- Dillo-dev - lists.mailman3.com

[patch] list.h cleanup
by p37sitdu＠lavabit.com Jan. 12, 2013

Jan. 12, 2013

http://www.inventati.org/noname/dillo/list.h.patch

1 0

Intermitent internet access
by jcid＠dillo.org Jan. 11, 2013

Jan. 11, 2013

Hi, FYI, I'm away from home and with intermitent Internet connection. -- Cheers Jorge.-

1 0

White on white for the active tab
by onepoint＠starurchin.org Jan. 11, 2013

Jan. 11, 2013

I am getting white text on a white background for the active tab. I haven't changed any colours form the defaults. Why would this be? Regards, Jeremy Henty

3 3

[patch] modified a_Dialog_choice5 to accept any number alternatives
by p37sitdu＠lavabit.com Jan. 11, 2013

Jan. 11, 2013

-5 LOC, nothing is changed visually

3 2

namespace comments in dw/*
by corvid＠lavabit.com Jan. 11, 2013

Jan. 11, 2013

p37sitdu wrote: > Copied comments are almost in every source file. Committed.

1 0

patches
by p37sitdu＠lavabit.com Jan. 11, 2013

Jan. 11, 2013

I have uploaded some of my patches that were not commited (yet) to http://www.inventati.org/noname/dillo/ so it is easier to track them. Descriptions are embedded inside.

1 1

drawing areas with negative coordinates
by corvid＠lavabit.com Jan. 9, 2013

Jan. 9, 2013

Just something I noticed. There aren't an overwhelming number of them, so it seems like...not too big of a deal. I have the impression that it generally means that something is being drawn while an allocation somewhere is still (-1, -1, 1, 1). diff -r 35fb41bd0cdb dw/fltkviewbase.cc --- a/dw/fltkviewbase.cc Wed Jan 09 16:40:25 2013 +0000 +++ b/dw/fltkviewbase.cc Wed Jan 09 17:17:36 2013 +0000 @@ -431,8 +431,14 @@ void FltkViewBase::queueDraw (core::Rectangle *area) { - drawRegion.addRectangle (area); - damage (FL_DAMAGE_USER1); + static int counter = 0; + if (area->y < 0) { + MSG("%d let's not draw (%d,%d) w %d h %d\n", counter++, area->x, area->y, + area->width, area->height); + } else { + drawRegion.addRectangle (area); + damage (FL_DAMAGE_USER1); + } } void FltkViewBase::queueDrawTotal ()

1 0

[patch] dClose
by p37sitdu＠lavabit.com Jan. 9, 2013

Jan. 9, 2013

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.

2 1

[patch] consistent file save UI behaviour for overwriting files
by onepoint＠starurchin.org Jan. 8, 2013

Jan. 8, 2013

The UI for downloading files is split between Dillo (which prompts for a file to save to) and the downloads DPI (which prompts for another file if the first file already exists). This leads to inconsistent behaviour for "Save link as file" depending on whether the URL is cached or not - if it is cached then any existing file is silently overwritten but if it is not cached we go to the downloads DPI which prompts before overwriting anything. 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. http://starurchin.org/dillo/patch/downloads-dpi_file_test_remove.html http://starurchin.org/dillo/patch/uicmd-cc_file_test_add.html Comments? Shall I push this? Regards, Jeremy Henty

2 2

dillo3.0.2 Irix
by corvid＠lavabit.com Jan. 8, 2013

Jan. 8, 2013

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.

1 0