As most of you will know, a common problem for any new operating system is hardware support. Drivers don’t grow on trees, and usually need to be written by manufacturers, which costs time and money. Luckily for Chrome OS, it uses the Linux kernel which makes the hardware support question far less problematic. Still, when it comes to printers, the situation is different, and here, Google is trying to achieve something which should’ve been done ages ago.
When you buy a digital camera or a USB peripheral such as keyboards, mice, and sticks, you don’t have to worry about drivers. Manufacturers of those types of products had the common sense to standardise in such a way that you rarely – if ever – have to do any driver hunting.
When it comes to printers, however, it’s all still a mighty mess. Thanks to CUPS, Linux distributions generally provide excellent out-of-the-box printing support, and since at least Windows 7, Microsoft has started to deliver printer drivers through Windows Update (without annoying crapware). Mac OS X used to go the brute force route (install all printer drivers by default) but adopted a more sensible approach with Snow Leopard, where it defaults to installing only drivers for printers on your network. Like Windows, Snow Leopard also started delivering printer drivers through Software Update.
This is all patchwork, however. What you really want is printers to operate in the same way as most USB peripherals – no drivers, built-in support in operating systems using open standards everybody can implement. It seems like Google is in fact working with printer manufacturers to beat some sense into them to achieve this goal.
Since Chrome OS is supposed to be open source and open standards, it’s plausible to assume that whatever Google is working on might benefit other projects as well. Printer support is always a problem, and not having to install crapware-infested drivers (or no driver at all for those printers that don’t work on Linux) seems like a dream come true.
Had to install an HP printer today for a customer. Suffice to say, every one of HP’s programmers need to be wrapped in a printout of the millions of lines of code wasted in bloatware they ship with their drivers and then forced to bungee jump with it.
Also, I hope Google work on sorting out the unbelievably atrocious CSS printing support in Webkit (and _every_ browser to boot). It’s literally stagnated for 10 years. Opera supports more print CSS than anybody else, and they still are only scratching the surface. I remember fighting with print styling in Firefox 1.0 and there has been _zero_ change since.
I was pleasantly surprised when I installed the final build of Windows 7, only to have it properly detect my wireless (!) Lexmark printer/scanner combination, and download the drivers from Windows Update…
…without crapware!
I was just thinking about a couple of issues—caused solely by stupid companies—Chrome OS is going to face.
• Those Lexmark wireless printers that use an app to configure the wireless.
• Idiotic broadband installers from Virgin and TalkTalk that _require_ IE.
Basically, the biggest issue Chrome OS faces is companies who insist on doing everything in a proprietary way with the brand hammer coming first. I hope Google has the clout to knock some damn sense into these companies and put them in their places.
They just fund open source projects with slush funds from their proprietary search engine.
But yes I hope that Google works with hardware companies as partners instead of telling them what their needs are which is what the Linux kernel team has been doing since its inception.
Yeah. The world’s a buncha shit, ain’t it? I admire your succinctness. For once, this is not sarcasm. I really kinda do.
Edited 2009-12-02 21:14 UTC
Goodness me, that’s the single biggest advance with Vista and 7 – no kidding. To get a driver, have your printer or scanner (or hardware) work with Windows software as-is and not have a ton of ridiculous software installed on your system.
I can remember installing a 3G dongle from O2 on XP and it completely took over and controlled XP’s entire networking and wireless system!
I once used a machine with four different wireless utilities loaded on it. What is wrong with these companies!?
TBH I don’t give a rats ass about printing. Printing should die.
But open standards and open source are very important to me. I really hope Google will demand open 3D drivers, but I won’t get my hopes up. Open network/wlan drivers sure thing. Open PowerVR drivers, I think not. A shame really.
LOL. I really think that you’re in the minority there. Granted, people shouldn’t be wasting trees on printing out e-mail and such, but there are plenty of legitimate uses for printing stuff. You can’t always have your computer with you, so being able to print stuff out can be a necessity. And sometimes you need to give stuff to people, and they need to have it printed. Not to mention, for more massive documents (like technical papers or books), it can be far easier on the eyes to have them in printed form.
I certainly think that the printer driver problems should die, but I don’t understand why you’d want printing to die.
Especially during my days of studying Psychology at university, I used to print out all of those massive scientific articles I had to read – and we had to read LOTS of them. Reading on a computer screen becomes uncomfortable for me after ~20 minutes, especially the type of intense reading that is required to understand highly scientific articles like these.
Any student doing a serous university study probably knows the problem.
Agreed. And this is a problem that doesn’t seem go get talked about that much which is well worth solving.
So… what is it about reading on the screen that is bad? A couple of things that come to my mind are:
1. White backgrounds. (Hi, OSNews.com!)
2. Unnatural aspect ratios. 16:10 is nice for watching movies, I guess. But for reading, wouldn’t you prefer 16:10? Or maybe narrower?
3. Portability. Ergonomics. What if the computer looked and felt like a paperback novel? (That’s just a starting point, of course.)
4. What else? Thom?
Agreed… white backgrounds are too bright and are a pain in the ass, can cause headaches, etc., but to be fair I think I’ve got so used to them by now that they don’t (usually) bother me as much. I still hate them, though. They’re even worse if you leave your monitor’s brightness and contrast to their ridiculously-high default levels.
Huh? Once I switched to a 20″ 16:10 monitor on my desktop, I’d never go back. I miss my 19″ CRT for some things (mostly the ability to switch resolutions natively, and certain wallpapers that were designed for them) but widescreen has some benefits that it lacked. There’s much more room.
Sure, when sitting right in front of the monitor at a normal desktop distance you can’t always see the entire screen if you’ve got a big monitor, but what’s moving your eyes from window to window going to hurt? The same could probably be said of a standard (non-widescreen) 22-24″ CRT.
On the other hand, if by widescreen you mean 16:9 (like Apple seems to have switched all their iMacs with), then I agree with you… that belongs nowhere but on a display meant for purely for watching widescreen movies and possibly modern, widescreen-designed video games. NEVER on a computer, where IMO you need MORE room, not less.
I’m thinking of upgrading to a 22-24 inch widescreen with the next-higher resolution (1920×1200)… my biggest problem at this point is actually resolution, not aspect ratio. Even on small screens (like small laptops and netbooks), the problem seems to be resolution rather than aspect ratio.
“Agreed… white backgrounds are too bright and are a pain in the ass, can cause headaches, etc.”
And nobody has thought of using this little button called “monitor brightness”?
Does that really fix the problem? I think that there are enough pieces to “the problem” that we can fix one or two without really being as comforatable as we are with paper. Thom’s complaint, which I’ll condense to “The Loo Issue” is not fixed by any little button. Even if monitor brightness fixes the difference between black letters on white paper, and “off pixels” on an “on pixel” background (and I’m not at all convinced tht it does), there is more to it than just that.
Like what? I usually set my monitor brightness to the lowest level and websites with white backgrounds look fine. I find the black ones less pleasant to read.
I don’t think I understand your post at all.
What is better for the eye? I really don’t know the answer. If you had asked me, I would have said white
What would you prefer? I think you accidentally said 16:10 twice. Or maybe you inverted your wording, and you meant to say ” you wouldn’t prefer” instead of ” wouldn’t you prefer”.
Uhm… Isn’t that kindle or Nook? Or do they not go far enough?
Brownish. Reflects less light than plain white.
Oops. How about:
2. Unnatural aspect ratios. 16:10 is nice for watching movies, I guess. But for reading, wouldn’t you prefer 10:16? Or maybe narrower?
Edited 2009-12-02 20:54 UTC
The single biggest problem?
The couch. The bed. The bathroom. While you could technically use a laptop for that, a laptop severely limits the number of positions you can sit, hang, chill, or lay down in.
A simple paper book or printed scientific article can be read in any darn position that I may find comfortable, while a laptop forces you in, say, a small number of positions. That’s annoying.
EDIT: Your comment does read like I am a problem with reading from screens
.
Edited 2009-12-02 22:00 UTC
Also, if you happen to have a wife/gf/equivalent, it seems that reading a book annoys her much less than sitting on a computer. Even if you would be reading about exactly the same thing.
Go figure.
What, exactly, do you mean by “equivalent”?
Probably husband/boyfriend.
Well, that’s kinda what I was getting at with the paperback form-factor comment. Though I will note that at 46, I’m finding naproxen sodium to be somewhat helpful with aches and pains. Especially those back aches one gets after long sessions with the bathroom laptop.
I find black backgrounds much worse. After just a minute or two I’m already seeing horizontal stripes everywhere I look.
Before ChromeOS is released a lot of new e-ink / OLPC / other ebook readers will be released. (I really hope ChromeOS computers will have a OLPC-like screen. Would be great for Google Books)
Much better solution than killing tree that sort of throw away reading.
Chrome OS is about the future, where all app are web apps and everything is super easy.
Somehow dead trees don’t fit well with that image.
I’m actually surprised I can read highly complicated stuff from screen these days.
I did/do however spend quiet some time investigating on what monitor I want to buy and if you do read a lot it really does matter.
Way to show how out of touch you are with the real world.
I guess you expect businesses to go back to writing shipping labels by hand?
Printing? YouDon’tNeedThat(tm).
I guess we are expecting printing to be handled by people in “secretarial” jobs, that could well sit on a windows machine. Most of the employees in a sane company shouldn’t need to print anything. Well, that’s the utopian hope anyway.
Some ideas:
– Company should buy everyone an ebook reader with good annotation capability
– All intra-company signing should be done with public key cryptography
– Why not handle actual printing by posting a pdf to an intranet website, from which it would be printed?
I have went to Dell printers, which I believe are
really Lexmark ‘IBM’ whether that holds true or
not I am not sure any longer.
HP has fell down on drivers.
I always smile when I see the Windows-CD that comes with the HP-printers and the large buch of stuff that is on it and their ‘minimal’ specs.
Because the hplip on Linux is so small in comparison.
Although I do think 25MB is still quiet large, it does support _all_ those devices from HP, not just one printer.
That does not mean hplip is a good solution. Quite the opposite in fact: hplip is an HP-specific non-solution to a problem that doesn’t exist. There’s no reason HP couldn’t ship “normal” CUPs raster drivers or Ghostscript IJS driver, which is the correct way to do it.
Most distros will install hplip/hpijs by default even if you don’t have an HP printer.
Most HP printers for business comes with so called stripped drivers have been for 10 fricking years. Of course if support guy lacks any talent or knowledge on subject customer gets shit, but is that printers fault.
Scanner support is also abysmal. While my HP all-in-one works with SANE, it’s still same mess as with the printer. Should this really be so complicated?
NO!!!
It’s nice to see that the penny has finally dropped with what they’ve taken on with ChromeOS, but I’m afraid that if they’re looking to come up with an all-new clean solution then they’re going to be disappointed. No one is going to use ChromeOS if they can’t use their existing printers effectively, and that unfortunately means that they will have to deal with the current mess ‘as-is’.
What about “Cloud” and °Secondary device” don’t you understand?
I don’t understand how those relate to printing. Are you implying that you would print documents you created on the chrome OS computer, on the other computer transferring them through the cloud?
I don’t like that idea of storing everything in cloud, but then again its implied that everything you do on chrome is in the cloud. So Its there weather or not you print it on another computer.
Its just a lot less convenient to have to switch computers.
Or are you saying that using the cloud means reducing the amount of printing?
I’m afraid trying to get around a desktop being a desktop isn’t going to help Google. Presumably people will still be able to handle documents and e-mails with ChromeOS, whether in the cloud or held locally, and a requirement of that is that people will often want to print these things.
You can’t avoid getting sucked into the desktop tasks that people will want to perform. A netbook not being able to print isn’t going to help it sell, that’s for sure.
ChromeOS is really a moment of truth for the Open Source status quo. (Of which I have been an adherent.) Is the current cluster f–k^W^Wbazaar style of advancement, where ideas and standards spend years battling it out before we ever agree on what colors the bike shed should be help us more than hurt us or hurt us more than help us?
Remember when Apple’s Rhapsody operating system was just something on the horizon? Did anyone else notice how they came in with essentially a completely new desktop OS, and steam-rolled over our (Linux’s) desktop efforts? Oh, to be sure they took advantage of OSS. But they eschewed the slow, ponderous, glacial decision making processes that we in the OSS community consider “normal”. And the zipped right past us. Sometimes centralized control and having a 600 lb gorilla on your team can work wonders.
Of course, Apple, Inc. severly limits their market by insisting upon high profit margins and upon Apple, Inc. being the sole provider of the hardware.
Note that Google is not so encumbered. They are leveraging the advantages of what OSS has achieved, just like Apple did. But they are not planning on getting bogged down in the quicksand that the mainstream of OSS (the Linux distros) do. They’re big and influencial enough to do that. A 600lb gorilla. And they probably won’t balk at paying actual money to get what they want. (Something that tends to horrify the mainstream Linux distros.)
It looks to me like all the advantages of OSS. All the advantages of Apple’s approach. Withough the constraints of either. And they don’t care who provides the hardware as long as it works with ChomeOS.
Of course, there are plenty of things which could go wrong. And many of us Linux folks will likely hate it, because it’s not really Linux as we know it. But… this is a real world experiment unlike any we have seen before. And we need to watch and learn as it proceeds.
I’ve voiced, in other posts, my practical (as opposed to philosophical) concerns about becoming totally dependent upon “The Cloud”. But others have claimed that there are good ways around that. Gears, for example. And others have voiced concerns about just how horrible apps like Google Documents are. So there are other problems as well.
But this issue of hardware compatibility is one that I’m betting Google will fly right past, leaving us FOSS folks wondering how we managed to spin our wheels for so many years, making such feeble progress, while somehow telling ourselves we were doing so well.
Maybe I’m just young and idealistic, but I still like what I see going on in the Bazaar. I certainly wouldn’t call it stagnant: I actually see change happening pretty quickly out there, certainly much quicker than in the Cathedral. If you’re talking about one solution to a given model becoming dominant and displacing its competitors, then sure, that can take a while, but that state isn’t really the goal. Ideas get thrown out quicker and evolve quicker in the bazaar, even if that means they have to compete with evolving solutions to the same problem. And that’s fine with me: I like having lots of solutions to choose from, having those solutions compete, and being able to see them evolving rapidly. That’s a good thing.
The big difference, I think, between Chrome OS and other FOSS efforts, as you point out, is that Chromium is backed by a 600 lb gorilla. That’s not really a lesson that we can learn, it’s just an obvious reality: if you want to be able to influence other players, you have to have some form of leverage yourself.
I think Chrome OS can be a huge success as an OS for secondary devices and web portals, without really having a take-away lesson for other FOSS projects. The big difference is that Chrome OS is backed by a major industry player, with tons of cash and a user-recognized brand. That’s not really something the FOSS community could emulate, even if they wanted to.
I’ll shut up now.
Why would we hate it? It will definitely move Linux forward, by providing certain commercial clout that will put some pressure on the hardware manufacturers.
I imagine it’s windows users that will hate Chrome OS in the first place, as it erodes their “everything is windows so I will never need to know anything else” vacuum shell.
I’m not usually one to make this kind of RMS-like quibble. But in case you haven’t noticed, ChromeOS uses a Linux kernel without being what we’re used to calling “Linux”. Which, in this case, might really be best called GNU/Linux. (Excuse me while I go wash my mouth out with soap.)
It’s not exactly what we’ve been pushing forward all these years. And it’s somewhat more controlled by an evil megacorp than many would be comfortable with.
There is plenty of room for dissent.
I think he means that Linux users won’t be using it. They may buy a netbook with it for their grammies but Linux geeks will be the last group to reduce their portable computer to a browser.
What kind of Linux geek would choose a browser over plasma for netbooks?
http://www.youtube.com/watch?v=gGPtYF4vs8k&feature=related
Google is making a GrammieOS. It’s primary appeal will be to people who want to put their Grammie on a computer that they never have to support.
Yer, and they still have a desktop that only amounts to single digits of overall percentage usage. Granted, they screwed over their existing user base, developer base and application compatibility to get the clean start they wanted but at least they had just enough of an existing user base (mostly graphic artists) to build off.
I doubt it. The problems of backwards compatibility with existing printers still remains with this brilliant, all-new approach no matter how much anyone says “They’re Google, I’m sure they’ll get past it!”
Hardware support is a complex set of of hurdles. You need to have a platform that is attractive enough for a manufacturer to take notice of, income in other words, and this brings with it the problem of attracting users with something compelling. You then need some way of developing drivers either for yourself or with manufacturers and creating a delivery system that works, and it needs to work well.
They have a ton of work to do in a year.
I’d go even further and suggest that ChromeOS is a moment of truth for the desktop status quo. I don’t believe ChromeOS is the exact answer, but I do see it as a disruptive technology that will introduce people to the idea that there are possibilities beyond the traditional desktop model.
Google is also not encumbered by an existing market and software ecosystem. This is why we can’t expect any progress from Microsoft or Apple– by challenging the status quo, they only defeat themselves. The OSS world has a lot of flexibility here, but sadly, has chosen not to exercise it.
One of the other major issues is the Javascript/HTML duopoly. There are many cases where web technologies are simply insufficient for certain classes of applications. Give us sandboxed versions of Java, C#, Scala, Python, Ruby, etc. and then we can begin to talk about delivering real applications through the browser.
This is, perhaps, the most aggravating part about dealing with fervent FOSS advocates. Despite the tremendous progress that Linux has made, there are still some glaring weaknesses. Some people are either so indoctrinated or so delusional that they refuse to acknowledge any weaknesses whatsoever. That type of attitude is nothing but detrimental to progress.
I moved this bit to the bottom because I wanted to comment on it separately. Honestly, I think it’s a wash. So far, what we’ve seen is that the bazaar is a great model for labor, for dissemination of ideas, for cooperative work, but a horrible model for anything requiring strong leadership. Sure, there are strong leaders for certain projects within the OSS world, but when it comes to pulling it all together, the end result looks like patchwork. Shuttleworth is the closest thing the OSS world has to a strong leader at the distribution level, and though he certainly puts his money where his mouth is, nothing truly beneficial ever seems to come from it.
I’d also like to add that working with AJAX is a major annoyance compared to developing local apps. Javascript and HTML were never designed to be used to that degree. The idea of working with AJAX in my spare time makes me sick. I’d consider a project that used a plug-in like Silveright but AJAX? Puke.
I think it has been a failure in its expectations due to the lack of expected volunteers. OpenOffice is a good example in that it never drew the collaboration that Sun was expecting. The vast majority of the commits have been from Sun employees.
The most popular open source projects like Firefox and Mysql are more like cathedrals that allow the public to contribute if they wish. However the vast majority of the work is done by full-time priests.
The bazaar model works better for smaller projects with 2-5 man teams where you don’t have to organize as many people and can keep a smaller codebase.
But I agree about leadership, as many open source projects are held back by infighting. There are also stronger forces towards splintering rather than consolidating. The other problem is that everyone wants to work on the cool stuff like mp3 players. It’s hard to find volunteers who will spend all day making sure excel macros can work in calc.
I kinda agree: I have no experience with Ajax — or web development at all — but from the little I know about it, it seems like a clumsy hack to bludgeon a popular and common tool into doing something that it was never meant to do and is not well-suited to doing. Ajax is great for boosting site interactivity, and seems to work wonderfully for building web-based collaboration and social tools, but it’s not going to turn the browser into a powerful-enough VM to replace Desktop PCs and the applications that run on them.
Those are all good points, though I think they’re more the weaknesses in a strengths-and-weaknesses list than un-overcomable and universal truths of the bazaar model that will always cripple it. Sometimes the Bazaar model works better, sometimes the Cathedral model works better. Lots of Bazaar-model projects have enjoyed a lot of success and released some pretty good software, and lots haven’t.
My inner programmer that knows he’ll have to maintain some of those things in the future really wants it be a nicely sandboxed Scala environment, with direct access to the DOM tree (so we could still use HTML, but ditch javascript).
Doesn’t cups pretty much take care of the whole printer issue already? Or is google going to reinvent the wheel again and say “we can do better!”
Sort of tired of seeing google thinking they can always do better. No. Sorry google. You can’t always do better. Really.
You still need the drivers. CUPS may help, but it doesn’t fix everything. It manages printers, but you still need the drivers to talk to them. The issue here is that it would work a lot better to a have common protocol for communicating with printers so that you don’t have to have a specific driver for each printer.
Now, with what Google is doing with Chrome OS, it may very well not use CUPS given how stripped down it is, but the issue that they’re trying to solve here is that every printer has its own special driver, so it can be a nightmare to properly support very many printers, let alone all of them.
CUPS provides centralized printer services for the platform which can then push documents out too local or network attached printers. It still needs drivers though. If it can’t talk to the printer then it’s not going to print anything of use. This is where the issue lies. If Google can convince hardware manufacturers to agree on a standard or provide drivers without crapware (even just drivers) for non-Windows platforms then they accomplish something that helps everyone. Heck, if the EU wants to do some real good; look at how limiting printer manufacturers are towards consumer platform choice. Frak it.. EU, look at all hardware driver practices if if we’re remotely lucky, it may leak over to this side of the ponds corporation sponsored governments.
The bigger problem it relates to is the closed source nature of drivers. The product is the hardware which is in demand regardless of the consumers prefered platform. There is no justification for driver source to be hidden. Got a bundled firmware in your NIC driver; put it on a flash chip on the device not in the driver – this is no more difficult than bundling it with the driver. Don’t have the patent or copyright control to release the secret sauce in your GPU; no problem, put it behind a generic hardware interface chip and let folk provide drivers against that shared interface. Don’t have the budget to develop for more than one OS; doesn’t matter, the Linux folks if not other platforms are outright begging for minimal specs to include support for your hardware and at no cost to your budgets.
But ney… how to talk to the hardware is an “trade secret”.. the unwashed masses can’t possibly be graced enough to allow them use of your hardware outside of your one approved OS.
This is peripheral hardware for general purpose computing. Locking it into a single platform through bad management decision making is freaking madness.
I’ll end my rant. Drivers are just one of those topics that cause unnecessary grief for no good consumer benefiting reasons.
If I could put 10 mod points towards that comment, I would.
Doesn’t cups pretty much take care of the whole printer issue already? Or is google going to reinvent the wheel again and say “we can do better!”
Nope, it’s a different thing. Cups is a printer queue manager etc, and it includes drivers for hundreds of printers. But what Google wants is that you can use exactly ONE standard driver to power all the up-and-coming printers so that you won’t anymore need a gazillion drivers.
That is a good goal, it’d be fantastic if all the printers in the future followed a single set of standards and would be operated with a single standard driver. However, given how stubborn and incredibly stupid most printer manufacturers are I doubt it’ll happen in the next, say, 100 years.
Remember all the times that Kirk would ask Spock a question, and he’d pop one of those little 3-1/2″ floppy looking things into his console before looking into his hooded viewer?
Those were printer drivers. It’s a little known fact that he started loading those shortly before the premier episode. (“The Corbomite Maneuver”) And if it had not been for the help from the Fabrini, he wouldn’t have been able to finish by the end of season 3.
I suppose it wouldn’t deflect the course of history too badly for me to reveal that most of his efforts during seasons 1 and 2 were devoted to Canon ink jets.
Edited 2009-12-02 23:39 UTC
We already have this…and have had it for TWENTY years.
It’s called postscript.
It used to be it took more powerful printers to do it, but these days every $49 printer has a powerful enough chip and enough RAM to do postscript.
Until Adobe allow companies to implement it royalty free, it will be yet another cost that the printer vendors do not want to carry if they can avoid it. Microsoft has already provided an alternative called XPS. Xerox already is manufacturing XPS based printers but the ultimate solution would be for Microsoft to submit XPS to a standards body so that it can be implemented royalty free.
Edited 2009-12-02 19:31 UTC
The internet commission will be rewarding you with a holiday gift basket that includes porno and blank cdrs.
Microsoft did submit it to a standards body: ECMA. But as we know they ‘own’ them. So I doubt anyone else had a good look at the standard and makes a lot of sense to them. Maybe those printer manufacturers.
Well unfortunately I don’t see Adobe releasing postscript into the main anytime soon – their hatred of open standards is notorious and hatred of open source even more so (anyone remember the abruptly cancel Framemaker that was about to be released then suddenly cancelled?).
If Microsoft were smart they would release XPS to some sort of consortium which would make printing and presentation subsystem a whole lot easier to maintain as vendors move away from hardware specific drivers to a generic XPS system and a uniform way for hardware to be controlled from the desktop.
Drafts of the standard were made available during its development, and public feedback was taken. Even if you don’t count MS’ initial release of the spec on their website, interested parties had a year to review and comment on the ECMA working drafts.
The final 1.0 spec has been available since June.
… in your’ PC’s main processor, that is
the only devices that actually have an internal some-hundred-megahertz RISC processor and rasterizer, are laser printers or some printer/scanner/copier/fax(sometimes) all-in-one’s – but i can hardly find any at that price, at least where i live…
what i can find at that price, OTOH, is only cheap plastic epsons /hp’s /canons etc, that carry the bare minimum of electronics in order to lower the price of the printer itself (compensating that with exorbitant priced cartridges, forced onto you by all means)
Yes, those some-hundred-megahertz processors are so cheap today, it’s incredible they try to cut a couple dollars on the final price to sell you basically plastic wrapped shit.
Any serious consumer, those that know what they are doing, will completely avoid these devices.
Black/white laser printers are under $100 CDN around here, and sometimes go on sale for under $70 CDN. And they show up as Postscript printers. Colour laser printers are under $400 CDN now, with sales pushing close to $200.
There’s pretty much no reason to use an ink-based printer anymore. Especially once you show people a cost-comparison for ink vs toner over even a single year.
Very good point. But if I’m not wrong, this would add to the price of the hardware due to licensing.
Well I’ve got a lexmark x2500m that was basically a paperweight in ubuntu.
I was lucky though. Found a way to use the driver for the x2550 that lets me print and use the scanner. Can’t check the ink levels though but I dual boot into 7 so it’s not too bad.
I’ll happily give up seeing the ink levels if HP would stop making that stupid childrens toy looking quarter screen print progress display required with the driver. I just want the printer and a driver. Taskbar popup messaged for low levels or error codes is fine but that brain abortion they’ve provided instead is terrible.
It might be Dell actually. I have a Dell at home and HPs at work. I seem to remember one of my users seeing a stupid poppup when printing though. Whichever of you companies, or both, need to get that bloat out of the driver code.
Once upon a time, there was a thing called PostScript. All printers that used it worked the same, from a humble office LaserWriter to a mighty typesetter. But the humble printers were too expensive for individual use because the evil father of PostScript wanted to screw as much money out of people as possible. The evil father didn’t realize that he could make just as much money licensing PostScript for 100% of the printers, rather than 1% of the printers at 100 times the license fee. Some companies, like NeXT, even ran the PostScript interpreter on the computer, so the printer could be made very, very stupid, and very, very cheap. In the end, NeXT was eaten by its half-brother, Apple, and Apple ate CUPS as well (but let everyone in the whole wide world use it for free), and CUPS, together with a *free* PostScript interpreter called Ghostscript, solved a lot of problems for people, unless, of course, they worshipped the evil god of the northwest, whose name was Bill! Bill didn’t like anything he couldn’t totally control, a lo! his followers were condemned to a thousand years of driver misery.
Most of the time, I can just connect a printer to my Windows or Linux machines, and the system will automatically configure it and everything will work. Given that, I don’t really see what additional, practical benefit I’ll derive from a single, universal printer driver. It’s a nice idea, but it won’t really change much for me. If device manufacturers believe that most users feel the same way I do — which may or may not be the case — then they probably won’t want to bother with significant changes to their drivers.
Not that I’m opposed to the idea, mind, I just don’t see a tremendous pay-out to 100% solving a problem that’s already been solved 95%.
Edited 2009-12-02 19:59 UTC
It’s not the 95% of times that it works that matters. It’s the 5% of times that it doesn’t. That’s what people notice.
Say a car were 99% safe. Only one in a hundred times that you drove it would the steering fail, sending you carreening off the road, or over the cliff. (I’ve been doing a lot of mountain driving lately, where “low shoulder” means about 5000 ft low.)
Edited 2009-12-02 20:12 UTC
Not that I’m opposed to the idea, mind, I just don’t see a tremendous pay-out to 100% solving a problem that’s already been solved 95%.
It’s not 95% solved. You just happen to be lucky with your printers. I know a whole load of people who have trouble getting their hardware to work under Linux, including me; there are no CUPS drivers for my Canon Pixma printer.
And I’d much rather choose a universal driver, similar to f.ex. how most USB keyboards, mice and storage works without any additional drivers, than the bloated 100Mb+ packages provided by the manufacturers. Less bloat is always less bloat, no matter how you spin it.
If licensing Postscript is such an issue, why not PDF?
Postscript was too complex (has programmatic loops for example), PDF started as a subset of Postscript. Surely some of the more restricted and better defined PDF variants (PDF-X) could be used by printers?
This is actually what the printer manufacturers are about to do (or are already on).
The term or phrase I have heard being used by representatives of printer vendors is “PDF as the spooling format”.
But the format of the printing data is only one aspect.
Another one is how to move the data to the printer.
Trivial with a networked printer (those usually understand IPP).
Unnecessary difficult with any locally attached printer.
Maybe something like emulating a network over USB/Bluetooth/whatever. After all each of those transport mechanism already has a specified extension/profile for networking and operating systems have generic drivers for that as well.
Edited 2009-12-03 00:13 UTC
A USB Printer Class driver is (checks)…under 1000 lines of code. Talking to a locally attached printer is the easy bit.
IMO the problem affects only the very low end printers, i.e. cheap-ass inkjets below 100 euros. If you go above that Postscript quickly becomes a feature.
Postscript does the job, it’s extemely easy to make a 15 year only Postscript printer work on a modern Linux or Windows system. There is a responsibility for consumers here: Buy a quality printer.
PostScript or PDF on printers would be really cool, but why not something even simpler? The computer could do all the rasterization with its great speed and send the printer a compressed PNG or JPG image.
This would take a bit more bandwidth (for images it would be exactly the same, and text compresses quite well), but would allow for really, really dumb and cheap printers. The driver could easily be ported to more obscure OSs such as Haiku. Also, it would allow the OS to use any rendering method it wants, so it is not forced into using e.g. PDF or PostScript.
The bandwidth is there. Making a format to stream compressed tiles would make it all much easier. Full-speed USB can do nearly 1MB/s, and USB 2 can really do 20+MB/s.
The bandwidth is there. They would need to do things like bring hardware and driver design and maintenance in-house, and have no BS attached.
But, if they did such a thing, I can’t believe it would not reduce long-term costs (it would greatly increase costs for the first two or three generations), as only format conversions for the hardware would really need to be written to add new hardware, and written to talk to a clean driver-internal API at that. So, the driver might work like input->analyze->transform->format/compile->print, where format/compile could be skipped for printers that directly understand PS, PCL, XPS, etc.. Driver features could be added in an orthogonal way, including OS support, making a clean, easy to maintain system.
I know it sounds like a pipe dream, and really, it is, but at a technical level, it is dead nuts simple. It would allow them to keep having dumber host-based printers, too, while also having excellent support for them, even across many platforms. We wouldn’t hate host-based devices so much, if they were supported right.
FI, you generally don’t hate on your integrated or cheap card NIC, because it uses 10% CPU maxing out the connection, where the server one uses <1%, do you? No, because it works every time, the NIC isn’t useful w/o the computer it’s on (local/network printer analogy), and that performance loss isn’t critical.
Back here in reality, if the idea comes up, someone will start a blame game, and the status quo will be upheld.
Edited 2009-12-03 05:12 UTC
Printer / scanner drivers (at least on the consumer end) are designed by a group tweakers with ADD who throw darts at a wall covered in sticky notes to select the next east asian no-name company to print the IC boards and write the drivers of their next printer.
Unified drivers? Incremental design? Do we look like Epson?
Postscript? What’s that?
I swear, 90% of printer and scanner companies rate somewhere between dirt and the gum stuck on my shoe. They are the bane of my existence. I’m looking at you Lexmark.
Good news to see some efforts in standarisation of USB devices,
Even though, in the old days we had more standarised printers as today. Because it was nessesairy. DOS based program all implementing their own ‘drivers’ lead to ‘IBM Proprinter compatible’ and ‘HP Deskjet compatible’ printers.
That was back then. Today, I would suggest all printers to use PostScript.
Also, I would like to ask google to do the same thing about other USB devices, like webcams. I want USB devices to ‘just work’ when I plug them in, Plug&Play, no matter what OS I am running. That would require standarised drivers for each hardware class.
All the pieces are there already. The printer just needs to use PS (or PDF), put the PPD file into ROM on the printer, and create a printer class in USB that will automatically download and install the PPD file when plugged in.
If there’s an updated PPD available, then it can be installed by the OS using whatever the normal update method is (Windows Update, Microsoft Update, Apple Software Update, distro repos, etc).
KISS at work. The PPD is just a text file. PS is well-known. PDF could work in a pinch. All that’s missing is a standardised printer class for USB that would tie it all together. Then it’s 1 simple OS driver for the USB side of things.
If all you want to do is print, that’s entirely true. The problem is that they can do a lot more than that, and the manufacturers all have different ways of doing those other things.
I would be curious to see some numbers about how many of the “printers” sold today are actually multifunction devices. The price level has gotten so low on basic color inkjets, that the manufacturers really push hard on the MFDs. Quick glance at NewEgg, I only see six new basic inkjets advertised — 3 HP and 3 Canon. Bump the price point to $50-$75, and you’re running into the MFDs.
The multiple functions could show up as multiple devices, as if the MFD had an internal hub connected to multiple independent devices. Some card readers use this strategy; while that can be confusing for card readers if the empty drives are always visible, it makes a lot of sense for an MFD.