Re: [london-hack-space] QR Tagging

[Adrian Godwin]

On Mon, Feb 21, 2011 at 11:40 AM, Ciarán Mooney <general...@googlemail.com> wrote:

 

* Each code is non-unique until someone registers it.

Does registration require a fancy phone  ? (I don't have one either)
Could there be a space-owned reader that could just be taken to an item to answer the 'whose is this' question ?

I liked the idea of the QR codes but agree it's completely unusable in practice. I'm glad you're trying to fix it.

-adrian

[Ben Blundell]

www.talesofthings.com

This was a page I built when I worked at UCL that does exactly this, especially if you put avery labels in the printer.

Android and iPhone clients are available. Dont judge me though. personally, i hate this work because of politics and the shitty design i was forced to use. nevertheless, it works and is supported and should be fine for this.

Users can comment on stuff and add media too....all that sorta crap you want from a social media esque site.

Ben

On 21 February 2011 11:40, Ciarán Mooney <general...@googlemail.com> wrote:

Hi,

Generally the space has a problem identifying ownership, large things
get left to be worked on, or semi-abandoned when that member doesn't
come as often. Other times people see something that looks cool but
can't find anyone to tell them about it, or people want to use an item
(dishwasher) and don't know to check for salt.

To help solve this Sam The Techie made some QR codes that could be
affixed to projects so they could be identified by people with smart
phones. This is a great idea but I think there were a few issues that
prevented uptake.

* No positive encouragement by members to use codes.
* Each QR code was generated for a specific project, and only done once.
* The codes could not easily be stuck to peoples projects.
* Code requires a smart phone to use. Those who lack one cannot access
information via the code easily.

As an extension to these QR Codes I'd like to print off a few
pre-generated QR codes printed onto sticky backed Avery labels. This
would be accompanied by brown labels with string, so the QR code label
could be stuck to the brown label, then tied around item. For
situations when you don't want to put a sticker on something.

The QR code itself would direct to a web-form which when filled in
generates a wiki page about that particular item. After the page is
generated the QRcode will redirect to the appropriate wiki page and
can be used by other people to look up who the machine belongs to,
it's current status, how to use it, etc.

Alongside the QR code a few printed human readable comments would be
useful. A shortened link for those who lack fancy phones (ie me), a
"Responsible member:..........." box, "Can be hacked, [YES][NO]?", and
"Contact details: ........". This would allow people glance at the
sticker and discover who they can ask about it, or inform of it's
imminent destruction. Without the need for a fancy phone.

This is for small to semi-large things, ie microwave, projectors, PCR,
tesla coil, trebuchet or other works-in-progress. Things that may have
indeterminate ownership. Other items like the Makerbot or laser cutter
are obviously owned by the space, but may benefit from people being
able looking up information and documentation about them easily.

Tables and chairs probably don't need to be tagged. (Although there's
no really good reason why not, just because it's boring.)

As with many things at the space this will require a consensus to take
root. People will hopefully use the system if they feel it benefits
them. Although as an initial seed you'll find me walking around asking
people to put them on their stuff.

I personally don't have enough coding experience to get this done on
any short time-scale, but I am willing to have a crack at it. I'd like
it to be written in python because that's the only language I half
understand.

In short the new QR labels would be,

* A code that can easily be stuck to a project by peeling off backing.
* A code that can non-destructively be attached with string and brown
paper labels.

* Each code is non-unique until someone registers it.

* A new set of codes can be generated as and when required.
* There is human readable text to make things quicker and easier for
alternative situations.

Most of these ideas are going to be transplanted onto the wiki page
soon, along with some of my thinkings behind the work-flow of the
programs.

Ciarán

--
(>) SECTION9 * Benjamin Blundell
(>) b...@section9.co.uk * www.section9.co.uk
(>) Code, Design, Graphic, Web 
(>) London Hackspace, Laboratory 24, Cremer Business Centre. London E2 8HD

(!) For Secure email use PGP Key-ID: 0A9C59AF @ http://pgp.mit.edu

[Ben Blundell]

We implemented a camera that uses a flash plugin. In addition, I built an OpenFrameworks read that works with webcameras. The problem with such cameras is bulk and lack of autofocus.

Its been done is the answer and it aint great.

RFID stickers are the solution imho and we can build these with a simple code lookup and some LCD thingy. that would be a portable reader we could use.

Ben

[Ciarán Mooney]

Hi,

> We have codes on the trebuchet already, do we need to do something now?

I'd like to "grandfather" the old tags in, so people like you who have
been super-considerate are not put-out by a new system.

> FYI, our QR code is just a weblink to the project page, so it is
> already unique, and I would expect this to be the case for other
> projects too.

The new system I imagine would work on something similar.

Ciarán

[Ciarán Mooney]

Hi,

> Can't we just have a printer which can print sticky labels on demand?

Yes, Jonty mentioned he has one, but it only works under Windows.

As a first instance I'd like to print out ~20 labels, which would be
fine done on an A4 printer. After the initial batch I would try to
keep the unused tags available, but would like people to
print-on-demand too.

Ciarán

[spooq]

We have codes on the trebuchet already, do we need to do something now?

Luke

On 21 February 2011 12:06, Adrian Godwin <artg...@gmail.com> wrote:
>
>

[spooq]

Ahh I see, you are pre-generating a bunch of stickers for people to use.

Can't we just have a printer which can print sticky labels on demand?

Luke

On 21 February 2011 12:12, Ciarán Mooney <general...@googlemail.com> wrote:
> Hi,
>

>> Does registration require a fancy phone  ? (I don't have one either)
>> Could there be a space-owned reader that could just be taken to an item to
>> answer the 'whose is this' question ?
>

> My idea is that the QR code just contains a link to a webpage, and
> also includes a printed version of that link below the code. You could
> just put that URL into your browser and fill out the same form.
>
> If you wanted to tag something first, and fill in the paper blanks,
> and then it can be logged on the wiki later. The QR code will be
> unique so once you've stuck it to your item it can't be used by
> someone else.
>
> Ciarán
>

[spooq]

FYI, our QR code is just a weblink to the project page, so it is
already unique, and I would expect this to be the case for other
projects too.

Luke

[Ben Blundell]

We ran a system like this at Animation 10. Lots of kids looking at very interesting PC bits and old Comp Science stuff. We had two cameras they could scan the item with.

http://www.flickr.com/photos/section9/sets/72157624543374684/

Ultimately, the reason kids liked it was they could create their own media on the fly with a qr code that would play back their message. There was no interest in the actual scanning.

Kids are a good basis for testing whether a system is fun and enjoyable to use. We've seen that scanning with iphones and android and the like is good for about 5 seconds before people just get bored. No-one cares about the page for your site when it takes a while to scan and then load on a small handheld.

That said, I agree wholeheartedly that some kind of device that a visitor can just pickup and use to say "ooooh what is this project about?" is a much MUCH cooler and more hackspace idea. I'd suggest focusing efforts on that.

Ben

On 21 February 2011 12:16, spooq <spo...@gmail.com> wrote:

On 21 February 2011 12:15, Ben Blundell <o...@benjaminblundell.com> wrote:
> This isnt the problem.
> We had this issue before. Who carries with them an android or iphone at all
> times or a camera phone that can scan qr codes?

I do. It is literally never further than a couple meters away from me,
and usually much closer :)

> And even if you do, can you
> be arsed to go grab it and then scan it?

Sure.

> Would it not be simpler to just
> build an rfid handheld scanner (or buy one) for the space to do this?

That could help the people who don't carry smartphones, sure :)

Luke

> Ben

>
>
> On 21 February 2011 12:12, Ciarán Mooney <general...@googlemail.com>
> wrote:
>>
>> Hi,
>>

>> > Does registration require a fancy phone  ? (I don't have one either)
>> > Could there be a space-owned reader that could just be taken to an item
>> > to
>> > answer the 'whose is this' question ?
>>

>> My idea is that the QR code just contains a link to a webpage, and
>> also includes a printed version of that link below the code. You could
>> just put that URL into your browser and fill out the same form.
>>
>> If you wanted to tag something first, and fill in the paper blanks,
>> and then it can be logged on the wiki later. The QR code will be
>> unique so once you've stuck it to your item it can't be used by
>> someone else.
>>
>> Ciarán
>
>
>

> --
> (>) SECTION9 * Benjamin Blundell
> (>) b...@section9.co.uk * www.section9.co.uk
> (>) Code, Design, Graphic, Web
> (>) London Hackspace, Laboratory 24, Cremer Business Centre. London E2 8HD
> (!) For Secure email use PGP Key-ID: 0A9C59AF @ http://pgp.mit.edu
>

[spooq]

On 21 February 2011 12:15, Ben Blundell <o...@benjaminblundell.com> wrote:
> This isnt the problem.
> We had this issue before. Who carries with them an android or iphone at all
> times or a camera phone that can scan qr codes?

I do. It is literally never further than a couple meters away from me,
and usually much closer :)

> And even if you do, can you
> be arsed to go grab it and then scan it?

Sure.

> Would it not be simpler to just
> build an rfid handheld scanner (or buy one) for the space to do this?

That could help the people who don't carry smartphones, sure :)

Luke

> Ben
>
>
> On 21 February 2011 12:12, Ciarán Mooney <general...@googlemail.com>
> wrote:
>>
>> Hi,
>>

>> > Does registration require a fancy phone  ? (I don't have one either)
>> > Could there be a space-owned reader that could just be taken to an item
>> > to
>> > answer the 'whose is this' question ?
>>

[Ciarán Mooney]

Hi,

Oni, thanks for that link. I'd like to have something very similar and
if you could get us a non-strings-attached copy of the code then I'd
be grateful!

However they don't have any human-readable part to the QR code, which
I feel is a crucial part.

> If it hasnt worked already, why would it now? Its the scanning that is the
> issue, not the labels or system per-se. Hands up who has actually done any
> scanning?

I don't know for sure, but ultimately people have not tagged their
stuff because it's not easy (not sticky), and people have not "read"
the tags because you need special software and hardware.

> We had this issue before. Who carries with them an android or iphone at all

> times or a camera phone that can scan qr codes? And even if you do, can you
> be arsed to go grab it and then scan it? Would it not be simpler to just

> build an rfid handheld scanner (or buy one) for the space to do this?

People seem to get hung up on the QR part, that is only one section.
Most of the work is being done by the web-form, backend, and
hand-written elements. This could just as easily be done without QR
codes, but I feel that for a lot of people the QR code would make
their life easier.

I don't think RFID is any easier than using QR codes, and certainly
not as cost-effective. I think it's better to avoid specilist hardware
that needs to be purchased, because no-one will buy it! Hence why I
would like human-readable parts to the label.

> If we're after simplicity, why not just write your name, and maybe the URL
> of a Wiki page, on a brown label and tie it to your item?
> RFID & QR codes are fun, and if you're automated industrial supply-chains
> they absolutely rock, but for working out who owns the big piece of
> laminated MDF leaning against the wall, they're a bit... over-engineered.
> Sorry. I know we're all geeks and everything... but I think this is one
> problem where lo-tech is the best option.

Your are very correct. My idea would just make this easier for people
to take part. Most of the problem is social engineering we need to get
a critical mass of tags around so people (a) know what they are, and
(b) think "I should add a tag". This works very well for the members
boxes for example.

> I should point out, this system was running on linux and was totally
> automated, from recording video and tagging to printing out the QR code on a
> proper sticker. It couldn;t be simpler at all. I have the code on git.

Could you point us in the direction of the code please? Sounds like
it's half the job done already.

Ciarán

[Ciarán Mooney]

> Ciaran, you need specialist hardware for QR Codes....the reader.

True, but the QR code is only one-part of the label.

> Here is the scenario you are proposing i think:
> 1) Take a sticky label from a pile. this has a QR Code on it
> 2) attach to object
> 3) attach some link, url or media to said QR code

Almost,

1) Take sticky label from pile, which has a URL short link and blank
spaces for your name, project name, and contact details, it also has a
QR code
2) Fill out blank spaces with your name etc.
3) Attach to object. (People can now look at the label and see it
belongs to you)
4) Go to the URL that is on the sticker. Or scan the QR code, which
contains the link.
5) Webpage is a form.
6) Fill in form.
7) Form generates wiki page
8) Back-end now forwards the short-link to the appropriate wiki page.

> This wont work because authoring is a pain.
> now, what if the QR code was, as it is now, a link to the wiki. This means:
> 1) create a wiki page
> 2) print off a QR Code that has that URL in it
> 3) attach this sticky code to the object

Basically half of what I was thinking.

> We still have the READING problem but authorship is a little easier.
> to do this you need the labels you are suggesting. That means a set of avery
> labels loaded into a printer. Avery labels come in various sizes, all on A4
> sheets. That means cost and waste if you are just printing one.

Agreed, however I'd have to be specific about which size avery label
it was designed for. Printing out 30-40 labels first would be cost
effective, seeing as they don't expire we are okay to do this.

> If you are going down that route, the dymo printer we had was far superior
> for this. It was dead easy to setup and run but of course, it costs.
> Critical mass i agree with but it has to be easy and fun to get it to work.
> This is not an easy problem as it took us a year to solve and we still
> hadn't got it right.

It costs, but Jonty has one we can use. We are back to the consumable problem.

The first iteration of this idea got things wrong, and right. This
will be the same. The QR codes that Sam made are better than nothing,
but they can be improved. I'm sure this will be too.

> You know what else works? Just putting some glue on the back of
> regular old paper, or poking a hole through it and attaching with
> string, or putting in a clear plastic bag, or blu-tack, or
> thumbtacks........ :)

It does, but no-one does this. I've asked myself and the only answer I
can think of is "It's a pain in the arse.". Personally I'd rather have
a pre-made sticker I can pick out of a bag, fill out and stick on.

> How about two rubber stamps
> - one with name/project/url/contact/"can be hacked" boxes
> - one with the above + space for a QR code sticker

I like it.

Ciarán

Ciarán

[Ben Blundell]

I should point out, this system was running on linux and was totally automated, from recording video and tagging to printing out the QR code on a proper sticker. It couldn;t be simpler at all. I have the code on git.

As I say though, its always down to the nice-ness, fun factor and ease of scanning and creating something about the project in use. The system as it stands is just a little too much faff.

In addition, I'd also built a handheld RFID reader:

http://www.vimeo.com/11784354

This was a wand you'd just wave over things and you'd get playback of what the item was about.

Creation was tricky but people loved hearing about the projects. Dead easy to make! and worked really well. This could be implemented into the Glados system or similar. 

Ben

[Lester Hawksby]

Possible simple(r) solution:

How about two rubber stamps
- one with name/project/url/contact/"can be hacked" boxes
- one with the above + space for a QR code sticker

The stamps could be used straight on projects, or on stickers, or on
paper tags to tie to things.

This would support QR codes if they catch on, but also work for people
who don't want to; would be readily recognisable; could be applied in
a variety of ways depending on what it was attached to; needs only
simple stationery to get started.

Even if we do do RFID, which to be honest sounds like technological
overkill to me, we will need a fairly standard visual identifier to
tell people where to scan - so a stamps approach could lead into that
OK too.

(I think ownership of phones that can read QR codes is pretty high
among Hackspace members. Not universal, but high. However, I don't
know many people who habitually carry an RFID reader, and going to get
an Official Hackspace RFID reader has its own problems)

Lester

[Ben Blundell]

Im generally pro RFID in this case for the reason that a roll of RFID stickers at 125khz isnt that expensive as i recall (i have a few in box already for people to trial) and we use RFID quite a bit in the space. Having an actual reader for people to play with looks good for visitors and can be intergrated with other things such as the card system we use already. No need for printing.

There are some good QR Code readers for iPhone and Android I accept but people need to be clued up and have to install them. 

I have the bits from sparkfun to create a magic RFID wand that can link to either babbage or lovelace or my touchtable to do what ever we'd like really. I think its more fun and there is some fun hacking we can do with that.

Im thinking ahead a little here to when we get visitors (like the kids popping over from that school). I think thats something we can build and be a little more proud of.

Ben

[spooq]

How big are the RFID stickers? Might be able to squeeze them into the stamp?

Luke

[Ben Blundell]

This isnt the problem.

We had this issue before. Who carries with them an android or iphone at all times or a camera phone that can scan qr codes? And even if you do, can you be arsed to go grab it and then scan it?  Would it not be simpler to just build an rfid handheld scanner (or buy one) for the space to do this? 

Ben

On 21 February 2011 12:12, Ciarán Mooney <general...@googlemail.com> wrote:

Hi,

> Does registration require a fancy phone  ? (I don't have one either)
> Could there be a space-owned reader that could just be taken to an item to
> answer the 'whose is this' question ?

My idea is that the QR code just contains a link to a webpage, and
also includes a printed version of that link below the code. You could
just put that URL into your browser and fill out the same form.

If you wanted to tag something first, and fill in the paper blanks,
and then it can be logged on the wiki later. The QR code will be
unique so once you've stuck it to your item it can't be used by
someone else.

Ciarán

[Ben Blundell]

Adrian, for the reason you have just outlined.

Pre-printing means you dont have to remember everything. You shouldnt have to. Keep it simple. People dont want to have to do such things. You need to have it so that you can do association between a code and  media really easy

Again, we tried exactly this with talesofthings.com You could print out a random QR Code, attach it to your object and then scan it with the iphone. If you were the first to scan it, you could do aurthoring there and then on your iphone and it would bounce back to the website.

Its been done.  before I left UCL I'd heard it met with some success.

On 21 February 2011 13:10, Adrian Godwin <artg...@gmail.com> wrote:

I don't really understand the need for print-on-demand.

How about (and I think this is quite close to Ciaran's procedure) ..

1. Print pile of labels with QR code and short URL. Use pronouncable but meaningless URLs - this makes it easier to remember them as you walk from the object to Lovelace.
2. Attach label to object. Ideally, note Dylan's point and write your name on it. Scribble out the QR if you don't want to bother with anything else.
3. Type URL into wiki (or scan QR) in the form hack.rs/label/blizqgh (where the last component is unique code) and write as little or as much wiki code as you like
4.Profit !

On Mon, Feb 21, 2011 at 12:47 PM, Ben Blundell <o...@benjaminblundell.com> wrote:

Ciaran, you need specialist hardware for QR Codes....the reader.

I agree authorship is really hard and the killer for this sort of project. Here is the scenario you are proposing i think:

1) Take a sticky label from a pile. this has a QR Code on it

2) attach to object

3) attach some link, url or media to said QR code

readers can then scan the QR Code to get the url, link or media attached.

This wont work because authoring is a pain.

now, what if the QR code was, as it is now, a link to the wiki. This means:

1) create a wiki page

2) print off a QR Code that has that URL in it

3) attach this sticky code to the object

We still have the READING problem but authorship is a little easier.

to do this you need the labels you are suggesting. That means a set of avery labels loaded into a printer. Avery labels come in various sizes, all on A4 sheets. That means cost and waste if you are just printing one.

If you are going down that route, the dymo printer we had was far superior for this. It was dead easy to setup and run but of course, it costs.

Critical mass i agree with but it has to be easy and fun to get it to work. This is not an easy problem as it took us a year to solve and we still hadn't got it right.

Ben

--
(>) SECTION9 * Benjamin Blundell
(>) b...@section9.co.uk * www.section9.co.uk
(>) Code, Design, Graphic, Web 
(>) London Hackspace, Laboratory 24, Cremer Business Centre. London E2 8HD

(!) For Secure email use PGP Key-ID: 0A9C59AF @ http://pgp.mit.edu

[Ciarán Mooney]

Hi,

> How about (and I think this is quite close to Ciaran's procedure) ..
>
> 1. Print pile of labels with QR code and short URL. Use pronouncable but
> meaningless URLs - this makes it easier to remember them as you walk from
> the object to Lovelace.
> 2. Attach label to object. Ideally, note Dylan's point and write your name
> on it. Scribble out the QR if you don't want to bother with anything else.
> 3. Type URL into wiki (or scan QR) in the form hack.rs/label/blizqgh (where
> the last component is unique code) and write as little or as much wiki code
> as you like
> 4.Profit !

You've hit it on the nail. So I'm not talking complete nonsense.

> I don't really understand the need for print-on-demand.

When we run out, or people want to pre-fill in the blank spaces. Not
essential to my idea, but "would be nice" to have the option.

[spooq]

On 21 February 2011 12:50, Lester Hawksby <lester...@gmail.com> wrote:
> Possible simple(r) solution:
>
> How about two rubber stamps
>  - one with name/project/url/contact/"can be hacked" boxes
>  - one with the above + space for a QR code sticker

Some projects could also integrate the QR code more directly, e.g.
things cut on the laser could just etch the code. Would be interesting
to see if readers could pick that up!

Luke

[Kieran]

Would it be a pain/wasteful to do both RFID + QR, just for the coolness of
having them both?

The QR code and RFID could point at the same pre-readied URL which indeed may
also be written on one of the stickers/in the space on the stamp.

[Mark Steward]

I think the reason this conversation keeps going round in circles is
there are two separate goals:

1. making something cool and interesting to newcomers to attach
metadata to objects
2. allowing people to quickly understand what a project is when they
come across it

The first is best left as a separate project, or improvement once
we've got a basic working system (which we don't).

The second only needs the short URL, say hack.rs/123. You can then
ask that people put contact details _or_ a URL, levelling out the
effort for setting up a wiki page.

Stickers for these URLs (with QR codes or without) can be printed in
advance, and redirected later. No complicated form entry system is
required. You could even print the URL on the label printer or write
it in pen if you don't like stickers.

The most pressing issue at the moment is lowering the barrier because,
as predicted, QR codes are too much effort.

Mark

[Adrian Godwin]

I don't really understand the need for print-on-demand.

How about (and I think this is quite close to Ciaran's procedure) ..

1. Print pile of labels with QR code and short URL. Use pronouncable but meaningless URLs - this makes it easier to remember them as you walk from the object to Lovelace.
2. Attach label to object. Ideally, note Dylan's point and write your name on it. Scribble out the QR if you don't want to bother with anything else.
3. Type URL into wiki (or scan QR) in the form hack.rs/label/blizqgh (where the last component is unique code) and write as little or as much wiki code as you like
4.Profit !

On Mon, Feb 21, 2011 at 12:47 PM, Ben Blundell <o...@benjaminblundell.com> wrote:

Ciaran, you need specialist hardware for QR Codes....the reader.

I agree authorship is really hard and the killer for this sort of project. Here is the scenario you are proposing i think:

1) Take a sticky label from a pile. this has a QR Code on it

2) attach to object

3) attach some link, url or media to said QR code

readers can then scan the QR Code to get the url, link or media attached.

This wont work because authoring is a pain.

now, what if the QR code was, as it is now, a link to the wiki. This means:

1) create a wiki page

2) print off a QR Code that has that URL in it

3) attach this sticky code to the object

We still have the READING problem but authorship is a little easier.

to do this you need the labels you are suggesting. That means a set of avery labels loaded into a printer. Avery labels come in various sizes, all on A4 sheets. That means cost and waste if you are just printing one.

If you are going down that route, the dymo printer we had was far superior for this. It was dead easy to setup and run but of course, it costs.

Critical mass i agree with but it has to be easy and fun to get it to work. This is not an easy problem as it took us a year to solve and we still hadn't got it right.

Ben

On 21 February 2011 12:31, Ciarán Mooney <general...@googlemail.com> wrote:

[Adrian Godwin]

Well, maybe. No doubt it depends on the person - I find I can remember something for a short while if it's pronouncable and I can repeat it in my head while I walk to the point of use. If it's a random bit.ly style URL it will be hopeless, I agree, so the randomly-generated string needs to be filtered through soundex or something to remove ones that are hard to remember.

Also, there's no reason why the system shouldn't work in both directions, is there ? Have a pile of preprinted, easy to use labels. Add a label printer when someone donates / steals / pledges one. The snag with the current system (unless I misunderstand it) is that the only preprinted labels available are already allocated to projects so you can't just use one, and you need a reader to find out what they're pointing to anyway.

-adrian

On Mon, Feb 21, 2011 at 1:15 PM, Ben Blundell <o...@benjaminblundell.com> wrote:

Typing short URLS and having to remember them is the quickest way for this to fail, guarenteed!

Ben

[Ben Blundell]

One thing to be said about QR codes is they are gaining recognition and people know that the thing in front of them is ready to scan.

I reckon that really, if you go down the QR code route, we need a dedicated small dymo (which apparently we have) with a dead simple webform (again, we had a one button video record at anim 10 and tales has a simple webform) and then just hit go and you get a sticker. What is embedded in the sticker is a URL (either a bitly or some other random one BUT DEFINITELY one that works over the web) and thats it. Job done.

the advantage here is multiple stickers can be added if people want to update or leave comments on things. 

It'd be nice if this could be integrated into the touch table.... but thats thinking further ahead.

It might be a good idea to suggest people install some QR Code readers... maybe have that on the sticker or a sign when people come in.

Printing to author > preprinted. The reason? pre-printed requires that extra authorship step that, while occasionally useful (say you are out and about and want to tag something there and then and you dont have a printer) isnt in our case because its all inside the hackspace.

Ben

[Ben Blundell]

Done repeat yourself. Trust me, it aint worth it :P

Ben

[spooq]

You know what else works? Just putting some glue on the back of
regular old paper, or poking a hole through it and attaching with
string, or putting in a clear plastic bag, or blu-tack, or
thumbtacks........ :)

Luke

[Adrian Godwin]

Oh, another small variant :

Extract QR code from random pile.
(or print if we have progressed far enough to do that)
Scan it using fixed reader connected to Lovelace (or your phone)
Enter details in wiki page
Stick / tie label to object.

Which simplifies the mobile reader problem into a fixed reader and avoids the remembering problem.

-adrian

On Mon, Feb 21, 2011 at 1:15 PM, Ben Blundell <o...@benjaminblundell.com> wrote:

Typing short URLS and having to remember them is the quickest way for this to fail, guarenteed!

Ben

On 21 February 2011 13:13, Ciarán Mooney <general...@googlemail.com> wrote:

[Ben Blundell]

If it hasnt worked already, why would it now? Its the scanning that is the issue, not the labels or system per-se. Hands up who has actually done any scanning?

[Ciarán Mooney]

Hi,

> I think the reason this conversation keeps going round in circles is
> there are two separate goals:
>
> 1. making something cool and interesting to newcomers to attach
> metadata to objects
> 2. allowing people to quickly understand what a project is when they
> come across it

Thanks for pointing this out.

> The first is best left as a separate project, or improvement once
> we've got a basic working system (which we don't).

Agreed. This is not what I intended.

> The second only needs the short URL, say hack.rs/123.  You can then
> ask that people put contact details _or_ a URL, levelling out the
> effort for setting up a wiki page.

> Stickers for these URLs (with QR codes or without) can be printed in
> advance, and redirected later. No complicated form entry system is
> required.  You could even print the URL on the label printer or write
> it in pen if you don't like stickers.

Agreed. My thinking was that a web-form would make things slightly
easier, which would hopefully increase adoption.

> The most pressing issue at the moment is lowering the barrier because,
> as predicted, QR codes are too much effort.

Indeed.

> Printing to author > preprinted. The reason? pre-printed requires that extra
> authorship step that, while occasionally useful (say you are out and about
> and want to tag something there and then and you dont have a printer) isnt
> in our case because its all inside the hackspace.

I'm misunderstanding your use of the work "authorship". I can't get my
head around the point you are trying to make. Could you please explain
what you mean again?

Are you able to provide us with a copy of the code that you wrote for
the "world of things" website?

Ciarán

[Ben Blundell]

Ciaran, you need specialist hardware for QR Codes....the reader.

I agree authorship is really hard and the killer for this sort of project. Here is the scenario you are proposing i think:

1) Take a sticky label from a pile. this has a QR Code on it

2) attach to object

3) attach some link, url or media to said QR code

readers can then scan the QR Code to get the url, link or media attached.

This wont work because authoring is a pain.

now, what if the QR code was, as it is now, a link to the wiki. This means:

1) create a wiki page

2) print off a QR Code that has that URL in it

3) attach this sticky code to the object

We still have the READING problem but authorship is a little easier.

to do this you need the labels you are suggesting. That means a set of avery labels loaded into a printer. Avery labels come in various sizes, all on A4 sheets. That means cost and waste if you are just printing one.

If you are going down that route, the dymo printer we had was far superior for this. It was dead easy to setup and run but of course, it costs.

Critical mass i agree with but it has to be easy and fun to get it to work. This is not an easy problem as it took us a year to solve and we still hadn't got it right.

Ben

On 21 February 2011 12:31, Ciarán Mooney <general...@googlemail.com> wrote:

[Ben Blundell]

ARGH! ><

Ok ok, point accepted from Mark. Two seperate projects is perhaps the key but unless i've missed something, a QR code needs to be scanned if its to be used. That means a scanner which means a smart phone. End of story!

You then have the authorship issue. Sure ok, pre-printed but with what URL? If you take an off the shelf QR code reader, it will translate the QR Code to a URL and say "do you want to go visit this?" clearly that URL has to be global and work which means something should eventually be at the end of it. 

that means someone needs to:

1) take preprinted

2) attach to their object.

3) either scan to convert the QR Code or read/remember/write down the url at the bottom

4) take the new website they have found and author something for it.

What i've said is we've tried this, with many users over and over and over again. The issue always boiled down to associating an arbitrary code. If you d the authorship first, there is no need for two of these steps i've just mentioned. You fill in your project, hit go and grab your new sticker. Want another one? just hit go again. 

This will work. Why will it work? Because I've already done it.

On 21 February 2011 13:35, Adrian Godwin <artg...@gmail.com> wrote:

On Mon, Feb 21, 2011 at 1:05 PM, spooq <spo...@gmail.com> wrote:

And as has been said, it is likely a pretty decent proportion of

Hackspacers have the hardware in their pocket already.

Yes, but quite a long way from an overwhelming proportion.

For me, a phone is very far from essential. I sometimes carry a cheap one, but more often leave it at home. I leave it in my coat, my bag etc. while I'm somewhere like the hackspace. They need to get a lot more functional or a lot more convenient to carry before I'll consider them must-have enough to modify this behaviour. And I am competely unable to sympathise with any different point of view, a shortcoming that I know is often echoed by people who wouldn't be without one.

-adrian

[Adrian Godwin]

[Ciarán Mooney]

Hi,

> What i've said is we've tried this, with many users over and over and over
> again. The issue always boiled down to associating an arbitrary code. If you
> d the authorship first, there is no need for two of these steps i've just
> mentioned. You fill in your project, hit go and grab your new sticker. Want
> another one? just hit go again.
> This will work. Why will it work? Because I've already done it.

Ah ha! I think I now get the point your making.

Hmm. You obviously have valuable experience, and I'd rather not waste
time if you know what works and what doesn't.

Just to see if I have understood properly. What your saying is that
people would write the information out about an item/project first,
then would like a label that to attach to an object. The unique
identifier on the label (along with other human readable info) is then
printed and affixed to the object. This worked.

If people had an object, affixed a label that had been pre-made and
then used some kind of identifier of the label (a URL in our case) to
link the object/label to an online document. Then this didn't work.

Have I got this right? It seems that it's a case of which direction
you approach the problem.

Could we adapt my idea to work in both directions?

Ciarán

[Ciarán Mooney]

Hi,

> In adddition, this was all written in Python (using cups with the dymo under
> linux - they have good linux drivers btw) and a simple algorithm to generate
> working short urls (we used our own in the end rather than bit.ly).
> If it comes down to it, yes, you could do away with the QR codes altogether
> and simply have, as you have suggested, a shorter link thats just 3 chars or
> so and have that. Does away with the QR codes and the fancy scanning and
> what not.  If i have understood rightly, that would lower the bar even
> further

Ooo, that sounds very useful.

We certainly could do away with the QR codes, but some part of me has
a soft-spot for them. Although if we lower this bar any futher we
might as well give people a handful of Sharpies and tell them to go
nuts. And we know where that will lead, have you seen the toilets
recently? :)

I'll try and summarise all the points made here on the wiki. This is
obviously a problem that is not as easy as I thought.

Ciarán

[Ben Blundell]

In adddition, this was all written in Python (using cups with the dymo under linux - they have good linux drivers btw) and a simple algorithm to generate working short urls (we used our own in the end rather than bit.ly). 

If it comes down to it, yes, you could do away with the QR codes altogether and simply have, as you have suggested, a shorter link thats just 3 chars or so and have that. Does away with the QR codes and the fancy scanning and what not.  If i have understood rightly, that would lower the bar even further

Ben

[Dylan Beattie]

> We had this issue before. Who carries with them an android or iphone at all times or a

> camera phone that can scan qr codes? And even if you do, can you be arsed to go 

> grab it and then scan it?  Would it not be simpler to just build an rfid handheld scanner

> (or buy one) for the space to do this? 

If we're after simplicity, why not just write your name, and maybe the URL of a Wiki page, on a brown label and tie it to your item?

RFID & QR codes are fun, and if you're automated industrial supply-chains they absolutely rock, but for working out who owns the big piece of laminated MDF leaning against the wall, they're a bit... over-engineered.

Sorry. I know we're all geeks and everything... but I think this is one problem where lo-tech is the best option.

-D-

[Ben Blundell]

Yes, that is pretty much it. QR codes are recognisable. This means people may or may not have an app that recognises them already installed. such apps are "looking" for URLS already. They are primed to automatically follow them.

So what we did was to say:

1) fill in the name of your object (or record a video in the case of the one we did for kids)

2) hit go

3) printer prints out a QR code with a UNIQUE Url or the form (www.talesofthings.com/<11 digit code>)

4) user sticks code on object

The end of the site basically strips the 11 digits off and returns the website in question with the data. The heavy lifting is done with a django ninx backend. Essentially thats the job done.

the barrier was so low, even 6 year olds were trying it out when we did this. one button press and you got your code. The scanners were set so that it would play whatever you've recorded by stripping off the 11 digit bit, doing a lookup and there you go.

Because authorship is done at the same time as QR Code generation, it cuts out the whole remembering step and makes it really really simple and ANY old QR code reader will work because it is an internet accessible URL.

In the case of the hackspace wiki, things could be different. People might say... hang on! I've already filled in all my project. I dont want to have to do it all again on a different page. This is a problem we had at talesofthings.com . We got around it by having the landing page store any other web media you wanted. This was OK for us but in the case of the hackspace, its not ideal

In such a case, what could be done is at AUTHORING time, allow the user to override the default QR Code with one they provide. That is a simple one field optional addition to the form. You still dont have to remember anything. 

Now, of course, if you just forget QR Codes most of this goes away because a simple URL made obvious like: hack.rs/125 or something is memorable and means no scanning. Sure it doesnt sound as cool and what not but it means one can actually write on the sticker and then author later, largely because they can remember sometihng so short. Sure there is potential for collisions here but maybe that is a better solution?

[Adrian Godwin]

On Mon, Feb 21, 2011 at 1:42 PM, Ben Blundell <o...@benjaminblundell.com> wrote:

ARGH! ><

Ok ok, point accepted from Mark. Two seperate projects is perhaps the key but unless i've missed something, a QR code needs to be scanned if its to be used. That means a scanner which means a smart phone. End of story!

Does it HAVE to mean a smart phone ? I appreciate your point earlier that it's not easy to build a dedicated mobile scanner, and presumably a bought scanner would be an industrial item that would cost more than a phone. Though might be less in danger of theft. But can't they be scanned by a fixed scanner, perhaps especially the touch table ?

The point is that by rearranging the workflow you get

1) take preprinted

3) either scan to convert the QR Code or read/remember/write down the url at the bottom

or print a new label
 

4) take the new website they have found and author something for it.

2) attach to their object.

 

By allowing for 3/4 to be done at the same location (whether a phone, the touchtable or lovelace) you avoid several other difficulties without adding new steps. Scanning a preprinted code becomes an exchangable step with printing a new one.

This is pretty irrelevant if wer actually have a suitable label printer already - it only makes a difference if that's not immediately available in a form that allows for the economical printing of a label on demand.

-adrian

[Ben Blundell]

Typing short URLS and having to remember them is the quickest way for this to fail, guarenteed!

Ben

On 21 February 2011 13:13, Ciarán Mooney <general...@googlemail.com> wrote:

[Ciarán Mooney]

Hi,

Ben, you are slowly bringing me around to your way of thinking.

An established system that is easy to use, ie "Print a tag that
corresponds to this wiki/project page", which then prints out a label
on demand is seeming more and more suitable. Hopefully Jonty has
spotted this thread, and I'll see about getting the label printer off
him.

As a first "blitz" to get momentum going we may need to print off a
large quantity of generic labels that we can affix to items, eg
dishwasher, hackspace monitors, projectors etc. In this instance we
may need a bag of labels, a laptop and someone (me) willing to go
around labelling and linking up the tags. This is essentially my
original idea.

After this we need to make it easy for people to print their own.

Ben, any chance of seeing this mystical code?

Ciarán

[Ben Blundell]

If the scanner isnt portable there is no need for a QR code because it cant be scanned. If the scanner is portable one can do authorship and printing without the need for a scan step.

If the goal is simply to point out that, yes this is something somebody is working on and yes, here is some info about it, then a sharpie and a box of stickers is best.

We have tried avery labels before in a standard printer. The problem is that:

a) alignment is an issue when printing from say, a web page. We tried loads of python to pdf based solutions and none of them were perfect. When you just want one (or even say, 5 of the same one) it becomes more hassle to align and setup and what not

b) Im willing to bet that avery labels and stickers are more per sticker than dymos

On 21 February 2011 14:37, Adrian Godwin <artg...@gmail.com> wrote:

On Mon, Feb 21, 2011 at 2:13 PM, Ben Blundell <o...@benjaminblundell.com> wrote:

Adrian, the last point you make is bang on with what we experienced.

OK, thanks for making that point. I hadn't appreciated it was so important.

In answer to the smartphone, yes, yes it really does have to be handheld. What else has an 

a) autofocusing camera (trust me, you dont DONT want a crap webcam)

b) has netacccess

c) is portable enough to get to the really awkward place someone has put the sticker?

For lookup, I absolutely agree with portability (manually entering some sort of easy URL if you don't have a suitable phone available is just a fallback, as Ciaran originally proposed. Or, indeed, reading the sharpie writing.)

The problem I'm trying to solve is that of creating the label in the first place, perhaps as a stopgap until we have a handy printer. Or if we run out of printer media. If it's not printed on demand, it needs to be scanned. But not necessarily after it's attached to the item - the printer doesn't need to be mobile because you print before you stick and so neither does the scanner when it's used purely as a method to associate a preprinted sticker with the wiki page that you're editing. It's really a vanishingly small difference : instead of writing the page, printing the label and attaching it, you write the page, scan the label and attach it. Except that you don't need a dedicated printer.

-adrian

[spooq]

On 21 February 2011 13:00, Ciarán Mooney <general...@googlemail.com> wrote:
>> Ciaran, you need specialist hardware for QR Codes....the reader.
>

> True, but the QR code is only one-part of the label.

And as has been said, it is likely a pretty decent proportion of
Hackspacers have the hardware in their pocket already.

> It does, but no-one does this. I've asked myself and the only answer I
> can think of is "It's a pain in the arse.". Personally I'd rather have
> a pre-made sticker I can pick out of a bag, fill out and stick on.

If that is too hard, I'm not sure how low the barrier needs to be for
people to author :P

>> How about two rubber stamps
>>  - one with name/project/url/contact/"can be hacked" boxes
>>  - one with the above + space for a QR code sticker
>

> I like it.

So do I, and I've got the rubber to make it happen, so if someone
wants to make a design, they can send it to me. If no-one does by
tomorrow, I'll make the design myself.

Luke

[Adrian Godwin]

[Ben Blundell]

the company itizen actually send out a batch of QR Codes when you sign up. They do the same thing we do essentially. They are smaller and slightly cooler but then again, they got a professional printer to do these. you need to scan them and set it up. as i recall, we still have more users than they do :P

Having spoken to users in the field about the various methods it was deemed that having to author and print and download a special app was just too much damn hassle. The more steps removed the better. This isnt a hacking issue; its a user interface issue and although it may seem not important to remove little things, it is these little things that really do make or break a system like this one.

Ben

[Ciarán Mooney]

[Ben Blundell]

Ok so talesofthings.com is closed source. This kinda sucks because it has everything in it, including all the mistakes we made along the way.

Animation 10 source code is all c++ and written in openframeworks. It is designed to work with video and uses a client server setup. It is probably not suitable for this because the emphasis was on:

1) recording a 30 second clip

2) compressing it asap!

3) generating and printing a QR code

4) monitoring an arduino button and switch for video effects, start / stop etc

we used the qrcodelib written in openCV to try and use a webcam to read QR codes. Again, the problem here was they were fixed focus. Had we not attached the camera to a stand the whole thing would have fallen apart.

In this case, you need something more web based. we used:

1) django for tales, cherrypy for animation 10

2) Google's graph API (which generates QR Codes) for tales and something else C++ based for animation 10

3) dymo's linux drivers linked with cups. To do the actually printing we made a python os call to the system to use cups' command line printing with a PNG of the actual QR Code.

Now i totally agree we'd need to put some out there first to get critical mass going. You could just have a load of dummy pages that maybe have aliases eventually to the project they are going to be talking about. We do already have a few and maybe thats enough but hey. 

the code is on my server which is packed away. It *was* on git but I'll try and dig it up.  

As i see it, what you would have is:

a form that that when filled in, generates a page on the wiki automagically and makes a call to google to generate the QR code using that new URL. At the same time, it makes a call to the printer to print said QR code.

for safety and future revisions, you could have a program that generates a random QR code with a url like hack.rs/2213asds for example. when hack.rs receives this code, there is a cgi/python/ruby whatever program that says actually, thisURL goes to this page on the wiki by doing a db lookup - this having been filled in by the user when the did the authoring. This is what talesofthings.com does in the main.

[Ben Blundell]

Adrian, the last point you make is bang on with what we experienced.

In the case of the kids, we had the label printer right there and dedicated. So yes, it worked great and that was what i was proposing to get the QR code bar lowered.

With talesofthings, there was a demand for QR Codes to be preoprinted. We'd tried using portable photo printers and even stealing existing QR codes (by hashing their contents!) but in the end, users wanted a pocket of QR codes they could attach at a whim to things they'd seen on the street or at home or whatever.

We allowed a random code with no associated page to be generated. When this was scanned for the first time using our app, the server checked the code to see if it was associated. if it wasnt you were given the option to add the details there and then on your phone. It works but it is clunky because there is the step where the QR itself has to go from the physical to the digital either by scanning it or typing it in. With the authorship first you DONT need to do that. Ultimately, that is the difference and it turned out to be important.

In the case of the hackspace this method is a little silly because you CAN setup something dedicated, therefore you should because its a simpler solution.

In answer to the smartphone, yes, yes it really does have to be handheld. What else has an 

a) autofocusing camera (trust me, you dont DONT want a crap webcam)

b) has netacccess

c) is portable enough to get to the really awkward place someone has put the sticker?

Sharpies! More sharpies! :D

[Sam Cook]

Quick question (as this thread is pretty fast moving): what are you trying to achieve? 

as mark says it seems there are two things here:

1. cool way of letting newcomers interact with the space and find out whats what 

2. info for members: can it be hacked, if not who's is it, what is it etc

with the first auto generated uber QR codes that massage you gently as the web page loads would be awesome. For the latter I want to be able to read it on the sticker, I don't want to have to get out a phone/log on to lovelace to find out that "This bit of interestingly shaped MDF is bilbo baggins but is an abandoned project please use"

The QR stuff is awesome and would be an excellent addition for large projects but for a lot of stuff all I want to know is "can it be used, who's is it, what is it"

S

[Ben Blundell]

for number 2 a stack of stickers with space to write is probably best then :P

We could add holograms! :D

Ben

[Ciarán Mooney]

Hi,

> 2. info for members: can it be hacked, if not who's is it, what is it etc
> with the first auto generated uber QR codes that massage you gently as the
> web page loads would be awesome. For the latter I want to be able to read it
> on the sticker, I don't want to have to get out a phone/log on to lovelace
> to find out that "This bit of interestingly shaped MDF is bilbo baggins but
> is an abandoned project please use"
> The QR stuff is awesome and would be an excellent addition for large
> projects but for a lot of stuff all I want to know is "can it be used, who's
> is it, what is it"

Personally I am trying to achieve number 2.

All the things you describe I included in my original idea, human
readable essential information.

Unfortunately a label only has limited space, hence I feel the need to
connect it up to the wiki where more detailed instructions can be
given. My example would be the washing machine, which although simple
does quite detailed instructions.

Unfortunately a label is never going to be able to tell you if an item
is abandoned or not. At least with a label with contact details you
can email the person saying "What's up with this? Do you plan on doing
anything with it, or can we chuck it or canabalise for parts?". At the
moment this process stops if no-one has any idea who it belongs too,
and people are very polite about not destroying stuff they don't
understand.

The QR code was a method of implementation, I realise that is the
title of this thread, but the idea is not dependant on them!

Ciarán

[Sam Cook]

We could add holograms! :D

so much for simplicity... 

stickers or stamps seems to be a good option for the simplicity, something I would use a darkroom for is making holograms... really need to look into getting the chemicals (we already have the lasers, although with well balanced RGB I think you can get full colour ones....)

S

[Dave Williams]

> Personally I am trying to achieve number 2.

/thread

[Jonty Wareing]

On Mon, Feb 21, 2011 at 02:20:53PM +0000, Ciar�n Mooney wrote:
> Hopefully Jonty has spotted this thread, and I'll see about getting
> the label printer off him.

I'll try and remember to drop it off in the space tonight.

--jonty

[Sam Cook]

I think keeping the stickers simple is probably the way to go: a spot for a URL if needed should be sufficient. 

For a lot of the HS owned stuff a big sticker saying something similar to "read page X before use" or "don't use without training" is going to be required as people may miss smaller stickers/labels. This is one of the problems w/ the makerbot is that people are willing to fiddle with it but often don't realise that it is probably the most delicate piece of equipment in the whole place....

[Ciarán Mooney]

Hi,

> I'll try and remember to drop it off in the space tonight.

Thanks! Step one completed.

Ciarán

[Ciarán Mooney]

Hi,

> Does registration require a fancy phone  ? (I don't have one either)
> Could there be a space-owned reader that could just be taken to an item to
> answer the 'whose is this' question ?

My idea is that the QR code just contains a link to a webpage, and
also includes a printed version of that link below the code. You could
just put that URL into your browser and fill out the same form.

If you wanted to tag something first, and fill in the paper blanks,
and then it can be logged on the wiki later. The QR code will be
unique so once you've stuck it to your item it can't be used by
someone else.

Ciarán