Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 


Forgot your password?
Close
typodupeerror
×
OS X Printer Software Apple Linux

With Push for OS X Focus, CUPS Printing May Suffer On Other Platforms 267

Posted by timothy from the long-predicted-effect dept.
CUPS is the popular open-source printing system that many projects have used successfully as a core, for desktop printing and as the basis of dedicated print servers. Reader donadony writes with word that Apple "has chosen to abandon certain Linux exclusive features, [while] continuing with popular Mac OS X features. The changeover is being attempted by Apple to set new printing standards that will not require 'drivers' in the future." However, as this message from Tim Waugh at Red Hat points out, all is not lost: "Where they are of use for the Linux environment, those orphaned features will continue to be maintained at OpenPrinting as a separate project."
This discussion has been archived. No new comments can be posted.

With Push for OS X Focus, CUPS Printing May Suffer On Other Platforms More

With Push for OS X Focus, CUPS Printing May Suffer On Other Platforms

Comments Filter:

  • Printer? (Score:4, Funny)

    by Anonymous Coward on Tuesday February 21, 2012 @08:56AM (#39109873)
    Is that what those big things full of paper next to the computer were? Haven't used one in years...

  • No! (Score:4, Funny)

    by Anonymous Coward on Tuesday February 21, 2012 @08:57AM (#39109881)

    We must maintain, at all costs, beloved technological anachronisms like printer incompatibilities and X11. Shame on Apple! Shame on them for trying to rid computing of all its cruft.

    • Breaking compatibility for market advantage is so noble of them, clearly we all must approve.

      • Re:OK, whatever. (Score:5, Informative)

        by HarrySquatter ( 1698416 ) on Tuesday February 21, 2012 @09:10AM (#39110029)

        They aren't breaking compatibility. They are simply moving features they don't need into a separately maintained project.

        • Re: (Score:3)

          by makomk ( 752139 )

          They're also apparently removing auto-detection of other CUPS printers on the local network entirely for everyone except Mac OS X users. I guess that's one way of making sure Macs are the only place where things Just Work.

          • Re: (Score:3)

            by printman ( 54032 )

            CUPS 1.6 (and the next 1.5.x) will support "Bonjour" discovery using Avahi on Linux and other platforms (in addition to mDNSResponder, which is also open source and works just dandy on the same platforms...)

            The problem with CUPS Browsing is that it relies on UDP Broadcasts, which are bad for Wi-Fi bandwidth and power consumption. CUPS Browsing also has issues with hostnames - how do you setup a network using CUPS Browsing without broadcasting IP addresses (printer@11.22.33.44?? Yeck), setting up a DNS serve

      • Re:OK, whatever. (Score:5, Insightful)

        by WrongSizeGlass ( 838941 ) on Tuesday February 21, 2012 @09:13AM (#39110065)

        Breaking compatibility for market advantage is so noble of them, clearly we all must approve.

        Not every Linux distro includes every package by default. If you want to install the CUPS 1.6 package, or the filters for CUPS 1.5.x that are not supported by OS X you are free to do so.

        I don't know if Apple will succeed with 'driverless printing', but if they do then every platform will benefit. Sometimes moving forward means letting go of some of the past.

        • Re:OK, whatever. (Score:4, Interesting)

          by Medievalist ( 16032 ) on Tuesday February 21, 2012 @09:28AM (#39110221)

          Meh. Maybe I'm just cynical, but I'd think "moving forward" would involve building a new product, not just hacking out chunks of one that's shared with one's competitors and spinning them off.

          You make a good point about choices, though. The ancient, spaghetti-coded Berkeley LPD still works on modern platforms, and it's probably significantly more efficient than CUPS (I haven't actually checked, but that's where I'd lay my bets).

          • Re:OK, whatever. (Score:5, Interesting)

            by rnturn ( 11092 ) on Tuesday February 21, 2012 @10:00AM (#39110631)

            "The ancient, spaghetti-coded Berkeley LPD still works on modern platforms, and it's probably significantly more efficient than CUPS (I haven't actually checked, but that's where I'd lay my bets)."

            If by efficiency you mean printer thoughput, I think you'd win your bet. I abandoned CUPS on the system that serves as the print server on home network. It turned out that most applications that generated PostScript output and sent it off to a CUPS client to print on a CUPS server resulted in turning my 20ppm printer into a 1ppm printer... if we were lucky. Ditching the CUPS server and reinstalling LPRng restored the printer's normal throughput. I still have the CUPS clients set up on various systems but as a print server CUPS was a dog.

            • Re: (Score:3)

              by lindi ( 634828 )
              The focus has been in PDF for quite some time: https://wiki.linuxfoundation.org/en/OpenPrinting/PDF_as_Standard_Print_Job_Format

          • I think it's fair to say that Apple don't consider desktop Linux to be a competitor. And they're not persuing the server market very hard. Which poses the question what competitor this is about.

            iOS is Apple's biggest OS these days. And mobile printing is a something that's ripe for improvement - and driverless printing would be particularly useful.

            So perhaps they've made the decision that the work they do on printing for iOS they'll keep for themselves, rather than do all the work for Google's Android comp

          • Maybe I'm just cynical, but I'd think "moving forward" would involve building a new product, not just hacking out chunks of one that's shared with one's competitors and spinning them off.

            I don't know why you'd think that. It seems to me that with all the FOSS available, you'd have to be crazy to start from scratch on a product like this. Even if you start from a fresh design, it makes complete sense to find a FOSS project that does similar functions and reuse any good code that does what you want it to do.

            I mean, why reinvent the wheel?

        • I don't know if Apple will succeed with 'driverless printing', but if they do then every platform will benefit. Sometimes moving forward means letting go of some of the past.

          If they do succeed, chances are they will wall it off with patents.
          Look at their iphone patent war bullshit - one of their main attack patents is "slide to unlock" wtf?

      • On the flip side, what about "not being content with their contributions to the project and demanding they maintain features that get them no (and even negative) return"? Would that fit under "noble", or "OSS fanatics once again shooting themselves in the foot"?

        Good gracious, between the complaining here, and on the "HJT Source released, but masses unsatisfied with level of OpenSourciness", I start to wonder why anyone bothers trying to release source as it only seems to inspire flames. Maybe Microsoft ge

      • Re:OK, whatever. (Score:4, Insightful)

        by Phreakiture ( 547094 ) on Tuesday February 21, 2012 @09:33AM (#39110275) Homepage

        On the other hand, there was printing on *NIX before there was CUPS. There can still be printing in a post-CUPS era.

      • Breaking compatibility for market advantage is so noble of them, clearly we all must approve.

        Actually, Apple isn't breaking compatibility, the CUPS developers are on behalf of Apple. One has to wonder why, Apple's requested features couldn't have been added to CUPS instead of also having to remove others that Apple doesn't use.

    • I don't really think think that CUPS 1.6 does anything like that. Apple merely removed some features which they don't need for their own systems. From what I read there is nothing in that release which would achieve driverless printing. Anyway, I find the article barely readable at best. I appreciate the difficulty of writing in a foreign language - English isn't my first language either - but most of this just doesn't parse.

      I would recommend to read this article [h-online.com] instead.

  • Job Security (Score:5, Funny)

    by SJHillman ( 1966756 ) on Tuesday February 21, 2012 @09:01AM (#39109925)

    If print drivers were to be eliminated across the board, half of our IT staff would no longer be needed. Fix the issues with stuck sensors, paper jams, etc and we'd be down to three people.

    • When dealing with printers, all you need is three people [youtube.com].

    • Like all
      custom trays?
      Staplers?
      document security at the printer?
      Stuff like imageRUNNER ADVANCE
      PCL?
      Document Scan Lock and Tracking?
      and other stuff the basic drivers do not have?

    • And if we mandated basic IT classes, we could cut that down to one person, right?

      Or we could keep those three people, and let IT focus on actually useful projects (you know, those hideously large projects that require 6 months just to get the design perfected?), instead of being everyone's technological cabana boy. There is nothing worse than having the network admin drop everything he is doing, so he can replace a toner cartridge on the network laser printer (come on people, it's not hard to figure out how

    • You can't eliminate printer drivers. it's an absurd idea. All they're doing is relabeling drivers as something else so that they can say there are no drivers. Some piece of software has to know how to convert from a generic format to the device specific format required by the printer, and that's the driver. Even in the distant day when all printers use a standards conforming Postscript the code that converts to Postscript and spools it out to the printer and checks whether the printer is busy or not is

  • From printer companies to print on their newly minted printers - who likes extra hdd space, it is so cheap!

  • So what is the fuss? (Score:5, Informative)

    by gweihir ( 88907 ) on Tuesday February 21, 2012 @09:05AM (#39109967)

    True to open-source fashion, the missing features get maintained by somebody else. If Apple makes more problematic changes, my guess is that eventually CUPS will just be forked.

    This is not a big deal. It would be with closed-source software were the vendor can force changes down user's throats.

    • The fuss is that "zOMG TEH APPLE IS EBIL!!!" ignoring the fact that all Apple is doing is setting up a second project to maintain the pieces they don't require.

  • Until... (Score:4, Insightful)

    by sycodon ( 149926 ) on Tuesday February 21, 2012 @09:05AM (#39109973)

    ...I can plug in a printer to my computer and without a single dialog box ever coming up asking/telling me about configuration, drivers, or anything else other than asking how many copies do I want, they need to keep trying.

    Printers have been stuck in the early 80s for the last three decades.

    • Re:Until... (Score:5, Informative)

      by truthsearch ( 249536 ) on Tuesday February 21, 2012 @09:20AM (#39110135) Homepage Journal

      That's how OS X works now. We've gone through a bunch of printers at my office, and a variety of brands. Each one just needs a wifi password set, then the desktop lets us print to it with no question. It just appears in the list of available printers.

      OS X comes with a long list of drivers installed. Apple would love to drop those, partly because it involves a lot of coordination with printer manufacturers. Little from the customer perspective would likely change.

    • I suppose you've never worked with a ZeroConf/Bonjour printer? That's pretty much what happens now. Except the plugin part. I plug it into the network and it just shows up under available printers.

    • That's also how it works with HP printers on Linux. For others, you may have to select your model from the menu.

    • Re: (Score:2)

      by jbolden ( 176878 )

      I suspect you didn't do much printing in the 1980s. Even a dozen years ago printing required manually mixing drivers and manipulating LPD settings. I had to do all sorts of preprocessing from the command line, and if I could get it to work reliable intermix it with LPD.

      In the early 1990s, I had to be careful about rip times and queue jobs coordinating between the ripping and the printer.

      In the 1980s you basically didn't have graphics and had to use continuous feed paper tearing it after you used it. Beca

    • Fedora 16 did this...sort of... with my HP Laserjet 3200se. Google that one and you'll see it's very very old.

      The problem? It chose the Postscript version (this printer adds Postscript support with a combined memory / processing upgrade). Mine has a 64MB memory upgrade, but doesn't have the Postscript upgrade. I had to change the driver to the non-Postscript version.

      But it was so close! Upon plugging in the USB cable, it identified the printer and confirmed I wanted to install it (a very simple dialog box

    • Re: (Score:2)

      by Jonner ( 189691 )

      ...I can plug in a printer to my computer and without a single dialog box ever coming up asking/telling me about configuration, drivers, or anything else other than asking how many copies do I want, they need to keep trying.

      Printers have been stuck in the early 80s for the last three decades.

      I've been able to plug printers into my Ubuntu systems and use them without any questions about configuration or drivers for several years now. Most of what makes this work is CUPS. Ironically, the exact same printers plugged into an OSX system typically require drivers to be installed, though OSX does this mostly automatically. Obviously, Apple is trying to improve that situation by pushing the "driverless" printing but I wonder why they haven't also done what Ubuntu and other distributions have and instal

    • Re:Until... (Score:5, Insightful)

      by uglyduckling ( 103926 ) on Tuesday February 21, 2012 @10:22AM (#39111027) Homepage
      Thanks for describing the OSX print subsystem. It's only in the Windows world that this isn't the case. Linux is variable, depending on how well CUPS is set up in the distro you're using.
      • My current (HP) printer worked on Windows 7 directly. I think I did later install drivers from HP though, to see what's new in them, but I wouldn't necessarily recommend them. I like the printer, I don't like the HP software. (And the installer was huge, and the update was even bigger (and still required the original installation))

      • Re: (Score:2)

        by sycodon ( 149926 )

        A lot of people are saying "works for me". But they are referring to some Linux distribution or OS X solution or perhaps HP printers with HP boxes, etc.

        The way I think printers should work is the same way hard drives work. For 99% of the hardware out there, regardless of OS or manufacturer, if you drop a HD into the box and start it up, the system recognizes it and offers to format it. No searching for the right driver or updated versions or installing this or that, etc.

        Printers need to be like this. While

        • A printer should be just another computer that accepts a document (pick something...PDF, whatever) and the usual parameters (num of copies, size, etc.) and prints it.

          Sure, in a world where the printers are a lot smarter (recognizing the media type and adjusting printing settings automatically - i.e. glossy photo paper = highest DPI, adjust ink levels automatically, etc., transparency = mirror the input, etc.) that would be great.
          If that's what Apple would be pushing for, awesome.

          Until then, I'll take my pri

        • Re: (Score:3)

          by jedidiah ( 1196 )

          Printers are far more interesting and diverse than hard drives.

          Although, there are already very well established standards that any printer manufacturer can use. This is not an unsolved problem.

          Companies simply continue to exercise what one might call "personal liberty".

          We simply don't need Apple to try and impose "yet another standard".

  • Why are printer languages not unified? (Score:5, Insightful)

    by chipperdog ( 169552 ) on Tuesday February 21, 2012 @09:11AM (#39110035) Homepage
    With the horsepower available to cheap microcontrollers and cheap memory today, why isn't Postscript (or even PCL) standard on all printers? That would reduce the printer drivers to a single ppd file. Head cleaning, alignment and such could be accomplished through carefully written PS.

    • Re: (Score:3)

      by Dupple ( 1016592 )

      Because you would need to licence PostScript from adobe. Then you would need to add the expense of RIP, suddenly it's not looking so cheap. PostScript can generate some pretty hefty files

      • Re: (Score:3)

        by tragedy ( 27079 )

        Are any of the patents on PostScript even still in force?

        • Re: (Score:2)

          by tlhIngan ( 30335 )

          Are any of the patents on PostScript even still in force?

          Probably not. Because you can pick up a Brother printer that supports Br-Script, which is Brother's implementation of PostScript (BR-Script3 is their PostScript3 compatible language). And yes, they even provide PPDs for the OS generic PostScript driver.

          The licensing fees go to Adobe are to license the trademarks. Which is why it's always advertised as Br-Script and unless you know, you may not realize it's PostScript-compatible.

          So if you want a PostSc

      • HP doesn't even license the official Adobe Postscript for most of their printers. They use PhoenixPage, a popular 3rd party implementation. The lack of PCL is to cut costs. Its cheaper to have the computer do the processing then in the printer like PCL5c/6 would require.

    • Re:Why are printer languages not unified? (Score:4, Informative)

      by gandhi_2 ( 1108023 ) on Tuesday February 21, 2012 @09:36AM (#39110311) Homepage

      Even printers that DO speak PCL and PS don't all work the same.

      Feed-tray options are one big reason.

    • Re:Why are printer languages not unified? (Score:5, Informative)

      by jbolden ( 176878 ) on Tuesday February 21, 2012 @10:01AM (#39110635) Homepage

      Postscript is proprietary. But there are languages like it which are open standards.

      The big issue with postscript as a printer file format is that the printer makes runtime choices. So for example printer fonts are used and fonts don't need to be included. Which effects both the look of the page and the spacing. Because computations can be done on the printer print times with postscript are inconsistent. That is why in commercial environments postscript is ripped to something like IPDS before being sent to an actual physical printer.

      So the very flexibility that makes postscript "driverless" is also what makes it a poor choice for document consistency. Adobe itself saw the problem in that when it switched the page definition standard to pdf which was from a printer language perspective a downgrade.

      • Re: (Score:3)

        by tibit ( 1762298 )

        Postscript is proprietary but is well documented, so that's not an issue in practice.

        As for the printer making choices, that's not informative because you've completely skipped the reality of it. Yes, the printer selects some defaults. Every print job has full control as to overriding the defaults, and any sane generic postscript print driver will not make its output dependent on the defaults -- precisely because, as you say, they may vary between printers. Heck, postscript guarantees that there is a set of

    • Re: (Score:2)

      by EXrider ( 756168 )
      Yeah, except every language creator out there wants to collect royalties on their proprietary solution, so every printer manufacturer out there wants to create their own proprietary PDL... and we're back at the same problem we have with print drivers.

      PostScript
      PCL
      HP-GL
      MS XPS
      Ricoh RPCS
      Kyocera KPDL
      Epson ESC/P
      The list goes on...
  • just do to CUPS like any other software project that turns to the dark side, fork it

  • So as open source goes... (Score:3)

    by 3seas ( 184403 ) on Tuesday February 21, 2012 @09:47AM (#39110443) Homepage Journal

    ...this is not news but rather simply an FYI.

  • This isn't really Linux vs. OSX (Score:3)

    by jbolden ( 176878 ) on Tuesday February 21, 2012 @09:48AM (#39110461) Homepage

    I'm reading this it doesn't sound like Linux vs. OSX so much as Apple having declared a new standard deprecating the old standard. Apple is typically aggressive about that sort of thing moreso than Microsoft. I think a fair description is that Apple is aggressively pushing the new standard, while the Linux community would prefer a slightly less aggressive push.

    For example avahi (the Linux equivalent of Bonjour) will now be essentially mandatory for CUPS discovery, unlike before where CUPS systems would discover each other independently. Making Bonjour / avahi mandatory is not breaking Linux, Linux has avahi every bit as much as OSX has Bonjour, it is simply moving CUPS aggressively towards a situation where discovery uses the new standard not the ad-hoc CUPS standard. (as an aside new versions of avahi using DNS-SD are required).

    The Linux community has a long tradition of complex dependency chains for full functionality. This is more unusual for BSD than for Linux and IMHO not really harmful to either. I think there is an interesting argument to be had about how aggressive to be about deprecating standards in the Unix software ecosystem and how much software should be independent. But this post confuses far to many issues to be helpful.

    • Re: (Score:2)

      by JBMcB ( 73720 )

      I don't see a problem with forking off discovery. It's not really a core function of CUPs and is probably better handled, and maintained, by a package that handles that functionality specifically.

    • Re: (Score:2)

      by tibit ( 1762298 )

      It's very good that they will enforce use of avahi/Bonjour for discovery. CUPS by default does UDP broadcasts for discovery. That sucks if you have a subnet with many machines on it.

  • Obligatory xkcd (Score:4, Funny)

    by MurukeshM ( 1901690 ) on Tuesday February 21, 2012 @10:01AM (#39110639)

    https://www.xkcd.com/927/ [xkcd.com]

    I'm not sure, but it seems relevant here.

  • In that when you go to a store, they will tell you, that doesn't work on a MAC, but it does on Windows.

  • Isn't this exactly what happens elsewhere, but in the other direction? After all, many people think that KDE, GNOME and other large programs are written for GNU/Linux and just happen to be ported elsewhere. Try to Google something about setting up Apache or bash and you'll find Linux this, Linux that even though neither are exclusive to GNU/Linux in the least.

  • Speculation (Score:3)

    by FellowConspirator ( 882908 ) on Tuesday February 21, 2012 @12:31PM (#39113133)

    The article is based on speculation. One of the bits of speculation, that CUPS would do away with PPD support, shows a lack of knowledge about how CUPS works on OS X and how the driverless print system (to support iOS devices) works. Namely, the PPDs are still required for the printer server (computer) to setup the printer with the appropriate features, color spaces, etc. CUPS requires a filter to translate the driverless print job (PDF or JPEG) to the raster protocol used by the device as specified in the PPD. For OS X, it's true that it's Quartz and Linux the filters will be different, but this is not so different than how it's been all along anyway.

    The one thing that does ring true, however, would be moving from CUPS' proprietary CUPS-to-CUPS automatic discovery protocol to Zeroconf (Bonjour). There's a whole number of reasons that would make sense (for Linux just as much as OS X).

Slashdot Top Deals

Think of it! With VLSI we can pack 100 ENIACs in 1 sq. cm.!

Close