[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Problems with gnunet-qr
From: |
Christian Grothoff |
Subject: |
Re: Problems with gnunet-qr |
Date: |
Sun, 15 Dec 2019 23:32:55 +0100 |
User-agent: |
Evolution 3.30.5-1.1 |
On Sun, 2019-12-15 at 22:19 +0100, Alessio Vanni wrote:
> Hello,
>
> Due to the fact that I don't have a device that can work with gnunet-
> qr,
> I decided to hack it in an attempt to add support for PNG files. The
> attempt was mostly successful, but I met some issues that made the
> program generate errors.
>
> The first error isn't particularily serious, but I thought I'd report
> it
> anyway: short options don't work, i.e. using "-d" rather than "
> --device"
> says that "-d" is not a recognized option. The issue seems to be
> caused
> by the '\0' in the "--verbose" option definition; changing it to 'V'
> fixed the issue. Maybe the getopt library should guard against '\0'?
OMG. That shouldn't even look like that. We have a special function for
verbose that should be used. Fixed in Git Master.
> The next problem happens after the QR code is processed, when the
> program tries to start the program associated with the URI. In the
> `gnunet_uri' function, the `sigpipe' global variable is used, but
> gnunet-qr never sets a non-NULL variable to it, meaning that the
> function making use of the variable ends up with a segmentation
> fault.
> I can't really figure out how this would be fixed, because I'm not
> practical with file handling using libgnunetutil and I don't know
> what
> the program tries to accomplish here.
>
I've initialized the sigpipe now.
> Lastly, still in `gnunet_uri', adding a GNS URI fails without giving
> a
> reason why. I can't tell if it's because the interface with
> gnunet-namestore changed or something else, all it says is that
> adding
> the URI failed.
Eh, good question. The URL format did not really change, but we did
change the crypto _last week_. So if you created the QR code two weeks
ago, that would not work with this week's code. That said, I didn't try
this recently at all. Schanzen: we probably should test this, clearly
gnunet-qr was not working as intended before...
> The QR code with the GNS URI was created using gnunet-bcd, then the
> QR
> code was "exported" into a PNG file (I took a screenshot of the QR
> code
> with scrot) which was then used by gnunet-qr.
>
> Since I'm already working on it I can probably try to fix these
> issues,
> but I'd like to request some guidance especially for the last two
> (the
> first isn't really an issue...).
>
> Thanks,
> A.V.
>