Downtime, bugs, and failures on any kind of service (email, web, XMPP, etc)

Some email that was sent to me via a burnermail.io account never reached me. The sender got no bounce, so the message just went into a silent black hole. Of course whenever this happens there is no way to know whether the msgs were lost by the forwarding service or the email hosting provider. Hence why I am asking.. would like to see if there is a pattern.

Worth noting that spamgourmet.com is very unreliable. Messages get dropped silently all the time. But until recently I was unaware of any possible issues with burnermail.

If you visit: https://12ft.io/$URL_to_pdf.pdf using a GUI browser, the raw PDF binary is dumped to the screen. There is no way to capture this. If you use wget it just gets an HTML wrapper. If you hit F12»inspect»element, you can derive a proper URL to a PDF and use wget on that. E.g.

wget 'https://12ft.io/api/proxy?q=https%3A%2F%2Fmswista.files.wordpress.com%2F2015%2F04%2Ftypesofmemory_updated.pdf'

But the PDF is corrupt. There is no user-side hack here. The service is broken. Apparently the server is doing a character set conversion as if it’s ascii text.

(BTW, that sample URL above works fine without 12ft.io. It’s just an example to demo the 12ft.io problem. Of course when a PDF is walled off and I am forced to use 12ft.io, then I’m hosed)

The admin is only reachable in Twitter and Gmail, nethier of which work for me. The is a Mastodon bot at @thmsmlr@bird.makeup but that’s only good for following him. No way to report it to him AFAIK. Hence why I am posting this here.

In both Lemmy 0.19.4 and Lemmy 0.19.5, you click the magnifying glass to open the search dialog. If you enter a search query and tab out of the field, whatever you typed is cleared. Even if you simply hit <enter> without tabbing out of the query field, the search form is refreshed and it tells you enter a query, as if you had not done so already.

Both versions have this problem with Ungoogled Chromium. The problem does not manifest on Tor Browser.

This bug was introduced with version 0.19.4 and still persists in 0.19.5: There are four possible timeline views:

• subscribed
• local
• all
• moderator view

That selector is broken in Ungoogled Chromium 112.0 but not in Firefox-based browsers. In UC, clicking “moderator view” highlights the button, the page refreshes, but the selector does not stick. It snaps back to whatever view is the default and remains trapped on that timeline.

This problem is replicated in both 0.19.4 and 0.19.5 instances.

If I use the cross-post feature to copy the post elsewhere, the form is populated just fine but then I have to search for the target community at the bottom of the form. As soon as I select the target community, the whole rest of the form clears. If another field is re-populated, the target community field clears. So only one field at a time can be populated.

Tested with Tor Browser.

Untested in Lemmy 0.19.5.

Lemmy 0.19.4 introduced a quite serious defect whereby if you are using Ungoogled Chromium (and perhaps stock Chromium), the form to create a new post accepts input but then the instant you tab out of the field, the whole field is cleared. Poof… just like that, all your work vanishes and no way to get it back.

Firefox-based browsers have no issue.

~~Lemmy 0.19.5 seems to have fixed it.~~ But there are other problems with both 0.19.4 and 0.19.5, so I suggest not upgrading past 0.19.3.

(edit) actually the problem manifests differently in 0.19.5. The form can be filled out and there is no data loss, but the “create” button is insensitive. It remains gray and behaves as if the form is still empty.

The “disobey”¹ onionmail server has been accepting my POP3 logins without issue for months/years. There has been “no new messages” for as long as I can remember and I have also not sent mail for a long time. Then I tried sending myself a message and I get “500 Mailbox full”. Yet my inbox is empty.

It’s quite disturbing because I have no idea when the admin apparently decided out of the blue to delete my account. It might have an automated removal, perhaps due to such sparse/rare traffic. But regardless, it makes it hard to trust any #onionmail server because they all run the same code. This same scenario occurred on another onionmail server as well.

Does anyone here use onionmail?

¹ a5dkbvgakon2lxmauleiizkv6i3s36wp6w3i32a3buc4xmtdnbttmryd.onion

Searching for this link from sopuli.xyz yields nothing:

https://mastodon.social/@madeindex/112441311556219112

So there is no way to participate in federated discussion from Lemmy if it originates using Mastodon software.

cross-posted from: https://sopuli.xyz/post/12874750

Mastodon used to show people the mirrored version of federated content which shielded users from Cloudflare’s discriminatory blockade. But something apparently changed. If I try to visit this mirror of a mastodonapp.uk status on layer8.space:

https://layer8.space/@tmmj@mastodonapp.uk/112387605497275701

it redirects to:

https://mastodonapp.uk/@tmmj/112387605489133663

which is apparently a shitty Cloudflare node that deceives us into thinking the account does not exist. If you are logged into the mirrored node, then it does not redirect and you can see the content. Of course, only if you have an account on the mirror which means anonymous viewing is no longer possible.

If I want to share that layer8.space link with other people, it would be an injustice to share the mastodonapp.uk link because it’s in a walled garden that excludes people. It would be like sharing a Facebook link with an audience that includes people outside of Facebook. So naturally I would share the layer8.space version because layer8.space allows all people to visit. But now this is impossible. Cloudflare’s stranglehold of control has been increased by this Mastodon move.

Worse, Cloudflare has started pushing error code 404, not 403. So CF is misrepresenting the error to suggest that the page does not exist. Cloudflare has carte blanche in fucking up the web. A 404 error is supposed to inform users that an object is not found, not that they are not authorised to access it.

The attached image is what Cloudflare-excluded people see when trying to visit this image:

https://files.mastodonapp.uk/media_attachments/files/112/387/580/865/787/635/original/f4442c8789ad52c2.png

This was the onion service for #underwood email back in the onionV2 days:

underwood2hj3pwd.onion

An onion v3 replacement never got publicised AFAIK, so the service died when Tor Project pulled the plug on v2. Anyone have the updated onion host?

#Mail2Tor offers these onion services:

• webmail - works (superficially… seems unreliable)
• IMAP - non-responsive for years (server: g77kjrad6bafzzyldqvffq6kxlsgphcygptxhnn4xlnktfgaqshilmyd.onion:143)
• SMTP - authentication problems for years

They list “updates” on the linked website have also have not changed for years. It seems someone is asleep at the wheel. It’s impressive that the webmail service runs at all despite the apparent neglect (though the reliability still seems questionable).

help plz

If someone would create a mail2tor account and test the IMAP and SMTP servers, it would help to get some confirmation so I can rule out whether the problem is on my side.