LR CC 2015.6 and Photoshop CC 2015.5 print bug (Mac)

[lhodaniel]

What happens to individuals who are using old printers? I'm still using a 7600UC and it is still going strong. So is a friend of mine (although he has Imageprint and can bypass the problem maybe.) Unlikely that Epson will update a driver for that, though it seems better on the Mac side. In Windows, the latest driver is for XP and was written in 2007! I find that unacceptable for a pro printer and that is why, when I do get another, it'll probably have Canon written on it.

Lloyd

[howardm]

I'm not convinced taht Canon is that much better.  Try locating an updated PRO9000/9500 driver.

[Doug Gray]

What happens to individuals who are using old printers? I'm still using a 7600UC and it is still going strong. So is a friend of mine (although he has Imageprint and can bypass the problem maybe.) Unlikely that Epson will update a driver for that, though it seems better on the Mac side. In Windows, the latest driver is for XP and was written in 2007! I find that unacceptable for a pro printer and that is why, when I do get another, it'll probably have Canon written on it.

Lloyd

I've run windows from XP to 10 with many vintages of drivers including an Epson 9800 and have never had a color shift like this.

[digitaldog]

I'm just trying to understand what the real issues are, since Adobe seems to be not forthcoming with working solutions.

At this point, I think that's the focus to take. What are the real issues and who's responsibility is it to fix?

[elolaugesen]

just one thing.  I have used the latest Epson Printer driver 9.33 for the 3880 since it came out.   should note that epson sneaked sneaked in an update to the firmware in late 2015...  anything in there???
I know this is not OSX upgrade but a hardware...  but.......  did they do this to other printers too???

at the time nobody knew what this firmware upgrade was for ?  I did not apply it as I did not want to mess up my printer just in case....

[Mark D Segal]

No, I agree it's totally fucked up and unsatisfactory. I'm just trying to understand what the real issues are, since Adobe seems to be not forthcoming with working solutions. This is a very serious issue that affects many professional users, and hurts their reputation and they are left dangling by Adobe (or the printer driver programmers if they have to make a followup move as well). This should be fixed without hesitation, and not left to users to analyze (although their reports can help).

Cheers,
Bart

Well, in a nutshell, yes. However it is a holiday weekend so I'm prepared to cut them some slack for a few days - IF they get serious about finding out what is REALLY going on.

What makes printer drivers an implausible cause is the sequence of events, which I think I have correctly here:
(1) Some time ago Apple released El Capitan. I presume that release included the new code that several of those Adobe posts mentions. I upgraded to El Capitan and there were NO issues with printing, or any other aspect of colour management.

(2) Some time after that, Epson issued an updated set of drivers for various current models primarily to include the new Legacy papers. There was, if I remember correctly also a firmware upgrade for the P800. I installed all of that. Still NO issues with printing or any other aspect of colour management.

Since then there have been no further updates of either Apple or Epson software or firmware for the P800 and SP4900 printers.

(3) A few days ago I upgraded to the latest versions of LR and PS, and lo and behold, my colour-managed print flow got whacked. And I know this statistically because I am doing research on the relationship between the accuracy of colour rendition on paper versus visual perception with reference files having known, repeatable colour values, so everything I'm doing is being measured with an i1Pro 2 spectro and i1 Profiler - i.e. high-end colour management gear, with great attention being paid to the "M" factor and all the rest of it.

Therefore, the logical, and very strong inference to be drawn from this sequence of events is that item (3) is the culprit. Now, if I were Chief Inspecteur Clousseau of the Surete Nationale and knew how to unearth obscure smoking guns causing simultaneous damage under the hood, perhaps I would have by now stumbled on parallel or hidden causality, but somehow I think the probability is low enough to dismiss any such thing out of hand. Remember when the dog bit and he said "so sorry, but it's not my dog". Speaking of which, keep up the good work Andrew - the time you are committing to this will pay off sooner or later! I think sometime after the weekend there should be traction.

[Mark D Segal]

just one thing.  I have used the latest Epson Printer driver 9.33 for the 3880 since it came out.   should note that epson sneaked sneaked in an update to the firmware in late 2015...  anything in there???
I know this is not OSX upgrade but a hardware...  but.......  did they do this to other printers too???

at the time nobody knew what this firmware upgrade was for ?  I did not apply it as I did not want to mess up my printer just in case....

I don't see how they could "sneak in" a firmware update. It gets posted on their website as such, and the user needs to go through a whole procedure to install it - nothing sneaky about that.

[elolaugesen]

Tried to find out what the  firmware was about?....  No answers anywhere....     Just throwing in all known changes for consideration.   Especially as adobe seems to be telling us to update to latest printer drivers. 

[Alan Goldhammer]

One of the advantages to Windows is that MSFT tries to keep things backwards compatible.  There has been no firmware update for the Win driver since October 2013.  The drivers will have a new version when Epson release new papers and they include the profiles and paper settings in the driver download.

[howardm]

if you *keep* your auto nozzle checks, the 3880 firmware version # is printed on it so you can compare pre-v9.33 values.

I also really doubt they snuck in a firmware update.  I think I recently threw out a couple of years worth of nozzle check output so  

[Mark D Segal]

One of the advantages to Windows is that MSFT tries to keep things backwards compatible.  There has been no firmware update for the Win driver since October 2013.  The drivers will have a new version when Epson release new papers and they include the profiles and paper settings in the driver download.

Has Epson not released new Windows drivers for the Legacy papers?

[digitaldog]

Here is a question that just popped into my head after email from someone: does the OS X version matter? I'm on 10.11.5. When did this so called API of Apple's change? Does anyone here with a pretty old version of OS X, say 10.10 encounter this bug?

[GWGill]

No. Whatever change to an operating system Apple introduces, Adobe is responsible for the behaviour of their own application in the context of that OS.

The fundamental purpose of an operating system, is to provide a stable set of API's to applications. An operating system that is unable or unwilling to do that, is a failure. Rapidly deprecating API's, or (worse yet) changing the behavior of existing API's, shows either great arrogance in dealing with those you have locked into your ecosystem, or a reckless disregard. Consider this simple fact: one bad change to an operating system will break thousands of applications and heavily inconvenience millions of users. Insisting that the application vendors work around the change will overall cost thousands of times more, and be much less certain of delivery, than the OS vendor doing the right thing in the first place.

Having said that, from what's been relayed here it appears to be a more subtle sort of issue. Apple has been deprecating and replacing a huge number of API's in OS X, triggering a great deal of inconvenience and many potential problems with applications. Switching development kits is something that has to be done to keep up with the API churn. The question is really whether Adobe have misunderstood something about the new API's, or whether Apple have not implemented them as expected. Either is possible, and pretty likely if Apples documentation of it's colorsync API's is as good as it was last time I looked for it (i.e. it didn't exist). I think it's reasonable to think that Adobe hasn't done the sort of sanity checking of  it's print path that one would hope for though.

 (As  a developer I can tell you that I have a dim view of Apples API churn. The fact that they feel the need to deprecate so many API's indicates to me how poorly they were thought out in the first place, and the last thing I want to be doing is spending precious development time working around OS changes, which adds nothing for the user, rather than actually improving my application.)

[GWGill]

In Windows, the latest driver is for XP and was written in 2007!

But if the Windows drivers are still working fine, and can be color profiled reliably, that's not a criticism - that's a boast! An operating system so stable, and drivers written so well that they still don't need updating!

[David Good]

Has Epson not released new Windows drivers for the Legacy papers?

Yes they have Mark.

[Mark D Segal]

OK thanks David, so referencing Alan's reply #48, the latest Windows driver update would have been fairly recent.

[David Good]

I had recently upgraded to print driver 6.63 (epson17589.exe) for Windows.

[Mark D Segal]

The fundamental purpose of an operating system, is to provide a stable set of API's to applications. An operating system that is unable or unwilling to do that, is a failure. Rapidly deprecating API's, or (worse yet) changing the behavior of existing API's, shows either great arrogance in dealing with those you have locked into your ecosystem, or a reckless disregard. Consider this simple fact: one bad change to an operating system will break thousands of applications and heavily inconvenience millions of users. Insisting that the application vendors work around the change will overall cost thousands of times more, and be much less certain of delivery, than the OS vendor doing the right thing in the first place.

Having said that, from what's been relayed here it appears to be a more subtle sort of issue. Apple has been deprecating and replacing a huge number of API's in OS X, triggering a great deal of inconvenience and many potential problems with applications. Switching development kits is something that has to be done to keep up with the API churn. The question is really whether Adobe have misunderstood something about the new API's, or whether Apple have not implemented them as expected. Either is possible, and pretty likely if Apples documentation of it's colorsync API's is as good as it was last time I looked for it (i.e. it didn't exist). I think it's reasonable to think that Adobe hasn't done the sort of sanity checking of  it's print path that one would hope for though.

 (As  a developer I can tell you that I have a dim view of Apples API churn. The fact that they feel the need to deprecate so many API's indicates to me how poorly they were thought out in the first place, and the last thing I want to be doing is spending precious development time working around OS changes, which adds nothing for the user, rather than actually improving my application.)

Hi Graham,

These are interesting and important insights coming from a developer perspective - much appreciated.

I think the current crop of comments trying to isolate the cause and responsibility for the problem being discussed here, at least to my mind, is not so much to find culprits to blame - just yet, but more importantly to understand the cause so that the problem can be fixed; over the holiday weekend this is turning out to be difficult, possibly on account of defensive behaviour, but there may be other factors. Priority number one is to get this issue resolved; then priority number two should be for the responsible company/ies to take actions substantially reducing the probability of repeat episodes. We've had several too many over the past 18 months or so. 

Your insight into AP degradation is most likely highly relevant, but to my mind not determinative. So in that regard, I have one question and one comment. The question is about the reason for the degradation. You mention the possibility of poor initial design. I really wonder about that. OSX on the whole is a fine operating system, but Apple has been putting a tremendous effort into continued evolution and updating to take advantage of new technologies, enhanced technologies and most importantly system security. Would you rule out the possibility that the changes they are making are in response to these factors rather than poor initial design? And a supplementary, is it not the case that Apple sends developers the new development materials well ahead of the changes being implemented to give other application developers time to adapt (and yes, I hear you about the time and cost involved)?

The comment I have is where responsibility ultimately lies for what a developer brings to market. Given the inter-relatedness and complexity of this industry, there needs to be clear protocol on who is responsible for what and when, otherwise there would be (perhaps is) chaos. I sense from what you are saying that Adobe, for example, remains responsible for properly proofing the performance of an update to its own product, regardless of prior changes that Apple may have made to the OS. This is my understanding; I'm interested to know whether you share that view - i.e. as an application developer, you would want to thoroughly pretest an update before issuing it, especially knowing what you do about how changes to the system environment under the hood could affect your product in unexpected ways.

[Alan Goldhammer]

OK thanks David, so referencing Alan's reply #48, the latest Windows driver update would have been fairly recent.

Correct.  I was unclear in my post about that.  I did install the latest driver but the firmware was unchanged, only difference was support for the new paper line.  I'm on Win8.1 so I cannot comment whether any change in firmware was required for Win10

[Alan Goldhammer]

Your insight into AP degradation is most likely highly relevant, but to my mind not determinative. So in that regard, I have one question and one comment. The question is about the reason for the degradation. You mention the possibility of poor initial design. I really wonder about that. OSX on the whole is a fine operating system, but Apple has been putting a tremendous effort into continued evolution and updating to take advantage of new technologies, enhanced technologies and most importantly system security. Would you rule out the possibility that the changes they are making are in response to these factors rather than poor initial design? And a supplementary, is it not the case that Apple sends developers the new development materials well ahead of the changes being implemented to give other application developers time to adapt (and yes, I hear you about the time and cost involved)?

The comment I have is where responsibility ultimately lies for what a developer brings to market. Given the inter-relatedness and complexity of this industry, there needs to be clear protocol on who is responsible for what and when, otherwise there would be (perhaps is) chaos. I sense from what you are saying that Adobe, for example, remains responsible for properly proofing the performance of an update to its own product, regardless of prior changes that Apple may have made to the OS. This is my understanding; I'm interested to know whether you share that view - i.e. as an application developer, you would want to thoroughly pretest an update before issuing it, especially knowing what you do about how changes to the system environment under the hood could affect your product in unexpected ways.

As I noted in an earlier post, MSFT worries about backwards compatibility with software and always has.  The same cannot be said of Apple and there have been a number of instances documented on LuLa where things have run afoul after new version of MacOS were released.  I had an interesting talk with my daughter's boyfriend yesterday about this.  He's a software developer and just got his PhD in EE and computer science.  He's done a fair amount of app development in the music area and told me that it's often difficult to decipher what Apple has done regarding APIs in new OS releases.  Ultimately, he and others on the team work through these issues but it's not always easy.

To answer your main question, Adobe is responsible for LR/PS and insuring that it works correctly with the OS.  As Graeme notes it can be a challenge to correctly implement a new software application if the documentation from the OS supplier is sketchy.  That's no excuse for releasing a buggy product.  Since the print module of LR is maybe the most critical portion of LR it's mandatory that Adobe get it right.  They failed in this regard.  We don't know what type of QA/QC they employ in testing the new release.  I hope this involved some actual printing!   Assuming it did, how much printing was done and what types of images were printed.  Ideally they would do a standard patch set in the way you have done in your paper testing and checked the results of each patch.  As one who has also done this type of research, it's not terribly hard to do.  I've printed and measured 100 patch sets to compare paper from different batches of Museo Silver Rag and it was pretty straight forward (and I manually scan using an i1Pro).

I hope the problem gets addressed quickly and it would be good to get an explanation of what went wrong during the development and QA/QC.  I doubt that Adobe will tell us what went wrong.