Posted by Albert Wiersch on 05/08/06 01:24

"Jukka K. Korpela" <jkorpela@cs.tut.fi> wrote in message
news:Xns97BDA733CB3jkorpelacstutfi@193.229.4.246...
> "Albert Wiersch" <donotreply@123donotreply123.com> wrote:
>
>> And what is wrong with notifying web developers, should they choose, that
>> the typeface they requested the browser to display may not be available
>> on a significant number of systems?
>
> If they wanted to know that, there would surely be better sources than a
> subjective checker that claims to be a validator to sell better,
> especially
> when the said checker's author apparently has no idea of the real problems
> with fonts.

It's a validator because it is in the common definition of the word, not
because it sells better. It's the type of program 99% of people want when
they want an HTML validator.

> It would be foolish to report rarely available fonts in a list
> appearing as the value of font-family as being erroneous or dubitable. The
> real problems arise when a suggested font _is_ available...

So you would prefer to ignore the issue when the developer prefers a certain
typeface to be used but is not generally available. When the developer does
specify a typeface that is available, you want to warn them of the problems
of changing typefaces. Feel free to enlighten us as to the problems of
changing typefaces that are common enough to warrant alerting a developer
to. I'd be happy to alert a developer to any issue that I can that is worth
alerting them to (this does not include alerting the developer to very
obscure issues when the alerts are likely to cause more trouble and
confusion than they solve).

> We've seen your advertizing before, and we've seen that you don't have
> much
> to say to arguments that point out that you have just invented "problems"
> and
> then proudly present your product as a valuable tool for finding them.

I don't invent problems. The program points out issues that are of potential
concern to the developer. Naturally, some issues are more serious than
others... but this doesn't mean the issues are "invented".

> To confuse people more, you list them in the midst of real problems and
> their
> detection which _would_ be useful _if_ your checker distinguished between
> facts and your fiction. Given that your checker warns about harmless
> spaces
> and makes a noise about meta description tags as if they were needed, its
> potentially useful messages lose their value.

What harmless spaces? Even if the spaces are harmless in some regards,
perhaps the developer wants to know about them because they want their
documents to adhere to a certain style... and such checking for "harmless"
issues can generally be turned off if the developer chooses. As for meta
description tags, CSE HTML Validator doesn't say anything about them being
needed. It does, however, recommend them. They are not worthless.

> Besides, presenting the
> checking of URL syntax (which checking is almost surely flawed - URL
> syntax
> is very tricky to handle properly) as a great benefit is disproprotionate;
> it's of little use to check the syntax only, when there are quite useful
> (and
> free) _link checkers_.

What URL syntax checking are you talking about? CSE HTML Validator Std/Pro
does do actual link checking. In addition, it does perform useful checks on
URLs.

Albert

[Back to original message]