Notices

Is mBarCode supposed to be able to handle adding contacts from QR codes?

Tested it on this: http://zxing.appspot.com/generator/
Created a Contact Information entry. mBarcode reads it fine but won't give an option to add the contact.

Is this me missing a plugin or is there simply no plugin to handle this?

i imaigne the resolution doubler would help with small or distant codes

edit: btw, uninstalling and resintalling everything did fix it, it even remembered the custom websearch i had added

Try this:
http://maemo.org/packages/view/mbarcode-plugin-qrcode/

Or uninstall and then re-install mBarcode, which will then pull in all the plugins. Sorry, we've been trying to work out how to get the dependencies right and those who upgrade and have at least one plugin already available don't get them all pulled in.

Hmmn. I had done so already. Ran into the first bug where you don't get the plugins properly pulled down due to some plugin already being installed and fulfilling the requirements. Only had the copy-button. So I made sure to uninstall them all (and purged mBarcode). That let the plugins be pulled down automagically when reinstalling mBarcode again from the DEV-repo.

So yes, I already have the qrcode plugin, but I can not get it to add a contact from the code generated on the above linked site. The string read is displayed as: "MECARD:Testname;TEL;666666;;" and mBarcode give me the following options:
*Rating from Informed Individual
*Give opinion with Informed Individual
*Open iMode data (Does nothing)
*Search the Web

Taping Open iMode data does nothing. I get the click from the n900 confirming my tap, but nothing happens. The other options seem to work as intended (not familiar with Informed Individual so can't say for sure, but at least these options do something).

mbarcode-plugin-qrcode is listed as version 0.0.4-2.
mbarcode is listed as version 0.2.2-4
(I'll omit the rest of them)

So, anyone else able to add a contact from that site I linked using mbarcode? If so, we need to start comparing our devices =)

Quick reply.
Contact Dragly, the developer of the plugin, and ask him to add the functionality.

Actually that's my fault - I never got round to implementing MECARD import code.

In fact I had also been meaning to move the vCard import back to a code-only process and not need to use an intermediate .vcf file, but it looks like I've not got round to doing that yet either.

So basically, if anyone can find either some code to decode MECARDs to fields or code to convert from a MECARD to a vCard (as either way, either writing a vCard or using the code-only method, both require a string containing a vCard) that would be useful.

And basically my apologies - I'm pretty busy at day-job work, so don't have an awful lot of spare time for coding (or am just knackered). Today is a public holiday, and while I'm heading to work as there's stuff to be done, I shouldn't be there all day so will try to do some of the things I've promised this afternoon/evening.

Right, I did some asking in #gstreamer about ways of switching video resolution on-the-fly but the news is not all that good.

First I'll see if I can change the filter caps to get the v4l2 device to change res on the fly, if this works then all's well and good and we can temporarily switch to higher res to send to the analysis code and use a videoscale element to downscale for display. But I have my doubts whether it will allow us to switch on the fly.

The videoscale element eats CPU, so we don't want to have to use it all the time. Likewise, sending full frame data to the decoders is slow(er), and iirc Dragly mentioned that larger res slowed down the UI updates, so we're a bit stuck there.

Anyway, I'll do some testing to see where the CPU eating/UI & decoder slowness tradeoff lies.

One option is to do a quick and dirty scaling by just taking every other pixel and sending this to the UI (and the decoders when low res data is sufficient), I'll have to do some testing.

Using the fcam stuff it isn't possible to change which pixels of the sensor are being used? Like get a low res full frame, search for where in there a barcode (1d or otherwise) seems to be, and then zoom in to the best resolution for that area?

I'll have a look at the fcam stuff and see how it works, but I guess it just sub-samples the full frame to achieve that effect, and all that subsampling will (probably) slow things down (especially for the video that we use).

I'll do some testing and report back.