monovergent

In words with left- and right-handed letters on the keyboard, like "word", sometimes I will instead type "owrd" because the command from my brain seemingly is processed faster by my right hand than my left. It rarely, if ever, happens the other way around. In recent years, I think this accounts for as many as half of my typos. Interested to see if this happens to other people and how I could try correcting this.

Searching gives me the impression there's a million ways to solve the same problem on Linux, and I find myself profiling answers into about four categories at a glance:

• Succinct: one or two-liner, a single config file, or just a few clicks
• Long-winded song-and-dance: Full train of thought interspersed between various commands and logs, several config files (some of which don't already exist), or installing an obscure package that is no longer maintained
• Specific to a desktop environment or version I don't have
• Just looks wrong

I'll usually just take solutions from the first category, which almost always works, save for differences between updates and versions. Solutions in the second category also seem to end with a 50% chance of the OP unable to solve the problem. If I'm desperate, I'll try the second one, but it often ends up not working, eventually leading me to come up with a much cleaner solution of my own.

Curious if anyone else does this too and if those one-liners are really better solutions or if it's just confirmation bias.

Sometimes I wonder what the thought process behind the gaming aesthetic was. RGB (*if tunable) itself is fine and adds a nice opportunity for personalization, but are those tacky fonts, crystal-facet enclosures, and overall showiness just tasteless or do any gamers actually prefer that look?

There were three suppliers of keyboards for the xx10-xx20 series of ThinkPads: NMB/LiteOn, Chicony, and ALPS. A list of FRUs by language and supplier can be found at: https://web.archive.org/web/20200807103606/https://pcsupport.lenovo.com/us/en/solutions/pd010629

Having used NMB, Chicony, and counterfeit keyboard, I'd like to share some of my thoughts for posterity, especially as it seemed that the subtle differences between NMB and Chicony were scarcely documented aside from the notion that "NMB is better". It fascinates me that there is a cottage industry for counterfeits of such a niche product as a laptop keyboard, and these counterfeits are also poorly documented.

ALPS: FRU NOs 45N20xx

Getting this out of the way first: I don't have any personal experience, but ALPS seems to have unanimously produced the least desirable, mushiest variant.

Chicony: FRU NOs 45N21xx

While not as sought-after as the NMB variant, it is still leagues ahead of any consumer-grade laptop keyboard. Tactile response is consistent between keys. I'm not sure how else to explain it, but it will satisfy you unless you decide to someday try the NMB variant. My only criticisms are the thin, high-pitched rattle noticeable when releasing the larger keys and the relatively mushy mouse buttons.

NMB/LiteOn: FRU NOs 45N22xx

This is the most sought-after model and is likely the subject when someone reminisces about the classic keyboard "feeling" better. Most of its advantages come down to tighter tolerances. The slack between keycaps and scissor mechanisms is minimal, virtually eliminating the rattling sound. Tactile response is consistently crisp, never mushy, and won't leave you guessing whether you made a key press. The gap between keycaps is also narrower. That said, the TrackPoint feels the same between Chicony and NMB. If photos are clear enough, the NMB variant can be identified as the circle around the Windows logo is merely smooth, but not indented.

Counterfeits

Counterfeit keyboards can have any combination of both real parts and the following, or only fake parts:

• Reprinting / resurfacing: This is the smallest offense as the underlying components could be genuine in this case. Signs include uniformly smooth keycaps, indentations of printed-over letters showing through, or an unusual layout for the language.

• Incorrect typeface: The correct font for ThinkPad keyboards is Helvetica, but counterfeits may use Arial. The difference is most evident on the the capital letter 'R', where the "leg" curves outward in Helvetica while it is a straight diagonal to the corner in knockoffs with Arial. The font for "ThinkVantage" should match the font of the ThinkPad logo. If not, the keyboard is certainly fake.

• Volume / Power button LEDs: Fakes sometimes use pure red LEDs under the mute buttons rather than orange LEDs. The clear window may be off-center. The genuine power button has a clear ring on the perimeter, but fakes may have a clear window near the center instead.

• Fake TrackPoint: may have poor sensitivity or be sunken in too far.

• Counterfeit keyswitch mechanism: poor or inconsistent tactile response or unreasonably loose keycaps.

• Metal backplate with large holes cut out, is very pliable, lacks paint around the edges, or lacks cutouts for drainage channels along the front edge.

• Caps Lock and Enter keys with a different texture from the rest of the keys

Out of sheer curiosity, I bought myself a clearly counterfeit keyboard to try out. While the listing pictured a keyboard with the wrong typeface, I was sent one with the correct typeface on the keycaps. However, the differences are still immediately apparent:

• Instead of the detailed FRU and barcode label, the back features only a quality control sticker and a label suggesting a 2025 manufacture date.

• The backplate has much larger holes under the keys and there are no cutouts where the drainge channels should have met the keyboard edge. The edges have no black paint and the shiny metal outline is visible after installing.

• The tabs at the front edge have poor tolerances and in order to slide the keyboard all the way in, I had grasp the edge (fingernails hooked into the lip, bearing against keys on the bottom row) and give it an extra tug.

• A healthy ThinkPad classic keyboard should be slightly concave to the user, but the backplate for this was completely flat so the middle bowed outward when installed. But with such pliable metal, I was able to just bend it into shape.

• The flex cable feels flimsy and I am afraid the connector will punch through the pull tab if I were to try removing the keyboard.

• The TrackPoint is recessed below keycap height. The cap is also nonstandard, combining full height with the narrow hole of a low-profile cap. It is stiff and awkward to use, barely remedied by setting the cursor sensitivity to max. I still can't tell if it's due to the mechanics of a recessed TrackPoint, or if the underlying component is just as bad.

• The volume rocker is uneven and only actuates with some force.

• Right and left mouse buttons feel different. The left button does not actuate when pressed next to the middle mouse button and feels as though it jams against something.

I'm also writing up this entire post on the counterfeit keyboard. The tactile response is not exactly mushy, but it is weak and slightly inconsistent, leading to an occasional missed keystroke. Overall, the keyboard feels on par with those in cheap keyboard folios for tablets - arguably better in the sense of key travel and tactile response, but it is evident what lengths they went to save costs. I could live with the keyboard, but the awful TrackPoint is a deal-breaker. That said, for $20, 12 of which probably went to warehousing, 3-day shipping, and packaging, I won't complain.

Are they factory seconds or new parts made on old molds?

I highly doubt it. At the very least, the metal backplate is fundamentally different.

I'm not sure what the thought process behind counterfeit classic keyboards is, but if it were up to me, I'd diversify my portfolio by making 7-row keyboards of a slightly higher grade for later models like the T480 as well, perhaps even one for the Framework laptop.

Strategies for obtaining a NMB keyboard

• Playing the lottery: Buying a used or for-parts ThinkPad on the cheap will almost always yield a genuine keyboard, but photos on listings will rarely give enough detail to determine which variant is installed. Avoid playing the lottery with listings for just the keyboard that say "compatible". Compatible can mean anything from the NMB variant (unlikely) to a counterfeit.

• If the listing shows a missing keycap and there is a red dome under where the keycap should be, it is a NMB keyboard. You may then choose to source a replacement keycap.

• Search for overlooked models compatible with the classic keyboard, like the T400s or W510.

• Search for keyboards in other languages with a similar enough overall layout. Also a good way to avoid counterfeits.

I've realized that the display size, in inches, is sometimes in the name or model number of electronics that are sold not just in the US. Do people outside the US also talk about buying 55-inch TVs, 14-inch laptops, and 27-inch monitors? Does it naturally roll off the tongue or does it seem strange to anyone?

If it's all inches, why didn't measuring screens in centimeters take off?

Got myself a Dell Latitude ~~E4310~~ E6410 and Thinkpad T510 for free, both with discrete Nvidia graphics soldered to the mainboard. I've installed Linux on them and just went with the nouveau driver since the proprietary Nvidia driver for such old cards is no longer in the Debian 12 repo. Not going to do anything cutting edge on them, but it does leave me wondering:

• I read that I could, with some effort, install the proprietary driver manually. Am I missing out on anything at all without them, or is nouveau mature enough and the graphics old enough that I wouldn't notice?
• Is nouveau with old discrete graphics better or worse than having just Intel's integrated graphics?
• Does power consumption vary significantly between nouveau and proprietary drivers?

EDIT

Answering myself after going down a rabbit hole with the T510:

1. The dGPU is the NVS 3100M, which does have some level of hardware acceleration support under nouveau, so at least it isn't draining power for zero benefit. However, the dGPU is unable to go past its lowest power state without manually manipulating /sys/kernel/debug/dri/0/pstate (I did not try to) and I suspect that this is what kneecaps 3D performance. There should be a marked difference, but I won't be doing any serious work on these machines, so I'm leaving everything as-is.

2. This situation is worse than having just integrated graphics due to the inherent power consumption of the GPU core while unable to benefit from higher power states and other optimizations.

3. Power consumption is probably less, but for much worse performance. At least it is a much better fallback than leaving at maximum.

• A later variant has the BIOS option to disable the dGPU, mine is an early variant with no options

A recent trip to the dumpster netted me a couple of old laptops (from around 2012). The batteries are completely flat and will not take a charge. I plan on using them as beater computers around the house, so battery life doesn't really matter but would be nice to have. The cheapest no-name batteries available for them are about $15 each. A used OEM battery with about 75% health is a dollar more.

For those who bought the cheapest aftermarket battery, was it worth it? Were the batteries surprisingly good or am I better off with a used but original one?

On Windows Vista and every subsequent version of Windows, if I search for a file and include the entire C:\ drive, I might very well have time to make tea or a sandwich while the search results come in. On Windows XP, using the search dialog with the animated dog, I can search the entire C:\ drive and expect it to be done in a minute or two, if not in seconds.

It can't just be nostalgia; I can replicate these results on period-accurate hardware today. What changed with Vista to make file searching so much slower, even with indexing enabled?

On occasion, I'll have to work with markdown files, sometimes with inline LaTeX. I'm surprised how limited my options are, or I'm looking in the wrong places. Pandoc does the job, but the lack of a integrated graphical workflow isn't my cup of tea.

Has anyone found a good graphical markdown editor that can handle inline LaTeX and doesn't pull a gigabyte of dependencies? Preferably also can render the final output to PDF.

A lot of recent medical advice says that hydrogen peroxide in first aid is counterproductive. Of course, what I'm about to say is one person's anecdote. But I find that if I just leave the occasional cut or scrape alone or wash it with soap and water, it'll tend to get a bit inflamed (very locally) and hypersensitive, which is very annoying when it's on my hands. On the other hand, If I just rinse it out and slather some H2O2 on the wound, it kind of chemically "cauterizes" the wound, prevents irritation later on, and heals just as well.

Am I just doing it wrong, or does anyone else find that hydrogen peroxide is good on minor wounds, despite recent medical findings? I don't mean to cast doubt on legitimate medical research, but I'd like to understand why H2O2 seems to work for me when research says it should be counterproductive.

Or historical exploits/trojans/etc. that deserve more attention? I've mostly heard about lucrative vulnerabilities that concern Linux servers, but what about the end-users on desktops? Or is the Linux desktop market small enough that we mostly just see one-off instances of users blindly running malicious scripts?

I recently figured out how nice Wine works for running old Windows games. However, many of them are fixed at 800x600 or another similarly low resolution. No big deal under X11 or Windows since the game will just stretch to fill the screen. But on KDE Wayland, the game just runs unscaled with black bars all around and none of the display settings seem to help. Is there an accepted way of setting the screen to a lower resolution but stretching it to fit the full display on Wayland sessions?