Review: HP's smallest laser printer – M140w and Linux set up
shkspr.mobi
shkspr.mobi
Isn't this pretty much synonymous with printer port at this point in time? I have a few other devices with this particular USB port like old old USB1.0 external drives. However, over the course of time, it has been pretty much regulated to the realm of printer port. Also, by not shoving USB-C on it means that it's accessible to a much larger range of users. Doesn't seem so strange of a decision to me.
"There's also an "ID copy" button which is specifically for making copies of ID cards and passports. No, I've no idea either. Seems like a feature in search of a user."
This is actually handy where it allows you to scan one side of the ID and then flip it over to scan the second side of the ID and then print both sides onto one piece of paper at once. The fact that the author can't figure that out or be bothered to test it for the review yet add this to the review does not give me much confidence in the reviewer.
I believe that was so it was clear why you couldn’t plug your hard drive into your printer and expect it to work. There were no B to B cables. Heck there were no A to A cables, right?
I think there were some other weird connectors designed to sort of handle the fact that some devices (phones) could act as either a host or peripheral.
All that said sticking with that connector makes a lot of sense. If you’re buying your first printer it’s unfortunate that you have to buy a cable. Better ton of people will be replacing a previous printer and already have that exact cable on hand. If they used USB-C those people would have to replace the cable.
ID copy? No idea.
What do you mean by no idea? Do you mean you have no idea what the reviewer's point was? Do you mean you still have no idea on what it's for even after I just described it?
B-B cable (whether full, mini, or micro) is at least innocuous and inert.
An A-A cable is necessarily non-compliant by definition.
Because the official spec says so, and as described above USBx was originally a strictly host-device protocol, the A plug is at the host, and the B plug is at the device.
An A-A cable is a host-host link, which from the USB standpoint doesn't make sense (pre type-C). In particular, the host is in charge of sending power, the device is supposed to be able to handle it. In particular if one of the hosts gets powered off, its 5V now gets powered (possibly without limitation) by the other, drawing way more than the 500mA specified by the spec, and damaging the cable if you're lucky (until that catches on fire and burns down your house) or the electronics (and killing your mobo).
Think of it as creating a plug-plug extension cord and connecting two of your house's outlets together (don't do that).
> In practice it would be equivalent to a device with two say micro-usb-ports.
micro-USB has both A and B variants (likewise mini).
What you're probably thinking of is a micro-B plug. And having a B-B cable is also not compliant though nobody does it because it's completely useless (because the device side is "passive", so nothing can happen).
> Unusual for sure, but necessarily non-compliant?
Yes necessarily non-compliant.
[0] "Smart" devices like smartphones and tablets can be both device and host which created an issue, this was acknowledged and support was added with "On The Go" (OTG) and mixed-compatibility receptacles (so USB mini and micro have AB receptacles, aka female, which can thus accept either an A plug — acting as host — or a B plug — acting as device). OTG cables are still asymmetric, with an A side and a B side, and A-A cables are still illegal OTG. Type C replaces this with a fully symmetrical cable.
Imagine two usb-to-serial adapters. Now, imagine connecting them to different PCs. Now connect a serial cable between those two adapters.
That is what I'm describing.
Only the usb-to-serial-adapters are actually inside the cable.
No USB host is talking directly to another USB host. No USB host is "powering" another USB host etc. Your comments are not relevant to the scenario described.
What you’re describing here is a multi-host device, like a KVM. That is completely unrelated to the original comment, it’s two unrelated host-device links.
Probably with what amounts to a broken device in the middle: both hosts will behave as hosts to the serial link, and USB is a host-centric protocol. So at best nothing will happen because they won’t be able establish anything, and at worst they won’t be able to cope with unexpected transactions and will get corrupted.
And again there seems to be some serious disconnect somewhere here because I've got no clue what you are trying to say. My made-up usb-to-serial adapter example will work as is - even with two hosts talking to each other (through a intermediary). Just as well as such a cable will work.
Why talk about corruption? Why would host-centric be a problem in this scenario?
The only reason I could find is that B connectors are a bit higher, so they would need to use a larger case.
There are plenty of printers that can print straight from USB storage, and I wouldn't be surprised if there are scanners/multi-function devices that can scan to USB storage as well. Having two USB ports to plug those storage devices into would cause _a lot_ of support calls because some people would insist on using the wrong port.
The same applies for record players able to record to USB storage, many external hard drives and other periopherals from the 2000's and early 2010's. Mini-USB was also quite common for smaller devices in those times.
After 2012-ish, many seems to have switched to micro-USB and now we're gradually moving to USB-C.
That being said, A-to-A cables do exist and some rare devices use them.
Also it's a needed feature because these printers are targeted for small business and front desk workers (doctors, dentists, gyms, etc). I've seen them do this many times in past. They need a copy of proof of insurance or drivers license you're giving them.
Don't blame me for their colloquialism.
Mostly because of its significant form factor: the “mini” and “micro” USB plugs people think of are pretty much all Bs (so the smaller version of this one), because before USB-C USB is a very directional protocol: there’s a host and there’s a device, and this relationship is literally hard-wired (through A and B plugs) because the host side provides the power.
That aside, “regular” Type B plugs still exist for things like displays (and probably non-C hubs with detachable host-side cable though that seems rare): in the relationship between computer and display the display is the device, and thus is on the B side, and space is not as a premium so the solidity of the plug is more valuable than the gains in size.
I have 2 of those (in the USB-3 variant with the weird bump for the extra pins) at the bottom of my display, as it has KVM features and can thus connect to 2 “upstream” computers (well 3 because it also has a USB-C input).
To summarise the document, the "square USB plug" aka USB-B was conceived when USB client devices were simple, had short traces, little noise, and thus were expected to emit a signal with very little signal loss. This allowed for a relatively long cable. USB-C was conceived to deal with complex devices, which incurred more signal loss, allowing for a shorter cable to keep the total signal loss in the system in check. So even today, A-to-B cables can be longer than C-to-C (even though it is only marginally).
The odd thing is that the spec defines legacy C-to-B cables at far below their true capabilities. According to the spec the loss budget for a C host device is 6.5dB, as opposed to the 10dB of an old A/B host. This should allow for a cable assembly with a channel loss of 11dB! And yet the spec insists on a mere 4dB. So technically someone could make a "non-standard" 11dB cable that should just work, given the guarantees by the rest of the spec.
[1]: https://usb.org/sites/default/files/USB_3.1_Loss_Budget_Rev_...
The original idea was to have a few sizes corresponding to a few sizes of devices. For large devices expected to last many years, like printers, scanners, multi-drive storage devices, many types of pro-AV equipment (e.g. one of those pan/tilt/zoom cameras permanently installed in some lecture halls), scientific equipment, and so on, the square plug is king.
You don't want to lose a $5000 device due to a failed connector.
I mind having a proprietary cable for each device. I don't mind having a few types of cables as they make sense.
The trick is that I exclusively install and use them as network printers. CUPS always works like a charm and I can print from windows, mac and my phones as well. Occasionally on linux I have to reach for foomatic drivers, but it's rare. I haven't plugged a USB cable into a printer since the 90s.
Current brother model has been going strong for 4 years now.
They simply work with most modern Linux distros and Android releases. There's no nickel and diming. They are reliable and maintainable.
When I need high-quality photos? I drive to the drugstore and use the kiosk.
Around the time I was looking to replace it - about 8 years ago - with a colour laser, I'd assumed HP was still popular & 'Linux friendly', but it coincided with a horrendous HP laptop experience (of the 'I am never buying from this vendor again' variety). Forums seemed to have a fair flood of people dissatisfied with HP support on our favourite OS, too -- while Brother was getting very positive reviews, and linuxprinting/openprinting showed consistently good-to-great support across their range.
I picked up a HL-3170CDW (colour laser) at around A$350 - which is comparable to the contemporary pricing for TFA's monochrome HP @ £220. I did a pass-me-down of that model a year or so ago, and went to the MFP version - DCP-L3510CDW (colour scanner, copier, laser printer) - for the same price. Print resolution is the same as TFA (600x600), but scan is at least 4x better. My only complaint is that they slightly changed their toner cartridge form between the models ... which felt like a very HP thing to do.
I won't buy a modern HP Printer, they're not well made or particularly user friendly.
I have some sub 99 dollar brothers which have given me a decade long service life.
If you want an HP LaserJet, there are a bunch of used ones that might keep running for decades of home use, so long as you buy them a cartridge every few years.
I currently run an M401n, with a little print server in front of it, and with the non-HPLIP "HP LaserJet Series PCL 6 CUPS" driver. (The print server is because I don't trust the firmware with network access.) It's not as imposing as my LaserJet 5N, but it works fine, so long as I use the PCL driver.
[0] - https://github.com/pdewacht/brlaser
[1] - https://github.com/pdewacht/brlaser/issues/79#issuecomment-7...
[2] - https://github.com/benpye/nix-config/blob/main/overlays/brla...
For printing if you use a modern version of Ubuntu or other distro it's pretty uneventful and network printers, especially those that support Apple print and advertise themselves on your network, all just work.
Making printers is not trivial. Paper is not as simple as it seems. It's super thin and light which is difficult for machines to handle at speed, and it also comes stacked in a near-solid block from which pieces need to be separated delicately but firmly. The materials science behind the rubber to grip with just the right amount of friction, time after time, accounting for wear and dirt, is mind-boggling. Paper is also noticeably affected by temperature and humidity, both of which change its properties nearly by the minute. This changing nature affects feeding, making the pieces floppier at times and liable to miss the handoff to the next set of rollers, or clingy at others and adhering to a plastic guide for longer than expected. Also the electrostatic properties vary which affects how much charge the drum needs to put on each piece to ensure the correct amount of toner particles adhere -- too much and you get squished smears, too little and the text is faint. And let's not forget the fixing process, which heats the paper to bake the toner into the crevices of the page, which may take twice as long due to extra moisture, but you wouldn't want to burn things if someone is sending through some extra thin paper for some reason. And of course as it's printing the printer is changing all of these variables with its heat and motion, changing the dimensions of every space in its interior. Every printer contains many environmental sensors so they can detect these changes and tweak their control mechanisms for each individual piece of paper. It's kind of akin to baking a perfect baguette every time.
Printers are miniature marvels, and there are still some wizards out there who know how to make them well. But we have to, like, patronize those wizards instead of the goblins of HP or else we will truly be buying our own way into the dismal future of crap that HP creates.
Carly Fiorina.
It was probably the peak for reliability, I worked for a company that sold a lot of them and the failure rate increased quite noticeably when they introduced the LaserJet 4.
Of course, that doesn't make it worth it for just anyone to start making printers. But we see time and time again that where they fail nowadays is in UI, integration, and business models.
It's similar to the post on the first page about USB-C hubs, and graphics cards, and earphones. Typically you buy them as OEM stuff - very very few go through the trouble and design earphone drivers (loudspeakers) from scratch (etc). That, and reference designs. Visiting Shenzhen is an eye opener, browsing Aliexpress too :).
I guess the key elements (scanner, toner dispenser, heater) aren't exactly commodity/interchangeable, so maybe better: are there open-source laser printer firmware?
I feel like the technology has been established for a few decades now, and the latest consumer iterations often add ridiculous "features" and outright user-hostility that maybe there's room for an alternate solution.
Edit: see rdw's answer elsewhere here, for why a laser printer is a pretty complicated piece of control system
There is no opensource car or house or motorbike, is there? How about an oven, microwave?
I know this is recently bought and returned one (nothing fundamentally wrong, just couldn't print on the cardstock I needed to use).
I'm still trying to work out how to cancel the ink subscription...
I don't think any domestic laser printer came with a USB cable in the last 20 years. It's not a conspiracy to make you use wi-fi - it's just manufacturers being cheap and not including the USB cable has been the standard for a really really long time.
>>" Even more annoyingly, it doesn't use USB-C - instead it's that awful square USB plug. Eugh!"
Again, that has been the standard in printers since forever.
Meanwhile, they won't look at discussions like this, and on other sites linking to similar discussions about this blog post. And thus, more dumbing down of computing. More reduction in capability, and ease of use.
I bet calling toner, ink, is due to a single person in Iowa, being confused for 12 seconds, and posting about it.
Meanwhile, they got it, are not still confused, but of course who updates a post to reflect that.
(The above is how I view the competences of marketing types)
- WiFi is optional (and I prefer not having it) BUT having a physical ethernet port should be universal;
- do not have nor require any specific "app" must be mandatory, we have made "driverless" printers, we have pdf and ps printers, we have a2ps, nothing more should be needed for printing;
- bit toner issue: no "chip" must be required, no issues with third party cartridge should be present, at least not those made by design choice by the printer OEM.
Long story short: in the past HP have made various nice enough printer, after Samsung deal overall quality is descending to a level I can't buy them anymore.
The rest are minor things to choose from like automatic double-side printing, covered paper so if one do not print for a certain period of time sheets do not get dusty, easy access to all paper path in case of jammed paper, easy cleaning/change for printer pad just in case etc.
Then you have the wide variety of baroque protocols used by Windows. I don't do Windows and I'd like to reduce the attack surface, but if you disable them, often standard IPP/CUPS or lpd stops working for no reason.
I'm surprised printers are not more commonly used by botnets, their security is abysmal and they usually stop getting firmware updates just a couple years after being discontinued, which is itself only a couple of years after being introduced.
It's so you can print (copy) both sides of the ID on the same page - press it, copy one side, then flip it over. And you get both sides on a single page.
Previously with an HP inkjet, I would dread printing because what if I use up the ink, and then need to print something urgently (e.g. a government form)?
The feeling is that now the printer is serving me, as opposed to previously where I felt I am its servant.
Some 1-star reviews I've written don't seem to bother manufacturers. I guess all publicity is good publicity.
Yeah, when my home laser printer dies, I'm getting a refurb 4M, I'll set up a little shrine in the garage and print a page once every 2 years.
Mostly its significantly shorter front-to-back, in exchange for modest width and height increases.
I'm imagining a micro printer/scanner where to scan a document you pass it through the same route as the printer. And the printer is the width of the page and a cylinder about 3 inches diameter containing the toner, electronics, developer, fuser, and feed rollers. If there's a little spare room, put in a lithium battery good for printing 50 pages or so while out and about.