One of the major problems Google is facing with its Chrome OS is printer support. It was reported earlier that Google is working with manufacturers on making printers act similar to for instance digital cameras, so you won’t need to fiddle with drivers any longer. It turns out Google has indeed been working on this, but I don’t think the solution they unveiled today is what everyone had in mind – despite that, I’m seeing the value here.
My understanding was that Google would work with manufacturers to eliminate the need for special purpose drivers, instead opting for an open standards-based approach, which has worked wonders for the world of digital cameras. This would give all platforms the opportunity to work with printers without having the need to download drivers, thereby eliminating one of the many reasons why I believe printers are from hell.
Google is indeed aiming for something like that, but being Google, they had to cram the internet in somewhere, and Google Cloud Print, as the name suggests, is no exception. Basically, Google is aiming to put the printer stack on the internet, so that not each and every platform has to create its own print stack and print drivers.
“With the proliferation of web-connected mobile devices such as those running Google Chrome OS and other mobile operating systems, we don’t believe it is feasible to build and maintain complex print subsystems and print drivers for each platform,” Google writes, “In fact, even the print subsystems and drivers on existing PC operating systems leave a lot of room for improvement.”
“Our goal is to build a printing experience that enables any app (web, desktop, or mobile) on any device to print to any printer anywhere in the world,” Google continues, “This goal is accomplished through the use of a cloud print service. Apps no longer rely on the local operating system (and drivers) to print. Instead, apps (whether they be a native desktop/mobile app or a web app) use Google Cloud Print to submit and manage print jobs. Google Cloud Print is then responsible for sending the print job to the appropriate printer, with the particular options the user selected, and providing job status to the app.”
And here come the problems. While Google is working to get manufacturers to support Google Cloud Print out of the box (i.e., the printer is Google Cloud Print-aware), legacy printers will continue to rely on the native print stack and drivers. In other words, this is a future-oriented solution, and doesn’t do anything to address the problem of printer makers being wholly and utterly incompetent morons who couldn’t create a desirable product if their life depended on it missing printer drivers for non-Windows platforms (my Lexmark printer has no Linux-drivers).
Another possible issue here is, of course, the whole internet-based thing. How will this work if you don’t have an internet connection up? The internet shits itself every now and then, and then needs to clean its trousers – will I still be able to print if that happens?
Whatever the answer, Google Cloud Print needs to fart shards of poisoned glass in my face before it can be classified as worse than printing today. The project’s code is not yet released, as it is not ready yet, but I’m happy Google is trying to move the world of printing forward after decades of incompetence from the likes of Lexmark, Canon, and HP.
That last paragraph was lovely.
I imagine for legacy printers what they could do is have some sort of daemon or program that runs on your computer that acts as a conduit to the existing print framework, then you would only have write a plugin to use the individual drivers (only on windows of course)
Will the printed page include ads related to the document contents? 🙂
Printer exploits. Your printer now has to be globally accessible by google. Awesome. Soon I’ll be able to relive the college days by printing out messages to random people signed by ghosts.
Plus, if my $50 printer gets a virus that just means the eventual beat down of it will be that much more enjoyable. Printers are from hell. There is no escape from hell for the printers. They’ll only fall down to a lower ring.
Do you know that internet-aware printer have existed for .. hmm 10 years at least. Plus your assumption is wrong because the process might be initiated by the printer which would get the document to print from the cloud, therefore could still work when printer is hidden behind a gateway.
Anyway to this day, printing is such a mess, I can only applaud Google’s decision to tackle that problem. Even if they do it for Chrome OS, it means almost every alternative OS might benefit from it… (eg. Haiku).
Yes, I used to exploit them in college for fun and profit ( if you really profit by stealing something you are already paying for anyways). Maybe you didn’t understand that part of my previous message, I wasn’t trying to focus on the history/details as much/
But this would mean doing en masse. There are security vulnerabilities in every non trivial piece of software. Look at the quality of printer drivers to see what kind o f glorious code will now also have access to the internet.
Google is taking a very roundabout and horribly stupid approach to this issue. The simple and obvious solution is (1) for print manufacturers to finally agree on an open, standard printing stack and (2) Google should give up the notion that EVERYTHING has to be done via the cloud and implement SOME things in the OS.
Because, like it or not, with a physical machine, you have local hardware the user wants to interact with. Such as speakers, headphones, printers, scanners, USB sticks and webcams, mice, keyboard, display etc. But instead of acknowledging this and adding support for them to the OS, Google is living in denial. They’re saying, “it’s ok, some of that stuff is bundled with the netbook so we only have to support that”. OK, let’s say that covers keyboard, touchpad and the display. But as you can see, there are lots of peripherals still not covered.
What are they gonna do? Bundle them with the notebook? You can bundle speakers, a webcam and you can work with audio jack headphones, but you have to draw the line somewhere. You can’t put a printer or a scanner or all the possible USB devices in there.
I don’t even understand why they WANT TO make that poor netbook do everything. If you have a device which is not capable of doing something you send the files to a machine that can. No need to torture it and everybody with completely roundabout “solutions”.
So, their solution is for everything to be “cloud-aware”. Unfortunately, this means smarter, more expensive firmware in hardware, means they stop being “peripherals” and become standalone hardware, means trouble with IPv4 and NAT traversal, and when and if it will be done it will mean a miriad of previously “dumb” pieces of hardware now exposed directly to the Internet. Most of them badly.
Granted, this is also the vision proposed by IPv6, but I’m not sure it’s such a desirable one. I’d prefer that most hardware stayed in a secure LAN bubble and communicated via Bluetooth or WiFi or Ethernet (implying explicit pairing, either physical or logical) to a PC or router which is the only one that’s Internet aware and who also does the appropriate filtering and protection.
Exposing a gazillion pieces of hardware of varying levels of firmware quality to the Internet is asking for trouble. We already have a gazillion pieces of software applications struggling with worms and remote exploits. Why do we want that in hardware too? Just because Google wants to drink the “everything in the cloud” Kool Aid? Well how about interacting with the display and keyboard via the cloud too then? Or is that TOO STUPID?
This being based on PDF might actually work fairly well. But I would like to see Google just as a broker for print jobs. The PDF should either be send directly to the printer or it should be encrypted when passing through Google, but I guess encryption is out of the question because it would need too much horsepower. At least this will be an open (source) system with hopefully lots of vendors/brokers/etc.
These printers will be fairly pricey in the beginning anyways (does Adobe get money from commercial vendors that implement PDF?)
I’d imagine a print server would be able to convert the old printers connected to it to “cloud printers”.
PDF is an open standard, and there’s no royalty to Adobe, so it would work. Of course, putting ethernet or wifi into a printer is going to make it a little more expensive. I think that it will basically be the equivalent of bundling a cheapo wifi router into the printer, or, as the article mentioned, building the printserver software into the cheapo wifi router. But in these days of mobile everything, we all want (and many of us have) wireless-enabled printing in our homes and workplaces anyway, so this revolution wouldn’t change much other than the software and service underlying the whole process.
I just used Newegg as a basis for a quick check of what’s currently on the market. Of the 192 inkjet printers (more likely representative of home use) for sale, 100 of them had network ports, wifi, or both. Since over half of them are already network enabled, I would say the revolution is already well underway
I just checked the laser printers as well, and 218 out of 288 had network ports
Edited 2010-04-16 21:58 UTC
Unfortunately, no. Having a network port today means the printer will “see” the local LAN and nothing more. The printer firmware is way too stupid for NAT traversal or cloud awareness, and I bet that adding such stuff will increase the price a lot.
So Google is going another way, by tackling printing daemons on the local LAN. But you can’t do that either, cause there’s too many of them, so instead they will bundle a proxy with Chrome, proxy which will try to act as a bridge between printing daemons and the cloud.
Problems:
* There are lots of printing daemons out there. Supporting them all is quite a chore. And if you’re not gonna support all of them, why bother?
* I don’t like the idea of Chrome opening up my computer to outside access, perhaps even without me knowing it. That’s what malware does.
Their alternative, of making smarter printers which are “cloud-aware”, is also half-baked. Come on, if you’re gonna work with the manufacturers, how about asking them to implement a common, open and standard printing stack instead of “cloud awareness”? Then the ChromeOS can print directly to printers or print daemons and only has to support one stack. And if they wanna do the cloud thing they can, but only for legacy printers or as a document sharing app with printing capabilities.
Edited 2010-04-19 13:45 UTC
I think you are reading further into my comment than anything I meant, but that is probably my fault for the broader comment about the revolution.
I was primarily replying to the parent’s comments about having to add networking hardware to printers and pointing out that that step is already there.
I think in an ideal world there would be no printer drivers with every printer being PDF native; unfortunately it reminds me of win modems and those cameras that don’t conform to USBVC by requiring proprietary firmware to be loaded into memory before initialising the camera. What ever standard you try to put in place there will always be some dick head from some company waiting in the wings to screw it up for the rest of us.
Edited 2010-04-17 03:36 UTC
Personally, what’s important to me when i buy say an iphone app, is that it should also work without any problems in offline mode. That would apply for any future printer i buy as well…
It’s called PCL5e and PCL5c.
Just avoid PCL6 “Enhanced” (another name for GDI – PCL6 Standard is just PCL5e and PCL5c renamed) printers, and just about anything from an Apple IIGS to a brand new machine with Windows 7 can print to a PCL printer.
Even quite a lot of HP inkjets use PCL5c.
Before anyone says PostScript… PCL is cheaper to implement, and can’t do as much, but still just as standard.
Yeah, wouldn’t it be nice if companies would use the standards/protocols that are already out there, instead of developing their own, incompatible ones?
Do we really need PJL, HPIJS, HPIJL, SPLIX, etc?
Between PS, PCL, PPD, and IPP, we have the entire stack covered, with “drivers” being nothing more than simple text files. Just stick those into the device firmware, and have them “installed” as part of the USB device detection process, or as part of the “connecting to network device” process.
It’s not rocket science. The tools are already there. All the building blocks are there.
And yet, none of the hardware manufacturers give a damn.
Well, PJL is part of the PCL spec – if not PCL1, then certainly by PCL2 or 3, IIRC.
This is such a good idea that I had it a while back, and was briefly tempted to start a company doing the same thing. Then I realized that I wasn’t Google and would face an uphill battle of working with all the printer vendors or convincing enough people to come onboard before it would be a viable idea. I think Google can pull this off, and we’ll all wonder how we went for so long with the status quo.
Well this solves printing for the Chrome OS, but obviously not for Linux.
The proxy idea is nice (not nice is requiring the Cloud for it), but wasn’t the CUPS able to print on Windows-connected printers in a similar way?
Well, in a way… ChromeOS IS linux…
but how this differs from CUPS?
It can receive printing commands, even from web, acats as a server, and deals with the drivers.
CUPS works 100% for the company I work for. We have (at this time of writing) 432 printers installed locally/remotely in just one server accross our stores/warehouses, with no problems. CUPS with IPP/LPD/SMB/FLEX(our own printing backend) works like a charm.
See my review: http://www.cups.org/newsgroups.php?s20504+gcups.general+v20508+T0
I know of something like this already. The companies I do IT support for have this set up in SAP. When they hit print, it prints to whichever printer is set up for that transaction. It can be a network printer across the room, a printer down the hall, or even a print that isn’t installed/mapped two states away. It’s great, as long as it works. The problems come about when the service goes down – suddenly, no one is able to print to the printer across the room. There goes everybody.
If google is indeed serious about this, they need to add some sort of backup – you send a print job, it goes through their service back to your printer. Then you have some sort of automatic setup that happens that makes a connection from your computer to that printer for x amount of time, and temporarily sets everything up. New print jobs go straight to that printer (taking less time) and the service still works if the network/google/etc goes down. Say you print something, turn your computer off, and then the next day your internet connection is down but you have to print something. This print service tries to connect, but when it can’t, it falls back to the temporary connection from the last successful print job, so you can still print. As long as google stays open about this system, I don’t see why it wouldn’t be possible to set it up in pretty much any operating system. And, of course, there is the old (current) way of doing things if that doesn’t work.
Anything they do will be welcome – Thom has already phrased it better than I in the article how bad it would have to be before it is worse than the current system.
SAP?
Uaaaah, I am feeling sorry for you.
What if someone wants to print something an he doesn’t have Internet access at that moment?
Do we really have to make something that is not perfect yet worse?
Adding more factors into the equation will make the life for sure better, right …
I don’t know about you, but I don’t want to send large pdf files over the internet to a server first to retrieve the response that is then sent to the printer directly next to me. Neither do I want to send personal documents to a server that I have no control over.
Depending on the internet to be able to write text (all these online apps) and to print, WTF?
That is worse than it was with the old client systems, because there you did not depend on a server many hundred miles away — and the connection inbetween.
Computing from the 60s with a new name and new clothes, but still the same shit.
PS.: I’d rather invest my time in finding a printer that has support for my used OS than using this pos.
Edited 2010-04-16 18:11 UTC
In a long time ago, computers were so expensive that only big megacorps could afford them. They then rented out computing time at insane prices, and had a dictatorial control over their users.
Now, computing power is everywhere. Anyone can do something using a cheap and independent computer, without worrying about charges or punch cards. Even if you’re a competitor of the firm who made the computer. And the new trend is to use said computer to contact distant servers owned by big companies who have dictatorial control over any data that is submitted to them, through an unstable internet connection.
Cloud computing – or proof by example that history repeats itself, no matter how smarter the people feel or how further science has gone.
Edited 2010-04-16 21:33 UTC
You forgot the most ironic part of the equation:
Bill Gates “sneaked” (stole, by the reckoning of the time) time on a mainframe at MIT (not sure of exact institution, but I think it was a University).
Microsoft is now pushing “the cloud” (belatedly of course, just like that “fad” TCP/IP and that other “fad” file sharing WITHOUT Netbios).
Just like their Windows Phone: they think if you rename junk (being polite, I’d prefer a four letter word that start with sh and ends with it) Se7en times, no one will notice it is still junk and still imitates the stuff your “competitor” put out last year. (or was it 2006?)
1.- User clicks the “Print” button.
2.- Print cloud service scan your document so good bye privacy.
3.- You get your document printed.
4.- You get mail spam the next day relative to the document you printed a day before.
No thanks.
You are so correct!
HP leads the charge there. “thank you for bundling all sorts of crap I don’t need, can I print my document now?”
Brother has decent drivers, at least the ones I’ve seen. Here’s a driver…..that’s it. Like 300kb. HP? Here’s your driver (500MB) and it’s now slowed your system to death. Sure you can dig out the actual 100kb of files that actually represent the driver you need, but unless you’re super geek, it ain’t gonna happen.
Don’t even get me started on Lexmark. There is no cause for that shit to be on the shelves.
Funny thing: There is just a update to the HP driver under Fedora 12. All HP printer drivers for Linux including the management tools (align, clean, …) are just about 11 MB. The hpijs package alone is just 4.8 MB. These are the official HP drivers and they are open source! It’s a strange phenomenon that manufacturers think that they have to bundle all kind of crap with their Windows drivers but under Linux they keep it straight forward and clean. There ARE GUI tools for Linux for the management tasks, but they are optional. The HP stuff is split into several packages in most distributions.
Why do manufacturers (or HP) want to torture the Windows users but give the Linux users exactly what they want, not more or less (if they support Linux at all, that is)?
‘cos they know windows users are suckers!
now that I have gotten this out the way I can give the more Politic answer:
the vast majority of windows users don’t know any better, nor do they have any expectation of what a driver is. The bundled crap usually has all sorts of links to ancillary services that may generate that extra bit of revenue. Until the users experience something different then they assume this is normal.
My experience is backwards to that of many people. I have consistently bought HP printers, and recommended them to others, based on my experience with HP printers under Linux.
On Ubuntu, you just plug in a HP USB printer, turn it on, and it is detected, driver installed and ready to print in a matter of 20 seconds or so. Other local machines (OS agnostic) can print to it over the LAN via CUPs as a Postscript printer, if one so desires. Works flawlessly every time. I have had the same trouble-free no-hassle experience with many HP printers over the past few years.
So I recommended a HP printer to someone who asked me recently, and they bought one, and they asked me to install it for them under Windows 7.
What a nightmare! It took a couple of hours all up, it insisted I used the supplied CDROM and start the install before I plugged in the printer, it proceeded to show me an advertising slideshow, eventually it began to install 8 applications, most of which were further advertising, and only one of which, the last one, was actually the driver. The install process littered the desktop with HP icons, and the taskbar keeps popping up about how I should go online to buy HP supplies.
I’m so grateful that I don’t run Windows.
Edited 2010-04-19 01:22 UTC
They appear to do this to Mac OS customers as well:
I set up a Photosmart 2575 (not a new or fancy printer, it is several years old). The “driver” is 150+ MB, installs all kinds of ink level and scanning utilities (because they couldn’t use Mac OS X Image Capture with a plugin, they have to have their own utility, right?)
And may your deity of choice help you if you upgrade from 10.5 to 10.6!
Then you are advised to “uninstall” the previous software and use what is built in to 10.6 (plus HP software that comes via Apple Software Update).
150MB? The Imagewriter driver was 50 Kilobytes!! And it jammed less and ran out of ink less frequently that these HP “innovations”.
You don’t even need a printer
http://holykaw.alltop.com/how-to-print-from-an-ipad-7
Error, Cloud on fire!
printer discovery, autoconfig, and driverless printing has been around long before this or even mDNS. hasn’t anyone ever heard of LPR? besides, if I wanted hard copy spam I’d just port forward a half dozen existing IP printing protocols to my LAN attached printer. why reinvent the wheel? or is my antiquated solution not fully buzzword compliant?
Browser: Mozilla/5.0 (iPhone; U; CPU iPhone OS 3_1_3 like Mac OS X; en-us) AppleWebKit/528.18 (KHTML, like Gecko) Version/4.0 Mobile/7E18 Safari/528.16
Call it the synergistic mini-cloud 3.0 solution.
“IPP has as one of its design goals not to invent new security features when it can use existing protocols. For example, the original RFC suggested that authorization take place (for example) via HTTP’s Digest access authentication mechanism or via SSL3. Also, IPP itself does not do encryption, but can make use of encryption at the SSL/TLS protocol-layer. Building a printing protocol on top of HTTP offers the advantage of an already well-tested mechanism for transferring files over the Internet.
IPP aims to define a protocol for end users’ most common printing situations over the Internet.”
http://en.m.wikipedia.org/wiki/Internet_Printing_Protocol#section_2
Browser: Mozilla/5.0 (iPhone; U; CPU iPhone OS 3_1_3 like Mac OS X; en-us) AppleWebKit/528.18 (KHTML, like Gecko) Version/4.0 Mobile/7E18 Safari/528.16
Which raises the question; iFolder/NFS and IPP, why do we keep using SMB? Seems like an odd thing to keep pushing things like SMB when it would be easier to create an all in one client for windows supporting NFS/iFolder and IPP with an easy to use interface.
Because backward-compatibility comes to bite us again? We’ve hit that stage where, though smb sucks, just about everything supports it already anyway so we keep using it to maintain compatibility with older machines and devices. There are already NFS clients for every os, yet since they are not part of a stock Windows installation everybody just uses SMB as a guaranteed fallback, and that doesn’t even begin to cover routers with built-in print servers and the like.
This is exactly why everyone and his brother still uses FAT–possibly the worst filesystem ever designed^H^H^H^H^H^H^H^Hcreated–for removable storage
For crap’s sake … it’s madness. They follow this stupid and unreasonable trend to move everything into the internet … god damn you stupid conformists! where is security? where is the reason?
Well, looks like they’re going to face some really serious problems in a near future.
So I have a cloud printer – how will it find it’s cloud? I don’t want one shipped with an auto-signin to google. I guess you could give your printer some sort of web configuration interface where you enter a auto-conf URL for the cloud of your choice but that seems a bit clunky.
I guess somebody nice could run a registry of clouds that the printer could use to present a menu on startup, but who’s going to run this registry? Google?
In Google Cloud Print your printer doesn’t find it’s cloud, the Cloud finds you! 😀
Edited 2010-04-19 13:14 UTC
This is idiotic and no one would care if it didn’t say “Google.”
the printer becomes a brick?
That’s incredibly lame how they are designing it around their own cloud instead of providing a basic network interface for printer manufacturers.
A basic network / usb interface that takes pdf files and provides basic information like ink levels and paper tray status would be much better. Not just to benefit ChromeOS but other systems.
They’re really setting ChromeOS up to bomb. No local storage is bad enough but depending on Google servers for printing goes too far. They can take their cloud vision and shove it.
they had to bring the internets into it.
Why couldn’t they just get printer manufacturers on board with implementing a standard interface. My Android phone can play music through my car’s BlueTooth interface because its standard. Why couldn’t they just come up with a bluetooth interface? How about a mass storage interface where your printer acts like a thumb drive… you drag files to z:\landscape and it prints them.
I don’t own printers anyway because I print 3 things a year and the ink dries up…. I just bring a thumb drive to Kinko’s and print it for 14 cents. Maybe with this I’d be able to print it directly to kinkos and just go pick it up rather than spending 20 minutes having them do it.
A bluetooth interface would get manufacturers on board because if done properly it could be an added-value feature for all systems.