Of Interest....

[Dronecatcher]

"Your browser may not be compatible with all the features on this site."

We should all be very concerned if some websites are implying that anything other than a Chromium-based browser is now obsolete. The prospect of an Internet that is only accessible with Chromium-based browsers should horrify everyone.

cheapskatesguide.org

I've experienced a large dose of this kind of thing this afternoon - after being excited to discover I could set up a Mastodon (Twitter alternative) account in Safari 4 - despite the glaring https in the URL and setting up all the preferences and adding graphics etc the illusion burst as soon as I clicked the feed....an empty page.
Same with InterWebPPC and TFF, same on Leopard with WebKit - no error dialogue just a void.

 

[redheeler]

It’s sad to see web developers become lazy and not want to maintain compatibility with older browsers, but it’s also not unexpected.

 

[Dronecatcher]

t’s sad to see web developers become lazy and not want to maintain compatibility with older browsers, but it’s also not unexpected.

That's true but I find it bizarre for a developer to make the sign up, configure, customise and login sections compatible but the actual feed not.

 

[B S Magnet]

Given earlier hints of this “blank” response appearing on interactive web pages not rendering (such as, for example, when viewing Archive-dot-org’s front page on TFF and IW-PPC stopped being possible many months back, and Interweb/FF45.9ESR/Nightly 52 on Intel not rendering the front pages of major journalism front pages relying heavily on XHR), it became apparent how the browser engine libraries used for handling/rendering XHR/JS have moved forward to such a level that there are now these graceless drops in render support for older browsers (whose JS-handling lacks means to parse said-latest XHR/JS).

If anything, the gap between A) the falling-off of JS development back-compatibility in current standards; and B) the hitting of hard limits on JS/XHR back-porting (to enable older browser engines to keep up gracefully, if slowly), has C) widened and produced this canyon of unsuccessfully-rendered “current” interactive pages by legacy browsers.

“Graceful degradation” is no longer prioritized on a widest-possible/greatest-common-factor reach so much as it serving a hindrance for advancing current industry standards-based deployment for generating active/interactive pages (including tracking user behaviour within a page and returning that telemetry to the server).

It’s, idk, very disappointing.

[As for Mastodon, I left social media a couple of years ago, including Mastodon instances, so I haven’t tried to, say, log in to my old account on any of my older (or even my newer) browsers in a minute. I do remember being, briefly, on at least one instance, one which used a different code base than Mastodon itself, which seemed to cause more issues on non-current browsers, but I no longer have the login to that instance, much less remember which GNU/Social-based engine they used.]

 

[Dronecatcher]

“Graceful degradation” is no longer prioritized on a widest-possible/greatest-common-factor reach so much as it serving a hindrance for advancing current industry standards-based deployment for generating active/interactive pages (including tracking user behaviour within a page and returning that telemetry to the server).

Thanks for developers explanation of the blank voids - shame there was no "your browser doesn't compute" warning before I went through the process of setting the account up.

I'm not particularly fussed about Mastodon - was just excited by the possibility that something contemporary would work in Safari 4!

The spec creep on browsers is accelerating - I lasted more than twice as long on Leopard as I did with Snow Leopard, then after SL, I lasted a year on EL Capitan before having to shift to High Sierra on which I'm now having to use multiple browsers to get by.

At this rate, before long we'll be on new hardware every year - unless you want to live on the fringes of society and be considered "unmutual."*

*

 

[redheeler]

That's true but I find it bizarre for a developer to make the sign up, configure, customise and login sections compatible but the actual feed not.

The feed page likely makes a huge jump in complexity with the amount of JavaScript for dynamic content loading. I’m sure they didn’t test any of the pages with the older browser recently, but only the feed page is so JavaScript-heavy and constantly evolving that the chance of something breaking goes way up.

 

[redheeler]

Given earlier hints of this “blank” response appearing on interactive web pages not rendering (such as, for example, when viewing Archive-dot-org’s front page on TFF and IW-PPC stopped being possible many months back, and Interweb/FF45.9ESR/Nightly 52 on Intel not rendering the front pages of major journalism front pages relying heavily on XHR), it became apparent how the browser engine libraries used for handling/rendering XHR/JS have moved forward to such a level that there are now these graceless drops in render support for older browsers (whose JS-handling lacks means to parse said-latest XHR/JS).

If anything, the gap between A) the falling-off of JS development back-compatibility in current standards; and B) the hitting of hard limits on JS/XHR back-porting (to enable older browser engines to keep up gracefully, if slowly), has C) widened and produced this canyon of unsuccessfully-rendered “current” interactive pages by legacy browsers.

“Graceful degradation” is no longer prioritized on a widest-possible/greatest-common-factor reach so much as it serving a hindrance for advancing current industry standards-based deployment for generating active/interactive pages (including tracking user behaviour within a page and returning that telemetry to the server).

It’s, idk, very disappointing.

[As for Mastodon, I left social media a couple of years ago, including Mastodon instances, so I haven’t tried to, say, log in to my old account on any of my older (or even my newer) browsers in a minute. I do remember being, briefly, on at least one instance, one which used a different code base than Mastodon itself, which seemed to cause more issues on non-current browsers, but I no longer have the login to that instance, much less remember which GNU/Social-based engine they used.]

Backwards-compatibility used to be a major part of a web dev's job back when a large-enough user base was still using IE6, but nowadays the near-seamless auto-updating of browsers and operating systems has caused a cultural shift in the industry away from caring about backwards-compatibility. Something else that happened in recent web development is the rise of web frameworks like React and Angular where it isn't as much the site developer's choice what to support, but that of the framework developers. If the framework stops supporting your browser then gradually issues start popping up all over the web on sites that use it.

 

[B S Magnet]

View attachment 2086921

Real talk: whenever I see him I can only think of the conniving Mr. Deveraux in Silver Streak.

 

[PowerHarryG4]

Thanks for developers explanation of the blank voids - shame there was no "your browser doesn't compute" warning before I went through the process of setting the account up.

I'm not particularly fussed about Mastodon - was just excited by the possibility that something contemporary would work in Safari 4!

The spec creep on browsers is accelerating - I lasted more than twice as long on Leopard as I did with Snow Leopard, then after SL, I lasted a year on EL Capitan before having to shift to High Sierra on which I'm not having to use multiple browsers to get by.

At this rate, before long we'll be on new hardware every year - unless you want to live on the fringes of society and be considered "unmutual."*

*
View attachment 2086921

That's surprising about El Capitan and High Sierra. I'm on Mavericks and I haven't noticed too many website compatiblity problems using Firefox 78. But I guess I do also have to use Chrome Legacy for the websites that don't function.

 

[mectojic]

If I can just get by with High Sierra for all needs for the next 5 years or so, I'll be satisfied. Hope there won't need to be workarounds. I do think Macs from 2009-2012 have enough processing power to handle modern web, compatibility aside.

 

[B S Magnet]

That's surprising about El Capitan and High Sierra. I'm on Mavericks and I haven't noticed too many website compatiblity problems using Firefox 78. But I guess I do also have to use Chrome Legacy for the websites that don't function.

I’m starting to see dependency creep failures in macports when trying to build the latest versions of ports for Darwin 17 (i.e., High Sierra).

As recently as, say, four months ago, I could build SDRangel from macports, but the last couple of revisions have failed, either on that port itself, or in updating a dependency port on which SDRangel relies.

I’ve also taken stock of how the Qt6 platform, the graphical basis for a lot of macports applications (Qt, that is) will only build from Mojave or later — frustrating, in that Mojave is basically my cut-off OS as one for backward and forward compatibility. Fortunately, for the moment, most Qt-based ports and applications rely on the latest revision of Qt5, but that change-over is nearing for a lot of stuff.

 

[originaldotexe]

That's surprising about El Capitan and High Sierra. I'm on Mavericks and I haven't noticed too many website compatiblity problems using Firefox 78. But I guess I do also have to use Chrome Legacy for the websites that don't function.

the issues with firefox 78 seem to just be illusion. those pages that claim to not be compatible all work just fine when ran under a user agent spoofer in my experience. im not aware of anything of much importance added since the last version of 78...