As if the Kernel and git weren't enough, Linus now invents the "Unsend" button.

When I decided to learn touch-typing, Dvorak seemed the natural choice, not due to speed but effort. I also played guitar, and the idea of my fingers moving 1 mile a day with Dvorak, versus the average 16 that a Qwerty typist churns through, seemed a no-brainer.

A hundred-some dollars later, I had a Northgate Omni Key Ultra with dip switches so it toggle to Dvorak mode, and had also sprung for the caps where keys had changed rows. Mavis Beacon and a month's time helped me learn the layout.

Don't have that keyboard anymore, but having known how to touch for nearly 30 years it's irrelevant, as all modern operating systems now support Dvorak as a "language". I do still have my clunky DAS Keyboard though, from back in the PC days, with no stencils on the keys. Between blank caps and the Dvorak layout, it was great security through obscurity.

Having switched to Apple in '09 with an MBP, I've gotten used to the tactile feel of that keyboard, and use a wireless Matias on my desktop rig when not on the laptop.

As soon as they also start playing the sports.

Apple has given up on being "cool" in favor of being "rich".

Had been very happy with them, since getting an MBP back in '09 and cutting the WinDoze cord.

Unfortunately, it's getting to be a little long in the tooth. Have already lost one USB port, getting spontaneous re-boots and logouts. It's annoying, and as old as the machine is, no sense throwing money at replacement parts from eBay.

Still need a good rig, though, and am not interested in something glued and soldered that I can never work on.

So, recently, spent around the same amount as that '09 MBP and built a Hackintosh.

It's not a laptop anymore, but gosh, does it run!

Intel Z370 (using onboard graphics), i7 8700, 32GB DDR4, M.2 drive, 28" 4K screen, Matias keyboard (over the last nine years, my fingers have gotten really used to the touch on that MBP)

Everything works, running High Sierra. Did have to spring for a Broadcom WiFi/Bluetooth card, but fortunately Intel lets you remove the one they included.

Still want a portable box for when I'm away from the desk, so am going to reload the MBP and see if that breathes some life into it. If not, will just find a middle-of-the-read laptop and build another Hackintosh. (It's just for on the road, so no need for super-cow powers.)

This is really too bad for Apple, because if they had just continued to make excellent hardware that folks could actually work on, I would have spent the money on them instead.

I almost always answer the phone, but do appreciate T-Mobile's spam filter which shows the name ID of suspect calls as "Scam Likely". Those tend to get ignored.

'doze? Wine. Much better than past days, but still not quite there.

OSX? Already Linux-adjacent, being based on BSD.

Given the vastly greater ability to modify 'nix to suit ones personal tastes (keyboard shortcuts, etc.), one could *mostly* make it work just like 'doze, but why?

Certain things would have to be relearned, such as paths that do not have colons in them with delimiters going the other way.

If one can handle such cosmetic changes, it doesn't really matter. 'doze, 'nix, Mac ... it's all just pushing buttons.

Over the past dozen+ years (going back to FF 1.something, when still on WinDoze), I have installed many add-ons. These days, around 30 survive, and while I'm sure some can be removed (for example, don't really *use* ColorZilla anymore), with nearly 90,000 users, my favorite by far, and the one most indispensable to daily use, is Tab Groups.

As a "Legacy" add-on, it will, due to Mozilla's mandate, not survive the upgrade to v57.

It's more than just a bookmark or history manager, and there is nothing like the functionality it provides in the new FF. Containers don't cut it; don't want a huge vertical "Tree" view. Want that familiar icon that helps me organize my tabs into logical collections, letting me switch to a different group, or being able to right-click on a tab and move it to another group.

Some months ago its author announced he would not be converting it to WebExtensions, and has released its source code to the wilds of GitHub. I, alas, do not have the free time required to dig in and figure out how to perform the conversion.

I currently have 35 groups, one just for /. That group has about a dozen things in it, related to exploration of various stories I've read on the site. It's very handy being able to organize my surfing in that manner. The groups do get pruned from time to time, if after a bit I fail to follow up on some page that's been saved to a particular group, or when cleaning out base search queries.

(There are "Containers" in the modern FF world, and one very nice thing about them is keeping cookies, etc. separate. That's a *good* idea. Tab Groups does not do that, but I hope its successor does. Unfortunately, any WebExtensions add-on I've seen which employ them falls far short in doing what Tab Groups can.)

Am in general pretty loyal to my technology, so while they're on the machine, don't use Chrome, Vivaldi, Opera, etc. Am on Mac these days, so M$ browsers are out (even in the WinDoze days, they sucked. Anyone remember IE's skinned cousin Maxthon, which sucked ever-so-slightly less?)

Thus, FF it is, and until something so radically better comes along that I needs must re-evaluate my choice of browser (as did it, rising from the ashes of Netscape, which I had used since v2 back in the 90's), FF it shall continue to be.

I totally understand the developer's recalcitrance to re-write his entire app. I also totally understand Mozilla wanting to push their browser into the future, but feel they are falling into the all-to familiar trap of not only desiring that I use their code, but deciding for me how I should use it, as well.

There is no technical reason why Mozilla could not allow the performance hit of having legacy extensions remain functional. Sure, it will add bloat to the program (think multiple sets of libraries being needed to accommodate XUL or XPCOM as well as WebExtensions), but Mozilla should give developers more time than they have, to catch up with the new way of doing things, or let new faces take up the old code and convert it, rather than simply throwing years of good work down the drain.

The whole idea of an add-on is that a third parties may add functionality to a program which the original authors have not coded in to the core. (Just imagine how painful using *nix would be, without bash scripts.) While WebExtensions is much more secure (it limits what the add-on can "do" in modifying the core code's behavior), many, many popular extensions have already been written - and vetted - by Mozilla, using the older XUL technology.

While I agree that moving forward, new extensions should be forced to use this more secure way of doing things, Mozilla should also recognize the contribution of earlier add-on authors and allow older extensions to run if the end user desires it .

Thus, until something comes along which can provide the functionality of Tab Groups, and is written using WebExtensions, I have sadly been forced to turn off FF upgrade notifications.

Typed in my first text editor from a printout - an issue of Dr. Dobbs, I think, don't recall exactly - for the venerable C64, which had no numeric keypad, making data entry of the 6 or 7 K worth of hex assembler code fairly tedious. Otherwise, loved that keyboard.

On the PC side, still miss my Northgate OmniKey Ultra - they offered a bag of keytops for Dvorak which I used before learning to touch on that layout. (Thank you, Mavis Beacon!) DIP switches on the bottom of the keyboard let you switch layouts.

Also on PC, miss using (but still have) a DAS Keyboard - flat black, clicky keys like an IBM, no stencils on the buttons. (Between that and having the Dvorak layout enabled via software, it was difficult for anyone but me to sit down and use the machine. Security through obscurity, anyone?)

But it's OK, I really don't mind not using those anymore, because the keyboard on my '09 MBP is still chunking along quite nicely. Took a little time getting used to it, but now anytime I sit at another machine (even the little bluetooth thing attached to my Samsung tablet) it feels weird.

Between the NSA, FaceBook, and Cortana, there will be no need for keyboards.

By combining predictive text technologies with routine snooping on every detail of our private lives, not only typing, but actual, human-to-human conversation will be rendered obsolete.

This will turn out to be a good thing; not having to bother with the tedious task of composition, everyone will have more time to stare at their Wall.

I miss the days when the only other person that truly knew what was inside my head was Santa Claus.

The Difference Engine, his collaboration with Neal Stephenson.

Holmes-era Scotland Yard employing a huge analog computer that runs on steam!

I tried to read the article, but it was too hard to think about.

Actually, it does raise some good issues, but one corollary of its central point - that as systems grow more complex, the potential for error increases - is that as time goes on, our understanding of how to solve a particular problem will change.

Putting Lovelace aside, we've only been working with software for less than a century. Many of humanity's technologies are quite a bit more mature, stable, and less prone to such error, as we've had time to work out the kinks.

Were this article to be re-visited in 2117, many of the "problems" it raised would no longer exist (though certainly other, newer ones that had yet to be "debugged" would).

Why don't we just have the current US CEO declare hurricanes (especially those with foreign sounding names) to be illegal immigrants, and thus not allowed to enter the country?

I for one would love to see our Border Patrol agents try to stop one; not to mention what kind of wall it would take, or how we'd get the storms to pay for it...

What I'll be doing less than a year from now is not updating FF.

Of the couple of dozen or so add-ons installed over the years, I only use a handful on a regular basis, and one of those (probably in the minority here) is Tab Groups. The author has already stated he's not converting it to the new system, and has released the source, but the voluminous free time it would take for me to do the port is not alas in my current repertoire.

If Mozilla wants even a chance at staying relevant, they should let the program run both WebExtension or XUML as needed.

No sense completely alienating your customer base; especially when you're no longer #1.

Just find one or more friends who also want to maintain good backups, build encrypted boxes for each location and be each other's off-site storage...

Seems the obvious choice would be the Libre Foundation, but my snarky humorous side would love to see Google take over and somehow use the stewardship in their never-ending fight with Ellison et all...