Jun 292015

But what about new dogs and old tricks?

A few days ago I saw a request for help on a usenet discussion group. The person was asking a question on behalf of someone else, who was using a free webmail service and wanted to be able to download emails from the webmail service to a directory on their computer.

In reply, I suggested the use of an email client, such as Thunderbird, which – if the free webmail service allows it (and in the cases mentioned in the request, they do) – can download the email using POP3. That stands for Post Office Protocol 3, and is a method of retrieving email from a mail server and storing it on your computer that has been around since 1988 (with its predecessors, POP and POP2 going back few years earlier).

I also mentioned IMAP as alternative method – though also using the same software. IMAP stands for Internet (or Interactive, depending what you read) Message Access Protocol, and is another way to access messages from an email service using software on your computer. The difference between POP3 and IMAP, however, is that with POP3 the email is usually downloaded from the server and stored on the computer, while with IMAP the email remains on the server, allowing it to be accessed from multiple computers and other devices. IMAP’s origins date back to 1986.

That means both predate webmail by a good few years – which (as its name suggests) is email accessed via a website. Although the web’s origins go back a little further, the basic web that we know and love today was invented in 1989, and it only really began to form as the 1990s progressed, and started to take off from the mid-1990s onwards. The first webmail system appeared in 1993 – and that was only a test system.

The use of an email client to access emails doesn’t quite fit the description given in the original request – which was to download and store the email to a directory (of their choosing) on the computer. With an email client such as Thunderbird (or another example is Messenger Pro, which I currently use), when using POP3 the email is indeed downloaded to the computer – but where it is stored on the computer isn’t usually chosen by the user, although most email software will allow a directory structure to be created within its store.

All of which means that the original request, and the reply I gave, amused me – a point I expressed in my reply by saying:

Ahem! These newly invented protocols and the innovative concept of email clients running on a user’s computer allow webmail to be used in a much more practical way! 🙂

I strongly suspect the person in question has only ever known about webmail, so the request specified a directory of their choosing because they didn’t know that a perfectly good method of handling email – downloading it, storing it on the computer, and accessing it in a meaningful way, using software designed for the purpose – already existed, and was around long before webmail was ever a thing.

The person was looking for a wheel, not knowing the wheel had already been invented.

In a different context, I’ve seen a similar wheel reinvention occur in the way some people reply to emails.

I’ve been online since the mid-1990s, and back then a basic netiquette (network etiquette) was already established and encouraged, covering various areas. The area that matters here involves quoting – including some content from the email being replied to, in order to provide context.

The basic notion was that given an email that said this:

We need to arrange a meeting ASAP. I can do 11:00AM on
Wednesday the 1st, or 11:30AM on Thursday the 2nd. Are
either of those okay for you?

The reply might look like this:

> We need to arrange a meeting ASAP.
> [...]

Thursday at 11:30 works for me. See you then.

The text that has been quoted in the reply to provide context is marked with an angle bracket – > – at the start of each line, and the reply follows it – that works because, in this case, the message was a very simple one.

You’ll also notice that the quoted text is a different colour to the reply. All those years ago, that wouldn’t have happened – all the text would be the same colour, and it was merely the presence of the angle bracket that indicated quoting – but these days, decent email software automatically identifies quoted text (by spotting lines that begin with an angle bracket, as well as a few other commonly used characters) and colours it.

And the other thing you’ll notice is that not all of the original message has been quoted – it has been snipped (indicated by the […] – which isn’t crucial, but it is polite). Enough remains to provide context.

For something a little more complicated, the reply should be interleaved. For example, the reply to this:

I have some further questions for you. Question 1?
Question 2? Question 3? Question 4? Question 5?

Should ideally look like this:

> [...]
> Question 1?


> Question 2?


> Question 3?


> Question 4?


> Question 5?


With the reply taking that form, each answer is provided immediately below the question it relates to – the question is providing the context to the response that immediately follows it.

As I said, when I first got online some twenty years ago, this was the approach to email that was already established and widely adopted. It’s practical, and it works.

As more and more people came online, however, and more and more software was made available to handle email – not least including webmail interfaces as discussed above – basics such as the above largely became forgotten, except amongst people like me; people who have been online a long time and learnt the old ways – and who, in the main, still use them!

These days, what many people do is ‘top post’ – that is to say the reply is put above the message being replied to. For the first example given above, the reply would look like this:

Thursday at 11:30 works for me. See you then.

> We need to arrange a meeting ASAP. I can do 11:00AM on
> Wednesday the 1st, or 11:30AM on Thursday the 2nd. Are
> either of those okay for you?

That’s straightforward enough – not a problem, you might think. However, now consider the second example, the reply to which would look like this:

Answer. Answer. Answer. Answer. Answer.

> I have some further questions for you. Question 1?
> Question 2? Question 3? Question 4? Question 5?

There are three things to note here.

  1. The original messages in both cases are now quoted in full – whereas in the versions given above,  some of the original message has been snipped (and indicated). This is only a minor issue, but it’s worth pointing out: The reply only really needs to provide context – the specific points that are being answered, which isn’t necessarily the whole email.
  2. While the questions are still quoted – the context is provided – they are all quoted together, below the answers, which are all given together, at the top. The recipient of that email has to keep looking down at the quoted text in order to remind themselves of the question/context (which might involve scrolling down) before looking back up at the answers (which might involve scrolling back up). For anything complicated, this can be very messy and can lead to misunderstandings.
  3. You’ll notice that the quoted text is no longer coloured – that’s because another relevant factor is the shift from using plain text for email (something else I still do) towards using HTML instead – which almost seems to have happened alongside the shift from interleaved replies to top posting. With HTML-formatted emails, colouring of the quoted text isn’t controlled by the reader’s software, but by the writer’s software: If you reply to my emails using HTML, I only see colour if you include it – which might be how you have your software set up, or it might not. And that’s only if I’m allowing my software to display HTML formatted emails, which is usually not the case.

In my profession, I sometimes send fairly long emails containing a lot of questions – and some of those individual questions themselves might require more than a couple of sentences in answer.

And most of the people to whom I send such emails have a tendency to top post. Like the person mentioned above, who has probably only ever known about webmail, it’s very likely that top posting is the only method these people have ever known.

However, over the last couple of years I’ve noticed an interesting phenomenon. A few of my clients have sent me replies that (using the multiple question example above) look like this:

My answers are in red below.

I have some further questions for you.

Question 1? Answer.

Question 2? Answer.

Question 3? Answer.

Question 4? Answer.

Question 5? Answer.

Note: Given that they’re using HTML email, the font isn’t a fixed width one as used above – that’s simply the effect of enclosing the example in <pre> and </pre> tags to highlight it.

What seems to be happening here is that people who have been brought up on top posted replies, and the use of HTML email, are suddenly realising that their normal method of replying is flawed – it doesn’t work when trying to reply to something complicated – and concluding that the way to do it is to interleave their reply with the original message, so that their answers are given context. And they’re adding colour to make their interleaved replies stand out from what they’re answering.

They’re trying to reinvent a wheel that was already well established – because it worked – over twenty years ago. Just like the person in the webmail example above appeared to be looking for a wheel that they didn’t know had already been invented.

These are new dogs trying to (re)invent old tricks.