I've noticed that if the ADSL line goes down NetSurf can't even access
a local web site served by WebJames (http://localhost/). The request
just times out. As soon as the link comes back up it's OK. This isn't
a problem with Oregano. I think it was OK when I specified the initial
page as a file. Could this be a "feature" in NetSurf or is it a
|_|. _ Richard Porter http://www.minijem.plus.com/
Just another quick reminder that the server that handles
netsurf-browser.org and the email addresses within the same domain will
be down for maintenance and a data centre move from tomorrow afternoon
to Monday morning. Any email sent to netsurf-brower.org addresses will
be dealt with by our secondary mail servers and not be lost, but they
won't be delivered until Monday.
Using version of 20th April (& 15th April) on Iyonix + OS 5.13 :
Selecting Google > Images > Search for e.g. "hat" > Click on first image >
Click on "See full-size image" at top of page.
Netsurf error appears "Netsurf has detected a serious error and must
This is repeatable, but not on all images.
I tried to send in a bug report but only managed to get "failed creating
/____Mail from John(a)JohnWoodhouse.plus.com
. . . using RISCOS 5.13 on an XScale powered Iyonix computer.
I occasionally encounter files of one filetype, named as if they were
another. For example,
is a .gif with a jpg extension, as are pictures 1-9 in that series. I have
also seen Windows BMPs filetyped as PNG, I think it was. I've complained
about the BBC one; although perhaps they would take more notice if I sent
them a .cmd file filetyped as a .jpg? I've never experimented with how
Windows behaves in those circumstances, but I know from the odd behaviour I
seen with .exes renamed as .bats and the like that its filetype handling is
. . . well, about as well implemented as the rest of the OS, frankly. :-/
Netsurf refuses to display these images, but also makes it difficult to save
them so that they can be retyped appropriately. Not a bug in Netsurf,
really, but are there any plans afoot to work around web sites that
misbehave in this particular way? I'm not sure I actually would want it to
just silently determine the correct filetype and then display it, but it
would be nice if one could at least save the file locally to examine it.
I'd put in a feature request, but I'm not really sure what the preferred way
to handle this situation would be.
Have you heard of the International Obfuscated Perl Code Competition? Each
contestant is shown ten Perl programs - five that have been written normally
and five that have been deliberately obfuscated. In order to win, you just
have to decide which is which.
Here's a summary of changes in the last two weeks:
+ Process form elements with display: none
+ Support insertion into textbox after optional trailing space
+ Fix click area for scaled down images used as links
+ Fix corruption of history data structure on cloning
+ Fix cookies treeview crash
+ Set minimum META refresh delay to 1s
+ Support full link target recommendation
+ Update User Agent string (updated UnixLib)
+ Squash memory leak
+ Improve stability
+ Many bug fixes and tidy-ups
Please test the latest version
Bugs should be reported to the bug tracker
Michael Drake (tlsa)
I connect my RISC OS network via an IPCop box¹ using dial-up.
The IPCop pages as displayed by NetSurf have continued to be better and
better displayed as layout improvements have been made.
Recently - and it may have coincided with recent work improving the
handling and 'accommodation' of duff meta tags - the IPCop 'base' page,
which has a meta refresh tag of the form:
<meta http-equiv='refresh' content='30;'>
seems to try and reload a page called
(isn't this drag&drop of URLs from the bar wonderful!).
I was wondering if the recent 'liberalisation' regarding metatags could
have broken the refresh handling.
If you have a page you visit with a refresh tag, could you try it and
report back before I raise a bug report?
¹ written up at: http://homepage.ntlworld.com/jrwmail/IPCop/
|| John Williams || Wirral, Merseyside, UK
Who is John Williams? http://www.picindex.info/author/
What is the correct behaviour when no content type is returned by a
site for a page? Should the browser default to text or HTML?
streetmap.co.uk returns a correct type of text/html for its title
page, but entering a search results in a page with no content type
which NetSurf renders as text. The same if a map page URL is dragged
in from another browser.
*curl -I http://www.streetmap.co.uk/
HTTP/1.1 200 OK
Expires: Wed, 11 Apr 2007 17:23:16 GMT
Date: Wed, 11 Apr 2007 16:53:16 GMT
Last-Modified: Mon, 02 Apr 2007 12:42:25 GMT
*curl -I http://www.streetmap.co.uk/newsearch.srf?
HTTP/1.1 200 OK
Date: Wed, 11 Apr 2007 19:32:26 GMT
Of course you'll have spotted its Microsoft shite.
David J. Ruck Phone: +44- (0)7974 108301 Email: druck(a)druck.org.uk
On the ROOL site, new tickets in the bug tracker are entered using a
hidden field carrying the author name. Unfortunately, attempting to add a
bug with NetSurf causes an error from the Web server. The server debug
output seems to indicate that the hidden field data was not sent at all,
or perhaps got sent in a malformed way that caused it to be discarded.
Until recently the page had a few errors that caused it to fail
validation - it was using ">" instead of " />" or an explicit closing
tag, to close an element definition in an XHTML document. The problem
remains despite fixing the page errors.
If you are logged in to the site you can attempt to create a ticket using
the following address:
Is there a known problem with hidden input fields in Netsurf? Or is there
something else wrong with the page that annoys the browser, which the W3C
validator doesn't pick up
TTFN, Andrew Hodgkinson
Find some electronic music at: Photos, wallpaper, software and more: