An experienced user doesn't need a change to the defaults. This is just to make it easier for the inexperienced user. But there's worse: in the draft HTML5 specification you can read:

No client-side scripting is needed in many cases, though an API is available so that scripts can augment the user experience or use forms for purposes other than submitting data to a server.

Clearly this use is based on advertising copy, but why let buzzword merchants determine the course of language evolution? You'd get the impression (or is that “experience”?) that there are no experienced people left in the industry. Maybe you would be right.

... Latin exportāre, < ex- out + portāre to carry.

†1. trans. (gen.) To carry (things or persons) out of a place; to take away, carry off. Also fig. Obs.

2. a. Comm. To send out (commodities of any kind) from one country to another.

But that's an old, worn-out magic word. Nowadays it appears to mean “process with software”. DxO Optics “Pro” has recently replaced the term “process” with “export”. That gives me the impression that the images will end up elsewhere, but by default “export to disk” writes back to the same directoryfolder where the input image resides. What does that have to do with exporting?

The real issue isn't even that it's a gratuitous change in terminology. Producing output images from DxO isn't primarily a question of where they end up: it's the processing that is involved. That's particularly evident with DxO, since it can take over 10 minutes to produce a single image. That's clearly (inefficient?) processing, not moving data from one place to another. So it's just plain misleading to call it “exporting”.

But in the Microsoft space, it's used to mean directory, a list of file names and information about where to find the corresponding file.

The term folder gives the incorrect impression that the files to which it refers are somehow physically contained in the folder. In most file systems, this isn't the case, and in UNIX-related file systems there's no requirement that the file be referenced in only a single directory, or that only a single reference to the file be present in any directory. The term directory means something akin to a telephone directory: entries consist of a name and a number, and sometimes additional information. There can be multiple entries for the same telephone number, and numbers can be in multiple directories—just like in a file system directory. Nobody would imagine that the telephones or the people they connect to are physically located inside the phone book, but the term folder encourages this assumption on the part of files.

As if this wasn't bad enough in the first place, consider the issues of different languages. From time to time I do translation work between English and German. German presents a cultural problem that they hardly use folders—they use hard-covered files instead:

The word for this kind of file is „Ordner“—and that's what Microsoft call directories, although the correct translation is “file”. And that's supposed to make it easier to understand?

This term should really be historical: formatting is just not needed any more except for floppy disks, which should also be considered historical. But now people apply it to a completely different function, that of setting up file system structures. There's a slight justification for this in the Microsoft world, since they never had a different term, but it's creeping into the UNIX world too, where it's unnecessary: there's already a good description for it, either “creating a file system” or the name of the program, such as newfs.

The problem is, the alternative terms (“create a file system”) are a bit of a mouthful. In addition, you can build file systems on media that don't need any kind of traditional formatting, such as flash memory. So I'm gradually coming to accept this change of meaning.

How did we get into this mess? Over the years, specific names have been assigned to punctuation marks. The term “slash” of the character / has been around for decades. More recently, at least in the places I've been, the reverse slash (\) was introduced and given the name backslash. I've never seen it on a typewriter, so I assume it was introduced some time in the mid-20th century for data processing purposes, along with a number of other strange and now obsolete characters such as the not sign (¬) and the “currrency” sign (¤).

One of the more unfortunate developments with Microsoft was that early on, before they had subdirectories, they chose the slash as an option delimiter in their command line processor. UNIX uses the hyphen for this purpose; thus, to list the contents of a directory, you might use the command ls, but if you want a detailed listing, you'll add the -l option. In the very early days, Microsoft chose the slash for this purpose; thus the command to list a directory sorted alphabetically would be DIR /O, or even DIR/O.

Later, Microsoft discovered the concept of multiple directories per file system, and they had a problem. They acknowledged that they wanted something UNIX-like, but the slash had been used for the option character, and they had also allowed it to be a field delimiter.

At this point it would still have been possible for them to change things, but they didn't; instead, they used the backslash for directory path delimiters. This has caused enormous problems ever since:

• They later adopted the C programming language, which uses the backslash to represent special characters. For example, \n represents the end-of-line character, \r represents a carriage return character, \t represents the tab character, and—very importantly—\\ represents the backslash character.
• The UNIX shells use the backslash in a similar manner, so you might see something like this:

  === grog@teevee (/dev/ttyp4) ~ 24 -> ls -l Mail\backup
  ls: Mailbackup: No such file or directory
  === grog@teevee (/dev/ttyp4) ~ 25 -> ls -l Mail/backup
  -rw-------  1 grog  lemis  450 Jul  7  2005 Mail/backup
  

  It's not just that the \ isn't recognized: it disappears altogether. To represent a backslash, you need to use two backslashes:

  === grog@teevee (/dev/ttyp4) ~ 28 -> ls -l Mail\\backup
  -rw-r--r--  1 grog  wheel  0 Oct  9 11:26 Mail\backup
  

• Finally, of course, there's the question of URLs. The web grew up in the UNIX space, and so the delimiters are slashes. But some browsers encourage people to get it wrong. Years ago I wrote a web page why\backslashes\are\not\slashes.html. This should link to the same page as http://www.lemis.com/why%5Cbackslashes%5Care%5Cnot%5Cslashes.html, but broken browsers take you to http://www.lemis.com/why/backslashes/are/not/slashes.html.

The Internet is a worldwide, publicly accessible series of interconnected computer networks that transmit data by packet switching using the standard Internet Protocol (IP). It is a “network of networks”...

Or do they?

Microsoft obviously has a different opinion: on the start menu of Microsoft “Windows” XP I can see it:

Image title: start menu Dimensions: 387 x 229, 18 kB Dimensions of original: 387 x 229, 18 kB Display this image: thumbnail    hidden   alone on page Display all images on this page as: thumbnails    this size Show for Saturday, 26 January 2008: thumbnails    small images    diary entry

firefox is the Internet? It's a web browser! And that's the problem: for many uneducated people, the world-wide web is the Internet. But there's no reason to take the least common denominator. Clear terms are important for understanding. Using the word “Internet” to describe the web is like using the term “telephone system” for describing an telephone-based information system. No wonder people are confused.

This (lack of) thinking seems to pervade Microsoft. You can see more evidence in the “Control Panel”:

Image title: control panel Dimensions: 359 x 215, 10 kB Dimensions of original: 359 x 215, 10 kB Display this image: thumbnail    hidden   alone on page Display all images on this page as: thumbnails    this size Show for Saturday, 26 January 2008: thumbnails    small images    diary entry

Here there are separate panels for the “Internet” and the network. This one probably occurred because of Microsoft's proprietary protocols, although I think most of them now run over the Internet protocols.

But that's only part of the story: the “Internet Options” appear all related to web browsing (what on earth is a “Temporary Internet File”?). If I want to set up Internet parameters on a Microsoft box, I can begin my search of the maze of twisty little menus, all different, at the “Network Connections” box, where they call the Internet protocols “TCP/IP”.

And, of course, Microsoft's web browser is called “Internet Explorer” (a term I'd be more likely to apply to things like nmap and wireshark). I've commented elsewhere about Microsoft's abuse of simple terms like “Explorer”.

To my immense disgust, this usage is creeping into the Unix space. In May 2016 I installed a Linux distribution that includes an “Internet” configuration page that only refers to a web browser. Network configuration is elsewhere.

How can that happen? Ignorance, as usual. In the ADSL days it became customary to have a central network component including an ADSL modem, a switch, NAT, firewall, and probably 802.11 access point. So they called it a modem. Now the successor devices no longer have the ADSL modem component, but the name has become established, and though it's no longer a modem, that's what they call it. I'm reminded of the smile of the Cheshire Cat.

A good example of the confusion this can create comes from Aussie Broadband support. On 7 April 2024 I had two mail messages from support personnel while trying to debug a problem. The first message referred to a domestic router as “modem”, and I replied:

On Sunday,  7 April 2024 at 14:57:34 +1000, Greg 'groggy' Lehey wrote:
> On Sunday,  7 April 2024 at 13:58:33 +1000, Aussie Accounts wrote:
>
>> Unfortunately we are unable to access the modem as our logs are show a
>> connection to a dell devices at the moment.
>
> Potentially the Dell device you see is the computer behind the router.
> There is no modem.  I haven't checked personally, but what I
> understand is:
>
> 1.  NTD connected to a Netcomm router (via Ethernet, of course).
> 2.  Router connected to a Microsoft computer and two phones.

Another support person read this exchange and responded:

In a fixed wireless setup, the NBN box/NTD/Network terminal device, often gets
called the modem.

What an excellent example of the confusion that sloppy terminology can cause!

The problem is that this word already has a meaning in this connection. In traditional operating system technology, open means “make the file available for processing”. This is a confusing term for the non-technical, since it relates to the way the operating system handles data. The FreeBSD man page open(2) doesn't help:

open -- open or create a file for reading or writing

What open really entails depends on the operating system, but typically it means:

• Look up the specified file in the directory and find its metadata (in UNIX this is an inode). Fail if it doesn't exist (Unix error is ENOENT). Unix open has the option to create the file if it doesn't exist.
• Establish whether the user can access the file in the manner requested. Fail if he may not (UNIX error EPERM).
• Set up data structures in the operating system to allow access to the data in the file.
• Return some information to the user which he can use to actually access the file. In UNIX, this is a small positive number, an index in the file table.

And that's all. No data is transferred to or from the file itself during the open process.

So open is a technical term. Why not let the non-technical use it for a different purpose? That's the way they built the Tower of Babel. There's already a perfectly good word, and one that is more descriptive. Why replace it with gobbledegook?

As if the situation wasn't bad enough, the term seems to have acquired further meanings beyond mail processing. It seems to be in general use for any action of displaying on the screen, and it seems that Apple, at any rate, uses “open” to mean “use”. If the file is a document, “open” means display it. If it's an “Application” (i.e. a program), “open” appears to mean “execute”. I think there's another meaning there too, but I forget.

Ask that same person about “free software”, and he won't have these problems. But the terms aren't interchangeable: free software is what it says. What's Open Source? I've come up with a lot of possibilities:

1. Free software
2. A community movement.
3. Carrying on from there, an attitude to society.
4. A way of developing software.

Richard Stallman has more extreme views, of course: “Open Source” is a way of developing software without conferring all freedoms.

About the only thing that most people don't consider is the availability of the source code—but that's where the name comes from. The proponents of the term had a number of reasons for it, mainly to impress commercial people who would take exception to words like “free”, but that doesn't seem to have worked. So, while I tolerate the term (and belong to the “community”), I don't use it.

People have told me that there's a movement afoot to retrofit the expansion “Programmable Video Recorder”. But then, nearly all VCRs are programmable, and that would just add to the confusion. What's wrong with “CVR” (“Computer Video Recorder”)? The name “DVR” (“Digital Video Recorder”) also suggests itself, but nowadays just about all video recorders are digital.

But on 2 December 2022 I found another abuse of the term: export is an old, worn-out bad word, so Luminar Neo calls it “share” instead, at least some of the time. Potentially the words don't have exactly the same meaning—I've seen Luminar use “export” too—but I don't know what else they might mean with the term.

I find this incredibly ugly. I suspect that my biggest objection is that it's a very narrow view of music, from the perspective of somebody who maybe really does listen mainly to songs. But the vast majority of pieces of music are not songs. Looking at Apple's indecipherable iTunes window, I find it almost nauseating to see individual movements of Mozart's wind concerti referred to as “songs”. It's even not appropriate for things like the individual parts of a song cycle like “Die schöne Müllerin”.

Yes, “piece of music” is ugly too, and far too long to pronounce. But what's wrong with the old word “track”? Other words have extended their meaning beyond the original. “Song” is doing it now, but by riding rough-shod over other terms.

Image title: DxO crash 1 detail 2 Complete exposure details Dimensions: 474 x 164, 8 kB Dimensions of original: 474 x 164, 8 kB Display this image: thumbnail    hidden   alone on page Display all images on this page as: thumbnails    this size Show for Saturday, 27 May 2017: thumbnails    small images    diary entry

Now doesn't that give you the warm, fuzzy feeling that things are under control? If so, let “Windows” check on-line for a solution. Then try to find a cause.

Finally most systems are coming to use the domain name system for accessing computers, though it has taken a long time, and for reasons I don't understand many Microsoft ports of UNIX programs, such as route, have been emasculated so that you can only give them IP addresses, and not DNS names. In the meantime, the average literacy level of users and registrars alone has dropped to alarming depths, and many people use the term web address when they mean “domain name”. Some domain name registrars even do so on their web sites, though the practice seems to be diminishing. That's all the more surprising when you think that people tend to call the web “the Internet”.

This is a calculated marketing move, of course, and they've even trademarked it, which says little good about the US legal system. I've written a separate rant about this one, so I won't go into more detail here.