Subtle changes, bugs and feedback

With faff...

1. Go to the users page and note their authorId, the 4-6 digit number in the URL. eg @Backstop is user 72105.
   
2. Go to the thread you want to search and start a search within it (RHS, just above following, leave the search field blank) eg a blank search of this thread
   
3. in the URL add &authorId=XXXXXXX to the end (note the capital I) eg a list of posts in thread 131364 authored by user 72105
   
   

Alternatively

1. Go to the thread and note the conversation id (in the URL, this thread is id=131364)
   
2. Go to the user's page and click through to the list of their comments (number in the top right, eg @Backstop's comments are here)
   
3. add &id=XXXXXXX to the end of the URL (eg a list of posts by Backstop in this thread)
   
   

All the details are here if you're curious.

That worked, thanks.

When I click 'following' I get this message:
'Not Found

Sorry, we couldn't find that. It might have been deleted.'

Any ideas?
I'm not ruling out user error.

Oooh, I deleted some stuff this morning post-merge. Let me clear some caches and see whether that resolves it.

Try now.

Sorted, thank you.

Getting page crashes and reload on posts with many images on iPhone.

This triggers it

https://www.lfgss.com/conversations/308000/newest/

Paging buttons still piquing my OCD.

Le Sigh.

I know what this is, but it is taking time to write the code to resolve it.

Explanation: LFGSS resizes large images but actually doesn't generate thumbnails today as that was a good way to also preload the images when you want to view them. This puts a strain on iOS devices which only reserve a very low amount of RAM for a page. When we increased the size of attachments people started taking photos and submitting them direct from their phone, and so they are larger. Meaning that a page with many attachments today may take a lot more memory than a page in the past. This causes an iOS tab to run out of memory, and then the process is killed, which triggers the browser to reload it.

The fix: I need to actually generate thumbnail images that are much smaller, and only show the full size if clicked.

Gotcha. Sounds like a royal pain! Hope it's not too much of a time suck to resolve.

Not sure if you announced a change and I missed it, but external links are now not showing the full address when you hover on them, instead they just give you the microcosm.app url. So now I can't see if something is potentially NSFW while browsing at work.

Obvious answer would be to stop browsing at work I suppose...

On that note I'd really like it to be a per user configurable option to have preview links for microcosm.app links, e.g. when enabled a user gets preview.microcosm.app links which takes them to an interstital page which shows what the final link is (which can still be a microcosm.app link if you want to mine what links are really being clicked...)

I've tripped the work claxons a few times clicking on some links with no idea in advance where they go.

Oh - LFGSS gone yellow - whats all that about ?
(been away you see)

Super @skinny just won TCR.

I've just noticed the yellow and wondered what it was for

Yay, TDF yellow.

In August.

and congratulations skinny

LFGSS gone yellow - whats all that about ?

And what colour does the logo change to when @hippy 's ragged gooch finally crosses the finish line?

Sometimes it takes ages, as in more than 10 seconds for a microcosm.sm/out link to resolve. I usually end up hitting refresh a few times when this happens.

I would estimate that is happening about 3% of times that I click an external link.

Chrome on Android 8, but I've had this for quite a while on microcosm. Probably over two years. Always been chrome on android.

@Velocio not getting the code to sign in on my work computer, using chrome, cleared the cache/cookies yesterday and its not sending the email.

Sometimes it takes ages, as in more than 10 seconds for a microcosm.sm/out link to resolve. I usually end up hitting refresh a few times when this happens.

It's the third party site.

microcosm.app resolves (and so is the status field) but the status field doesn't show the delay in resolving the DNS and connection to the third party until only after that connection has been made.

not getting the code to sign in on my work computer, using chrome, cleared the cache/cookies yesterday and its not sending the email.

On August 9th your email bounced in such a way from Yahoo that it caused the email gateway to consider the email address dead.

I'll remove the bounce knowledge and it should start working again.

At least once a day, I get this HTTPS fail. Couple of refreshes and it goes away. Not sure if it's all the corporate MITM proxy's fault.

1 Attachment

Very odd, but I have seen something similar with other Cloudflare sites when accessing from Chrome too. I am wondering whether there is something odd at our side and so have asked internally.

It would be great if you were using Chrome and could confirm it still happens once per day, and when it does a screenshot of the Chrome page (which is more informative than the IE generic error page).

^That's a Firefox build, but I'll see if it happens in Chrome.

Ah Firefox.

In which case this could be a TLS1.3 issue as your workplace proxy possibly isn't happy with it.

You can disable TLS1.3 in Firefox and see if that solves it... but it's probably better to leave it enabled and just to deal with the refreshes when this happens.