Re: A niche for the Hurd

[olafBuddenhagen]

Hi,

On Wed, Oct 22, 2008 at 12:20:07PM +0200, Arne Babenhauserheide wrote:
> Am Mittwoch 22 Oktober 2008 02:50:32 schrieb olafBuddenhagen@gmx.net:

> > http://sourceforge.net/mailarchive/message.php?msg_name=20080909073154.GB82
> >1%40alien.local
> >
> > (Just skip the text mode browsing stuff -- the relevant part is the
> > long monologue at the end... I really should put these ideas into my
> > blog.)
> 
> Definitely! 

Well, yes, that was rather a rheroric "should" :-)

There are so many many things I want to write about (the ones in the
above message only being a tiny fraction of them) -- but it always takes
me so long to write my ideas up... The whole reason I started the blog
was hope that I could *quickly* dump some of my ideas -- and on that
score it has been a total failure so far :-(

> Maybe you could also add it to the Blogs in the Hurd wiki, or a link
> to your Hurd related entries, so people can find them via the Hurd
> wiki? 

Hm... How would one go about linking individual entries?

Blogger.com does have a tagging feature, which I haven't used so far...
I wonder whether that could be used to automatically pull only those
entries tagged as Hurd-related.

> Netrik sounds very nice, too, and works quite well (I just installed
> and tested it in my qemu Hurd and on my Gentoo machine).

Does it? ;-)

I obviously use it quite a lot myself, and while I love the UI and the
colors, I must admit that I find it quite painful in many other
regards... There is just too much important functionality missing still
:-(

> It can do most umlauts in my Gentoo machine (äöü), but still seems to
> stumble on the "scharf s" (ß). 

Well, this is actually a bit more complicated: The truth is that at the
time I did most work on netrik, utf8 was not yet used very widely. As a
result, netrik always assumes one-byte character sets (iso-8859-* and
similar), and does no character set translation at all -- it only really
works for viewing iso8859 sites in an iso8859 environment... utf8 sites
in iso environment, or iso sites in utf8 environment, is total failure.

utf8 sites in utf8 environment (the most likely situation nowadays)
almost work (thanks to utf8 being mostly transparent), but not quite: As
netrik is not aware of the multi-byte sequences, it miscalculates screen
sizes. Lines being wrapped too early is not really a problem by itself,
but as a result old lines are not completely overpainted when scrolling,
resulting in ugly stray characters often hovering near the line end.
Even worse, link positions do not match, resulting in really really ugly
mis-highlighting...

Last but not least, non-ASCII character entity references (ß etc.)
are broken in netrik on utf8 environments. The result is that on some
sites you get a mix of working umlauts (those encoded in the source
directly), and non-working ones (those represented through entities) --
this is probably what you are seeing there.

Anyways, this is all rather off-topic here...

-antrik-