[Netsurf-develop] Re: [iyonix-support] Re: RiScript, NetSurf and the Iyo
by John-Mark Bell
On Sun, 7 Aug 2005, David Llewellyn-Jones wrote:
> In case it's any help (and sorry if this has already been covered), but I
> believe the reason NetSurf hangs the machine after RiScript is due to a
> problem with NetSurf and certain fonts. RiScript adds the fonts it creates as
> RISC OS fonts, and some of these cause NetSurf to crash. I get the same
> problem running NetSurf after armTeX, which also adds fonts to RISC OS.
Yes and no. In the case of TeX, the font logo10 is the issue. I've no idea
about RiScript, but expect it to be the same issue _iff_ it freezes the
machine repeatably.
Here's a very simple BASIC test program (the result of an investigation by
Adrian Lees and myself some time ago) which exhibits the same problem
(this _will_ freeze your machine, so I wouldn't recommend running it):
SYS"Font_FindFont",,"TeX.logo10\EUTF8",160,160,0,0 TO fh%
DIM string% 8, block% 64
string%!0 = 245:REM Character 245 (as UTF-32)
string%!4 = 0:REM Terminate string
block%!0 = 0:block%!4 = 0
block%!8 = 0:block%!12 = 0
block%!16 = -1
SYS"Font_ScanString",fh%,string%,&421a0,&7fffffff,&7fffffff,block%,0,4
Note that the use of a UTF-32 encoded string is pretty well irrelevant - a
UTF-8 encoded string exhibits the same behaviour.
I couldn't see anything obviously wrong with the font data (although it's
entirely possible (if not likely) that I've missed something).
I understand that Castle have been informed of this issue.
John.
17 years, 7 months
Re: [Netsurf-develop] NetSurf and an odd error message.
by Anthony Hilton
In <URL:news:local.netsurf> on Sat 06 Aug, Dr Peter Young wrote:
> On 5 Aug 2005 Anthony Hilton <ajh(a)tinshill.f9.co.uk> wrote:
>
> > In <URL:news:local.netsurf> on Fri 05 Aug, Dr Peter Young wrote:
> >>
> >> On 4 Aug 2005 Dr Peter Young <pnyoung(a)ormail.co.uk> wrote:
> >>
> >> >
> >> > This is with 1 August build, though I had it with some of the July
> >> > ones too; only happens sporadically. I think it only happens when I've
> >> > made quite a bit of use of Firefox beta 3, but I'm not entirely sure
> >> > of this. RO 5.09.
> >> >
> >> > I try to run NetSurf, and I get a thin window in the middle of the
> >> > screen, only just readable, saying "UnixLib detected recursion of
> >> > signal SIGSEV. Exiting."
> >> >
> >> > I then get a continuous hourglass, and need to do a hard reset to get
> >> > back control. stderr doesn't seem to be informative, but I have filed
> >> > it away in case.
> >>
> >> Following a post on c.s.a.misc, RiScript seems to have been identified
> >> as the culprit; something to do with fonts, which is way beyond my
> >> understanding. It happens repeatably if you try to run NetSurf after
> >> either RiScript 4.xx or 5.
> >
> > [snip]
> >
> > Intrigued I loaded Riscript5, then Netsurf (2005-08-01 18:15) and browsed
> > Drobe and downloaded the latest Netsurf build without the above symptoms.
> > (RISC OS 4.39, SA RPC)
>
> Yes, it seems to be 100% here; I've posted on the Iyo list, where
> someone has already posted with the same problem. BTW, how much memory
> do you have? This may be at the root of the problem, but it's way
> beyond my understanding.
Sorry, I had intended replying to the list, and am now posting this to the
list as well.
I have 96+2Mb in this RiscPC
Anthony
PS can the list software add a Reply-To: header so replies go to the list
instead of (just) to the original sender?
--
ajh(a)tinshill.f9.co.uk
17 years, 7 months
Re: [Netsurf-develop] NetSurfand an odd error message.
by David Pitt
In message <d7d43b954d.pnyoung(a)pnyoung.ormail.co.uk>
Dr Peter Young <pnyoung(a)ormail.co.uk> wrote:
[snip]
>I try to run NetSurf, and I get a thin window in the middle of the
>screen, only just readable, saying "UnixLib detected recursion of
>signal SIGSEV. Exiting."
>
>I then get a continuous hourglass, and need to do a hard reset to get
>back control. stderr doesn't seem to be informative, but I have filed
>it away in case.
>
That error happens here if NetSurf is run after RiScript 5.00 which has
seen the RiScript User Manual PDF.
I do not know if this is a NetSurf problem but for information if
nothing else there follows a copy of a post to the Iyonix group.
<quote>
The problem would indeed seem to be to do with fonts. Removing the
GillSans font from !RiScript.Resources.!Generated and unticking the
"Convert Type1 fonts into RISC OS fonts" option to stop it getting back
in there solves the problem here.
To confirm this, this is the last line from NetSurf's stderr after a
crash :-
rufl_init.c rufl_init 135: scanning 11 "Fonts.GillSans"
The GillSans font is in the RiScript manual UM50/pdf, the font appears
in !Generated on opening that file. It is self inflicted injury,
obviously in computing it is a really bad idea to read the manual!
It is Iyonix only, all is OK on the RISC PC.
</quote>
I hope that is useful, I could supply any further information required.
--
David Pitt.
Castle Iyonix, RISC OS 5.09.
17 years, 7 months
[Netsurf-develop] Netsurf and Riscript
by peter walder
Trying to access the RiScript site www.riscript.nl from Netsurf causes
my machine to completely lock up and requires a reboot. I am using the
Netsurf build from 3.8.05 running on a Risc PC with Riscos 3.7.
Any help much appreciated.
Peter Walder
--
Peter Walder
pjwalder(a)blueyonder.co.uk
17 years, 7 months
[Netsurf-develop] NetSurfand an odd error message.
by Dr Peter Young
This is with 1 August build, though I had it with some of the July
ones too; only happens sporadically. I think it only happens when I've
made quite a bit of use of Firefox beta 3, but I'm not entirely sure
of this. RO 5.09.
I try to run NetSurf, and I get a thin window in the middle of the
screen, only just readable, saying "UnixLib detected recursion of
signal SIGSEV. Exiting."
I then get a continuous hourglass, and need to do a hard reset to get
back control. stderr doesn't seem to be informative, but I have filed
it away in case.
Any ideas?
With best wishes,
Peter.
--
Peter \ / \ Prestbury, Cheltenham, Glos. GL52
Anne \ / __ __ \ England.
and / / \ | | |\ | / _ \ http://pnyoung.orpheusweb.co.uk
family / \__/ \_/ | \| \__/ \______________ pnyoung(a)ormail.co.uk.
17 years, 7 months
[Netsurf-develop] View Source (2005-08-03 17:15)
by Andy Wingate
Although 'View source' now gives a different filename for each use,
NetSurf is displaying the 'source' rather than my text editor.
If I 'View source' via the menu, I can hold down shift to get my
editor to take it. But that doesn't work for F8.
--
Andy Wingate <URL:http://www.sparse.net> OpenPGP key 0xC642BF8A
Stop searching. Happiness is right next to you.
17 years, 7 months
[Netsurf-develop] Re: mis handling attachments
by David W Mills
In message <20050802031840.0062788111(a)sc8-sf-spam1.sourceforge.net>
netsurf-develop-request(a)lists.sourceforge.net wrote:
[snip intro and text re attachments]
> >
> >
> Try hitting the "refresh button"
Okay..that works, but I still shouldn't have to do that. Hence my original
post.
David
[snip]
17 years, 7 months
Re: [Netsurf-develop] Problem site
by John-Mark Bell
On Tue, 2 Aug 2005, Anthony Hilton wrote:
> There is also a problem with the home page (Netsurf not handling frames or
> similar to the float problem Keith Hopper mentioned?) - the buttons
> (links to timetables, journey planner and live bus times) which appear top
> right in other browsers float over the left side menu in NS 2005-08-01 18:16
This is due to the site's use of absolute positioning. NetSurf does not
yet support this.
John.
17 years, 7 months
[Netsurf-develop] Another problem site
by Keith Hopper
Hi,
A quick look at -
http://persistent.info/archives/thor/index
looks a bit odd down the LHS with various bits being over-written with
others. The document itself is valid xhtml 1 transitional - but the
principal CSS (http://persistent.info/styles-site.css) does have a parse
error (an extra comma on line 57). The W3C CSS validator seems to accept
everything else as valid styling, so I am half convinced that this might be
a Netsurf bug (ie not recovering fully from the extra comma) as a float is
styled for the content on line 70 - which appears to be what is going wrong
with the rendering - no float.
Over to you!
Regards,
Keith
--
City Desk
Waikato University
[PGP key available if desired]
17 years, 7 months
