Has Apple Abandoned CUPS, Linux's Widely Used Open-Source Printing System? Seems So (theregister.com) 120
The official public repository for CUPS, an Apple open-source project widely used for printing on Linux, is all-but dormant since the lead developer left Apple at the end of 2019. From a report: Apple adopted CUPS for Mac OS X in 2002, and hired its author Michael Sweet in 2007, with Cupertino also acquiring the CUPS source code. Sweet continued to work on printing technology at Apple, including CUPS, until December 2019 when he left to start a new company. Asked at the time about the future of CUPS, he said: "CUPS is still owned and maintained by Apple. There are two other engineers still in the printing team that are responsible for CUPS development, and it will continue to have new bug fix releases (at least) for the foreseeable future." Despite this statement, Linux watcher Michael Larabel noted earlier this week that "the open-source CUPS code-base is now at a stand-still. There was just one commit to the CUPS Git repository for all of 2020." This contrasts with 355 commits in 2019, when Sweet still worked at Apple, and 348 the previous year. We asked Apple about its plans for CUPS and have yet to hear back.
since it is open source (Score:5, Insightful)
Why fork? Submit to Apple open source repos (Score:3)
Re: (Score:3)
Re:Why fork? Submit to Apple open source repos (Score:4, Informative)
Apple just deleting their repository? Oh come on man.
Forking sounds a bit premature and counter productive. Just submit to Apple for now. **IFF** they act unreasonably then consider forking.
Re: (Score:3, Insightful)
That was then. "There was just one commit to the CUPS Git repository for all of 2020." If two engineers were actually working on it now, ...
Its two engineers are **responsible** for it. What bugs are being ignored? What commits are being rejected?
... there would be more than one commit this year, so that may have been true at the time, but it's clearly not now.
No. The project may have moved from a development mode to a maintenance mode. Sometimes a standard becomes mature and its library complete. That's also a common time for the original dev to move from the old familiar to the new shiny.
Re:Why fork? Submit to Apple open source repos (Score:4, Insightful)
We actually can look at activity:
https://github.com/apple/cups/... [github.com]
So no pull requests have landed successfully this year, but many have been attempted. That would seem to suggest the time has come for another fork to become the reference fork since apple doesn't seem to be thinking about it.
Re: (Score:2)
We actually can look at activity: https://github.com/apple/cups/... [github.com]
So no pull requests have landed successfully this year, but many have been attempted. That would seem to suggest the time has come for another fork to become the reference fork since apple doesn't seem to be thinking about it.
Or the submission did not meet quality standards. As Linus is a quality gatekeeper for the kernel, Apple may be the quality gatekeeper for cups.
Re: (Score:3)
Did all the pull requests stop meeting quality standards at the same time the lead developer left?
The community who submit pull requests didn't have a major change. But the developers who merge the pull requests did under a major change. What's more the more likely cause of pull requests not being merged?
Re: (Score:3)
Re: (Score:2)
Sounds like an argument to not have it tied to apple and MacOS. If it's a cross platform printing library, Apple could always freeze the version they are pulling into MacOS. That's no reason to freeze development on the library though.
Re: (Score:3)
Not to mention that Apple are also in the middle of a CPU architecture transition, so I'm guessing updating CUPS is pretty low on their list.
Who knows, maybe they're switching to something else and that's the reason CUPS hasn't been updated in such a while. Is CUPS used in/by iOS or does iOS works with something else?
Re: (Score:3)
> Is CUPS used in/by iOS or does iOS works with something else?
libcups is in iOS, but the spooler is a completely different implementation on top of it.
Re: (Score:2)
What if there is a theoretical signed/unsigned mismatch that doesn't actually manifest any bug? Is that so critical it can't wait?
Re: (Score:3)
Well, it looks like 'OpenPrinting' will be new 'origin' for most, so it has already happened.
It is one thing to have some open/closed for not being ready yet/being too crappy to start with, but to have *zero* accepted for over a year is telling. Upon review some were pretty straightforward little fixes or translation corrections, and just hang out in limbo.
Re: (Score:2)
Re: (Score:3)
Also the departure of the original developer may have had to do with the change from development to maintenance mode. The standard mature, the code complete and highl
Re: (Score:2)
A maintainer's job is not just to say no, but to make sure that things keep moving forward, while minimizing disruption and breakage to existing users. The fact there are people submitting PRs suggests that there's still people actively putting effort into it, who no doubt would not have an issue reworking things in order to meet quality requirements for merging. Some invariably will just walk away when asked to make changes, but most want to see their efforts reflected upstream in a way that's acceptable t
Re:Why fork? Submit to Apple open source repos (Score:4, Insightful)
Of the code that I've been responsible for maintenance of since before 2015, there have been zero (0) commits in 2020, and there were zero (0) commits in 2019. A small number in 2018. And zero (0) commits in 2017 and 2016.
As with CUPS, none of the code I'm maintaining is even known about by PHBs, and it isn't web frontends, so there is nobody to thrash the code. it just runs and runs. Almost as if transistors wired as digital switches are deterministic!
Re: (Score:3)
I'm sure your VCR firmware and pager to FAX gateway still work great.
But people still user printers. And new ones enter the market all the time. And new apps that print things keep getting written. CUPS won't be done until all the trees are gone.
Re: (Score:2)
New printers are unlikely to need an update to cups, but a new or updated filter. Most likely a ghostscript update.
Re: (Score:2)
Unfortunately, the state on Linux printing is far from "it just runs and runs".
Re: (Score:2)
I'm sure your VCR firmware and pager to FAX gateway still work great.
But people still user printers. And new ones enter the market all the time. And new apps that print things keep getting written. CUPS won't be done until all the trees are gone.
That's just ignorant. "I don't know anything about it, therefore the reasons must be good!" This is why I don't listen to randoms "talk about technology."
If I had a 15 yo print server that had never been updated I could plug in a new printer and go wild, unless I bought a really shitty printer.
Re: (Score:3)
Of the code that I've been responsible for maintenance of since before 2015, there have been zero (0) commits in 2020, and there were zero (0) commits in 2019. A small number in 2018. And zero (0) commits in 2017 and 2016.
As with CUPS, none of the code I'm maintaining is even known about by PHBs, and it isn't web frontends, so there is nobody to thrash the code. it just runs and runs. Almost as if transistors wired as digital switches are deterministic!
Does your code get a steady string of about 20 bug submissions and pull requests per year which you've arbitrarily started ignoring recently?
The thing about maintenance is that you still need to do maintenance. Just because *your* project is perfect doesn't mean other projects are. There are plenty of projects in a maintenance state that still none the less get plenty of releases to close out bugs.
Re: (Score:2)
Re: (Score:2)
a reference to that open-source sustainability metric whose acronym I can't remember.
CHAOSS [chaoss.community].
Re: (Score:3)
Why fork it? Submit changes to Apple's open source repository.
Why the hell do you want to make changes? Maybe after ~20 years this simple, basic tool is finally finished. Maybe thrashing the code on the clock was more about that clock than the code?
Re: (Score:2)
CUPS is a simple, basic tool? It is apparent that you do not have the slightest clue about anything to do with printers or printing.
Re: (Score:2)
Some of the firmware I write... talks to POS printers in restaurants.
You're right though, I don't have the slightest clue about printers. I only know what the software does, the printer itself is a black box to me.
Re: (Score:2)
Apple is deprecating support for print drivers in macOS, pushing the AirPrint and IPP Everywhere protocols. There's a new Printer Working Group.
https://www.pwg.org/ [pwg.org]
Older and specialty printers will be a problem for some time, I'm sure.
Re: (Score:3)
Going with AirPrint and dropping print drivers is a very Apple-like decision. It should "just work", after all.
Are specialty printers even able to connect to Macs? And are the majority of those specialty printers owners using them with Macs?
Re: (Score:3)
Um, the Printer Working Group has been around since 1991, and was the organization behind the creation of the Internet Printing Protocol workgroup in the IETF.
Re:since it is open source (Score:5, Informative)
Already done [github.com], and Michael Sweet is very active [github.com] in this fork.
Re: (Score:2, Informative)
And unlike Apple's, it has open tickets they haven't got to yet, and they're actively thrashing the code with commits like "Cleanup comments/coding style to match contribution guide" which is work that should have happened, and then been closed (following the open/closed principle as this is systems code not a game) over a decade ago.
Re: (Score:2)
which is work that should have happened, and then been closed [..] over a decade ago.
So you are complaining that the new maintainers did not travel back in time and fix things 10 years ago?
Re: (Score:2)
Re: (Score:2)
It is already forked by the main author, Michael Sweet:
XCS [github.com]
It is labeled as an experimental fork of Cups.
It still unclear what will happen in the future. Will Apple pick up Cups development again, or do they leave it as it is? Will Michael Sweet keep on with this effort and will all Free Software projects have to switch from Cups to Xcs?
There is no rush for decisions, the current situation is fine, but in some years time something will have to happen.
Re: (Score:3)
Oh wow, wrong link...
https://github.com/OpenPrinting/cups/ [github.com]
Re: (Score:2)
The founder fork pretty much always supersedes the original. And there is nothing good about Apple having control of Linux drivers. So long and thanks for all the fish.
Re: (Score:2)
fork it and put it on gitlab or github, build a circle of good and well known & trusted developers and the problem is largely solved, printing has not changed that much, just keep it maintained, patched and drivers updated for new printer models
Let me explain: Many years ago, Apple hired the main developer of cups, presumably for a good salary, acquired the copyright (it's still GPL licensed, but I think Apple has the copyright), and paid the main developer to continue developing cups which he did very well.
Some time in 2019, he left Apple, did no more submits, and is quite willing and capable of building a circle of ONE good and well known and trusted developer.
how does Airprint work? (Score:2)
the last few years i've used airprint laser printers and my apple stuff just prints to them with no set up or whatever
Re: (Score:3)
AirPrint is a set of proprietary Apple extensions over IPP [wikipedia.org]. (Essentially it adds discovery of available IPP servers using Apple's proprietary discovery protocol, but there are a bunch of AirPrint extensions to the IPP part as well.) So, yeah, CUPS is probably dead, because the New Hotness (as far as Apple cares) are printers that support AirPrint, which isn't open source and as far as I can tell isn't an open spec either.
It's also very common for Apple to just kind of quietly stop supporting things without
Re:how does Airprint work? (Score:5, Interesting)
For example, "Time Machine," their backup solution, still exists, but they no longer sell the "Time Capsule" - basically a special Apple home NAS device - required to actually use it. And it also doesn't work on APFS devices, since it requires their old HFS+ file system to work, and all current Macs use APFS. Which means that it's not technically dead, but it may as well be.
Not exactly. When Apple disco'd Time Capsule they allowed any vendor to offer it. So all you need is a Synology or QNAP for example and you're good to go.
https://www.synology.com/en-us... [synology.com]
Re: (Score:2)
Re: (Score:3)
Time Machine backs up to any external disk. Time Capsule was just Apple's own version of a backup disk integrated with a router. The hardware project was abandoned because so many competing routers and backup disks existed, and there was no special advantage to combining the two.
Re: (Score:2)
It's the same reason why Apple stopped making AirPort routers.
Re:how does Airprint work? (Score:4, Informative)
I must live in bizarro world, then. Time Machine still works perfectly well on a current system paired with a plain old hard drive set to HFS+
Re: (Score:2)
I must live in bizarro world, then. Time Machine still works perfectly well on a current system paired with a plain old hard drive set to HFS+
I use a combination of one SD card + one hard drive to have two backups of important but small computers. And I wish Sandisk would create some SD card that you can insert into a MacBook without sticking out.
Time Machine is just fine thanks (Score:2)
basically a special Apple home NAS device - required to actually use it.
The Time Capsule was nice because it was a WiFi hub and Time Machine backup disc.
But Time Machine still works great today - either by connecting any local external hard discs and opting to use it for Time Machine, or you can also set up Time Machine over a network shared drive. I do both at my house and have not had any issues.
The same thing likely applies to CUPS: it's not officially dead, but it's not going to see any more updates fr
Re: (Score:2)
Time Machine never worked great. The best thing you can do for a Mac is leave that garbage disabled.
Re: (Score:2)
Time Machine never worked great. The best thing you can do for a Mac is leave that garbage disabled.
Now that is stupid. I have seen NAS drives where it didn't work well, but whatever we have at work works perfectly fine, and at my home it works perfectly fine for me.
Re:how does Airprint work? (Score:4, Interesting)
You cannot Time Machine back up to an APFS disk from Catalina or before. You can with Big Sur.
Re:how does Airprint work? (Score:5, Informative)
Yes, it does still require HFS+ for the Time Machine storage volume, but lets not create confusion here. First, the service doesn't care that you're backing up an APFS or HFS+ volume. Second, when Time Machine creates a backup repository on your backup drive, it creates an HFS+ Sparse Bundle. That could live on an APFS volume happy as a clam. So in other words, Time Machine does use HFS+, but does not care what filesystem you have as long as it can create the sparse bundle on it.
(FWIW, I have Time Machine running on an APFS system and the backup is on an APFS disk as well, so I know this works!)
Re: (Score:2)
(FWIW, I have Time Machine running on an APFS system and the backup is on an APFS disk as well, so I know this works!)
Off topic and out of curiosity, is the APFS a Hard Disk or an SDD? I'm just curious because it's sort of my understanding that APFS isn't so great for spinning media...
Re: (Score:2)
Off topic and out of curiosity, is the APFS a Hard Disk or an SDD? I'm just curious because it's sort of my understanding that APFS isn't so great for spinning media...
The other way round. HFS+ was created long time before SSD, so some assumptions about speed are wrong, and AFPS fixes that so it gets a speed up on SSD drives compared to HFS+. On spinning drives, that speed up is not there because HFS+ always knew how to run fast on spinning drives.
Re:how does Airprint work? (Score:4, Informative)
> Essentially it adds discovery of available IPP servers using Apple's proprietary discovery protocol, but there are a bunch of AirPrint extensions to the IPP part as well.
Um, DNS-SD and mDNS and ZeroConf (Apple marketing name Bonjour) are IETF standards - nothing proprietary there. IPP Everywhere uses the same protocol. Similarly, none of the IPP stuff is proprietary (I know, I wrote the PWG specs...) Just the marketing is proprietary... :/
So no printers or scanners at home (Score:2)
I find I save considerable memory and cpu by disabling cups and cups-browsed on my underpowered 10 year old laptop.
Re: (Score:2)
*@*-samtop:~$ ps aux | grep cups; uptime /usr/sbin/cupsd -l /usr/sbin/cups-browsed /usr/lib/cups/notifier/dbus dbus:// /usr/lib/cups/notifier/dbus dbus://
root 1290 0.0 0.0 105660 13228 ? Ss 12:11 0:00
root 1351 0.0 0.0 303676 10736 ? Ssl 12:11 0:00
lp 1395 0.0 0.0 86420 5792 ? S 12:11 0:00
lp 1396 0.0 0.0 86420 5920 ? S 12:11 0:00
* 10147 0.0 0.0 14428 1048 pts/1 S+ 14:10 0:00 grep --color=auto cups
14:10:54 up 2:00, 1 user, load average: 1.00, 1.25, 1.12
Odd, I'm not seeing that at all on my 7 year old laptop.
It is however running a modern OS.
WFS (Score:2)
Re: (Score:2)
Theres a number of ways of integrating into AD networks with OSX, including automounting of shares etc. Its just not obvious as its sysadmin kinda stufff. The problem is a lot of companies have sysadmins that wouldnt know which ways up on a mac (hint: go to the spotlight, search for 'directory utility' and then do the same search on google)
Re: (Score:2)
Re: (Score:2)
...I just want my Mac to act like Windows. Is that too much to ask?
First, Command-Shift-G. Get to know it. Make it your friend. It may save your life sometime.
Second, I feel your pain, don't get me wrong. I, too, hate having to take something written down as \\1.1.1.1\shared_folder\customer17\complaint3\logfile.txt and turn it into smb://1.1.1.1/shared_folder/customer17/complaint3/logfile.txt. However, the format that Apple is using as a URL is more "correct" than Microsoft's format. The question wou
Re: (Score:2)
Re: (Score:2)
Its just a different way of doing things man. Its like the command-cvx vs ctrl-c vx copy paste things. I've had numerous times windows people say "Man, why did mac chose to be different?". Well considering that the Mac had its keys first kinda seems that accusations back to front (and you get used to just shifting your finger going between OS anyway) Neithers really better or worse. The Apple was is those things belong in the menu at the top of the screen. The windows way has that stuff packed onto the top
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: WFS (Score:2)
Re: WFS (Score:2)
How do you enter the user name in that Microsoft proprietary UNC path?
On the Mac:
smb://domain;user@1.1.1.1/share
On Windows you have to go to a command prompt: /user:domain\user
net use z: \\1.1.1.1\share
(or authenticate with the IPC$ share then go back to Explorer)
Always use the command line on Windows because itâ(TM)s way faster than Explorer and allows you to change the authentication user
Re: (Score:2)
You can just click > type "cmd" > then run "sftp user@hostname"
Re: WFS (Score:2)
I just save an alias folder to my server and macOS automatically connects when I click on it.
Re: (Score:2)
Re: WFS (Score:2)
You can put the alias as a favorite on the Finder sidebar.
What does one use on Linux? (Score:2)
Using Samba between Linux and Linux is perverted and lacks many native things. Or converts back and forth each time.
And NFS, while seeming to be standard, apparently can't and shouldn't be used over a non-always-online connection, plus some other quirks.
sshfs, while being nice, also feels very non-native. I mean, FTP behind the scenes? That's a paddling!
None of those seem viable for /home/$user/. Especially if it should also be mountable on Android and be compatible with some offline-caching.
Re: (Score:2)
Re:What does one use on Linux? (Score:4, Informative)
NFS is designed to be reliable, in its default state if your connection to the server dies it will wait for the connection to come back... Any reads or writes that were in progress will wait until the server returns.
Samba is not, if the server dies then reads/writes will fail.
NFS can be configured - see the hard, soft and intr mount options.
A few years ago i was playing music (from an nfs share) on my laptop, i took my laptop away for the weekend and obviously the music stopped... When i brought it back and reconnected it to the original network, the music resumed from exactly where it had stopped.
Maybe its a solved problem? (Score:3)
Also an original developer leaving when the project goes from development to maintenance mode is a thing that happens as well. Its a good time to handoff something. Move from the old familiar to the new shiny thing.
Re: (Score:2)
maintenance mode.
Then maybe they should maintain it because there's been a steady string of pull requests and bug reports come through for a long time. They only stopped responding to them in 2019.
Re: (Score:2)
He will pathologically defend any decision or criticism of Apple to the digital death. Don't bother.
Re: (Score:2)
drnb is shill level +9000. He will pathologically defend any decision or criticism of Apple to the digital death. Don't bother.
LOL -- silly little troll. I am actually platform agnostic. Windows, Linux, macOS, Android, iOS ... don't care. I have no reluctance to criticize Apple when they actually do something wrong. Here its just premature.
Re: (Score:2)
LOL -- silly little troll. I am actually platform agnostic. Windows, Linux, macOS, Android, iOS ... don't care. I have no reluctance to criticize Apple when they actually do something wrong. Here its just premature.
You're a liar too. Funny.
Exoect to see more of this... (Score:3)
Apple which has never been comfortable being compatible for things they don't have to, will likely keep moving away from standard *NIX system environment. Standby for the 'enhanced' Apple command shell....
apple only printers coming only 30% ink markup (be (Score:2)
apple only printers coming only 30% ink markup (better then HP)
What I don't get is ... (Score:2)
Why the obsession with an always changing code base?
If your code does not approach a perfect state, with fewer and fewer changes, until it's not worth more effort anymore, I'm sorry, but you're doing something wrong.
Something like moving goalposts instead of doing that in a fork or rewrite.
PAPPL Printer Application Framework (Score:2)
Read TFA and comment #7:
https://www.phoronix.com/forum... [phoronix.com]
"Michael Sweet has been developing PAPPL [*1], probably as a replacement for CUPS."
*1) https://www.msweet.org/pappl/ [msweet.org]
why should Apple have it, available elsewhere (Score:2)
I can install CUPS from the usual package managers of software brought over from Linux, do I need Apple's CUPS for anything?
It's 2020. You don't need to print. (Score:2)
Re:Superseded, I would say. (Score:4, Interesting)
You don't know me, but just two weeks ago I was digging into figuring out CUPS once again, because I needed to build a fake printer in order to print to a PostScript file, because I have one stupid application that won't use the native Print Dialogue (meaning, no save-to-PDF was available). Luckily I didn't have to get the point of writing my own LPD filters, but it was a PITA.
What's interesting is I thought I used to be able to save jobs from the print queue of a paused, real printer. You can still QuickLook to see the interim format in the queue, which I think is probably PDF, but no way to actually save that PDF or PS that I can tell. That's too bad.
Re: (Score:3)
So it sounds like they've decided new and even some old printing features aren't that important. And they may be right. I looked at US business paper consumption, and it dopped sharply between 2006 and 2009, and has remained about 20% below the 2006 level since.
Now the big drop was in 2009, which was probably due to the global financial crisis of 2007-2008. But after recovering slightly in 2010, paper use has dropped back to 2009 levels and remained there even as US business has grown.
I suspect what's go
Re: (Score:2)
Re: (Score:3)
If Apple holds the copyright then they own it, regardless of what license it is released under.
Re: licence (Score:2)
Stop using "ownership" for information like it is a real thing in reality!
Go ahead: Try to define that word for that context, on a level of *physics*. ... /opinion/ ... does not qualify.
Hint: Unless you stumbed over causality and the speed of events, you haven't actually thought about it, and your
Re: (Score:2)
Since they own the copyrights they have rights no one else has. Specifically, they can change the license if they wish. By using the Apache license they grant others rights to use and change the software, but they could curtail those rights at any time. Hence they "own" the code.
You can make the same argument you are making with land: how can anyone "own" a mountain or a field? Native Americans in the United States once made this argument too, and you can see how that worked out for them. Ownership may be
Re: (Score:2)
If Apple holds the copyright then they own it, regardless of what license it is released under.
They own it, however, it doesn't really matter because they have granted free license to use it, and that can't be revoked.
Re: (Score:2)
Someone please correct me if I am wrong
It means they own it, and while you may have license to the code, they could force you to call your fork by a different name.
If they didn't own it, if ownership was unclear or something, or they actually abandoned the name then you could still call your fork "CUPS."
But you'll still be able to say it is compatible with CUPS.
Re: (Score:2)
It means they own it, and while you may have license to the code, they could force you to call your fork by a different name.
Not relevant from a copyright perspective, though you're not wrong, simply because I'm sure they have the trademark.
There are other implications to ownership regarding copyright.
Copyright does not protect someone else from using a name you're using.
Re: (Score:2)
If this is correct, then there is no way Apple "owns" it.
They own it. They simply have no control over it.
They are free to use whatever license terms they like with the code that they own; other people who are using it are bound by the Apache license.
The fact that the license is very permissive doesn't mean they no longer own it and are free to do with it things that the Apache license prohibits: patent non-assertion clauses, copyright removal, etc.
Re: (Score:2)
They are free to use whatever license terms they like with the code that they own; other people who are using it are bound by the Apache license.
I think I vaguely remember Apple had to give special terms to some printer manufacturers who couldn't stand the thought that some other manufacturer could "steal" their code. So they could make changes for their specific printer that couldn't be used for any other printer.
Re: FOSS project loses lead dev, goes stagnant (Score:2)
Not any different than non-open business projects. See: The graveyard of projects abandoned by Google.