Background: we've noticed that some Pitch customers (especially in the US and Australia) cannot access our domain name. We suspect that this is because their ISP or network router has a security feature that is intercepting DNS queries and returning bogus results. Any experience with DNS blocklists also appreciated

EU based and it worked.

UK access is working (using Chrome private tab on Linux OS)

Works here (Finland, FF on Linux)

Heavily locked down FF on MacOS in Sweden. Works fine.

UK Firefox mobile works

Works fine from Uruguay

Worked fine from Norway (wifi provided by Eduroam)

works from Wisconsin in the US with an iphone. My ISP is AT&T and I'm using the fiber modem/router combo they issued me

Works for me - UK, Scotland, BT Business ADSL

Works in Brasil

Worked from Antigua

Worked from Texas

Worked from Québec, Canada, with NextDNS and Bell ISP

Ok from Ontario, Canada ISP: Rogers

Worked for me. Country: United States ISP: Verizon Router: Asus RT-AX88U

Seems fine. Groton, MA Verizon FIOS

Thanks everyone for the help, I appreciate it. Looks like it's an localized issue for users of certain ISPs using Plume Homepass or similar invasive dns-based content filters

...Part of my motivation is that, as it's pretty common, I just don't enjoy designs that change every year. Also, 80% of my time spent on Chrome goes to the same handful of webs/apps. I'd appreciate some visual consistency between them, and given they're so few, it seems at hand.

I would say the closest analog to this that I've seen is https://www.datarabbit.com/.

Some of the best examples are on https://twitter.com/ryrobes

Looks like I missed out on it! Let's see

For hacker news, https://github.com/djblue/portal/blob/master/src/examples/hacker_news.cljcis one way to browse it via https://clojuredocs.org/clojure.datafy/nav😂 This should work with all tools that support nav.

Nice! I adapted that ns and ran it through the CIDER inspector. This is the homepage:

I can inspect the nth number, but that's just a java Long, useless in itself. I reckon that one wants to tell this Inspector that I want to nav the sub-item at index N? Is that what Portal does? i.e. offer a "nav this item" command

Yeah, in Portal when you double click an item, it will invoke nav and display the returned value. I think the default implementation of nav just returns the value passed in, so it's always safe to call 👌

You do need to provide the parent collection, key and value for nav. So you need to keep track of that info for when a user does try to nav.

It occurs to me that this is this line of thought is how the World Wide Web was created. You have a generalized piece of software (Web browser vs. REBL/Portal/CIDER) communicating via a standard format (HTML vs EDN) using a standard protocol (HTTP vs. NREPL). The datafy/nav protocols in clojure are a general step in the direction of hypermedia controls (links and form controls).

Yes, with a bit of luck a community like ours could help getting the Web back to its roots :) Some problems are less clearly solved than others (forms, form validation) but it seems doable

imho, using the web as a foundation is the problem. glomming more stuff onto a web doesn't make it less of a sticky mess.

What if one used REST (or graphql, w/e) APIs only? So a Portal view would be like a SPA, except that generic, data-centric and hacker-friendly

The problem isn't how apps get data (over the network or otherwise), it's that UIs keep being built on an old, rickety OO foundation. Fundamentally, there's no proper data model for views and that cascades and infects every other aspect of UI programming.

The Web/HTTP was the motivating project of Fielding's REST dissertation so technically the Web does use REST APIs only. One of my pet ideas (conjectures?) is that the addition of datafy/nav made clojure itself a REST application (in the Fielding sense).

There are a number of browser extensions that manipulate Facebook pages so maybe have a look at the source of those? Facebook Purity is one I use, but there's also Social Fixer (which my wife uses), and a few others. One thing I'll say about them is that Facebook regularly makes changes to break such extensions by trying to defeat JS-level manipulation of their rendered pages...

manipulate Facebook pagesThis is a good lead. I'll look into those. I kind-of have a fix for the underlying issue that I had (not being able to select country in the selector). The fix is to use the search bar search and just edit the URL as it's free text. They can refactor the free-form URL search but they can't refactor the Facebook ID from their Cassandra. :thinking_face: And they use this facebook ID instead of city names. This way I circumvent all the HTML parsing for this. And this makes it a little more robust than class-based selectors.

Scicloj has a bunch.

https://scicloj.github.io/docs/community/groups/

https://clojure.org/community/user_groups

@U05MPHVTGPK would you mind sharing the info on the distributed systems group? That sounds interesting.

https://www.the-paper-trail.org/page/dist-sys-slack/

http://distsysreadinggroup.slack.com