[Haskell-cafe] Re: Hiding side effects in a data structure

[Jon Fairbairn]

Peter Hercek <peter at syncad.com> writes:

> Jon Fairbairn wrote:
>  > A hyperlink of the form <a
>> href="http://.../long-research-paper.html#interesting-paragraph">
>> interesting bit</a> is far more useful than one of the form
>> <a href="http://.../long-research-paper.pdf">look for
>> section 49.7.3</a>.  It may not seem significant, but when
>> one is attempting to learn some new part of Haskell it's
>> really off-putting.
>
> Pdfs are not that bad.

No, they (or at least links to them) typically are that bad!
Mind you, as far as fragment identification is concerned, so
are a lot of html pages.  But even if the links do have
fragment ids, pdfs still impose a significant overhead: I
don't want stuff swapped out just so that I can run a pdf
viewer; a web browser uses up enough resources as it is. And
will Hoogle link into pdfs?

> The above definitely works OK on windows, not sure about linux
>  pdf viewers.

Works perfectly on my Fedora 7 systems.

While this would be a definite improvement over having to
search through the pdf, the delay and the fact that pdfs
aren't as good as html for on-line viewing are still enough
of an overhead that it's discouraging. If I'm using PHP (an
execrable language), I can type the name (or something like
the name) of any function into the search box on the PHP
manual webpage and get useful (albeit often extremely
irritating from a Haskell programmer's point of view)
results straight back.  Even including my language
designer's distaste for PHP, this can make writing a wee bit
of PHP a less onerous event than writing the same thing in
Haskell -- definitely not what we want!



-- 
Jón Fairbairn                                 Jon.Fairbairn at cl.cam.ac.uk