Subtle changes, bugs and feedback

I had the same problem with that CRC link, same set-up.

I thought it may be an affliliate link problem but clicking through the affiliate link page all seem to work.

@Velocio in case you haven't had an update.

Are you sure you don't have an adblocker so aggressive that it is preventing the link be opened? Or that you aren't using a hosts hack so total that it prevents the destination being reached?

What extensions are you using?

I'm using Chrome 49 with uBlock Origin 1.6.6.

When I click the CRC link above, CRC choose to send the link via DoubleClick and the domain ad.doubleclick.net , once they've done that it redirects to CRC. This is a CRC thing, they're redirecting some referrer traffic via DoubleClick to get metrics and traffic analysis.

So if your adblocker or hosts file is killing DoubleClick, then perhaps the browser detects the link goes nowhere and closes the tab.

I use uBlock Origin so that things like DoubleClick are blocked by default, but once in a while when I click a link that goes via it I have the choice as to whether to allow it just that once (which I usually do because at that time I'm following an affiliate link and realise this is how those small web sites make some of their money).

I was using adblock plus with default settings and block list. If I disable the extension then the link works correctly.

Would you recommend uBlock over adblock?

Definitely.

It's faster, uses less memory, can be as simple to use or you can embrace it more and gain way more control.

I like how it's not black/white about things... you can leave it in a "prompt me" mode.

I also like how I can basically turn it into NoScript. I've found for example that the Guardian web site has some of the most invasive trackers out there, so I set that to only allow "first party scripts"... literally nothing not on their domain and page will work. Which makes the site a joy.

I've been moving LFGSS to be first party as well... it's not totally there yet but most things will work with only first party scripts enabled. The things that don't work are embedded maps and videos in comments or on events.

But yes... uBlock Origin is the current state of the art in adblocking.

Find it here: https://chrome.google.com/webstore/detail/ublock-origin/cjpalhdlnbpafiamejdnhcphjbkeiagm?hl=en

I've swapped over to origin. Is there any boxes/setting you'd recommend changing? Also, are google-analytics bad or do they just tell you things?

Not really.

I pretty much leave the defaults.

I've whitelisted a few sites, including LFGSS... but these are ones that I truly know every single request for and have total confidence that there aren't trackers (beyond Google Analytics which is blocked by most people anyway).

To whitelist something you click the icon and press the big power button and then reload the page. Simple as.

Looking down my whitelist it's basically domains I own, sites I run, and a few Google ones (which I run in a different session, tracking cookies sandboxed from the rest of my browsing).

I'm currently running Ublock and Noscript (from when I was using Adblock and Noscript and Ublock replaced Adblock).

Overkill do you think or worth using both?

I find uBlock enough by itself.

But... I caveat that with my browsing habits very likely being different from yours.

If you run a single non-incognito window for everything, I'd probably still consider NoScript.

What I do (desktop):

• Chrome = Things I trust a lot: Gmail and Google drive, etc and LFGSS. Nothing else.
  
• Chrome Incognito = Things I trust a little: All social media (Twitter), sites I sit on most of the day (Hacker News), some news sites (BBC)
  
• Firefox in forced Private Browsing mode = Everything else.
  
  

I literally receive a link in an email, and will copy it to the right browser to open it. I only click links that stay within the realm of trust in which I'm working.

On my Android phone the above is replaced by:

• Chrome = Things I trust: LFGSS
  
• Firefox = Things I trust a little (uBlock Origin works, but no private browsing by default so I'm selective about what I open in Firefox)
  
• Dolphin Zero = Everything else (all browsing data thrown away after viewing each page)
  
  

I find compartmentalisation is pretty good even when I do stray into horrible stuff, but this doesn't necessarily protect me against malware which NoScript would. But, as a Linux user running uBlock I think I've mitigated that risk enough to browse without worrying too much about it.

If I were not compartmentalising my browsing by trust, then I'd definitely be a bit more paranoid about things.

The risks I see are mostly an untrusted site, somehow exploiting something in a trusted site... hence compartmentalising.

All good sense.

But no one got time for that shit. Just steal my life already.

Cheers. Given I'm on Windows and non-incognito most of the time I'll stick with both.

Hi all, I'm sure this has been covered before but I have a problem: want to login on iPad in chrome - can't. The normal sign on things happen, arrow goes right, accepts my details, but then when I go back to the homepage I'm not logged in.

I type this from Safari on my iPad so it's not the hardware, I thought it might be being logged in on two browsers at once but I've logged out and closed safari whilst trying to login on chrome, and still have the same problem.

Please help!

Cookies enabled?

@Velocio

Do you know that the quote function is still borked for quite a few people and definitely for me on across chrome?

Action:

Select text in a post and hit reply.

Symptom:

A reply box opens with the quoted text in it, but with the "quote chevron" at the end of the text, instead of at the beginning.

I know this is a problem for others as well due to the number of posts I see that have malformed quotes in, of this form.

Please help>

E.g.

Same behaviour in vivaldi.

Do you know that the quote function is still borked for quite a few people and definitely for me on across chrome?

It's borked for everybody on all platforms :-)

Quote attribution was better on Usenet 30 years ago than it is on this forum.

What gives with this?

It's a pretty clear and annoying bug that has widespread impact and can't be too hard to grok, surely.

Why has @Velocio always been silent on it?

Given a how-to, I'd be happy to get a dev environment up and hack on it myself, but I know that that is not straightforward.

David, throw us a fricken bone here?

Why has @Velocio always been silent on it?

Because if he fixes it, he's admitting that there is a fundamental flaw in his vision that fora should be different from how they were for the first few decades of internetz. If forum posts are ephemeral, it probably doesn't matter that quote attribution (and with it the ability to follow the thread of a discussion easily) has been discarded. The archive of rec.bicycles.tech suggests that fora are not ephemeral, they have a long tail where strangers from the future can see how generous Sheldon and Jobst were in giving their time to educate people.

No, that is tosh. In fact it borders on petulance.

We just need a chevron moved from the end to the beginning. My money is on a regular expression that needs tweaking.

To be clear, I make no comment at all on the theory and implementation of quote attribution on microcosm, other than in the minor and prosaic bug I describe above.

To be clear

It would have been clear the first time if you had quoted the part of my post to which you were responding :-)

It's a pretty clear and annoying bug that has widespread impact and can't be too hard to grok, surely.

Why has @Velocio always been silent on it?

Given a how-to, I'd be happy to get a dev environment up and hack on it myself, but I know that that is not straightforward.

David, throw us a fricken bone here?

I haven't always been silent on it, I recognise it's a bug. But, I'm not a specialist in JavaScript and when Microcosm ran out of money some bugs are harder for me to fix than others, and this one is one of the hard ones for me to fix. I've asked in the past that anyone skilled at JS to look at it, no-one has... so that's where we are.

You're welcome to look into it and create pull request, the code is ultimately in this repo: https://github.com/microcosm-cc/microweb/

Because this is purely client side you don't even need to spin up a full dev environment with database, virtual machines, load balancer, API, etc... all you need do is save a complete web page that has a comment form on it and then edit the JS files according to your debugging.

It's probably this section of the file: https://github.com/microcosm-cc/microweb/blob/master/core/static/js/Markdown.Editor.js#L1526-L1668

If it's not that, then it's maybe this: https://github.com/microcosm-cc/microweb/blob/master/core/static/js/simpleEditor.js#L130-L132

My money would probably be on the latter, but therein lies my failure to fix it... it looks perfectly fine to me. I'm baffled by it.

Not sure what the other messages about, those seem a bit rage-fest. There's no petulance, and I don't understand the context of the last message. If you want to help out that would be great. If you do feel you need a full dev environment I'll pop round and set one up for you, you will need a base Linux system with Virtual Box, git, Go1.6 and vagrant installed.

Oh, and the comments form is here:
https://github.com/microcosm-cc/microweb/blob/master/core/templates/forms/create_comment.html

And the editor itself was called pagedown, the demo is here: https://pagedown.googlecode.com/hg/demo/browser/demo.html

The old repo here: https://code.google.com/archive/p/pagedown/

An updated Git repo here: https://github.com/ujifgc/pagedown