- Dillo-dev - lists.mailman3.com

TLS connect error: "an EOF was observed that violates the protocol"

by a1ex＠dismail.de

Hi, Here are 2 separate issues: 1) socks5 proxy on 127.0.0.1:8080 not working. This is a proxy over ssh, for example: ssh -N -D 8080 user(a)example.com in dillorc: http_proxy="http://localhost:8080/" console output when trying to connect to a site: Connecting to 127.0.0.1:8080 CONNECT through proxy failed. Full reply not received: (nothing) ** WARNING **: CCC: call on already finished chain. Flags=CCC_Ended This setup works fine under Firefox. I tried doing a tcpdump while attempting the connection, and there is no activity. So, while trying to research this, I ran into another issue: 2) Any time I go to a page on this site, Dillo crashes with the following: Nav_open_url: new url='https://dillo-dev.auriga.wearlab.narkive.com/WT0JYUZq/dillo-won-t-resolve-u…' Dns_server [0]: dillo-dev.auriga.wearlab.narkive.com is 149.248.211.108 Connecting to 149.248.211.108:443 TLS connect error: "an EOF was observed that violates the protocol" Tls_close_by_key: Avoiding SSL shutdown for: https://dillo-dev.auriga.wearlab.narkive.com/WT0JYUZq/dillo-won-t-resolve-u… fd 6 is done and failed dillo(13689) in malloc(): write to free mem 0x60383b59140[24..31]@32 Abort trap gdb output: Program received signal SIGABRT, Aborted. thrkill () at /tmp/-:2 2 /tmp/-: No such file or directory. in /tmp/- This is on OpenBSD-current amd64 with LibreSSL 3.9.0, running an unmodified fresh checkout of Dillo master. Also tested on OpenBSD 7.5 with the same result. I tested the site with: https://www.ssllabs.com/ssltest/analyze.html?d=dillo-dev.auriga.wearlab.nar… There seem to be some handshake failures during the simulation. This probably is not be the fault of Dillo, but maybe there is a more graceful to handle this, rather than crashing. Even a test with curl has issues: curl: (35) LibreSSL SSL_connect: SSL_ERROR_SYSCALL in connection to dillo-dev.auriga.wearlab.narkive.com:443 Is this site just broken/misconfigured? Regards, Alex

Aug. 27, 2024

Re: Firefox bookmarks plugin

by Diego

Good work and very interesting idea. If I can suggest a change I would add this lines to dpidrc at Makefile the installation # Default dillo bookmarks dpi #bookmarks=bookmarks/bookmarks.dpi # ffbm: A Dillo plugin to sync bookmarks with Firefox bookmarks=$(DPI_DIR)/$(BIN) maybe it's time to patch dpidrc to make it more descriptive I sugest this dpi_dir=/usr/local/lib/dillo/dpi # standard dillo distributed DPIs #bookmarks=bookmarks/bookmarks.dpi #cookies=cookies/cookies.dpi #downloads=downloads/downloads.dpi #vsource=vsource/vsource.filter.dpi # protocols DPIs proto.file=file/file.dpi proto.ftp=ftp/ftp.filter.dpi proto.data=datauri/datauri.filter.dpi # dpid uses this line when no line for a protocol is found in this file proto.*=misc/ProtocolNotSupported.filter.dpi Thanks for all the work cheers! Diego. El dom, 25 ago 2024 a las 21:55, Rodrigo Arias (<rodarima(a)gmail.com>) escribió: > > Hi, > > Here is another small plugin to sync bookmarks with Firefox Sync: > > https://github.com/dillo-browser/dillo-plugin-ffbm > > It uses the ffsclient program to connect to the Firefox Sync server and > a few lines of shell script for plumbing. It replaces the builtin > bookmarks plugin when installed, so all menu buttons work as expected. > > Best, > Rodrigo > _______________________________________________ > Dillo-dev mailing list -- dillo-dev(a)mailman3.com > To unsubscribe send an email to dillo-dev-leave(a)mailman3.com

Aug. 25, 2024

Issues with HTTP multipart/form-data file upload

by Xavier Del Campo Romero

Hello Dillo dev community, I am testing support among web browsers for slcl [1], a JS-less minimalist web storage solution. slcl relies on "multipart/form-data"-encoded HTTP requests to upload files to a server. Whereas Dillo helped me to uncover a few wrong assumptions on my code, I have realised a few issues on Dillo itself related to filue uploads that should be considered. Issue #1: While Dillo is fine uploading small files (up to a few MiB), things go wrong with larger files, so much that memory usage increases up to multiple GiB and can even lock the system up. This is because Dillo is designed to send requests always from memory, which means file contents must be dumped into memory first, and this might be unfeasible for large files. Suggestion: Instead, Dillo should ideally send file contents on-the-fly, so that memory usage is kept to a minimum regardless the file size. Issue #2: Even if Dillo generates a 70-byte, random boundary string (yet mostly filled with '-', similarly to Firefox [2]), it ensures it is not found anywhere inside the file contents. Again, this can be a serious bottleneck in the case of large files, as it requires to scan the whole file for a match. Suggestion: Define *all* of the 70 bytes in the boundary string as random, and assume they would never be found inside a file. The chance of accidental collision is so low that it is not worth the effort into checking them. Suggested patches: - 0001-dialog.cc-Generate-more-random-boundaries.patch) Issue #3: Dillo only supports uploading 1 file at a time. This is mostly because it relies (probably temporarily?) on the a_Dialog_save_file function [3]. However, this is not a limitation on the HTTP protocol, and other implementation such as Firefox or Chromium-based browsers support this. Suggested patches: - 0002-dialog-Add-a_Dialog_select_files.patch - 0003-WIP-multi-file-uploads.patch Conclusions: Dillo seems designed to always send requests from memory, so it is not straightforward to break this assumption in order to support large file uploads. 0003-WIP-multi-file-uploads.patch is an incomplete first step into fixing this, but it surely needs deeper design changes. I did not put more effort into these patches for the time being because, after seeing the potential complexity behind this task, I thought it was a better idea to ask the community for feedback and guidelines. Thank you very much for reading. Best regards, Xavier Del Campo Romero [1]: https://gitea.privatedns.org/xavi/slcl [2]: https://github.com/dillo-browser/dillo/blob/8a360e32ac3136494a494379a6dbbac… [3]: https://github.com/dillo-browser/dillo/blob/8a360e32ac3136494a494379a6dbbac…

Aug. 25, 2024

Firefox bookmarks plugin

by Rodrigo Arias

Hi, Here is another small plugin to sync bookmarks with Firefox Sync: https://github.com/dillo-browser/dillo-plugin-ffbm It uses the ffsclient program to connect to the Firefox Sync server and a few lines of shell script for plumbing. It replaces the builtin bookmarks plugin when installed, so all menu buttons work as expected. Best, Rodrigo

Aug. 25, 2024

Re: [PATCH] Dark theme for downloads.dpi

by Rodrigo Arias

Hi, On Sat, Aug 24, 2024 at 06:40:33PM +0200, a1ex(a)dismail.de wrote: >Hi, > >On Sat, 24 Aug 2024 15:39:01 +0200 >Rodrigo Arias <rodarima(a)gmail.com> wrote: > >> I'm also thinking we may want to also include the icons, so we avoid >> the problem of having a single icon that must support dark and light >> backgrounds. In this case, maybe a layout like this may be more >> useful: >> >> ~/.dillo/theme/<name>/style.css >> ~/.dillo/theme/<name>/icons.png > >I agree that themed icons would be very nice, so your layout here makes >sense. > >Might be a bit of work to get Dillo to use icons.png instead of the >built-in pixmaps.h though :) > >Anyway, it sounds like this will be a pretty complicated patch. >Not sure where I can help, but can try if desired on smaller parts. Not sure if it would be an easy patch. You may want to review the issues in case you find some that seems more approachable. https://github.com/dillo-browser/dillo/issues I would also appreciate if someone can test this PR by doing some daily google searches over a week or so with the cookies disabled for google: https://github.com/dillo-browser/dillo/pull/251 The question is if it is better to leave the ucbcb=1 parameter in the URL by default or not: search_url="Google https://www.google.com/search?ie=UTF-8&oe=UTF-8&gbv=1&ucbcb=1&q=%s" I tried to skip the consent dialog by denying it by default so it doesn't appear, but I get mixed results even without cookies disabled. Some times it appears again and others it doesn't, not sure how they determine when it should appear. Maybe it is better to leave it out. Testing this PR would be helpful and doesn't require much effort, in case you are interested. I can only do limited tests before Google fingerprints my IP and then it never ask me consent again. (And yes, I'm aware of the implications of using Google, but it is the only search engine that gives me any result in very weird searches like part numbers of unknown electrical components) Best, Rodrigo.

Aug. 25, 2024

Re: [PATCH] dark mode for tooltips

by Rodrigo Arias

Hi Alex, Thanks for the patch. On Sat, Aug 24, 2024 at 10:39:21PM +0200, a1ex(a)dismail.de wrote: >Hi, > >As part of an overall dark mode for Dillo, we could also consider the >tooltips. Yes, I think a theme should affect all the elements of the browser and probably some other interactions. Best, Rodrigo.

Aug. 25, 2024

[PATCH] dark mode for tooltips

by a1ex＠dismail.de

Hi, As part of an overall dark mode for Dillo, we could also consider the tooltips. Here is an example which I think looks reasonable: --- a/src/tipwin.cc Sun Aug 11 22:21:59 2024 +++ b/src/tipwin.cc Sat Aug 24 22:17:21 2024 @@ -40,7 +40,7 @@ static void recent_timeout(void*); */ TipWin::TipWin() : Fl_Menu_Window(1, 1) // will autosize { - bgcolor = fl_color_cube(FL_NUM_RED - 1, FL_NUM_GREEN - 1, FL_NUM_BLUE - 2); + bgcolor = FL_WHITE; recent = 0; tip[0] = '\0'; cur_widget = NULL; @@ -50,8 +50,8 @@ TipWin::TipWin() : Fl_Menu_Window(1, 1) // will au void TipWin::draw() { - draw_box(FL_BORDER_BOX, 0, 0, w(), h(), bgcolor); - fl_color(FL_BLACK); + draw_box(FL_BORDER_FRAME, 0, 0, w(), h(), bgcolor); + fl_color(FL_WHITE); fl_font(labelfont(), labelsize()); fl_draw(tip, 3, 3, w() - 6, h() - 6, //Fl_Align(FL_ALIGN_LEFT | FL_ALIGN_WRAP)); Note that 'ui_main_bg_color' needs to be set to the desired dark color first. Regards, Alex

Aug. 24, 2024

Re: [PATCH] Dark theme for downloads.dpi

by a1ex＠dismail.de

Hi, On Sat, 24 Aug 2024 15:39:01 +0200 Rodrigo Arias <rodarima(a)gmail.com> wrote: > I'm also thinking we may want to also include the icons, so we avoid > the problem of having a single icon that must support dark and light > backgrounds. In this case, maybe a layout like this may be more > useful: > > ~/.dillo/theme/<name>/style.css > ~/.dillo/theme/<name>/icons.png I agree that themed icons would be very nice, so your layout here makes sense. Might be a bit of work to get Dillo to use icons.png instead of the built-in pixmaps.h though :) Anyway, it sounds like this will be a pretty complicated patch. Not sure where I can help, but can try if desired on smaller parts. -Alex

Aug. 24, 2024

Re: [PATCH] Dark theme for downloads.dpi

by Rodrigo Arias

Hi, On Sat, Aug 24, 2024 at 02:15:13PM +0200, a1ex(a)dismail.de wrote: >That would be great. Maybe we could just have 2 separate theme css >files for light and dark mode, for example 'light.css' and 'dark.css' >in ~/.dillo might be easiest. I think I would place them under ~/.dillo/theme/ so we can select which one you want from the menu, based on the name you put to the .css file. >How do you envision this will work with the existing 'ui_*_color' prefs >in dillorc? Would we just override those if theming is enabled, or >should we get rid of them entirely and have everything controlled by >the theme? It would be nice to keep it working for some time, so users can slowly update to CSS themes. Maybe we can just auto-generate a legacy.css theme with the dillorc options. I'm also thinking we may want to also include the icons, so we avoid the problem of having a single icon that must support dark and light backgrounds. In this case, maybe a layout like this may be more useful: ~/.dillo/theme/<name>/style.css ~/.dillo/theme/<name>/icons.png Which also makes it easy to install new themes by just cloning a remote git repository. Best, Rodrigo.

Aug. 24, 2024

Re: [PATCH] Dark theme for downloads.dpi

by a1ex＠dismail.de

Hi Rodrigo, On Fri, 23 Aug 2024 01:12:38 +0200 Rodrigo Arias <rodarima(a)gmail.com> wrote: > I'm wondering if we could add a mechanism to define themes without > having to rebuild Dillo. > > It would be nice to define some CSS-like styles that define the Dillo > interface colors and button styles too, and be able to switch among > them at runtime. As we already have the CSS parser, we could just use > CSS rules that apply to classes. > ... > I wanted to have at least a light and dark theme, so I can easily > switch from one to another quickly, probably with a single button > like the dark reader extension for Firefox: That would be great. Maybe we could just have 2 separate theme css files for light and dark mode, for example 'light.css' and 'dark.css' in ~/.dillo might be easiest. How do you envision this will work with the existing 'ui_*_color' prefs in dillorc? Would we just override those if theming is enabled, or should we get rid of them entirely and have everything controlled by the theme? > Here are some FLTK themes in Rust which also change the FLTK color > map: > > https://github.com/fltk-rs/fltk-theme/tree/main Thanks, there are some nice looking themes there, and might be a good source for ideas. I also found this, which is written in C++, but targets FLTK 1.4: https://github.com/MoAlyousef/Fl_Theme Regards, Alex

Aug. 24, 2024