~rwa

.bay.de

https://clttr.info

mostly "ops", sometimes doing some "dev" just for the fun of it

Mastodon

Gemini Space

Recent activity

Re: gmni: SSL error 296 3 months ago

From René Wagner to ~sircmpwn/gmni-discuss

Seems like other clients (which use LibreSSL for example)
are also hit by this.

The assumption is that it's caused by ed25519 based 
ciphers which currently seem not be supported by all libs.
BearSSL should support ed25519, but i'm not sure what
this "32-bit" should mean:
https://bearssl.org/todo.html

...
EC Curve25519 (32-bit) 	Done (0.3) 	
EC Curve25519 (32->32 only) 	Done (0.3) 	
EC Curve25519 (with __uint128) 	Planned
...

gmni: SSL error 296 3 months ago

From René Wagner to ~sircmpwn/gmni-discuss

Hi,

when trying to get gemini://discovery.geminiprotocol.com/ with gmni it fails with 

SSL error 296
Error: SSL error

I'm not sure why this happens, the capsule seems to work fine with other clients.

I've tested as of commit 93f30522 build against BearSSL 0.6 on ArchLinux.
Can anyone confirm this issue?

Regards
René

[PATCH gus] stop fetching when 5 consecutive requests failed 3 months ago

From René Wagner to ~natpen/gus

Sometimes capsules stop working during a crawl which results
in requests waiting for a timeout. This is especially annoying
when we try to fetch a huge amount of pages from these capsules.

The setting is only in effect during the current crawl and will
start over fetching these capsules on a new crawl.
---
 gus/constants.py |  1 +
 gus/crawl.py     | 17 ++++++++++++++---
 2 files changed, 15 insertions(+), 3 deletions(-)

diff --git a/gus/constants.py b/gus/constants.py
index 90ad2a6..d5b8a44 100644
--- a/gus/constants.py
[message trimmed]

[PATCH gmni] fix display of message on TOFU_FINGERPRINT_MISMATCH 3 months ago

From René Wagner to ~sircmpwn/gmni-devel

Previously the message was never displayed to users
leaving them with a simple "Error: certificate is untrusted".

This also fixes the display of line numbers in the message.
---
 src/gmnlm.c | 2 +-
 src/tofu.c  | 3 +++
 2 files changed, 4 insertions(+), 1 deletion(-)

diff --git a/src/gmnlm.c b/src/gmnlm.c
index 91e86f0..ee808d0 100644
--- a/src/gmnlm.c
+++ b/src/gmnlm.c
@@ -1022,7 +1022,7 @@ tofu_callback(enum tofu_error error, const char *fingerprint,
[message trimmed]

Re: [PATCH gmni] show proper message on TOFU_FINGERPRINT_MISMATCH 3 months ago

From René Wagner to ~sircmpwn/gmni-devel

Of course, the commit message was somewhat short.

----
@@ -1106,7 +1106,7 @@ tofu_callback(enum tofu_error error, const char*fingerprint,
 		free(host);
 		break;
 	case TOFU_FINGERPRINT_MISMATCH:
-		snprintf(prompt, sizeof(prompt),
+		fprintf(browser->tty,
 			"The certificate offered by this server DOES NOT MATCH the one we have on
----
Actually show the message to the user with fprintf.
The "prompt" variable put togehter with snprintf is only used when reaching
line 1040 which in this case isn't reached due the return in line 1035.

[PATCH gmni] show proper message on TOFU_FINGERPRINT_MISMATCH 3 months ago

From René Wagner to ~sircmpwn/gmni-devel

---
 src/gmnlm.c | 2 +-
 src/tofu.c  | 3 +++
 2 files changed, 4 insertions(+), 1 deletion(-)

diff --git a/src/gmnlm.c b/src/gmnlm.c
index 930ab19..1292bb6 100644
--- a/src/gmnlm.c
+++ b/src/gmnlm.c
@@ -1106,7 +1106,7 @@ tofu_callback(enum tofu_error error, const char *fingerprint,
		free(host);
		break;
	case TOFU_FINGERPRINT_MISMATCH:
		snprintf(prompt, sizeof(prompt),
[message trimmed]

[PATCH gus] robots.txt sections "*" and "indexer" are honored 3 months ago

From René Wagner to ~natpen/gus

We no longer use the "gus" section for ease of implementation.
It's probably barely used anyway.
---
 docs/handling-robots.md |  1 -
 gus/crawl.py            | 16 ++++------------
 2 files changed, 4 insertions(+), 13 deletions(-)

diff --git a/docs/handling-robots.md b/docs/handling-robots.md
index 4a80383..e1c602b 100644
--- a/docs/handling-robots.md
+++ b/docs/handling-robots.md
@@ -4,7 +4,6 @@ robots.txt is fetched for each (sub)domain before actually crawling the content.

GUS honors the following User-agents:
[message trimmed]

Re: [PATCH gus] add verbose search to robots.txt 3 months ago

From René Wagner to ~natpen/gus

Hi Natalie,

do you object to this patch or just overlooked it? :)

kind regards
René

Re: [PATCH] correctly handle robots.txt 3 months ago

From René Wagner to ~natpen/gus

Hello again :)

After letting my thoughts settle for a few days i decided
that geminispace.info will not use the "gus" section in
robots.txt anymore.
It will still honor * and indexer sections like before.

This allows for a straight forward implementation without
the need for hacks around the robotfileparser. 

I don't know if this change has any practical impact. 
Unfortunately robotfileparser has no method to dump the
content of the robots.txt after reading and i don't
want to bother with it anymore.

Re: [PATCH] correctly handle robots.txt 4 months ago

From René Wagner to ~natpen/gus

Good evening ;)

> Ahhhh, I didn't realize that flaw existed in the prior implementation.
> I wonder if we could modify the original implementation by inserting a
> small conditional in front of the overrides to check if the
> more-specific value is even present in the robots.txt file (if it's
> not, then we should skip the override, since it would just errantly
> override with the value of `*`).
> 
> Hmm. That's my initial thought for the "best" way to handle it, but I
> think I first need to check if there's even an easy/supported way to
> check if a user-agent is explicitly specific with Python's robots.txt
> functionality.
>