Most active commenters
  • csdvrx(10)
  • rhn_mk1(4)

←back to thread

1680 points etbusch | 26 comments | | HN request time: 0.93s | source | bottom
Show context
sryie ◴[] No.31434782[source]
I recently received my first framework laptop after being a loyal Thinkpad user for years. I am loving it so far. I run Ubuntu 22.04 daily and have not had any issues with battery life or the lid (but I do typically leave it plugged in during lunch and overnight). The expansion cards are brilliant and the keyboard is comparable to my old t-series. The aspect ratio is great for coding and I'm happy to see upgradeability is being taken seriously as promised. If I can get 5-10 years out of it like my old ThinkPads (all while upgrading piecewise along the way) I will be a fan for life.
replies(6): >>31434810 #>>31434877 #>>31435998 #>>31436540 #>>31436697 #>>31437014 #
Goronmon ◴[] No.31434877[source]
I recently received my first framework laptop after being a loyal Thinkpad user for years.

I get excited about different laptops occasionally...and then I remember that I won't have a trackpoint if I switch to a different brand, and I get disappointed. Literally happens every few months.

replies(14): >>31435070 #>>31435167 #>>31435374 #>>31435965 #>>31436141 #>>31436258 #>>31436464 #>>31436740 #>>31436868 #>>31437107 #>>31437319 #>>31440279 #>>31441931 #>>31442229 #
pedrocr ◴[] No.31435374[source]
Yep, same here. And with increasing urgency as Thinkpad quality control seems to have fallen off a cliff. Framework seems uniquely positioned to fix this though. Someone just needs to do a compatible top cover that takes Thinkpad keyboards. I'd take a stupid one without touchpad at all as I just disable it anyway. That shouldn't be too hard, it's mostly getting the plastic right and adapting the connector to the motherboard.
replies(2): >>31436698 #>>31442478 #
1. csdvrx ◴[] No.31436698[source]
> Yep, same here. And with increasing urgency as Thinkpad quality control seems to have fallen off a cliff.

Not really, they are among the rare laptops to still offer S3 for Linux.

And the X1 Fold is a technical marvel (working on Linux support right now, if I'm successful it may become my next toy device to try to use Linux on as a daily driver)

> Framework seems uniquely positioned to fix this though. Someone just needs to do a compatible top cover that takes Thinkpad keyboards.

This. I will buy one as soon as they make a thinkpad like keyboard [+] or the possibly to disassemble and mount a genuine Thinkpad keyboard.

+ : A keyboard qualifies as a "thinkpad keyboard" if has all of the following:

- PageUp above Left, PageDown above Right: to me, that's the most important thing ever!

- PrintScreen between right Alt and right Ctrl: very important too

- Delete above Backspace

- A trackpoint between the {G,H,B} keys with 3 buttons below the Spacebar: I'm not a trackpoint fanatic but I appreciate the precision it offers when I need it, and badly felt its absence when I tried a macbook (no, can't do!)

replies(5): >>31437715 #>>31438733 #>>31441897 #>>31444105 #>>31455784 #
2. pedrocr ◴[] No.31437715[source]
> Not really, they are among the rare laptops to still offer S3 for Linux.

The features are great but my complaint was about quality control. My T460s has had every single part but the chassis replaced, some multiple times, and still failed. A new T14s had to have the keyboard replaced because it randomly missed keystrokes. It then started having the screen randomly start flickering after resume. A new X1, top of the line 4K spec, has the internal screen randomly lose sync. The days of Thinkpads as dependable machines seem gone.

replies(3): >>31438020 #>>31440689 #>>31440897 #
3. csdvrx ◴[] No.31438020[source]
> A new X1, top of the line 4K spec, has the internal screen randomly lose sync. The days of Thinkpads as dependable machines seem gone.

I believe it's all due to the large hardware and firmware changes.

Take for example USB-C: we don't know yet how to make study ports. My X1 had its motherboard replaced due to a dead port.

Or look at ACPI S0ix: it's only since last year that it's become comparable to S3 in power consumption (and S3 is no longer officially supported since Intel 11th gen)

The keyboard too changed: the layout is the same as the xx30 series, but there's less travel.

Likewise, the screens are now 2k or 4k with thinner bezels, and intel HUD ("Xe graphic") is quite different from the previous generations: even if it's handled by the same i915 driver on Linux, GUC/HUC are more important, and disabling PSR no longer makes sense.

Change is constant, but I believe pre pandemic and post pandemic Thinkpads are very different beasts.

4. Liskni_si ◴[] No.31438733[source]
> A keyboard qualifies as a "thinkpad keyboard" if has all of the following:

> PageUp above Left, PageDown above Right: to me, that's the most important thing ever!

> PrintScreen between right Alt and right Ctrl: very important too

That's not a proper ThinkPad keyboard at all. That's the new 6-row fake which has 10 fewer keys than a proper ThinkPad keyboard, which is this one: https://laptopkeys.com/uploads/704_1348778226_Lenovo%20t410s...

replies(2): >>31440101 #>>31440135 #
5. rkagerer ◴[] No.31440101[source]
Youch does nobody else go haywire that Fn and Ctrl keys are in the 'wrong' spots?
replies(2): >>31440144 #>>31440378 #
6. csdvrx ◴[] No.31440135[source]
That's what available now: except on the T25, the old layout is no longer found on modern Thinkpads.

This modern layout has advantages: for example, the space between the keys makes it more comfortable to use with nails, so I no longer have to keep them short.

replies(1): >>31440488 #
7. csdvrx ◴[] No.31440144{3}[source]
There's a bios option to swap them, and custom firmwares doing the same for the thinkpad bluetooth keyboard
replies(1): >>31441617 #
8. pmlnr ◴[] No.31440378{3}[source]
That is the correct one. FN first, ctrl next. See DEC, and IBM.
replies(1): >>31442225 #
9. Liskni_si ◴[] No.31440488{3}[source]
> This modern layout has advantages: for example, the space between the keys makes it more comfortable to use with nails, so I no longer have to keep them short.

Layout and the shape of keys are orthogonal concepts.

But yeah, you're right that there aren't many options these days, and the T25 is getting old. :-(

10. loosescrews ◴[] No.31440689[source]
I have had similar experiences with the X1 Extreme. The biggest issue I have had is that the repair process almost always breaks something new. The first one spent so much time getting repaired that I actually bought a second one so that I could at least have one functional laptop. The second one is a newer generation, but the quality issues are similar.
11. bitwize ◴[] No.31440897[source]
The sense I'm getting is that my 2014 T450s was one of the last few "acceptable" ThinkPads.

If I need a new laptop, it will be a Framework.

12. rkagerer ◴[] No.31441617{4}[source]
Ah awesome!
13. wiml ◴[] No.31441897[source]
I'm curious what your use of PrtSc is that is that important to you. I don't think I've pressed that key since the 1990s. Am I missing out?
replies(2): >>31442091 #>>31442236 #
14. Macha ◴[] No.31442091[source]
On Linux I have it bound to flameshot, which you can consider the equivalent of Snipping Tool on Windows or Cmd-Shift-4 on OS X.
15. csdvrx ◴[] No.31442225{4}[source]
Personally, I like it as it is, but I can understand someone preferring the control key in the bottom left (though I would suggest using Caps as Control, in which case having Fn on the easier to reach spot and therefore first would still make more sense)
16. csdvrx ◴[] No.31442236[source]
It's a key that's very useful on Windows for screenshotting.

Personally, I use it for various color inversion profiles (ex OLED: make all the dark greys pitch black, night time: black and red) + for ShareX.

17. rhn_mk1 ◴[] No.31444105[source]
> PrintScreen between right Alt and right Ctrl: very important too

Please don't. Unless you want users of language layouts that make use AltGr to suffer.

Imagine typing away a message, accidentally slipping your finger from AltGr onto the PrintScr (actually SysRq), and triggering a sysrq reboot in linux. Regularly.

It's a choice between triggering crashes _all the time_, or disabling sysrq and never being able to debug the legit ones.

replies(1): >>31451019 #
18. csdvrx ◴[] No.31451019[source]
> Please don't.

Please do!

Because even with the most generous interpretation of your issue, it seems fully self-inflicted, by a lack of typing skills compounded by refusing to configure the keymap or the sysreq bitmask, and asking instead for that to become everyone problem by having the key moved!

> accidentally slipping your finger from AltGr onto the PrintScr

What about learning to touchtype? And until them, typing in a well lit room?

> It's a choice between triggering crashes _all the time_, or disabling sysrq and never being able to debug the legit ones.

That's a false dichotomy. You are not triggering crashes, you are instructing your computer to reboot (sysreq B) which it does.

It should not be blamed on the computer, but on your lack of attention, and the lack of adaptation, so I'd even call that a self inflicted problem.

If you can take the time to configure your laptop to use a non standard layout, you can certainly take the extra time to learn proper typing instead of bothering the vast majority of those who are happy with this layout.

If you can't take that time, you can certainly apply one of the many possible counter measures, like moving sysreq to another key (cf dumpkeys and loadkeys), or just disabling the sysreq reboot function (0 disables sysreq, 1 enables it, but you can have a finer control if you read the documentation, ex: 128 is the bitmask for the reboot/poweroff) which would let you debug the "legit ones" - though if your linux has legit crashes, you may have bigger problems!

replies(1): >>31455701 #
19. rhn_mk1 ◴[] No.31455701{3}[source]
> fully self-inflicted, by a lack of typing skills compounded by refusing to configure the keymap or the sysreq bitmask

So, git gud? Sorry, I'm not buying it. People have different motor skills, you know. You can't always make your body physically perfect.

Changing the mapping might work (although I doubt it, it's a deep kernel mechanism that probably avoids such complexity), but requires having the knowledge that it's even possible and how to do it. Sadly, laptops don't come with the instructions. And why should they? Machines should be made well in the first place.

Oh, and setting a mask doesn't help because b, c, e, i, k, o, r, u, all have nasty consequences.

replies(1): >>31460062 #
20. josephcsible ◴[] No.31455784[source]
All of your attributes of a "thinkpad keyboard" are downsides that would make me less likely to buy it. Keyboards should be as close to 104-key ANSI as possible.
replies(1): >>31459542 #
21. csdvrx ◴[] No.31459542[source]
> Keyboards should be as close to 104-key ANSI as possible.

Yes, bring on the numpad on a laptop... not!

Form determines function: ANSI 104 had its time, when we were using desktops!

Now except for gaming and CPAs, numpads have outlived their usefulness.

replies(1): >>31459645 #
22. josephcsible ◴[] No.31459645{3}[source]
I don't mean a laptop has to have all 104 keys. I mean the subset they do have should be arranged the same way that they are on ANSI 104. E.g., I want a laptop with keys up to 80% at https://commons.wikimedia.org/wiki/File:ANSI_Keyboard_Layout...
23. csdvrx ◴[] No.31460062{4}[source]
> So, git gud?

If if's bothering you as much as you said, YES, stop complaining, and start acting on your complaints!

I've already given you all the pointers.

Now I'll help you more if you need.

> Sorry, I'm not buying it.

Neither am I. I get the feeling you want to complain more than you want to actually solve your problem. But as this is HN, I'm giving your comment the most positive interpretation possible.

> People have different motor skills, you know. You can't always make your body physically perfect.

So you don't want to try or, due to physical limitations, can't train better fine motor skills to be on par with about 90% of the regular population? Not very plausible, but why not!

Still, this leaves remapping Sysrq or configuring the bitmask, so I'll guide you though the keymap fixing if you need (even if I hope I won't have to, and that you'll be able to learn by yourself with the right pointers)

> Changing the mapping might work (although I doubt it, it's a deep kernel mechanism that probably avoids such complexity),

With computers, there is no place for philosophical doubts: you try it, and note the results of the experiment: either it does work, or it doesn't work. And if it doesn't, you can make it do so by reading the code, understanding then changing it.

So first, did you try it? If not, why? If you did, what did you observe?

BTW if you didn't, let me remove some of your doubts: dumpkeys and loadkeys are all that you need to change the sysreq mapping: the "deep kernel mechanism" links an action with a key through a table, defined in software.

This is just like how the same key can trigger a Y or a Z (US vs German keyboards) - and yes, you can change that too if you don't like it.

To have a look at this tablet, outside X or Wayland (ex: chvt 1), do:

dumpkeys > current.map

Edit it with your favorite editor to move Sysrq to where you want ex (ex: Insert key?).

You can also add any other changes you want (like, keep both your alt as regular alt, and instead make something else the 3rd level key - say the right ctrl key?)

> requires having the knowledge that it's even possible and how to do it.

Yes, this is called having agency. But here, I gave you the knowledge! Do you have another complain/excuse? Or are you willing to try to fix the problem now?

BTW regarding "agency", I don't use Linux as a daily driver- I prefer Windows, not just because it's less elitist, but due to the better terminal options and the greater hackability of its GUI. You don't have to use Linux if you don't like it! There are many things I dislike in Linux myself.

> Sadly, laptops don't come with the instructions.

You'll find most of the instructions you want (and more!) on the Arch wiki.

But if it doesn't exist or if it's not accessible enough, what about writing some?

Personally, I'm preparing a tutorial to help people with a specific tablet (great hardware, but bad software and configuration OOB, so most people hated it, which I find sad)

Maybe you could do the same, as other people may be inconvenienced by the same problem you are having, and would benefit from your solution?

> And why should they

Because you or someone else (say me!) cares enough to want to hack they hardware to do their bidding? Because it fun?

> Machines should be made well in the first place.

Different people want different things.

Some tastes can't be reconciled.

> Oh, and setting a mask doesn't help because b, c, e, i, k, o, r, u, all have nasty consequences.

Do you really want/need me to also write your bitmask for you? Select the ones you won't want, and mask them out

But again, you should take the easy way out: just remap Sysrq to another key that's away from your fingers, and call it a day! You could have done this remapping in less than half the time it took you to write this complain!

replies(1): >>31472362 #
24. rhn_mk1 ◴[] No.31472362{5}[source]
> start acting on your complaints!

I just did: I raised an objection to a bad idea for anyone who might be misled by it. Also I won't buy a computer with this flaw.

> So first, did you try it? If not, why?

I did not, because I did not know how. I also don't want to know how to alter my computer to achieve a basic minimum of functionality, because a minimum is what is assumed. Either the OS or the hardware should have sane defaults.

> Different people want different things.

I suggest you remap your keys (when you use Linux) to fit your special need then ;)

replies(1): >>31473346 #
25. csdvrx ◴[] No.31473346{6}[source]
>> So first, did you try it? If not, why?

>I did not, because I did not know how.

Now you do

> I suggest you remap your keys (when you use Linux) to fit your special need then ;)

When there is no physical key, like to the left and right of the up arrow where pagedown and pageup would be, I can't invent them out of thin air.

replies(1): >>31473393 #
26. rhn_mk1 ◴[] No.31473393{7}[source]
There's typically the "emulate a right click" button or a Windows button in between AltGr and Ctrl. Those don't have the faults of a SysRq, so they seem like good candidates for a new hypothetical keyboard.