January 2009 - Dillo-dev - lists.mailman3.com

Crashes in dw::core::Layout::createImgbuf()

by Johannes.Hofmann＠gmx.de

Hi Jeremy, On Tue, Jan 20, 2009 at 06:20:32PM +0000, Jeremy Henty wrote: > > For the last few days I have been getting occasional crashes inside > dw::core::Layout::createImgbuf() when opening a new tab. It does not > seem to be predictable: if I restart Dillo and reopen all the URLs I > was just looking at it is fine. > > The gdb backtrace often shows ridiculous values for this: eg. the > attached example has this=0x6. I've seen other impossibly small > values for this, but sometimes the value seems more sensible. > > Any ideas? I'd like to investigate further but I don't know what to > look for. Pointers are welcome! I've never seen these errors yet. Can you try to use "hg bisect" to find the changeset where the error was introduced? I know it's difficult with sporadic errors... Or maybe try valgrind? Regards, Johannes

Jan. 20, 2009

Crashes in dw::core::Layout::createImgbuf()

by onepoint＠starurchin.org

For the last few days I have been getting occasional crashes inside dw::core::Layout::createImgbuf() when opening a new tab. It does not seem to be predictable: if I restart Dillo and reopen all the URLs I was just looking at it is fine. The gdb backtrace often shows ridiculous values for this: eg. the attached example has this=0x6. I've seen other impossibly small values for this, but sometimes the value seems more sensible. Any ideas? I'd like to investigate further but I don't know what to look for. Pointers are welcome! Regards, Jeremy Henty

Jan. 20, 2009

dillo CSS and charset

by jcid＠dillo.org

On Tue, Jan 20, 2009 at 07:29:46AM +0900, furaisanjin wrote: > 2009/1/20 Jorge Arellano Cid <jcid(a)dillo.org>: > > Yes, there're bugs and a timing issue. > > I'm looking at them... > > Ok, thanks. This is timing issue example page. > > http://www.picfun.com/app00.html > > HTTP server response is > > HTTP/1.1 200 OK > Date: Mon, 19 Jan 2009 22:21:11 GMT > Server: Apache > Last-Modified: Sun, 11 Jan 2009 01:58:25 GMT > ETag: "15e47-43cb-53585240" > Accept-Ranges: bytes > Content-Length: 17355 > Content-Type: text/html > > the page contains > > <META http-equiv="Content-Type" content="text/html; charset=Shift_JIS"> > > At first time charset conversion is something wrong but if I go > somewhere and come back, charset conversion looks OK. I think this > type of page doesn't contain CSS or contains embedded style. So I'm > guessing repush timing is different. Yes, I have a combinatorial case table of things to manage, some work some doesn't. The idea is to get to a design that handles them cleanly. BTW (@all), the "tools" button is 60% done, but stopped. As it also uses repush to enable/disable CSS parsing, is better to fix the above issues before testing it. -- Cheers Jorge.- ______________________________________________________________________ This email has been scanned by the MessageLabs Email Security System. For more information please visit http://www.messagelabs.com/email ______________________________________________________________________

Jan. 20, 2009

dillo CSS and charset

by furaisanjin＠gmail.com

2009/1/20 Jorge Arellano Cid <jcid(a)dillo.org>: > Yes, there're bugs and a timing issue. > I'm looking at them... Ok, thanks. This is timing issue example page. http://www.picfun.com/app00.html HTTP server response is HTTP/1.1 200 OK Date: Mon, 19 Jan 2009 22:21:11 GMT Server: Apache Last-Modified: Sun, 11 Jan 2009 01:58:25 GMT ETag: "15e47-43cb-53585240" Accept-Ranges: bytes Content-Length: 17355 Content-Type: text/html the page contains <META http-equiv="Content-Type" content="text/html; charset=Shift_JIS"> At first time charset conversion is something wrong but if I go somewhere and come back, charset conversion looks OK. I think this type of page doesn't contain CSS or contains embedded style. So I'm guessing repush timing is different. Regards, furaisanjin

Jan. 19, 2009

dillo CSS and charset

by jcid＠dillo.org

On Mon, Jan 19, 2009 at 11:29:27PM +0900, furaisanjin wrote: > Hi, > > 2009/1/19 Jorge Arellano Cid <jcid(a)dillo.org>: > > Please clarify me whether 2.1 behaviour is right or wrong! :-) > > I think the behavior is wrong. > > > And if you can send an example, better. > > This is one example (this page is news archive so the page may be > removed some day). > > http://www.so-net.ne.jp/news/cgi-bin/article.cgi?gid=mai&aid=20090119-570-O… > > This is HTTP server response. > > HTTP/1.1 200 OK > Date: Mon, 19 Jan 2009 14:23:11 GMT > Server: Apache > Cache-control: private, max-age=300 > Content-Type: text/html; charset=SHIFT-JIS > Transfer-Encoding: chunked > > And the page contains > > <meta http-equiv="Content-Type" content="text/html; charset=shift-jis" /> Got it! Yes, there're bugs and a timing issue. I'm looking at them... -- Cheers Jorge.- ______________________________________________________________________ This email has been scanned by the MessageLabs Email Security System. For more information please visit http://www.messagelabs.com/email ______________________________________________________________________

Jan. 19, 2009

[bug]: hang in Css_parse_declaration()

by onepoint＠starurchin.org

On Mon, Jan 19, 2009 at 05:06:39PM +0100, Hofmann Johannes wrote: > On Sun, Jan 18, 2009 at 09:57:58AM +0000, Jeremy Henty wrote: > > The following HTML makes dillo hang and burn 100% CPU. > > > > <html> > > <body> > > <span style="{}">a</span> > > </body> > > </html> > > Should be fixed in tip. It is! And the new default font size is nice too! Jeremy Henty

Jan. 19, 2009

[bug]: hang in Css_parse_declaration()

by Johannes.Hofmann＠gmx.de

Hi Jeremy, On Sun, Jan 18, 2009 at 09:57:58AM +0000, Jeremy Henty wrote: > The following HTML makes dillo hang and burn 100% CPU. > > <html> > <body> > <span style="{}">a</span> > </body> > </html> > > The bug does not appear if the span contains no text, or only > whitespace. > > The stack trace always ends inside Css_parse_declaration(). The line > number varies, but seems to be always one of 809, 823 or 963. > > You can see the same bug online at > > http://www.nation.co.ke/News/-/1056/515354/-/u1acuw/-/index.html Should be fixed in tip. Thanks for the report, Johannes

Jan. 19, 2009

Re: A couple of patches

by Johannes.Hofmann＠gmx.de

On Sun, Jan 18, 2009 at 01:23:00AM +0000, corvid wrote: > Johannes wrote: > > On Thu, Jan 15, 2009 at 10:41:59AM -0300, Jorge Arellano Cid wrote: > > > > * snip* > > > > > > BTW, I have a question. How can a minimal font size be specified? > > > Currently I have: > > > > > > body {font-size: 18px !important} > > > > > > in style.css, but often pages come with smaller fonts. e.g. > > > http://www.linux.org.uk/Portaloo.cs > > > > > > For me is important because I have a problems reading tiny fonts, > > > and my working laptop has a high screen resolution. > > > > I've just made a change that uses the existing prefs.font_factor > > value to scale the base font. > > If I use 14 instead of 12, it's a lot more readable. > (Or is the 12 important for some CSS reason involving picas?) No, we can easily change the default fontsize to 14. If nobody objects, I will make that change later today. Cheers, Johannes

Jan. 19, 2009

dillo CSS and charset

by furaisanjin＠gmail.com

Hi, 2009/1/19 Jorge Arellano Cid <jcid(a)dillo.org>: > Please clarify me whether 2.1 behaviour is right or wrong! :-) I think the behavior is wrong. > And if you can send an example, better. This is one example (this page is news archive so the page may be removed some day). http://www.so-net.ne.jp/news/cgi-bin/article.cgi?gid=mai&aid=20090119-570-O… This is HTTP server response. HTTP/1.1 200 OK Date: Mon, 19 Jan 2009 14:23:11 GMT Server: Apache Cache-control: private, max-age=300 Content-Type: text/html; charset=SHIFT-JIS Transfer-Encoding: chunked And the page contains <meta http-equiv="Content-Type" content="text/html; charset=shift-jis" /> Regards, furaisanjin

Jan. 19, 2009

dillo CSS and charset

by jcid＠dillo.org

On Sun, Jan 18, 2009 at 10:33:39PM +0900, furaisanjin wrote: > Hi all. > > I created new subject thread since the original one becomes quite long and. > > I found one difference between dillo-2.0 and dillo-css. there was > a_Decode_charset_init calling at Cache_parse_header in cache.c when > HTTP server Content-Type was detected but there isn't this in > dillo-css. Because if HTTP server Content-Type is same as META > content-type, there isn't any chance to call a_Decode_charset_init. > This is one reason why the page is drawn properly. Yes, there were cleanups on charset decoding. Please clarify me whether 2.1 behaviour is right or wrong! :-) And if you can send an example, better. -- Cheers Jorge.- ______________________________________________________________________ This email has been scanned by the MessageLabs Email Security System. For more information please visit http://www.messagelabs.com/email ______________________________________________________________________

Jan. 19, 2009