[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Sheflug] HTML mail (was Re: Linux)



On Fri, 2002-09-20 at 17:25, Will Newton wrote:
> Writing in block paragraphs is:
> a) Good practice in English
> b) A lot easier than learning and using HTML correctly.

Excellent point. Made in list format, not block :P And a good job it
didn't include any one of '> ', ': ', '|', 'will:', or any of the other
'quote' mechanisms, because my e-mail client could have reflowed it in
anyway.
a) I've often seen points start like this, continuing onto the 
next line b) but having the second point broken in :(

> > - occasionally people insert stuff other than text, such as 'ascii art'
> > in an attempt to work around the limitations of ascii.
> 
> HTML does not solve this, unless you mean adding image links which is another 
> can of worms.

I mean: adding lists, right-formatting, perfect-justification,
indenting, quoting, bullet points, etc. Anything basic rich text
controls enable you to do - bold, italic, etc.

> No client will ever please all of the people all of the time, it's often an 
> aesthetic judgement. But to call a quoting algorithm a hack in the same 
> breath as HTML is perhaps over-stating the case.

Not at all. In HTML, "<p>This broken\n paragraph</p>" is exactly
equivilent to "<p>This\n broken paragraph</p>". In text, the two are
different. Text wrapping and quoting breaks layout, and makes e-mails
hard to read.

> Cutting and pasting is OK if you have rectangles a-la emacs. :) It's what 
> tabs were invented for.

Cut out every cell individually - mmmm, yummy :) Do you mind if I send
you some text spreadsheets I need in table cells? 

> > Etc. Highlighting, track changes. You can't do that with text.
> 
> Neither can you do it properly with HTML. You can use colouring like that, 
> but what if author A wants to use red for his headings etc. You also do not 
> have any versioning, and I would not want to diff HTML. Again a case of HTML 
> being shoe-horned into a job it isn't really up to.

You've just missed the point. Sure, for true collaboration you might
want authorship and versioning, but people having to put up with text
e-mail clients will quite happily use the digital equivilent of
highlighting markers to make their changes. Works on paper, works in
HTML, utterly impossible in text.

> Well why use HTML if <p> and <br> will do it? ASCII 10 will do both their 
> jobs just fine.

Because <p> and <br> are structure, \r and \n are content. See above
broken paragraph example. You can wrap HTML exactly how you please and
it still works, wonderful.

> Links are a good thing, but I think you will find it was inflicted by the 
> invention of the printing press more than anything else.

Quite possibly. But the point is, we can do better :)

> remains, HTML is not a great leap for HTML. It adds a big heap of cruft that 
> is perhaps not necessary, and not much that would be really useful.

You call it unnecessary, yet people still use it and go even further -
emailing word documents around, all sorts. All because the basic e-mail
client is not capable enough for the range of expression people need. 

> Cookies? In email? I think I need a lie down. Frames? Aaargh! :)

Again, you've missed the point. You called the HTML parser complex - I
pointed out a fully fleged browser fits on a single density floppy. A
parsing library for a text-based e-mail client need only be a few Kb.

> I would have far less problems with HTML if everyone used XHTML

Agreed. The majority of the problem is due to poor implementations.

> > That's an argument against proprietary monopolies, not a technical
> > argument against a format.
> 
> This particular format is IMO beyond redemption though. :)

Bah. Anyone can be saved :) 

> > Granted, but then, that is the nature of structured documents.
> 
> I would contend that some are easier than others. e.g. I can quite easily 
> read a lot of LaTeX, but troff rots the brain.

The number of people who want to read the source are a complete minority
however. People interested in how precisely efficient the markup is are
dull bores, penny-wise pound-foolish. For the majority of users, this
isn't an issue. 

> HTML is a big ugly beast to parse without introducing some form of insecurity 
> such as a buffer overflow.

Cadangers. You can parse it with an event-firing state machine.
Javascript problems, yes, privacy problems yes, these are all
implementation issues. To say HTML causes buffer overflows is nonsense.

Of course, we're not arguing about Javascript though - the DOM is the
business of the implementation. 

Cheers,

Alex 'Flamewar' Hudson.

Attachment: signature.asc
Description: This is a digitally signed message part