- Dillo-dev - lists.mailman3.com

[patch] dClose

by p37sitdu＠lavabit.com

Merged a_Misc_close_fd, Dpi_close_fd and Http_socket_close into dClose and moved it into dlib. Replaced close calls with calls to dClose.

Jan. 9, 2013

drawing highlighting

by sgeerken＠dillo.org

On Tue, Jan 08, corvid wrote: > > I was looking at draw() calls for the view, and I saw how there was a > call for every time that the mouse moved when highlighting even though > the text highlighted didn't necessarily change. > > Here's quick code to demonstrate that there are plenty of redundant ones. > I don't know the iterator code well enough to know whether that's the > right place to do anything about it, though... Your patch makes sense, and I've committed it (slightly changed). The purpose of Iterator::highlight is to change a state, and so it should be checked, whether the state has changed at all. Sebastian

Jan. 9, 2013

font_factor

by johannes.hofmann＠gmx.de

On Tue, Jan 08, 2013 at 11:55:42PM +0000, corvid wrote: > > I wrote: > > In styleengine, we use font_factor for calculation of larger and x-small > > and so forth, but we don't apply it when given an ordinary font size like > > 11px. > > Just to double-check: That's not intended behaviour, right? > sorry for the delay... This is intentional, but we know that there is a usability issue. In CSS px, mm, pt etc. define absolute sizes [1]. So if we want to adhere to the CSS spec strictly I think we can't just scale them. On the other hand web designers just use px without thinking about the implications. Also other browsers allow scaling of those fonts afaik. I think we have the following options: * ignore CSS and apply font_factor also to absolute font sizes * implement a zoom feature that displays everything larger (also images etc. * check other browsers for more alternatives [1] http://www.w3.org/TR/CSS21/fonts.html#font-size-props

Jan. 9, 2013

[patch] consistent file save UI behaviour for overwriting files

by onepoint＠starurchin.org

corvid wrote: > Jeremy wrote: > > These two patches move the UI logic from the downloads DPI into > > Dillo. This puts all the UI logic in the same place and ensures > > consistent behaviour for "Save link as file" whether or not the > > URL is cached. > > Seems good to have the logic in dillo. OK, it's in. > I imagine a later step would be to teach UIcmd_save_file_check() > about file permissions, errno possibilities, etc. I guess so, although in the case of the file not existing you would want to check whether the missing component was the last one in the path or not. Regards, Jeremy Henty

Jan. 8, 2013

dillo3.0.2 Irix

by corvid＠lavabit.com

Jan wrote: > Nothing new with this patch. I found out it is not random, it > depends on the length: > > I can open: > > file:/tmp/12345678901234567890123.jpg > > but > > file:/tmp/123456789012345678901234.jpg > > returns a 404 32-byte limit? The first thing that popped into my mind was that sufficiently ancient systems would have path length limits like 32, but I don't see why that would apply here... dpi/file.c's File_normalize_path() has a "Dstr *ds = dStr_sized_new(32)". You could change the number and see whether anything happens. Although if Dstr should have problems resizing itself on irix, I'd be very surprised that dillo could even run at all.

Jan. 8, 2013

Patch for xembed updated

by Johannes.Hofmann＠gmx.de

Hi, On Sat, Jan 05, 2013 at 10:56:00AM +0100, podarcis wrote: > Hi, > > I have updated my patch (no removal of existing code, added > comments, trimmed the source,..). > > Let me perhaps explain from the beginning: > > My starting point was, that when embedding dillo from an external > program using the --xid parameter, dillo's window didn't show up in > the embedder. Thanks for the updated patches and the background info. Here is some background info on the xembed stuff from my side: dillo2 supported xembed more or less automatically as it was based on gtk2. The claws-mail dillo plugin used this feature to display html. When porting dillo to fltk we had to implement xembed on our own or break support for the claws-mail plugin. That's why we added xembed.cc. > > I have now found, there's a one-line fix for that problem: > > -> dillo_xembed_patch_1.txt I tend to commit this, but I still need to double check that it doesn't cause any problems. The advantage is that it will give some initial encouraging result if someone tries out --xid without a fullblown xembed implementation. > > After that was working again (by the way: it worked in dillo2 > already) I tried to make resizing of the embedder window work too: > > -> dillo_xembed_patch_2.txt > > A test program for all this: > > -> dembed.cxx I would prefer a minimal fltk-based example program in the test directory, that implements the embedder side of xembed. It shouldn't be too hard to do (e.g. by looking at the tabbed sources). It would be cool if you would like to work on that. BTW. we use a pretty restricted set of C++ features in the dillo source tree (e.g. we don't use iostream). Cheers, Johannes

Jan. 8, 2013

dillo3.0.2 Irix

by corvid＠lavabit.com

Jan wrote: > sometimes you get a 404 error when you try to open a html file, > this behavior seems to be random. Hmm... I wonder whether this would show anything interesting (run 'dpidc stop' before reinstalling DPIs) diff -r 7750054f37e5 dpi/file.c --- a/dpi/file.c Mon Jan 07 21:39:19 2013 +0100 +++ b/dpi/file.c Tue Jan 08 18:42:22 2013 +0000 @@ -671,6 +671,8 @@ struct stat sb; if (stat(filename, &sb) != 0) { + MSG("Couldn't stat %s: %s\n", filename, strerror(errno)); + /* stat failed, prepare a file-not-found error. */ res = ENOENT; } else if (S_ISDIR(sb.st_mode)) {

Jan. 8, 2013

dillo3.0.2 Irix

by diegel＠work.de

Thanks for your hints. The browser is able to render html. about:splash looks normal and sometimes you can open html files with the open file menu, sometimes you get a 404 error when you try to open a html file, this behavior seems to be random. I set VERBOSE. This is the output when I start dillo, try to surf home (don't work) and open a JPEG in the /tmp directory (works). src/dillo paths: Cannot open file '/usr/people/diegel/.dillo/dillorc' paths: Using /usr/local/etc/dillo/dillorc paths: Cannot open file '/usr/people/diegel/.dillo/keysrc' paths: Using /usr/local/etc/dillo/keysrc dillo_dns_init: Here we go! (not threaded) Disabling cookies. ** WARNING **: preferred sans-serif font "DejaVu Sans" not found. ** WARNING **: preferred serif font "DejaVu Serif" not found. ** WARNING **: preferred monospace font "DejaVu Sans Mono" not found. sorcerer:/mnt/build/dillo-3.0.2> src/dillo paths: Cannot open file '/usr/people/diegel/.dillo/dillorc' paths: Using /usr/local/etc/dillo/dillorc paths: Cannot open file '/usr/people/diegel/.dillo/keysrc' paths: Using /usr/local/etc/dillo/keysrc dillo_dns_init: Here we go! (not threaded) Disabling cookies. ** WARNING **: preferred sans-serif font "DejaVu Sans" not found. ** WARNING **: preferred serif font "DejaVu Serif" not found. ** WARNING **: preferred monospace font "DejaVu Sans Mono" not found. ** WARNING **: preferred cursive font "URW Chancery L" not found. ** WARNING **: preferred fantasy font "DejaVu Sans" not found. Nav_open_url: new url='about:splash' Nav_open_url: new url='http://www.dillo.org/' a_Capi_ccc : OpStart [2B] Info=10250a38 Flags=0 a_Dpi_ccc : OpStart [2B] Info=10250a60 Flags=0 a_IO_ccc : OpStart [2B] Info=10250a88 Flags=0 a_Capi_ccc : OpStart [1B] Info=10250ad8 Flags=0 a_Http_ccc : OpStart [1B] Info=10250b00 Flags=0 a_IO_ccc : OpStart [1B] Info=10250b50 Flags=0 Dns_server [0]: www.dillo.org is 134.102.206.165 Connecting to 134.102.206.165 a_IO_ccc : OpSend [1B] Info=10250b50 Flags=0 a_Capi_ccc : OpSend [1F] Info=10250ad8 Flags=0 a_Capi_ccc : OpSend [2B] Info=10250a38 Flags=0 a_Dpi_ccc : OpSend [2B] Info=10250a60 Flags=0 a_IO_ccc : OpSend [2B] Info=10250a88 Flags=0 a_IO_ccc : OpSend [1B] Info=10250b50 Flags=0 Nav_open_url: new url='file:/tmp/orfanelle.jpg' a_Capi_ccc : OpAbort [1B] Info=10250ad8 Flags=0 a_Http_ccc : OpAbort [1B] Info=10250b00 Flags=0 a_IO_ccc : OpAbort [1B] Info=10250b50 Flags=0 a_Capi_ccc : OpAbort [2B] Info=10250a38 Flags=0 a_Dpi_ccc : OpAbort [2B] Info=10250a60 Flags=0 a_IO_ccc : OpAbort [2B] Info=10250a88 Flags=0 a_Capi_ccc : OpStart [2B] Info=10250bf0 Flags=0 a_Dpi_ccc : OpStart [2B] Info=10250a38 Flags=0 a_IO_ccc : OpStart [2B] Info=10250a60 Flags=0 a_Capi_ccc : OpStart [1B] Info=10250ab0 Flags=0 a_Dpi_ccc : OpStart [1B] Info=10250ad8 Flags=0 ** ERROR **: [Dpi_read_comm_keys] No such file or directory Dpi_blocking_start_dpid: try 1 [dpid]: a_Misc_mksecret: f3f0000d dpid started a_IO_ccc : OpStart [1B] Info=10250b00 Flags=0 a_IO_ccc : OpSend [1B] Info=10250b00 Flags=0 a_Capi_ccc : OpSend [1F] Info=10250ab0 Flags=0 a_Capi_ccc : OpSend [2B] Info=10250bf0 Flags=0 a_Dpi_ccc : OpSend [2B] Info=10250a38 Flags=0 a_IO_ccc : OpSend [2B] Info=10250a60 Flags=0 a_Capi_ccc : OpSend [1F] Info=10250ab0 Flags=0 a_Capi_ccc : OpSend [1B] Info=10250ab0 Flags=0 a_Dpi_ccc : OpSend [1B] Info=10250ad8 Flags=0 a_IO_ccc : OpSend [1B] Info=10250b00 Flags=0 [file dpi]: (v.2) accepting connections... a_IO_ccc : OpSend [2F] Info=10250a60 Flags=0 a_Dpi_ccc : OpSend [2F] Info=10250a38 Flags=0 a_Capi_ccc : OpSend [2F] Info=10250bf0 Flags=0 a_Capi_ccc : OpSend [2F] Info=10250bf0 Flags=0 a_IO_ccc : OpSend [2F] Info=10250a60 Flags=0 a_Dpi_ccc : OpSend [2F] Info=10250a38 Flags=0 a_Capi_ccc : OpSend [2F] Info=10250bf0 Flags=0 a_IO_ccc : OpSend [2F] Info=10250a60 Flags=0 a_Dpi_ccc : OpSend [2F] Info=10250a38 Flags=0 a_Capi_ccc : OpSend [2F] Info=10250bf0 Flags=0 a_IO_ccc : OpEnd [2F] Info=10250a60 Flags=0 a_Dpi_ccc : OpEnd [2F] Info=10250a38 Flags=0 a_Capi_ccc : OpEnd [2F] Info=10250bf0 Flags=0 a_Capi_ccc : OpEnd [1B] Info=10250ab0 Flags=0 a_Dpi_ccc : OpEnd [1B] Info=10250ad8 Flags=0 a_IO_ccc : OpEnd [1B] Info=10250b00 Flags=0 The tests provide some output. The cookies test for esample outputs endless: [cookies dpi]: Disabling cookies. [cookies dpi]: (v.1) accepting connections... [cookies dpi]: cleanup corvid wrote: > Do the dw* programs in dillo's test directory do anything? > > Given that you aren't getting any repush messages, I'm guessing that > they'll be reasonably okay and that it's something to do with IO. > > I'm not the expert on that code, but in the interest of giving something > to try in the meantime, I might start by setting breakpoints on > IO_fd_write_cb() and IO_fd_read_cb(), or setting VERBOSE in src/chain.c. > > > _______________________________________________ > Dillo-dev mailing list > Dillo-dev at dillo.org > http://lists.auriga.wearlab.de/cgi-bin/mailman/listinfo/dillo-dev > > -- Jan Diegelmann

Jan. 8, 2013

dillo3.0.2 Irix

by corvid＠lavabit.com

Jan wrote: > I try to port dillo 3.0.2 to Irix: > > fltk 1.3 compiles fine with gcc 4.7. test/demo runs without errors. > > dillo compiles also with gcc 4.7, there is only a problem with setenv in > dpi/downloads.cc, that can be resolved using putenv. > > dillo starts correctly, but is not able to display any website or file: > > > dillo > paths: Cannot open file '/usr/people/diegel/.dillo/dillorc' > paths: Using /usr/local/etc/dillo/dillorc > paths: Cannot open file '/usr/people/diegel/.dillo/keysrc' > paths: Using /usr/local/etc/dillo/keysrc > dillo_dns_init: Here we go! (threaded) > Disabling cookies. > ** WARNING **: preferred sans-serif font "DejaVu Sans" not found. > ** WARNING **: preferred serif font "DejaVu Serif" not found. > ** WARNING **: preferred monospace font "DejaVu Sans Mono" not found. > ** WARNING **: preferred cursive font "URW Chancery L" not found. > ** WARNING **: preferred fantasy font "DejaVu Sans" not found. > Nav_open_url: new url='about:splash' > Nav_open_url: new url='http://www.dillo.org/' > Dns_server [0]: www.dillo.org is 134.102.206.165 > Connecting to 134.102.206.165 > Nav_open_url: new > url='file:/usr/people/diegel/dillo/www.dillo.org/index.html' > > I also tried with --disable-threaded-dns but same result. > > Any idea how to debug this problem? Do the dw* programs in dillo's test directory do anything? Given that you aren't getting any repush messages, I'm guessing that they'll be reasonably okay and that it's something to do with IO. I'm not the expert on that code, but in the interest of giving something to try in the meantime, I might start by setting breakpoints on IO_fd_write_cb() and IO_fd_read_cb(), or setting VERBOSE in src/chain.c.

Jan. 8, 2013

Gopher DPI

by p37sitdu＠lavabit.com

Added telnet:// URI support (type '8').

Jan. 8, 2013