Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.debian.project > #9270 > unrolled thread

GitHub Open Source Survey 2017

Back to article view | Back to linux.debian.project

Contents

  GitHub Open Source Survey 2017 Chris Lamb <lamby@debian.org> - 2017-06-06 19:30 +0200
    Re: GitHub Open Source Survey 2017 "Andrew M.A. Cater" <amacater@galactic.demon.co.uk> - 2017-06-06 22:00 +0200
      Re: GitHub Open Source Survey 2017 Adam Borowski <kilobyte@angband.pl> - 2017-06-06 22:20 +0200
        Re: GitHub Open Source Survey 2017 Christian Seiler <christian@iwakd.de> - 2017-06-07 14:00 +0200
          Re: GitHub Open Source Survey 2017 Andrey Rahmatullin <wrar@debian.org> - 2017-06-07 14:20 +0200
          Re: GitHub Open Source Survey 2017 Henrique de Moraes Holschuh <hmh@debian.org> - 2017-06-07 16:10 +0200
            Re: GitHub Open Source Survey 2017 Adam Borowski <kilobyte@angband.pl> - 2017-06-07 17:00 +0200
              Re: GitHub Open Source Survey 2017 Andrey Rahmatullin <wrar@debian.org> - 2017-06-07 17:10 +0200
    Re: GitHub Open Source Survey 2017 Martin Bagge / brother <brother@bsnet.se> - 2017-06-14 12:50 +0200
      Re: GitHub Open Source Survey 2017 Chris Lamb <lamby@debian.org> - 2017-06-14 13:10 +0200

#9270 — GitHub Open Source Survey 2017

Hi -project,

GitHub recently published the announcements of their 2017 Open Source
Survey:

  http://opensourcesurvey.org/2017/


I was wondering whether any fellow developers had read this and whether
they had evaulated it in the context of the Debian project. To kick this
off, let me quote some of their key insights:

 * Documentation is highly valued, frequently overlooked, and a
   means for establishing inclusive and accessible communities.

 * Negative interactions are infrequent but highly visible, with
   consequences for project activity.

 * Open source is used by the whole world, but its contributors
   don't yet reflect its broad audience.

 * Using and contributing to open source often happens on the job.

 * Open source is the default when choosing software.


Best wishes,

-- 
      ,''`.
     : :'  :     Chris Lamb
     `. `'`      lamby@debian.org / chris-lamb.co.uk
       `-

[toc] | [next] | [standalone]

#9271

On Tue, Jun 06, 2017 at 06:23:30PM +0100, Chris Lamb wrote:
> Hi -project,
> 
> GitHub recently published the announcements of their 2017 Open Source
> Survey:
> 
>   http://opensourcesurvey.org/2017/
> 
> 
> I was wondering whether any fellow developers had read this and whether
> they had evaulated it in the context of the Debian project. To kick this
> off, let me quote some of their key insights:
> 
>  * Documentation is highly valued, frequently overlooked, and a
>    means for establishing inclusive and accessible communities.
> 

Debian gets this. It's relatively well documented. 


>  * Negative interactions are infrequent but highly visible, with
>    consequences for project activity.

Much more visible here because we don't tend to hide problems,
potentially. It has had HUGE consequences for a few things but,
in general Debian works well
> 
>  * Open source is used by the whole world, but its contributors
>    don't yet reflect its broad audience.
> 

We do better than many - and we know where we're under-representative.

>  * Using and contributing to open source often happens on the job.

Or in universities / bedrooms ... we understand FLOSS better than most.
> 
>  * Open source is the default when choosing software.

Tell me again, what part of closed source software is part of Debian
main? :)

All best

Andy C

> 
> 
> Best wishes,
> 
> -- 
>       ,''`.
>      : :'  :     Chris Lamb
>      `. `'`      lamby@debian.org / chris-lamb.co.uk
>        `-

[toc] | [prev] | [next] | [standalone]

#9272

On Tue, Jun 06, 2017 at 07:52:57PM +0000, Andrew M.A. Cater wrote:
> On Tue, Jun 06, 2017 at 06:23:30PM +0100, Chris Lamb wrote:
> >  * Documentation is highly valued, frequently overlooked, and a
> >    means for establishing inclusive and accessible communities.
> 
> Debian gets this. It's relatively well documented. 

That is, that we do lack documentation?

It's indeed one of bigger problems -- especially that it's not fun to fix.

> >  * Open source is the default when choosing software.
> 
> Tell me again, what part of closed source software is part of Debian
> main? :)

Well, there's that sophistry that "non-free is not a part of Debian".  Or
that you can sanely run x86 without at least {intel,amd64}-microcode.  Or,
on most laptops, wifi firmware.


Meow!
-- 
⢀⣴⠾⠻⢶⣦⠀ A tit a day keeps the vet away.
⣾⠁⢰⠒⠀⣿⡁
⢿⡄⠘⠷⠚⠋⠀ (Rejoice as my small-animal-murder-machine got unbroken after
⠈⠳⣄⠀⠀⠀⠀ nearly two years of no catch!)

[toc] | [prev] | [next] | [standalone]

#9273

Am 2017-06-06 22:19, schrieb Adam Borowski:
> Or
> that you can sanely run x86 without at least {intel,amd64}-microcode.

Well, on some systems you can install BIOS/UEFI updates that will
load newer microcode very early in the boot process. In that case
you really don't need the {intel-amd64}-microcode packages, and
you could potentially run just Debian main without any non-free
software on the disk.

Regards,
Christian

[toc] | [prev] | [next] | [standalone]

#9274

[Multipart message — attachments visible in raw view] — view raw

On Wed, Jun 07, 2017 at 01:49:54PM +0200, Christian Seiler wrote:
> > Or
> > that you can sanely run x86 without at least {intel,amd64}-microcode.
> 
> Well, on some systems you can install BIOS/UEFI updates that will
> load newer microcode very early in the boot process. In that case
> you really don't need the {intel-amd64}-microcode packages, and
> you could potentially run just Debian main without any non-free
> software on the disk.
But they are generally updated less often than the Debian packages.
Also, it's a good idea to think about the difference between a system with
the microcode updated by UEFI and one with the microcode updated by a
Debian package wrt open sourceness and such stuff.

-- 
WBR, wRAR

[toc] | [prev] | [next] | [standalone]

#9275

On Wed, 07 Jun 2017, Christian Seiler wrote:
> Am 2017-06-06 22:19, schrieb Adam Borowski:
> >Or
> >that you can sanely run x86 without at least {intel,amd64}-microcode.
> 
> Well, on some systems you can install BIOS/UEFI updates that will
> load newer microcode very early in the boot process. In that case
> you really don't need the {intel-amd64}-microcode packages, and
> you could potentially run just Debian main without any non-free
> software on the disk.

Provided that the system vendor issues timely updates and you check for
new updates (and install them) frequently.  That would be doing it
monthly or at most once every two months for recent systems...  All of
you, owners of Intel Skylake, Intel Kaby Lake, and AMD Ryzen systems,
are doing just that, aren't you?

Because if you aren't, you are at a much higher than usual risk of data
corruption and system misbehavior.

And, unfortunately, even if you do, you could easily be still at risk:
Too many system vendors are BAD at issuing regular firmware updates, and
there are those that issue BIOS updates but stop updating the microcode
inside.

So, and I *am* sorry to say this, users in the general case *are* much
better off with intel-microcode and amd64-microcode installed.

That would be the only reason I actually bother taking care of those two
packages: it is not like I get paid to do it, and it is not relevant to
my job either.  And it takes an annoying amount of effort to try to
track down what a microcode update might be fixing[1].


[1] only to find out it *is* indeed fixing critical defects, so you know
you will not be able to talk yourself out of doing the job when the next
update is released, either :-(

-- 
  Henrique Holschuh

[toc] | [prev] | [next] | [standalone]

#9276

On Wed, Jun 07, 2017 at 11:07:19AM -0300, Henrique de Moraes Holschuh wrote:
> On Wed, 07 Jun 2017, Christian Seiler wrote:
> > Am 2017-06-06 22:19, schrieb Adam Borowski:
> > >Or
> > >that you can sanely run x86 without at least {intel,amd64}-microcode.
> > 
> > Well, on some systems you can install BIOS/UEFI updates that will
> > load newer microcode very early in the boot process. In that case
> > you really don't need the {intel-amd64}-microcode packages, and
> > you could potentially run just Debian main without any non-free
> > software on the disk.
> 
> Provided that the system vendor issues timely updates and you check for
> new updates (and install them) frequently.

I thus prefer for someone to be [un]paid to do that work for me.

> And, unfortunately, even if you do, you could easily be still at risk:
> Too many system vendors are BAD at issuing regular firmware updates, and
> there are those that issue BIOS updates but stop updating the microcode
> inside.

I have yet to see a system vendor who issues BIOS/firmware updates at all
after 1-2 years after manufacture.  If they don't sell the given piece of
hardware anymore, there's no money to be made by keeping it updated.

And, even if they do issue such updates, they tend to be providen as Windows
executables, which is unfun if you don't even have Windows on the machine in
question.

> So, and I *am* sorry to say this, users in the general case *are* much
> better off with intel-microcode and amd64-microcode installed.

Why would this be a bad thing (beside needing non-free microcode at all)?
If I need to run non-free stuff, I prefer it to be where I can see it.


Meow!
-- 
⢀⣴⠾⠻⢶⣦⠀ A tit a day keeps the vet away.
⣾⠁⢰⠒⠀⣿⡁
⢿⡄⠘⠷⠚⠋⠀ (Rejoice as my small-animal-murder-machine got unbroken after
⠈⠳⣄⠀⠀⠀⠀ nearly two years of no catch!)

[toc] | [prev] | [next] | [standalone]

#9277

[Multipart message — attachments visible in raw view] — view raw

On Wed, Jun 07, 2017 at 04:57:17PM +0200, Adam Borowski wrote:
> I have yet to see a system vendor who issues BIOS/firmware updates at all
> after 1-2 years after manufacture.  If they don't sell the given piece of
> hardware anymore, there's no money to be made by keeping it updated.
> 
> And, even if they do issue such updates, they tend to be providen as Windows
> executables, which is unfun if you don't even have Windows on the machine in
> question.
FWIW my laptops and desktop MBs are usually updated with a flash tool
built into the BIOS setup interface with a file from a flash drive. But
it's a small sample, and, incidentally, my desktop got a
binary-file-with-a-windows-flasher "100 Series ME (Version 11.6.27.3264)
Update Tool" just two days ago (I didn't flash it yet).

-- 
WBR, wRAR

[toc] | [prev] | [next] | [standalone]

#9286

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

On 2017-06-06 19:23, Chris Lamb wrote:
> GitHub recently published the announcements of their 2017 Open
> Source Survey:
> 
> http://opensourcesurvey.org/2017/
> 
> 
> I was wondering whether any fellow developers had read this and
> whether they had evaulated it in the context of the Debian project.
> To kick this off, let me quote some of their key insights:

Thanks. I had not read it before and still hasn't read all of it yet.

Spotted a detail in Fig. 3 that wasn't really addressed as far as I
can tell.

Well over 50% have marked CLAs of some type of importance. Isn't that
high?
They might be marked high with the sentiment of "I don't want them and
I deeply care for them to go away" I assume but still.

At the same time license is the most important type of documentation
when it comes to choosing the project to use it seems.

- -- 
brother
http://sis.bthstudent.se
-----BEGIN PGP SIGNATURE-----
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=di7Z
-----END PGP SIGNATURE-----

[toc] | [prev] | [next] | [standalone]

#9288

Martin Bagge / brother wrote:

> Well over 50% have marked CLAs of some type of importance. Isn't that
> high?

Yes, I would have guessed much lower. I would hope it was due to a
philosophical reaction to assigning copyright to $also_commercial_corp
rather than it being merely a hoop to jump through before a PR will be
accepted, but the latter might be skewing the stats.

Can we get more clarity from GitHub itself here..?


Regards,

-- 
      ,''`.
     : :'  :     Chris Lamb
     `. `'`      lamby@debian.org / chris-lamb.co.uk
       `-

[toc] | [prev] | [standalone]

Back to top | Article view | linux.debian.project

csiph-web