Complete QR Code Labels Overhaul

[tankerkiller125]

Right now the QR code labels are extremely limited, and I think that it needs correcting. Below is a list of key categories of things I think need correcting.

Labels for All
Labels for every type of item, location, etc. should be supported.

Layout Templates

• Should allow users to create unique templates for their specific labels and save that information.
• Should support multiple-templates
• Should have some pre-built templates (in both US Letter, and A4 sizes).

Options

• Should be able to configure what shows up on a label (just QR code, QR Code + ID, QR Code + ID + Write in line, etc.)
• MicroQR Code support (ability to print tiny labels for tiny objects)

In general, I really like the way Snipe-IT handles label generation, and I think it would be a good example to try and follow in terms of actual generation and options.

[icanc0]

How would a template be structured?

There's this library that does decently flexible PDF generation: https://github.com/johnfercher/maroto

My original plan was "code as templates", a simple function that takes in parameters + data and outputs a Maroto PDF. Except it's hard to make it pretty since we're directly manipulating PDFs. Considering the learning curve for Maroto, an extra dependency, and basically requiring a recompilation, I gave up on my last PR attempting to do this.

Any input on the overall architecture? We can do a custom generation workflow, but it's gonna be a lot of work. Wonder if Snipe-IT has something that we can learn from.

[tankerkiller125]

I've always been about to use F12 and edit the HTML and CSS for the barcodes on Snipe-IT, so I'm 90% sure it's just HTML.

Quite honestly I don't see the value in using PDFs for this, it's just an extra unneeded step IMO. But maybe there is a use case for it that I'm not aware of.

[victorhooi]

Are you sure about the HTML part? Or it's possible there's multiple ways to print labels in Snipe-It?

If you look at my comments in #55 (comment) - I added several screenshots there that walk through the user-flow for printing and customising labels.

If you go the Snipe-It demo page - and try to print a label it does appear (at least to me) to be a PDF file. I did try what you suggested (F12 - or opening up Chrome DevTools), and it does seem like a PDF file:

Or is there another method you were using to print labels in Snipe-It?

To my mind - a PDF is definitely superior when you want to actually print the output to paper/labels. You get better control over how things are positioned, as well as how they are rendered to the printer - which is pretty important for printed content.

Also, consistency/control over printing aside - a PDF file can be easily shared around, particularly to non-tech savvy people, and it's basically idiot-proof for them to get it wrong when printing =). (e.g. I output the labels I want printed, and send it to say, my parents, who know how to print a PDF file).

Finally, I find that despite best intentions, HTML/CSS and how it looks when printed can change between browser, or even between browser versions (e.g. I test between Chrome Stable, and Chrome Canary - and there are rendering differences. If you just want the text on a page, it may not be a big deal if say, the offsetting or the kerning change - but when you're printing labels, it may be the difference between half the text getting chopped off each label, or the label being entirely blank).

Having direct label printer support is also pretty useful, particularly at any kind of real-world scale - if you need to pump out 50 or 250 labels, you aren't going to sit there loading in label sheets to the manual feed tray, then peel off individual stickers from an A4 sheet - you're going to spool it to a label printer, then have them come out on a nice pre-cut roll ready to use. That's why Snipe-It and InvenTree have inbuilt templates or layouts for all of the popular label printers:

Even just at home - I find the Brother P-Touch labels invaluable for just pumping out ad-hoc labels quickly, without mucking around with loading in sticker sheets, or peeling them off, then worrying about wasting the other 23 labels on that sheet of paper.

And for a work context, real label printers (e.g. Brady) have much better resistance to fading, peeling, temperatures etc. (Had a fun time once, where we decided to be cheap, and use standard Brother office printer labels on some fiber/Ethernet cabling - it was near the hot exhaust side for some servers, and within 6 months, all the text had faded - which made the labels completely useless. Brady makes labels that are good for say, 15 years outdoors in UV.).

[icanc0]

I believe the original intent for PDF generation on the backend is so we can have full access to everything on the backend without going through a ton of APIs. PDF generation also ensures that the browser environment isn't going to affect how things are printed, it will print exactly what is shown on your computer.

We could also just use a templating system on the backend, but that comes with its own hurdles.

[tankerkiller125]

Are you sure about the HTML part? Or it's possible there's multiple ways to print labels in Snipe-It?

Maybe there are multiple options, I just tested and it's 100% raw HTML at least on the version I'm running (which honestly might be old at this point) Version v6.1.2 - build 10938 (master) and I don't see any options to use PDF in this version.

In general, regardless of PDF or HTML, in my view a template system for the page layout is 100% required to make it as easy as possible for people to get started with printing. It looks like the easy way to pull this off overall might be to use something like https://github.com/speedata/bagme to first create the page in HTML, and then convert on the fly to PDF.

[victorhooi]

Yeah, as mentioned, I simply tested on the Snipe-It public demo for the screenshots I attached to:

https://snipeitapp.com/demo
https://demo.snipeitapp.com/login

At time of writing it's version 7.0.7 build 14172 (ge50296870).

(I think 6.1.2 might be quite a few years old?)

I couldn't find any HTML label printing functionality - but not sure if I missed it somehow?

If you look at the Snipe-It issue tracker (circa 2018), you can actually see a bunch of issues about people complaining about the printing being off on their printers etc.

grokability/snipe-it#6161

but printed with a shift to the left on a different copier (both Ricoh copiers, both have the same settings, same tray, same browser, same version, etc)

You can see how this is an impossible situation for us to troubleshoot, right? If it prints fine on one and not the other, the issue is with the printer settings somewhere. That page is just HTML.

grokability/snipe-it#2555

grokability/snipe-it#3526

I assume this is around the time (2018-2020) when Snipe-It moved to PDF style label printing?

I do like the idea of a template system (although definitely get some out-of-the-box templates for the most common label brands - e.g. Avery/Brady/Brother - happy to contribute there, and do some testing for you with real labels).

Inventree has a pretty good writeup of how they do their label printing system:

https://docs.inventree.org/en/stable/report/labels/#extend-base-template

[saschabrockel]

Can somebody tell me how I can create QR codes in the current version v0.12.1 anyway? I saw this Reddit post https://www.reddit.com/r/selfhosted/comments/1aqmgo0/feel_like_ive_been_looking_for_something_like/ where somebody has at least a label under the QR code but when I hover in the app to a QR code I have no chance to add any text.

I mean we have this label overview function but how can I really say I want a list of all locations for example and print them all at once? Am I missing something?

[saschabrockel]

Oh damn... I want to do it for the locations like you did. How did you even find my post here 😄

Is it necessary to have a Brother Label Printer then? A friend would print them for me and I don't know what kind of machine he has.

[danielrosehill]

Haha. I was just trying to keep up to date with how the project is progressing.

Just wrote up an updated "how I did it" type post. It might help. You could do it without a dedicated label printer but if you have a large inventory it would be a bit of a PITA.

Another option (as I mentioned) is to go down the NFC route. In which case you could even potentially just use your phone to create the tags on the fly:

#163

[OBoudreaux]

I love the idea of being able to print location labels as well. Not so useful for "Office Shelf" but super useful for the mystery "attic box 21"

[niyisurvey]

I posted this on another thread.. perhaps the incorrect thread:

It's sooooo nuts you asked this TWO days ago. Ive been trying to do this for over a year, tried googling for a solution again today, found the new repo, and found your comment :)

Ive come perhaps halfway in finding a -bit- of a solution:

=======================

First install jq.

Then go to your swagger page (http://homeboxip:7745/swagger/index.html)

Then go to [Authentication] ,
User Login.

Enter your details then go to "try it out" then edit your sign in details, then "Execute". This gets you your bearer token.

Then try this:

curl -s -H "Authorization: Bearer [PUT UR TOKEN HERE]"
http://[HOMEBOX IP ADDRESS]:7745/api/v1/locations | jq -r '.[] | "(.id) - (.name)"

I get this then:

3b8e03b3-5b0b-4b2d-887f-b6f809976fe7 - Bedroom
256e39d3-90f4-4353-a6eb-41e56cdc05f0 - Drawers Column 1
26c5027b-d806-44dc-b640-1606de80b61b - Drawers Column 2
fad7a858-73ad-4633-bab9-8f58b8397f4c - Drawers Column 3
5b83de80-77fb-4427-a78d-e7cb38c351b1 - Fridge Drawers 1
389fcef3-af73-4d41-9887-565ddba332b7 - Hallway
ae1e85a8-9214-4034-a1f5-849e842449e6 - Kitchen
9514ae77-4bfd-46fd-8b9c-ea8164a5ba31 - Left Drawers
e262d913-1865-4d9b-90ec-3742d431668d - Left Drawers Column 1
5b2e9f15-e778-411e-936e-14975cd5e878 - Left Drawers Column 2
5a89aec4-18f1-49b6-a13b-62cc736e57a8 - Left Drawers Column 3
99950b3f-a6ae-49f4-97a1-a72518e3c8da - Left Drawers Column 4
4e88a948-aeec-4ad0-93b4-f6aa4b5066fc - Living Room
9f635d5a-da89-4442-b701-c2ffb981a2d6 - Office
d52bb865-294e-4520-b9c0-4ad9e883454d - Right Drawers 1
d13352fd-8e43-43e7-9e61-c59292c6a250 - Right Drawers 2
1d6060a9-df3d-4b68-96c4-5d33108e6b2c - Tray 1
681ae668-64bb-4635-8086-635c5812d331 - Tray 1
c7931019-cd07-439b-bb3e-40fab5c40fb2 - Tray 1
f9a3

ETC.....

Now this string is basically the bit that goes after the http://[HOMEBOX IP ADDRESS]:7745/location/

EG:

http://[HOMEBOX IP ADDRESS]:7745/location/08bce1db-c0bc-4b00-8ec2-67ec333e8702

This is as far as I have got, for now. I have 99 locations, so Im just goings to use a qr code generator or something with the URL.

I know there must be a less grubby way and more elegant way of doing this, but I'm not a coder and don't really know what I'm doing, besides what's in the docs.

==========

Update: then I went to google sheets, put the url in one column and then in the next column:

Extensions, App script, and put this in:

function generateQR() {
const sheet = SpreadsheetApp.getActiveSheet();
const data = sheet.getDataRange().getValues();
data.forEach((row, i) => {
if (i === 0 || !row[0]) return; // Skip header/empty rows
const url = https://api.qrserver.com/v1/create-qr-code/?size=500x500&data=${encodeURIComponent(row[0])};
sheet.getRange(i + 1, 2).setFormula(=IMAGE("${url}"));
});
}

It makes the barcodes 'too' big, but then you can adjust the columns or rows and it automatically makes them all the size you want.


This is my hall way.. the drawers are "TROFAST" from ikea. I guess they are meant for kids: https://www.ikea.com/gb/en/cat/trofast-system-700556/

Now I have stuck the plastic adhesive ID card holders onto the drawers, printed the barcodes, put them on the flash cards, and popped them in

  

========

[stegl]

"Enter your details then go to "try it out" then edit your sign in details, then "Execute". This gets you your bearer token."

is this right to always say " Base URL: demo.homebox.software/api "

i cant find it where to change the url and it doesnt work...

thanks in advance for any help

[tbgilbert]

I would be happy to support a bounty on this. Its the only thing I do not like about the package.

[dcnoren]

I'd also happily support a bounty on this. Is this possible?

[corwin007x]

If I may, would it be possible to define the sheet and label measures in non-freedom units as an option please?

[tankerkiller125]

You will use freedom units and you will like it!!!

All joking aside though, this is something we need to look into being a switchable thing.

[realchrisolin]

FYI, looks like this was added back in January in https://github.com/sysadminsmedia/homebox/releases/tag/v0.17.0

Added different measurement unit types to label generation page

[rickthegeek]

There are some Brother printers that connect directly to the network via Ethernet, can a label generation / QR function etc be added to directly print the asset label as the item is added? Sadly, I don't know enough about programming to do any work on this...

[OBoudreaux]

It would be nice to customize the text where it says "Homebox" to any text of your choice.

[TheCodeingPadawan]

There is one quick fix that would be good. The text needs to be black by default. Being an off black/gray may look nice, but thermal printers hate it. They have to try and dither to approximate the colour, and that kills the fine text and lines.

Just did it on a 4x6' label, with a Jadens JD-168BT label printer. The dithering made the id too light and spotty, and the text looks like the printer just gently whispered to the paper. The QR code, however, was perfect and crisp.

Further more, There also needs to be a function where if a ID is not yet set to an asset, Scanning the QR code should bring up a page to add a new asset with the ID of the label, Rather than just getting an error saying that ID is not found. This will allow you to print a bunch of blank labels to then add the data in as you go, sticking them to your things and then entering them into the inventory.

It would be nice to also have an option for random ID's, rather than sequential. In my opinion, It just looks better than having a bunch of zeros and then 1, 2, 3...

[agryson]

FYI the dithering issue should be fixed in the next version now that #751 is merged

[PAS-BC]

Having just purchased a Brother QL-800 with continuous paper DK rolls I am really looking forward to progress on this topic :)

[PAS-BC]

Very happy to see this update so I installed ver 18.

My QL-800 is connected to my PC which is running Linux Mint.
I usually have to fire up VirtualBox with Windows 10 to get access to the Brother labeller application.

Seeing that HomeBox is browser based, I hoped the 'Print from Browser' would just work.
I had to reduce size of label to 17% to fit on 'one page' of my DK-22210 (29 mm continuous label) roll.

The label picture showed perfectly on a browser page, with QR code and text looking excellent.

I was pleasantly surprised when the QL-800 spat out a 90 mm long label.

Sadly the label was blank.

EDIT:
I tried again, after connecting to HomeBox via a browser running in VirtualBox (i.e. Windows environment) and the label printed just fine.
Might be an issue with the CUPS driver when I try direct from the PC with Mint.

[tankerkiller125]

As a note there are some environment variables to control some of the label settings. It does use pixels though, so there may be some trial and error to do there.

[PAS-BC]

Where would I start to look for places to change?
Is this where I would create a new 'personal' version to trial things - is 'forking' the correct term?

Total NOOB in this environment though have has involvement in programming over the last 50 years.

[tankerkiller125]

If you are using the docker compose installation, then you already have some environment variables (assuming your followed our doc site). You can just add some more environment variables to customize labels. The relevent environment variables start with HBOX_LABEL_MAKER_ and can all be seen in https://homebox.software/en/configure.html

[PAS-BC]

I am using docker-compose, so thank you for reminding me about the printer options available.
The environment variables are just the standard ones right now and the labels print perfectly if I use a Windows environment.

So I expect my issue is with trying to run the QL-800 printer via CUPS, as happens with my Mint flavoured PC i.e. it's a deeper problem than tweaking some pixels.

I tried using a different browser, Firefox, in lieu of my regular Brave, but that made no difference - the print job is sent to the print queue and just stays in the 'Pending' state.

I can print from PC via CUPS to an old Canon MFP via network adaptor, and to a network attached Brother Laser, but the USB fed Brother labeller is borking.

Off topic, but I discovered that the problem of getting '404' when trying to open an Attachment in HomeBox using Brave is 'resolved' by using Firefox - of course that doesn't explain why Brave can't find (404) the attached file to either open it or download it.

[mygrexit]

This is the only downside to this great piece of open source software. How can we start a bounty?

[dscottboggs]

Hello, new user to Homebox, just getting it set up and assessing it for my purposes. I wanted to offer some feedback on the label generating feature:

• The generator assumes you're only going to have 1 of each asset. I may have several of something and want to be able to label each. Maybe, an option could be to specify a specific series of asset IDs, in addition to being able to specify a range. That way, I could create an initial asset of some item, then pile it up alongside its duplicates, and when I've gathered them all, print however many labels I need of the duplicates.
• When an asset ID is scanned that doesn't exist, an error is shown. It would be nice if that would trigger a "Create this asset" form instead, where the item would be associated with that asset. This would take the guesswork out of assigning a sequential ID to assets as they're entered. I may take a look at what it would take to put together a PR for this one, if that would be welcome. It seems like a small change.

Thanks for the awesome software and for taking the time to receive feedback on the feature.

[klausmcm]

For my (python) projects I'm currently depending on https://jinja.palletsprojects.com/en/stable/ (along with https://doc.courtbouillon.org/weasyprint/stable/ to convert the html to a pdf). This addresses all the Layout Templates points. I also created functions that I made available on the template to create one dimensional and two dimensional barcodes (QR, datamatrix, aztec). These functions take the string to encode and desired barcode dimensions as arguments and generate a base64 decoded image of the barcode that can be put into the html template. This should address the Options points.

[fslef]

It would be great if we could print the labels for all locations as well (in addition of assets)

[Viper00o]

You can

[fslef]

How can I do this? I see an option to get the code per individual location, but I don't see an option to print all location labels at once.

[fslef]

So @Viper00o ?

[Viper00o]

You are correct. I didn't understand that you meant all at once. You can do them all but only one at a time.

[IlyaVassyutovich]

Hello. I'm very grateful for your work!

Is it possible to remove Homebox and $Location from label template? I have small (height-wise) labels and these parts are neither needed (for me) nor fit :)

Screenshot and corresponding vue-code

homebox/frontend/pages/reports/label-generator.vue

Line 453 in e4cc79e

<div class="text-xs font-light italic">Homebox</div>

I don't really want to fork Homebox with all build-release hassle.

[luisriverag]

An option to add a field with user/contact details would also be convenient for items left in shared spaces

[agryson]

I've been looking at this a bit and think the simplest approach, compatible with the design guidelines of being simple and easy to use, is to have 2 or 3 basic templates that one chooses from as an env variable (e.g. compact, detailed, ...) that then tries to fit within the given height/width that are already available as vars.

Compact would be for example QR code with ID below by default.
Detailed could be like today with id, name, location to the right.
Etc.

That should cover ~80% of cases and it's relatively easy to add a new template if user feedback asks for it without going full blown custom.

This could then be extended with customisation down the road if needed of course.

[agryson]

So I played around a bit (I haven't developed in years, and never in Go).
Not all barcodes were working but I got:

• Aztec
• DataMatrix
• QR (default)
• PDF417
  

It works... but I have no idea what I'm doing.

[danielrosehill]

@agryson Ah yes ... the barcode type rabbit-hole. I remember it fondly!

I can't remember off the top of my head why I concluded that data matrix had a slight edge over QR .... but I think it had to do with the density of the code and the amount of ECC that could be fitted into it. But basically there wasn't any meaningful difference.

The other ones, as you noted, are kind of funny looking but don't really have much advantage over QR or data matrix for this use case. PDF417 is quite space inefficient IIRC. And Aztec doesn't fit much data at all.

I think I probably got to whatever answer I ended up with by asking ChatGPT something like: "what barcode symbology do wharehouses use for tracking inventory?" and then rolling with whatever it recommended.

For what it's worth: I think that a bigger concern than barcode type is actually buying decent label tape that doesn't fade!

After six months to a year I found that my labels which had to be exposed to sun were already hard to read so putting the inventory on them is definitely a good idea (I think that was my little contribution to the project, actually!)

In the end, I moved over to NFC, kind of for this reason. But I think they're both decent systems!

[agryson]

Yeah, having done it, I don't think there's much value compared to the complexity it would bring - but I'll run tests on smaller codes and see if datamatrix might have an edge there.
Honestly this is more of a test to learn a bit of the project (and Go!) before I try to tackle trying to create different templates (my cozy/compact idea from above).

[tankerkiller125]

As someone who works with a lot of warehouse type enterprises, Aztec is extremely common for shipping related things (UPS, FedEx, etc. all use it). PDF417 is also pretty common in that space as well. While PDF417 is in fact wide, the fact that it has data correction is why it's used. While Aztec from my understanding is more for the automated machines or something along those lines. Most of the places we've worked with though are moving towards RFID, especially because they can use passive reading to get a general location inventory, or get the fact that it went through Warehouse Door 3 or whatever.

For the average home user though QR Codes for most applications will probably be the preferred option.

[klausmcm]

Using the same module size, the data matrix barcode results in a smaller barcode (based on some testing I did with default error correction parameters). In the context of this project, PDF417 (or linear 1 dimensional barcodes) may be useful for labels that need to be short but can be wide.

Could the type of label be configurable? In my own label project, I use jinja2 templates and I include functionality in the templates to convert strings into base64 strings of the barcode image. I have functions that can generate data matrix, aztec and code128 barcodes.

[agryson]

So, (and bearing in mind that I'm doing this mainly for fun and education - can't guarantee I'll ever have production-grade code!) I think the approach I'd take here (chronologically) is:

1. 🛹 Provide 2 basic templates through env var for the 80% of users with sensible defaults
   a) Cozy (current)
   b) Compact (small as possible, prob just QR+ID)
2. 🚲 Provide basic customization through env vars (on top of what labels does today)
   a) Let user define what is line 1, line 2 etc. (id, name, location, My cUsTom TeXt, none, ...)
   - Cozy = Code + 4 lines max
   - Compact = Code + 1 line max
3. 🚗 Provide extended customization through env vars to:
   a) Extend Cozy/Compact to include more templates, maybe 4 or 5 in total, some of which are appropriate for 1D / wide codes
   b) Provide option for code to only encode ID, not URL
   -> This obfuscates my home server, reduces data to be encoded, allows for numeric barcodes, etc.
   BUT -> Restricts scanning to doing it from the app - may need work on mobile UX for that
   c) Expand range of codes available now that we have the option to only encode 6 numbers
4. 🚀 Create a UI tool in Group Preferences that lets us move away from env vars to saving these things in DB <- Probably a job for someone else!

Default would always be QR + Cozy.

[realchrisolin]

I also like how Snipe-IT handles label generation and have a couple ideas on how it could be better implemented in HomeBox, ranging in complexity and effort, that I'm willing to try writing. To be brief, I think modifying the /reports/label-generator page to function more like the Snipe-IT /admin/labels page makes the most sense but I'm open to other suggestions. Currently, the label-generator page only supports asset labels (no location labels) and has no support for different printers / label presets while the dimensions for the Avery 5260 are hardcoded in. Existing "print-ahead" label functionality can be preserved, while support for dynamic asset and location label generation can be added.

As this feature continues to be fleshed out, I'm envisioning changes to the labelmaker package to support "horizontal" and "vertical" label orientation to control how the label is constructed and a global "settings" page for label generation that sets the default label dimensions / orientation / fields that are passed to the labelmaker API when generating a label from the item or location page (similar to how this is implemented in Snipe-IT as seen below)

All that said, I'd like to get an update on the current direction / status of this initiative. It's been a year since this discussion was opened and there has been a number of improvements around QR code generation since with a few outstanding PRs in the works.

[realchrisolin]

Is that being discussed / tracked in another issue? Is there a tentative ETA on when that'll happen? It sounds like any of what I proposed should come after those changes.

[fslef]

A note on this one, this will be changed when we perform the entity merge for items and locations (just by the nature of the merge)

OK, but be careful, printing this 2 types of label serves two different needs. You must be able to choose to print either items labels or locations labels, if you are mixing all labels in one sinple page, the current function will lose all its purpose and will be useless.

[realchrisolin]

I'm inclined to agree with @fslef. In lieu of additional technical detail, merging the two seems counterintuitive and likely to cause more issues than it solves.

I came up with a working proof of concept: realchrisolin@ac128ff

[realchrisolin]

added PDF generation proof of concept too realchrisolin@abd85b6

[tankerkiller125]

The merging of location and items into entities is being done to deal with feature partity (custom fields, pictures, maintenance, etc.), in theory they can still be split for things like labels with some filtering, but generally speaking on the DB they will be the same table.

[Br0kenSilos]

Hey, great software. I realize that the labels feature is still under heavy development. I just want to toss an idea in there.. And maybe I have missed it somewhere.. But it would be nice to be able to select which labels to print. Either on the item page (like a check box to enable or disable label printing for the item) or possibly on the label printing preview page which check boxes to select which ones to print or not. This option should probably an Select All toggle.

I really appreciate all the work that has gone into this project. You guys rock!