Canonical Pursuing A Hardware/Software Survey For Ubuntu Installations

[microcode]

This could be affected by sampling bias: people who don't use Ubuntu because it doesn't support their exotic hardware very well will not appear in the survey.

[KRiloshart]

They ask you to send the data on USB Stick by snail mail to them.


OptIn would be okay, OptOut... meh.

How can they ask me to send data without internet connection?

[DanL]

This could be affected by sampling bias: people who don't use Ubuntu because it doesn't support their exotic hardware very well will not appear in the survey.

Ubuntu is surveying the people who use it. Why would they care about people not running it? It's not intended to be a general Linux user survey.

[DanL]

Otherwise, I hope it crashes and burns. Linux is already messy enough as it is; we don't need any more single-distribution-only driver patches that users of other distributions need to reverse engineer.

I don't understand how you got this idea out of a user survey. But then, I don't follow the "logic" in about half of the comments on Phoronix.

[Jumbotron]

Sounds kinda like what Mozilla does with Firefox and Google does with Chrome. Generally for it....as long as improvements that come from such a survey that are generic to Gnome or the Linux kernel and not JUST Ubuntu specific will be shared with the community at large.

[dstaubsauger]

The survey will offer you a few choices:

1) you plug a USB pendrive to save your answers, then it will install a minimal Ubuntu on your pendrive with a script that uploads the results as soon as it finds network connectivity. Booting from the pendrive on a connected machine is all you have to do.

2) the survey provides you with a unique 8192 digits contribution code that you can use from a different, network connected machine, by simply typing it as reply to the question about network connectivity. That way they will know your replies are about a different machine than the one you're using to answer.

3) you print out the survey results and fax them in: +44 20 7630 2401

4) same as 3, but you send the results by paper mail to:
Canonical Group Limited
5th Floor, Blue Fin Building
110 Southwark Street
SE1 0SU
London, United Kingdom

5) you capture the survey screens with your smartphone camera and send them by email using your smartphone connectivity

6) you don't fill the survey. They will know how many do not have network connectivity by subtracting the number of received surveys from the number of Ubuntu DVDs they sell.

That was comprehensive! I bet they have already written the continuous integration tests for all these methods!

[lucrus]

That was comprehensive! I bet they have already written the continuous integration tests for all these methods!

I don't think so, because there actually are a few other methods: I listed only the ones users have been begging for during the last 10 years of Ubuntu. But if you take a look at the remaining less common methods, it's clear they're not compatible with the continous integration tests. What's not so clear is whether users didn't beg for those because they're not compatible, or the other way around...

7) you take the survey without a computer, by calling the Canonical IVR service (not compatible because not all countries use DTMF tones)

8) you connect to Canonical even without a network interface, simply using their new software loopfan device (module "loopfan" in kernels 4.18+) that emulates a Powerline Ethernet adapter and thus mananges to send bytes over the public power lines by driving your PSU fan (not compatible because of voltage differences between countries and because it seems to fry your motherboard, but that could also be caused by some obscure APIC bug, so blame Intel, they melt down everything after all and try with "noapic" next time)

9) this method is unknown: the user who reccommended it was lacking network connectivity. We need to assume it's not compatible.

10) you buy limited edition custom Ubuntu DVDs, where the survey is pre-filled for you and printed on the cover (theoretically compatible, but it raised privacy concerns)

[Sonadow]

I don't understand how you got this idea out of a user survey. But then, I don't follow the "logic" in about half of the comments on Phoronix.

Based on this portion in the actual story written by Michael himself:

Of course, privacy-minded individuals may be against this though there are also valid uses for this particularly for being able to gauge their investments into improving certain GPU drivers and also lobbying of hardware vendors for better Linux support based upon usage statistics.

So who's the one without logic here?

[DanL]

So who's the one without logic here?

How would lobbying hardware vendors lead to Ubuntu-specific patches? And if so, why would you think that the patches would be closed source and that other distros would have to "reverse engineer" them?

[Sonadow]

How would lobbying hardware vendors lead to Ubuntu-specific patches? And if so, why would you think that the patches would be closed source and that other distros would have to "reverse engineer" them?

By having vendors ship binary *****.ko kernel modules that are compiled only for the kernel versions Ubuntu ships with?