) \ 3) * 4 + 3)
i2 = 0
For i1 = 0 To (UBound(ABytes) - 1) Step 3
B2(i2) = ANums(ABytes(i1) \ 4)
i2 = i2 + 1
B2(i2) = ANums((ABytes(i1 + 1) \ 16) Or (ABytes(i1) And 3) * 16)
i2 = i2 + 1
B2(i2) = ANums((ABytes(i1 + 2) \ 64) Or (ABytes(i1 + 1) And 15)
* 4)
i2 = i2 + 1
B2(i2) = ANums(ABytes(i1 + 2) And 63)
i2 = i2 + 1
Next
For i1 = 1 To i1 - LenA
B2(UBound(B2) - i1 + 1) = 61 ' add = signs at end if
necessary.
Next

For i1 = 0 to UBound(B2)
B2(i1) = Chr(B2(i1))
Next
sRet = Join(B2, "")

ConvertToBase64 = sRet
End Function


Sub Class_Initialize()
Qt8 = Chr(34)
Set IE = WScript.CreateObject("InternetExplorer.Application ")
Set FSO1 = CreateObject("Scripting.FileSystemObject")
IE.silent = True
IE.Navigate "about:blank"
While(IE.ReadyState 4)
WScript.Sleep 20
Wend
IE.document.body.innerHTML = "TEXTAREA ID=" & Chr(34) & "T1" & Chr(34) &
" wrap=off/TEXTAREA"
PPI = IE.document.parentWindow.screen.logicalXDPI
End Sub

Sub Class_Terminate()
IE.Quit
Set IE = Nothing
Set FSO1 = Nothing
End Sub
End Class

"J. P. Gilliver (John)" wrote

| Newspapers and magazines use italic, fonts - even images, for that
| matter! Even bold and underline, though granted not as much as WP
| documents.

Yes. Save them if you find them useful. I don't. Most
webpages are so badly designed these days that I just
read articles with CSS disabled, anyway. 13px Verdana.
About the same that I use in Notepad. So reading the
article in Notepad is usually more comfortable for me.

| What kind of nut would use wingdings?
|
| It's pretty universal these days - been with Windows since at least '95,
| I think - so why not use it? It contains a lot of genuinely useful
| symbols, and being TrueType they scale nicely. The example above is a
| trivial one: using ")" in Wingdings will give a little telephone symbol,
| which some people like to use as just a little light relief before their
| telephone number.

You MS Office people are so parochial. Wingdings
came out in several versions, made by MS. For years, if
I remember correctly, MS Word got later versions than
Windows did. So Word users merrily sent their astrological
symbols to each other, having no idea it was a limited font.
And they sent them to Mac users, with no idea Mac users
couldn't see them.

I don't see a telephone or a ")" above. It wouldn't matter,
anyway. I'm reading in plain text and only see what Verdana
can print. That's an example of the problem with wingdings.
They're fine if you're going to print it out. But you can't
assume anything being universal online or in software. Apparently
you assumed here that I'm reading newsgroups as HTML.
But why would I do that?

For me the emojis are similar. People can't be bothered
to write 3 words but they expect me to look up a tiny
image that seems to show a crying face with a donut?
Or maybe that's a spare tire? Is that even a face? Maybe
it's a dripping orange with a spare tire? And when I finally
figure it out it will probably mean "Oy".

Emojis are for semi-literate teenagers sending texts. I
don't see most of them in my browser. I just get little
boxes with the hex code for the emoji. (Imagine here that
you see a face with tongue sticking out.

"Ken Blake" wrote

| I don't see any point to .doc, either, in most cases.
| Word processors are for writing business letters that
| will be printed. I actually keep a script on my desktop
| to convert doc to txt.
|
| Word processors are for writing *many* different kinds of documents, not
| just business letters.
|

Letters that need formatting and maybe graphics,
and will be printed. I never have occasion otherwise
to fire up a 400 MB monstrosity of bloat and complexity.

In my experience, people who use MS Office get
accustomed to it and become very profficient with Word.
As a result they tend to overuse it and not realize
that everyone isn't sitting in front of Word.

It starts itself at boot, so you don't see what a resource
hog it is. It's forced on college students, so they end up
thinking Word = Windows. It hides the location of files.
It encourages you to use it for email, even though it
creates monstrous HTML emails with nonsense tags
starting with "MSO", which are invisible except to
people reading their email in MS Office. It's a contained,
parochial world of MS Office functionality and convenience,
designed to be seamlessly usable for people who work in
offices doing word processing, but also designed in such a
way that someone can be a serviceable office worker
without knowing how to use their computer.

I get people sending me email from Word. People send me
notes as doc or docx. It's like dealing with people from
AOL or people using Macs. They don't actually know how
to use a computer because all they've ever used is MS
Word. So they don't realize that Word is not the same thing
as Windows, and that many people can't even open their
files. And those other people probably have no idea why
they can't open the files. So it does no good that Libre
Office or free MS Office readers are available.

(Lately I've noticed that Google Docs is supplanting
MS Word. People send links and expect me to sign up
with Google so that I can read their file. Huh?! They
don't understand how crazy that is. Signing up and signing
in with Google was effortless, so what's my problem?)

It's fine if you use Word a lot and like it. It's great if
what you're doing needs to be printed out. For just about
any other purpose it's like the old saying: If all you
have is a hammer then everything looks like a nail. I use
Libre Office when necessary, for things like work receipts.
But I would never send a receipt to a customer as doc or
docx or even RTF. There's no reason to expect they can
read that. Those are limited MS formats. I always convert
to PDF before sending.