- Dillo-dev - lists.mailman3.com

Re: Save link as... silent failure on gopher protocol

by Diego

Hello, I'm sending two patches with an idea to solve this problem. It's a proof of concept that I'm offering to develop and complete, although I have very little free time. The first patch works with the latest dillo release (3.3.0) but also works with the current repository. It implements changes in dillo, download.dpi, and the file and data DPIs as examples. The second patch works with the latest gopher.dpi repository. The main idea is that the DPIs that handle protocols save the URL and write the file themselves, coordinating with download.dpi and dillo. To do this, they implement a new DPI command: `save_url` d_cmd = a_Dpip_build_cmd("cmd=%s url=%s destination=%s", "save_url", url, dl_dest); Dillo initiates a download, sending it to download.dpi as before. If it's HTTP* or FTP*, it downloads directly as it does now. Otherwise, it will send a `save_url` command to Dillo to coordinate. This allows Dillo to notify the user, verify the security of the change, etc.Currently, Dillo only checks if there's a DPI that handles the requested protocol. (A new function (`a_Dpiapi_is_service_handled(server_str)`) queries `dpid` without executing the DPI. No code is modified in `dpid`. Modifying `dpid` or executing the DPI would allow querying the capabilities of the DPIs, but it doesn't seem necessary.) - If a DPI is present, it sends the `save_url` command to the DPI that coordinates with `download.dpi` via a secondary channel so that downloads can continue even if all Dillo windows are closed. - If the protocol isn't handled, it notifies download.dpi through the secondary channel (this reduces the amount of code). The next step would be to implement `save_url` in other protocol DPIs. A possible change to download.dpi once all is done is to leave download.dpi it as just a GUI for DPIs and extract the code to download HTTP to a DPI. (There's already ftp.dpi) One last thing: can you recommend a simple Git manual? Saludos Diego. El lun, 25 may 2026 a las 21:23, Rodrigo Arias (<rodarima(a)gmail.com>) escribió: > Hi, > > On Mon, May 25, 2026 at 09:25:58AM -0700, Ben Collver wrote: > >Good day, > > > >I am using dillo 3.3.0 on Slackware64 15.0. > > > >If I open a gopher directory, right-click on an item, and click > >Save link as..., this opens a file save dialog. If i click OK, it > >dismisses the dialog. I expected it to save something, but it > >creates no new files. > > > >Gopher URI: gopher://gopher.pglaf.org/1/4/0/2/8/40282/ > >Top item: 40282-0 > > > >I found the following workaround: > > > >* Left click 40282-0 > >* Click Save button in browser toolbar > > > >This creates a new file as expected. > > Yes, this is a known limitation of the downloads in Dillo. Using the > "Save link as" to fetch a new resource will use wget under the hood, > which only allows http(s). You can see the message in the console: > > > UIcmd_save: /tmp/40282-0.txt > > ** WARNING **: Ignoring download request for > > 'gopher://gopher.pglaf.org/0/4/0/2/8/40282/40282-0.txt': not in cache > > and not downloadable. > > Browsing to the target file makes the gopher plugin fetch the file and > place it in the Dillo cache (you can see it in about:cache). Then saving > the current page (or going back and clicking the "Save link as" entry) > works because it fetches it from the cache. > > It is planned to solve this some time in the future, but involves > complex changes. In the meanwhile, you can use this workaround. > > Best, > Rodrigo. > _______________________________________________ > Dillo-dev mailing list -- dillo-dev(a)mailman3.com > To unsubscribe send an email to dillo-dev-leave(a)mailman3.com >

July 3, 2026

Re: [PATCH] Fix header parsing bug with malicious header fields

by Rodrigo Arias

Hi, >diff --git a/src/cache.c b/src/cache.c >index 154bef5c..5bd8a5dc 100644 >--- a/src/cache.c >+++ b/src/cache.c >@@ -287,7 +287,7 @@ static int Cache_bufsize(CacheEntry_t *e) > * Inject full page content directly into the cache. > * Used for "about:splash". May be used for "about:cache" too. > * >- * The @param data_ds buffer is copied into the entry buffer, so it is >+ * @param buf is copied into the entry buffer, so it is the > * responsibility of the caller to free it. > */ > void a_Cache_entry_inject(const DilloUrl *Url, >@@ -765,9 +765,8 @@ static Dlist *Cache_parse_multiple_fields(const char *header, > j--; > field = dStrndup(header + i, j); > dList_append(fields, field); >- } else { >- while (header[i] != '\n') i++; > } >+ while (header[i] != '\n') i++; > } > > if (dList_length(fields) == 0) { Thanks!, merged in two commits with a comment: https://git.dillo-browser.org/dillo/commit/?id=7ed29a43c18a960946fe47f37004… https://git.dillo-browser.org/dillo/commit/?id=9b7b799c7e47ca6e2b91aec82bd5… Best, Rodrigo.

June 21, 2026

Re: Nordstjernen web browser

by Bobby Hiltz

> Banned, > Rodrigo. Probably the best option. It is totally possible that that email account is a bot too. The commits haven't stopped! I am annoyingly anti-AI, and it appears Lobsters had the same feelings about this project (https://lobste.rs/s/adztcr/nordstjernen_web_browser) And the claims? Stable? Written from scratch? Private? LEGENDARY? Good call. Thanks, Bobby

June 16, 2026

Re: Nordstjernen web browser

by J. M. Castroagudín Silva

On Mon, Jun 15, 2026 at 7:09 PM Rodrigo Arias <rodarima(a)gmail.com> wrote: > > Hi, > > Hmm, from the repo: > > >Minimalism: The whole engine is about 116,000 lines of C (plus a thin > >C++ Qt shell) — small enough for one person to read and audit > >end-to-end. > > >Don't ask "do you want me to proceed?", "should I continue?", "ready to > >commit?" — just do it. The user interrupts if they disagree. > > >Don't leave traces that the code has been written by AI LLM, in commit > >messages and git author and git committer. And yet, there are countless commits made by claude: https://github.com/nordstjernen-web/nordstjernen/commits/main/ Amazing, the amount of resources burned for a technology so damn abysmal It would be funny if it wasn't so sad :( > > Burn baby burn! > > https://www.theguardian.com/environment/2026/feb/13/elon-musk-xai-datacente… > > Banned, > Rodrigo. > _______________________________________________ > Dillo-dev mailing list -- dillo-dev(a)mailman3.com > To unsubscribe send an email to dillo-dev-leave(a)mailman3.com -- Saúde, J. M. Castroagudín

June 15, 2026

Re: caniuse for Dillo?

by Rodrigo Arias

Hi, On Mon, Jun 15, 2026 at 10:51:54PM +0400, Artyom Bologov wrote: >Hi y’all, > >I want to make sure my site opens in Dillo in perpetuity, so I’m >interested in what features of HTML and CSS Dillo does or does not >support. Thanks for taking that into consideration! >Is there any information on that? Can it be made accessible on >the Web somehow? Should it be relegated to caniuse or MDN? We don't have an equivalent list (at least not updated). I think the simplest method is to just try from time to time in Dillo and text-only browsers like links. As a general rule, test your site without CSS and avoid new CSS features like var() for sizing/margins (or use a fallback). Thanks!, Rodrigo.

June 15, 2026

Re: Nordstjernen web browser

by Rodrigo Arias

Hi, Hmm, from the repo: >Minimalism: The whole engine is about 116,000 lines of C (plus a thin >C++ Qt shell) — small enough for one person to read and audit >end-to-end. >Don't ask "do you want me to proceed?", "should I continue?", "ready to >commit?" — just do it. The user interrupts if they disagree. >Don't leave traces that the code has been written by AI LLM, in commit >messages and git author and git committer. Burn baby burn! https://www.theguardian.com/environment/2026/feb/13/elon-musk-xai-datacente… Banned, Rodrigo.

June 15, 2026

caniuse for Dillo?

by Artyom Bologov

Hi y’all, I want to make sure my site opens in Dillo in perpetuity, so I’m interested in what features of HTML and CSS Dillo does or does not support. Is there any information on that? Can it be made accessible on the Web somehow? Should it be relegated to caniuse or MDN? Best of luck, -- Artyom Bologov https://aartaka.me

June 15, 2026

Re: dpi problem

by Rodrigo Arias

Hi, On Mon, Jun 15, 2026 at 02:33:24AM +0200, Teddi Vazquez wrote: >Hello, I'm trying to add the plugins for gopher and gemini, when i run >the "make install" the next message appears in both: > >echo 'proto.gopher=gopher/gopher.filter.dpi' >> /root/.dillo/dpidrc Don't install with root (sudo) use just "make install" as your user. >dpidc stop || true >** ERROR **: [Dpi_read_comm_keys] No such file or directory >** ERROR **: main: Can't read dpid's port number These errors are expected when there is no dpid daemon running, which is fine if you don't have dillo open. After that "make install" it should work. Otherwise let us know. Best, Rodrigo.

June 15, 2026

Re: Nordstjernen web browser

by a1ex＠dismail.de

Andreas Røsdal <andreas.rosdal(a)gmail.com> wrote: > Hi, I am making the Nordstjernen web browser. > > Implemented in C. It's like Dillo, small and simple. > > https://nordstjernen.org/ > https://github.com/nordstjernen-web/nordstjernen > > I hope it's interesting for you, the internet needs more small independent web browsers. I'm all for more independent web browser alternatives, but there appear to be a couple of issues which would make this a non-starter for many people in this community: 1. Proprietary "source-available" style license. This does not make the project attractive for potential contributors, and users may also be resistant to participate in such a model. 2. Appears to be largely "vibe-coded". You make statements like "clean-room web browser written from scratch", but then there is this in 'Agents.MD': "Don't leave traces that the code has been written by AI LLM, in commit messages and git author and git committer." Some would argue that it can't be "clean-room from scratch" when the AI models you are using have been trained on other peoples work... Regards, Alex

June 15, 2026

Nordstjernen web browser

by Andreas Røsdal

Hi, I am making the Nordstjernen web browser. Implemented in C. It's like Dillo, small and simple. https://nordstjernen.org/ https://github.com/nordstjernen-web/nordstjernen I hope it's interesting for you, the internet needs more small independent web browsers. Andreas

June 15, 2026