Printing

where do I find a PDF for a monochrome printer

Uh, this is RISC OS. There’s probably more mono than colour. ;)

Try the HP range. The old DeskJets (500?) were mono, with the colour (550C?) coming along a little later.

There’s also most of the Epson fx and lq range. HP LaserJet 2 as the bog standard for laser compatibility (which, amusingly, even works with mine today! (LJ6 for better resolution)).

Or do I misunderstand you?

Yes. ;)

The dumper processes what it is asked to do, but if there’s a choice of strip types, something must be controlling that choice. I think this is Printers using the PDriver_SetDriver SWI. Which I presume would be controlled by a PDF.

Try the HP range. The old DeskJets (500?) were mono, with the colour (550C?) coming along a little later.

There was a 500C, which could do mono or colour, but not both at once :)

There was a 500C, which could do mono or colour, but not both at once :)

Some while back, 1990’s??

Black cartridge and a CMY (no “Key” so the attempted “black” was more “dark mud” IIRC)
Feed rollers used to polish themselves, so a bit of abrasion was required at intervals.

Yeah, the ‘black’ was more of a greeny-brown. And I think you’re right about the rollers too!

FX80 was the first thing I looked for, before posting here. Not in my collection of PDFs because I have Gutenprint. Eventually I found it in a HardDisc4 download. I was able to see that my dumper had been given a strip type 1. More work tonight.

It doesn’t help that I don’t have a memory of lots of printers. I suppose I was never very interested in them. We had a Brother M1109 which was small but so heavy it appeared to be made of collapsed matter. It served us well for years. In well over 30 years we have only had 3 printers.

It doesn’t help that I don’t have a memory of lots of printers.

Given the headaches (see below), that’s probably just as well. ;)

Epson fx80 – workhorse, pretty much any dot matrix worth looking at supported this.

Epson lq – like the fx, but IIRC the matrix was 24 pins so stuff looked a lot nicer.

HP DeskJet – the standard for early inkjet printers. Originally monochrome, gained colour in varying degrees (such as the “can print in colour or black but not both at the same time”).

HP LaserJet – the standard for laser printers. IIRC LJ1 was 150dpi and LJ2 was 300dpi. Any laser worth looking at back then supported LJ2.

PostScript – supported by printers with scary price tags.

Everything else – something that was either really specific (like a golfball or daisy wheel printer, plotter, etc) or some piece of crap that wasn’t work looking at.

The primary problem is that pre-Windows, every program required its own drivers, and it wasn’t feasible for the smaller manufacturers to roll out drivers for all sorts of programs. Therefore there was milage in supporting something common.

An example (referenced above) was a laser printer I used to have. I think it was an Epson GQ3500 or something like that. It did not support LaserJet, but it did support the LQ (yup, treating a laser as a dot matrix!). Licencing issue? Thankfully they came to their senses with the later model (GQ5000?).

Then there was stuff like the Kyocera f800 that included its own page description language that was pretty powerful but not like anything else (and thus supported by practically nothing).

In the more modern era, a lot of cheap printers were classed as “Windows printers”. This is because the printer was just a print engine, all the clever stuff was done via the driver. Accordingly, they didn’t tend to do anything useful plugged into a non-Windows box.

A hybrid is stuff like… PCL6e I think they call it. It’s a version of HP PCL understood by a few budget printers. Sadly, this PCL isn’t documented.

Finally, coming into modern times, people wanted to print from their various devices. Apple led the way with AirPrint, which “just works” (but often does really dumb things like blowing up a small picture to full page size). Apple being Apple, there is practically no control over what happens. Pick a printer, pick the number of copies, prod the Print button. There’s another group with a weird name (I forget) bringing this functionality to other devices. Moria? Mopria?

More recently, IPP is trying to create a protocol for an “it just works” printer.
Modern WiFi printers will support at least AirPrint. There’d be no point in having WiFi otherwise. More recent ones will do IPP as well.

It served us well for years.

Back then, printers were built to last.

I have a Panasonic KX-P1180 under the other desk. It’s covered in dust and hasn’t been used in thirty years. I have a feeling that a clean and a drop of oil and a new ribbon, and my main problem will be trying to find a parallel port to plug it into.

My HP3630 inkjet? I’m actually surprised it’s still running after four and a half years. Maybe because I’m an Instant Ink subscriber? Or is that too cynical, even for me? ;)

The FX-80 PDF makes the Printer Manager offer me three types of grey scale, all of which result in strip type 1 when a monochrome or colour file is printed. (Funny, I don’t remember the FX-80 supporting grey scale – the pixels it printed were either black or nothing…) It was a surprise that it didn’t offer type 0. I think my best chance of being offered the other types of strip will be to reduce the strip types that my dumper offers.

I’ve got a convincing rendering of a simple sprite and simple Drawfile to a grey scale, which is actually R = G = B on an RGB 24-bit output. PWG-Raster doesn’t offer any other way of rendering a grey scale AFAIK.

Strangely, asking Paint to print my sprite in landscape mode produces a wrong result. I suppose it must be my fault, but it’s going to be a challenge to diagnose that, and it’ll have to wait until tomorrow.

Discussion of the way to send large data payloads is off topic, so I’ve started a new topic “URL_Fetcher and AcornHTTP” in Code Review.

PWG-Raster doesn’t offer any other way of rendering a grey scale AFAIK.

There is an sgray8 format, which doesn’t appear in PWG 5102.4-2012. A most curious omission.

There is an sgray8 format, which doesn’t appear in PGG 5102.4-2012.

The chart on page 30. There’s an sgray_16 version as well.

Perhaps, since the sgray ¹ example is one bit, and the srgb uses eight bits, they’re thinking you can work out the eight bit greyscale method, it’s just like r with no gb. ;)

¹ Bugs me to spell it with an ‘a’.

Two progress items to report.

First is that a couple of us have successfully printed out some documents using IPP. It’s a process that my last employer described as “knife and fork”, though; it’s not ready for general use by non-techy users yet.

Second is that I’ve just released an app called GetIPPInfo that should collect IPP info from printers, to try to help deal with some of the inconsistences in implementations of IPP in various printers. I’ve announced it in the Announcements forum. I would very much appreciate it if those of you who have IPP printers would use the app to gather the info and email it to me.

It’s a process that my last employer described as “knife and fork”, though

I talk a little about this on my blog, today’s entry. While a fair bit of manual work is needed, it can and does work, and honestly I think huge swathes of the problems are limitations in the print system and printers themselves being a pain in the intergluteal cleft.

What Dave has managed this far is amazing. I take my virtual hat off.

an app called GetIPPInfo

Looking at this in an editor on my phone, it seems to support both http and https, going by strings in the executable. If this is the case, does it use the URL fetcher?

If so, there is a chance that certain HP printers will not respond correctly because they refuse to accept HTTP/1.0 even though there’s no sensible reason for that (my own low level code is written with HTTP/1.0 capabilities and I just bung in a /1.1 identification and it works fine.
So if you have an HP that does not work (such as the HP 3630 inkjet), please leave a note here. I can link you to an alternative to collect data from it, but it would be useful to know how many HP models are objectionable (and, by extension, how many RISC OS users have such a beast).

Looking at this in an editor on my phone, it seems to support both http and https, going by strings in the executable. If this is the case, does it use the URL fetcher?

It does indeed.

…I can link you to an alternative to collect data from it…

Thanks, Rick, that’s good.

I must look some more at the HTTP module. I don’t think the difficulties are insuperable. (Whose famous last words are they?…)

It is also important to interrogate your WiFi capable printers for another reason. As As I mention on my blog, WiFi printing is actually two different protocols, that are actually quite similar.

The first, supported on practically everything, is Apple’s AirPrint. This uses IPP as a transport to get data to the printer, but the actual data is a closed “URF” format (though it has been reverse engineered to some degree).

The second, that Dave’s IPP printer driver works with, is the IPP Everywhere specification, which uses an open and documented bitmap format called “PWG”. This is likely to be supported in some printers as of 2016-2017ish, and it would be strange to find a printer that had been sold this decade that did not support it.

As PWG is similar to URF and it uses the same transport mechanism, there’s a possibility that a firmware update will add such capability. My short-lived Epson printer claimed AirPrint, not IPP Everywhere. It got a firmware update, it then handled both.

So if your printer is PWG capable, or AirPrint URF only, do be sure to send Dave your results so he can build a picture of what sort of devices us RISC OS users actually use.

If the HTTP module can be persuaded to send arbitrarily long payloads, presented to it in arbirtrarily-sized lumps, I can see the way to provide a little app that sits on the icon bar, to which a PWG-Raster file from my PDumperIPP module could be dragged. That would make IPP printing a two-step process, easy enough for non-techy users to do.

So if your printer is PWG capable, or AirPrint URF only, do be sure to send Dave your results so he can build a picture of what sort of devices us RISC OS users actually use.

Very much yes. It would be easy to create a URF dumper module. Most of the code from the IPP module would be the same, other than populating the page headers.

But let’s support the non-proprietary protocol first, and see how many printers are still excluded. The IPP info will tell us.

That would make IPP printing a two-step process

I was thinking, using my SpewPWG (because my printer is annoying) that it might be possible to take a leaf out of PrintPDF’s book and simply notice when a specific file appears (and is closed) to then load it into memory and spool it to the printer.

So long as the user only has the one printer (for now!) that might even manage to make it an automatic thing!

Dave,

I have tried your GetIPPInfo application on the following HP Printers:

HPLaserJetPro254DW
HPDeskjet3637
HP B110 Mulitifunctional

None give any data in the file apart from:
GetAttrs 0.01 (2022 April 2)
http://’It’s Ip Address’:631/ipp/print

I have tried it with the alternative port as well.

The ported IPP tools, by Chris Gransden, does work with two printers to get the information.

HP Laser Jet Pro 254DW supports airprint-1.6
HP 3637 supports airprint-1.4,ipp-everywhere

The HP B110 doesn’t give any details but I am sure I did get some at one stage.

Update: Looks like the HP B110a doesn’t do IPP/Airprint according to the manual I just downloaded so my memory playing games again :-)

I’ve just tried GetIPPInfo on my two printers. One is an M277 multifunction, whose status page includes Airprint, but not IPP. The other is an HP P3015 which doesn’t mention either. The info from both is:

GetAttrs 0.01 (2022 April 2)

http://M277/ipp/print

GetAttrs 0.01 (2022 April 2)

http://P3015/ipp/print

Does this suggest I need Rick’s modified version?

Well, I guess the good news is that it’s not just my HP printer that is fussy. It’s HP printers.

The bad news? It’s a bit of a pain in the arse, but then, printers are sent to test us…

https://heyrick.eu/blog/files/findipp_002.zip

It can run in one of two modes.

The first is a blind network scan where it will look for every device from x.×.×.1 to x.×.×.254 to see if there’s a response to a standard IPP information request.
Syntax: FindIPP
(note – this needs Inet$EtherIPAddr to be correctly set to an actual IP address rather than a hostname, in order to determine the LAN addressing, like 192.168.1.x)

The second mode is to provide it with the IP address (not name) of a specific device, which will be examined.
Syntax: FindIPP -ip <ipaddress>

It may spew a lot of data, so best run in a TaskWindow. The report will likely contain your WiFi AP’s SSID and your printer’s serial number as well as its IP addresses (IPv4 and IPv6). Possibly cartridge serial numbers as well. Private information such as passwords (AP or printer admin) do not get handed out. Whether or not you choose to edit out such information (don’t delete anything, overtype it with ‘X’) is entirely up to you.

If you experience problems, or don’t get data back, you can try adding the -dump command line parameter which will also perform a raw dump (Edit-style, like ^A=“[0A]”) of what is actually received before attempting to decode it.

Notes:

1. This works by performing a blind scan. The “correct” way is to use Bonjour, but we currently have no mDNS client. As such:
   1. Devices that use odd paths (like “/my/furry/kitten” which is completely valid) will not be correctly scanned – Bonjour provides the correct URI and, as far as I can determine, is the only way of knowing it.
   2. Additionally, some of you may see a phantom device, if your ADSL box, router, or something else is able to act as a printer server. Please report these too, as we’ll likely need to tell what is and isn’t a real printer. I can’t speak for other routers, but mine (Orange Livebox Play) does respond to IPP proding but does not announce itself as a printer via Bonjour).
   3. Really, I think we may need Bonjour at some point. ;-)
2. Some printers may need to be scanned twice. I’ve set the timeout fairly slow, but if it’s sleeping it’ll need longer to wake up.
3. Do not use hostnames, I don’t perform any resolving. IP addresses only.
4. If your printer is set to require security, you’ll likely get an HTTP “426 Upgrade required” response. My software does not support SSL/TLS so you’ll either need to temporarily disable this in the settings, or (if it doesn’t work with Dave’s program (ie is an HP)) report here that it is a “difficult” beast.

Data for a HP Office Jet PRo 7720 Series
GetAttrs 0.01 (2022 April 2)
http://192.168.0.111:631/ipp/print: 404
http://192.168.0.111:631/ipp: 404
http://192.168.0.111:631/ipp/printer: 404
http://192.168.0.111:631: 404
https://192.168.0.111:631/ipp/print

Hi Rick,

Thanks for the FindIPP app.

It finds all three of my printers and gives reports.

Even the HP B110a that doesn’t mention IPP or Airprint now reports version 2.

It seems it has to have ipp/printer at the end of the address not ipp/print.

Also using the correct URL now allows IPPTools to get information from it.

I’ll look at the data and clean it up and free it from any sensitive information and send it to Dave later this evening as I have to be somewhere..well should of been there an 1hr ago but time go by so quickly when you are enjoying testing things :-)

Well, my app hasn’t been a conspicuous success, has it? :-(

I think it has a problem dealing with things that just don’t respond, or maybe respond with an error that I don’t handle.

Well, my app hasn’t been a conspicuous success, has it? :-(

Seems like people that have gotten involved so far are using HP printers.

In other words: uh oh.

;)

or maybe respond with an error that I don’t handle.

I notice you try https after http fails. I would imagine that fetching status with plain http should return a 426 error to require encrypted.
(though this is only for fetching status, print job requests will just be ignored if encrypted is required)

But, yeah, there’s the lovely disconnect between what’s supposed to happen and what actually happens. :<