Re: Replacement Email Client

[Kenneth Parker]

On Sun, Oct 25, 2020 at 7:00 AM Teemu Likonen  wrote:

> * 2020-10-25 06:51:36-04, Kenneth Parker wrote:
>
> > On Sun, Oct 25, 2020, 6:40 AM  wrote:
> >> alpine
>
> > +1
>
> It would be useful to add some information how the suggested client
> (Alpine) serves the purpose that was asked by the original poster.
>

Fair Enough (and sorry it took me so long.  The system I use alpine on is
remote, and not easily reached through my Android Phone.

When I bring up an email, it shows me a summary of the email structure.
Its default is to show the rendered html part, but the option can be
changed,  here is a short excerpt (

||  Date: Fri, 29 Jun 2018 12:55:31 -0400
||  From: "redac...@gmail.com [hercules-os380]" <
hercules-os...@yahoogroups.com>
||  To: hercules-os...@yahoogroups.com
||  Subject: Re: [hercules-os380] b8000
||  Parts/Attachments:
||1   OK  38 lines  Text
||2 Shown496 lines  Text
||  
||
||
||
||  On 27 June 2018 at 03:53, 'Redacted redac...@gmail.com wrote:
||
||  >> Hercules is already doing exactly that. How far away is it from
being a
||  >> debugger?
||  >>
||  >> > MVSDDT uses (E)STAE; while the error path for 0Cx conditions is
||  >> > longer, it probably makes the code cleaner.
||  >>
[snip rest.  It was html, properly rendered to text]

But now, if I use the h command (for Full Headers), I get Full email
Headers, and then the html source of this email! (h again toggles back).

I believe the original poster was interested in html forms.  If in the
email itself, you get the Source of the Form Itself.  If it's in a link,
you see the link with the usual html "a" element.  (afraid to copy it here,
lest gmail Executes it!)

I will try to be more specific in the future.  ;-)

Kenneth Parker

Re: The .xsession-errors problem

[David]

On Wed, 28 Oct 2020 at 00:45, Andrei POPESCU  wrote:
> On Ma, 27 oct 20, 07:55:00, Greg Wooledge wrote:
> > On Mon, Oct 26, 2020 at 11:07:37PM +, Tixy wrote:
> > > On Mon, 2020-10-26 at 18:35 +0200, Teemu Likonen wrote:

> > > > It seems that ~/.xsession-errors file can still grow to infinity in
> > > > size. Sometimes it grows really fast. This is nothing new: we have all
> > > > seen it and talked about it. What do you do to maintain this file?

> > > Don't do anything here. The file is created fresh at each boot and is
> > > 30 lines long [...]

> > Something that you're doing, or something that was done for you, is
> > clearing that file.  Your case is not the default.  By default, that
> > file is never cleared, and just keeps growing.  Most people prune it
> > manually whenever they notice it getting bigger than they like, which
> > is usually somewhere between "once a year" and "never".

> On my system the file is rotated (renamed to .xsession-errors.old), on
> every login as far as I can tell.

> Didn't find (yet) what is doing this (using lightdm, LXDE and minimal
> Xorg).

I had a curiosity about this, because some people are reporting that they
need to manage their growing .xsession-errors file by various methods,
but I never have seen this.

I see the same behaviour as Andrei, and I also use parts of LXDE.

I investigated, guided by this teaching from Reco:
https://lists.debian.org/debian-user/2020/04/msg00583.html

I found that the process that renames the file to .xsession-errors.old
is the binary /usr/bin/lxsession owned by the user, with the parent
process lightdm owned by root.

/usr/bin/lxsession is a component of LXDE, so this won't apply to
users of other GUI providers.

The rename occurs when the user logs in from lightdm. The filename
.xsession-errors is defined in the script file /etc/X11/Xsession

Re: Replacement Email Client

[David Wright]

On Tue 27 Oct 2020 at 22:22:16 (+), Jeremy Nicoll wrote:
> On Tue, 27 Oct 2020, at 21:04, Dan Ritter wrote:
> 
> > - Use a good MUA and resign yourself to occasionally sending an 
> >   email to a browser. Despite your protestations of "logging
> >   in", having a browser display your email requires no such
> >   thing. The MUA saves the email to a file, then hands the file 
> >   to the browser to open it.
> 
> If the email also contained umpteen attachments each of which 
> was (typically) an image, then the MUA needs to create a folder
> containing the html and the unpacked images.  Depending on 
> the way that the html referred to those images, the MUA also 
> needs to revise the form of the image references inside the 
> written-out html so that the written-out image files can be found
> by the browser when it processes the html.  It's not quite as simple
> as you make out, especially if the code has to work on multiple 
> platforms / file systems.
> 
> In my experience such mails are usually forwarded jokes etc, and
> another problem is that they often contain forwards of forwards
> of forwards (etc) with nested sets of attachments (as a succession
> of technically naive users just forward the mail they received, 
> rather than creating a new original one with just once set of 
> contents).

I've found those sorts of emails (loosely coupled images) are easy to
deal with. In mutt, for example, press v for the attachments menu,
and again on any multipart or message/rfc822 that needs opening,
exposing the attachments within.

The problem is where the image attachments are tightly integrated
with the text, so that they really need displaying together.
I'm with Joe on that: I'd want to use a browser, that's what
they're designed for.

> Some email clients also don't assign separate names to the 
> attachments, so when they get written out the MUA has to 
> invent a name sequence (and of course modify the html 
> accordingly). 

Cheers,
David.

[HS] diffuser le son d'Android via Pulseaudio ?

[Gaëtan Perrier]

Bonjour,

Savez-vous s'il est possible de diffuser le son d'un mobile Android sur les
enceintes d'un PC en utilisant PulseAudio ?
Je sais le faire depuis mon PC portable, mais depuis android je n'ai pas
trouvé.

Gaëtan


signature.asc
Description: This is a digitally signed message part

[SOLVED] Re: PC wont boot to GUI

[David]

On Wed, 28 Oct 2020 at 11:46, John Figie  wrote:

> Thanks for your help, this solved my problem.

[...]

> Reboot and life is good now :-)

Great, thanks for confirming and providing a summary of your solution.

Re: PC wont boot to GUI

[John Figie]

David

Thanks for your help, this solved my problem. I did however need to figure
out how to connect my computer to the internet:
(additional help from
https://serverfault.com/questions/21475/starting-network-connection-from-ubuntu-recovery
)
I did open my sources.list file but there was no need to add "contrib" and
"non-free" as these were already in the list.

# ip linkthis returns the name of the ethernet on my machine it was enp2s0
# ip link set enp2s0 up
# dhclient enp2s0

# ip addr this now shows that I have an ip address on my network
given by my router

Then run

# apt install firmware-amd-graphics

Reboot and life is good now :-)

Regards,

John Figie


On Tue, Oct 27, 2020 at 7:01 AM Greg Wooledge  wrote:

> On Tue, Oct 27, 2020 at 01:18:42PM +1100, David wrote:
> > You need to install the firmware.
> > You do that by running this command as superuser:
> >
> > # apt install firmware-amd-graphics
>
> You may have to add "contrib" and "non-free" to your sources.list first,
> and then run "apt update".
>
> See  if
> you need instructions for editing the file.
>
>

Bug Report? - T14 Microphone Issue - Possible Missing Kernel Config?

[Jonathan]

Good Evening,

I was attempting to get my microphone working on my T14 AMD as it is the 
only non-functioning piece I was aware of. Browsing possible solutions I 
was shown the following could fix the issue if added to the kernel config:


CONFIG_SND_SOC_AMD_RENOIR=m

CONFIG_SND_SOC_AMD_RENOIR_MACH=m


I wasn't sure how to properly post a bug report or if this was even a 
bug. I apologize if this is an incorrect way to present this information.



Thank you


Jonathan

Re: Replacement Email Client

[John Conover]

Patrick Bartek writes:
> > >  
> > >> On 10/25/20 8:28 PM, Patrick Bartek wrote:  
> > >>> I'm not referring to viewing HTML emails. I already can do that in
> > >>> Claws-Mail using its Dillo plugin. I'm talking about filling in
> > >>> forms, etc. that are part of the HTML email and sending just the
> > >>> data without "replying" in the normal sense.  This is beyond
> > >>> Claws' and Dillo's capabilities.  I have to use a real browser,
> > >>> log into that particular web mail account (like gmail), click on
> > >>> that particular email, etc. to do so.
> > >>>

Has anyone:

1) ReBoot your machine.
2) login and launch claws-mail with the Dillo plugin, 
   and exit claws-mail
3) lsof -Pni > tempfile

and near the end of tempfile is 10 process running from Dillo, even
through claws-mail has exited, (all listeners, something like 50 MB of
memory.)

John

-- 

John Conover, cono...@rahul.net, http://www.johncon.com/

Re: Qualité du chiffrement des archives 7z

[kaliderus]

Rebonsoir,

Merci pour vos retours et tuyaux, mon interrogation était bien sur le
chiffrement embarqué par 7z, le moyen de transfert via ssh étant déjà
utilisé.
https://www.7-zip.org/7z.html précise effectivement l'algorithme mis en oeuvre.
Je crois que je vais m'orienter vers la solution à base de GnuPG.
Intéressants tes liens Sébastien, content de voir des solutions de
déduplication intégrées à Debian.

A+ la liste

Le sam. 24 oct. 2020 à 14:40, Sébastien Dinot
 a écrit :
>
> kaliderus a écrit :
> > J'envisage de faire quelques sauvegardes avec 7z et de les envoyer sur
> > un serveur sur lequel j'ai la main, mais pas sur le réseau
> > intermédiaire.
>
> Pourquoi ne pas utiliser un outil de sauvegarde qui chiffre les données
> localement, avant de les transmettre ?
>
> J'utilise pour ma part borg-backup, mais duplicity ou restic font de
> même.
>
> Et s'il s'agit juste de sauvegarder un fichier ou une simple archive,
> autant utiliser GnuPG, il est fait pour cela (et pour la signature
> cryptographique).
>
> Pour ma part, je n'utilise le chiffrement avec 7z (chiffrement
> relativement fort, mais symétrique) que pour transmettre des données
> confidentielles à des tiers qui n'ont pas de clé GnuPG et ne savent pas
> se servir d'un tel outil. Dans ce cas, je leur transmets l'archive
> chiffrée via un service d'échange de fichier quelconque (par exemple
> https://transfer.sh/ ou https://dl.free.fr) et le mot de passe
> permettant de déchiffrer l'archive via SMS.
>
> Et si les données confidentielles en question, sont textuelles et de
> faible taille, j'utilise le service https://privatebin.net/, toujours en
> envoyant le mot de passe par un autre vecteur.
>
> Sébastien
>
> --
> Sébastien Dinot, sebastien.di...@free.fr
> http://www.palabritudes.net/
> Ne goûtez pas au logiciel libre, vous ne pourriez plus vous en passer !
>

Re: Replacement Email Client

[Jeremy Nicoll]

On Tue, 27 Oct 2020, at 21:04, Dan Ritter wrote:

> - Use a good MUA and resign yourself to occasionally sending an 
>   email to a browser. Despite your protestations of "logging
>   in", having a browser display your email requires no such
>   thing. The MUA saves the email to a file, then hands the file 
>   to the browser to open it.

If the email also contained umpteen attachments each of which 
was (typically) an image, then the MUA needs to create a folder
containing the html and the unpacked images.  Depending on 
the way that the html referred to those images, the MUA also 
needs to revise the form of the image references inside the 
written-out html so that the written-out image files can be found
by the browser when it processes the html.  It's not quite as simple
as you make out, especially if the code has to work on multiple 
platforms / file systems.

In my experience such mails are usually forwarded jokes etc, and
another problem is that they often contain forwards of forwards
of forwards (etc) with nested sets of attachments (as a succession
of technically naive users just forward the mail they received, 
rather than creating a new original one with just once set of 
contents).

Some email clients also don't assign separate names to the 
attachments, so when they get written out the MUA has to 
invent a name sequence (and of course modify the html 
accordingly). 

-- 
Jeremy Nicoll - my opinions are my own.

Re: Replacement Email Client

[Joe]

On Tue, 27 Oct 2020 15:23:11 -0400
rhkra...@gmail.com wrote:

> On Tuesday, October 27, 2020 12:15:46 PM Joe wrote:
> > On Tue, 27 Oct 2020 07:43:43 -0400
> > 
> > Greg Wooledge  wrote:  
> > > [1]I used to read slashdot regularly, and on slashdot, the front
> > > page had a bunch of news stories and a poll.  The poll was
> > > written as a vanilla HTML form.  If you participated in the poll,
> > > it would send you to a new instance of the home page, because a
> > > form *must* load a new page.  Doing that would lose my place,
> > > showing a new set of stories, even if I hadn't finished reading
> > > the ones on the previous instance.  
> > 
> > It doesn't have to be like that. Nearly all of my web applications
> > just use the one page, though of course it does have to be reloaded
> > after a submit. Anything I want to be persistent, I need to arrange
> > through hidden controls, appearing as parameters in the reloaded
> > page. If someone is showing you a large number of random entries on
> > a page, then of course it may be too much trouble to do this, but
> > it is certainly possible.  
> 
> Is what you describe doing something you do on a web page or in an
> email?

A web page, for my own use only. I don't send HTML email. No, it's off
the original topic a bit, but that looks unlikely to have a satisfactory
answer.

I'd be horrified at the idea of a new, full HTML/JS renderer being
embedded in an email client. It takes years to work the worst of the
security bugs out of a purpose-built web browser, which is one of the
most complex pieces of everyday software around.

-- 
Joe

Re: aptitude safe-upgrade vs apt-get upgrade.

[Andrei POPESCU]

On Ma, 27 oct 20, 13:03:32, David Wright wrote:
> On Tue 27 Oct 2020 at 15:05:36 (+0200), Andrei POPESCU wrote:
> > On Lu, 26 oct 20, 09:55:00, John Hasler wrote:
> > 
> > I believe someone demonstrated quite recently on list that dpkg has some 
> > limits in the number and/or combination of packages it can deal with at 
> > once, so APT might have to pass them in smaller chunks and/or specific 
> > order (in case of Pre-Depends: maybe?).
> >  
> > > Dpkg is safe but can be rather frustrating.
> > 
> > As far as I'm concerned it does just fine what it was designed to do.
> 
> Years ago, I used to do mega installs with dpkg, copying the contents
> of /var/cache/apt/archives/ from one machine onto a caddy, and then
> installing them all on another.
> 
> I'm probably the person who recently demonstrated dpkg's (in)ability
> to cleanly install 1558 packages simultaneously without help:
> 
> https://lists.debian.org/debian-user/2020/07/msg00032.html
> (its References are all dead links because they fall in the previous month.)

Indeed, this is the experiment I was thinking of.
 
> This followed a demonstration of apt-get's reversibility:
> specifying 271 "top-level" packages that resulted in 1558 being
> installed altogether, and all 1558 being cleanly purged again.

Somehow I'm still under the impression that this is supposed to work 
with dpkg only.

You might want to post your findings to debian-dpkg for further 
investigation.

Kind regard,
Andrei
-- 
http://wiki.debian.org/FAQsFromDebianUser


signature.asc
Description: PGP signature

Re: Replacement Email Client

[Dan Ritter]

Patrick Bartek wrote: 
> On Mon, 26 Oct 2020 21:23:59 -0400
> Carl Fink  wrote:
> 
> I don't know if there is a  or not.  Someone else suggested
> that.  It may be javascript.  Never checked email's code all that
> closely.  Next time I get one of those type emails, I'll look.
> 
> > I think you're getting normal email with a link to a form, but (as
> > you say) Dillo doesn't let you click the link, so you never realize
> > what it is. Those "rate us from 1-5" things are normally five
> > different links to the same online poll, with a parameter telling the
> > page what rating you selected. If You opened the links in an
> > HTML-aware mailer like Mutt, you could just select one of those to
> > open the form in a browser.
> 
> Yes, sometimes when I use a browser for these HTML emails and click on
> something, a new tab with a new URL opens.  Sometime not. Next time,
> I'll check the code.
> 
> Not that it matters all that much.  All I want is a lightweight email
> client that works with HTML emails, too, so I don't have to switch back
> and forth to a browser, login, etc. to get things done.  

I don't think you're going to get it.

You can't get everything right with a web page these days with
anything less than a second-rank browser. (First rank:
Chrome/Chromium, Firefox, IE/Edge/Chromium; second rank: Opera,
Brave, derivatives of first-rank and derivatives of old versions
of first-rank. Third-rank: everything else, including anything
that doesn't support modern JavaScript and HTML5.)

No browser in the first or second rank is light-weight.

QED.

You can:

- Use a webclient, including a very lightweight webclient like
  Rainloop. Rainloop literally does not store anything on the
  server side; it runs in your browser. It's great for small,
  well-organized IMAP accounts; terrible if you hoard millions
  of messages.

- Use a good MUA and resign yourself to occasionally sending an 
  email to a browser. Despite your protestations of "logging
  in", having a browser display your email requires no such
  thing. The MUA saves the email to a file, then hands the file 
  to the browser to open it.

- Write your own and make the world a better place!

-dsr-

Re: Replacement Email Client

[Patrick Bartek]

On Mon, 26 Oct 2020 21:23:59 -0400
Carl Fink  wrote:

> On 10/26/20 6:16 PM, Patrick Bartek wrote:
> > On Sun, 25 Oct 2020 20:45:50 -0400
> > Carl Fink  wrote:
> >  
> >> On 10/25/20 8:28 PM, Patrick Bartek wrote:  
> >>> I'm not referring to viewing HTML emails. I already can do that in
> >>> Claws-Mail using its Dillo plugin. I'm talking about filling in
> >>> forms, etc. that are part of the HTML email and sending just the
> >>> data without "replying" in the normal sense.  This is beyond
> >>> Claws' and Dillo's capabilities.  I have to use a real browser,
> >>> log into that particular web mail account (like gmail), click on
> >>> that particular email, etc. to do so.
> >>>
> >>> I'm getting the sense that I may not be able to find a client that
> >>> can do that.  
> >> I don't remember ever getting an emailed form that was anything but
> >> a link to a web page. Who is sending you emailed inline forms?  
> > The ones I respond to are known to me and are legit --
> > organizations, businesses, government agencies, etc. -- that I do
> > business with.  To respond, I must switch to a web browser, login to
> > my email account (like gmail), find that particular email, and
> > enter the requested data either by keyboard, drop-down menu,
> > buttons, etc., then SUBMIT it.  This happens all within the email.
> > I never get forwarded to another web site. I always stay on the web
> > mail page. As far as I can tell only the data is sent. The email
> > itself is not replied to.  That is, there's nothing in the "Sent"
> > folder.  
> 
> Meaning no offense, I doubt it. I have been using email since before
> Gopher, and I have literally never received an HTML  email with an
> embedded form (that I opened, at least). I find it hard to believe
> that you get many of them.

I don't know if there is a  or not.  Someone else suggested
that.  It may be javascript.  Never checked email's code all that
closely.  Next time I get one of those type emails, I'll look.

> I think you're getting normal email with a link to a form, but (as
> you say) Dillo doesn't let you click the link, so you never realize
> what it is. Those "rate us from 1-5" things are normally five
> different links to the same online poll, with a parameter telling the
> page what rating you selected. If You opened the links in an
> HTML-aware mailer like Mutt, you could just select one of those to
> open the form in a browser.

Yes, sometimes when I use a browser for these HTML emails and click on
something, a new tab with a new URL opens.  Sometime not. Next time,
I'll check the code.

Not that it matters all that much.  All I want is a lightweight email
client that works with HTML emails, too, so I don't have to switch back
and forth to a browser, login, etc. to get things done.  

B

Re: Mounting a USB device

[Mick Ab]

Thanks for the replies.

It seems to me that the situation is as follows :-

Filesystems in /etc/fstab which have the noauto option are not
automatically mounted at boot time, so if these filesystems are already
plugged into USB ports at boot time, they would subsequently have to be
manually mounted in order to be used.

Filesystems which are plugged into a port after the system has been booted
are automatically mounted.

On 27 Oct 2020 18:20, "Kenneth Parker"  wrote:

>
>
> On Tue, Oct 27, 2020, 11:51 AM Mick Ab 
> wrote:
>
>> If a filesystem in /etc/fstab has a noauto entry, can that filesystem
>> only be mounted manually using the mount command or
>> is there any chance that it will be automatically mounted by
>> usbmount ?
>>
>> The filesystem is used in a USB port.
>>
>
> I have a dislike of Gnome, because it seems to mount  *every*  Filesystem
> I have, even ones that I consider sensitive.
>
> But it doesn't occur until the GUI comes up.  (I set the SystemD Default
> Target to multi-user and only type "systemctl start graphical.target" after
> I finish my "Apt Ritual").
>
> Not sure what Gnome Package does this.  Any Gnome Experts here?
>
> Kenneth Parker
>
>>

Re: Mounting a USB device

[tomas]

On Tue, Oct 27, 2020 at 02:20:14PM -0400, Kenneth Parker wrote:
> On Tue, Oct 27, 2020, 11:51 AM Mick Ab  wrote:
> 
> > If a filesystem in /etc/fstab has a noauto entry, can that filesystem only
> > be mounted manually using the mount command or
> > is there any chance that it will be automatically mounted by
> > usbmount ?
> >
> > The filesystem is used in a USB port.
> >
> 
> I have a dislike of Gnome, because it seems to mount  *every*  Filesystem I
> have, even ones that I consider sensitive.
> 
> But it doesn't occur until the GUI comes up.  (I set the SystemD Default
> Target to multi-user and only type "systemctl start graphical.target" after
> I finish my "Apt Ritual").
> 
> Not sure what Gnome Package does this.  Any Gnome Experts here?

Not a Gnome expert, but I play one on TV (well, my favourite search
engine [1] does that, actually ;-)

It seems that Nautilus, the file manager is the one doing it [2] (or,
controlling it, actually). It can be disabled, it seems.

Cheers

[1] No, not that one with the G.
[2] 
https://askubuntu.com/questions/18926/how-to-control-gnomes-auto-mounting-capabilities

 - t


signature.asc
Description: Digital signature

Re: Mounting a USB device

[Nicolas George]

ghe2001 (12020-10-27):
> > dme is faster.
> 
> What's a dme?  It's not on my system, and aptitude claims it doesn't exist.

It's not dme, it's dme

ssecem ~ $ apt-file search '/bin/dme'
herbstluftwm: /usr/bin/dmenu_run_hlwm 
klibc-utils: /usr/lib/klibc/bin/dmesg
suckless-tools: /usr/bin/dmenu
suckless-tools: /usr/bin/dmenu_path
suckless-tools: /usr/bin/dmenu_run
util-linux: /bin/dmesg

Take your pick.

Regards,

-- 
  Nicolas George


signature.asc
Description: PGP signature

Re: Mounting a USB device

[Reco]

Hi.

On Tue, Oct 27, 2020 at 07:06:20PM +, ghe2001 wrote:
> ‐‐‐ Original Message ‐‐‐
> On Tuesday, October 27, 2020 12:58 PM, Nicolas George  wrote:
> 
> > dme is faster.
> 
> What's a dme?  It's not on my system, and aptitude claims it doesn't exist.

/bin/dmesg, probably. A part of util-linux package.

Reco

Re: Replacement Email Client

[rhkramer]

On Tuesday, October 27, 2020 12:15:46 PM Joe wrote:
> On Tue, 27 Oct 2020 07:43:43 -0400
> 
> Greg Wooledge  wrote:
> > [1]I used to read slashdot regularly, and on slashdot, the front page
> > had a bunch of news stories and a poll.  The poll was written as a
> > vanilla HTML form.  If you participated in the poll, it would send you
> > to a new instance of the home page, because a form *must* load a new
> > page.  Doing that would lose my place, showing a new set of stories,
> > even if I hadn't finished reading the ones on the previous instance.
> 
> It doesn't have to be like that. Nearly all of my web applications just
> use the one page, though of course it does have to be reloaded after a
> submit. Anything I want to be persistent, I need to arrange through
> hidden controls, appearing as parameters in the reloaded page. If
> someone is showing you a large number of random entries on a page, then
> of course it may be too much trouble to do this, but it is certainly
> possible.

Is what you describe doing something you do on a web page or in an email?



> 
> I do use the very occasional smidgen of JS to replace things that have
> been left out of HTML, such as making a radio button group invoke a
> submit when changed, but on the whole I believe client-side scripting
> to be the work of the devil.

Re: Pourquoi n'y a-t-il pas FreeCad 0.19 en testing ou sid ?

[Iznogood]

Bonjour,

j'utilise FreeCAD régulièrement avec les mises à jours sous forme de
.AppImage. Le pack est complet et fonctionnel.

Cordialement



Le Tue, 27 Oct 2020 09:16:45 +
benoit  a écrit :

> Sent with ProtonMail Secure Email.
> 
> ‐‐‐ Original Message ‐‐‐
> Le mardi 27 octobre 2020 09:41, BERTRAND Joël  a 
> écrit :
> 
> >
> > FreeCAD est un logiciel un peu tordu à compiler. J'utilise cet outil en
> > rolling release et il y a des dépendances qui ne sont pas dans Debian
> > avec quelques incompatibilité entre OCC et OCE, sans compter d'autres
> > bizarreries avec vtk et consorts qui dépendent aussi des deux larrons
> > précédents.
> >
> > Pour utiliser Kicad (5.99) et FreeCAD (0.19) sur la même machine, je me
> > suis résolu à compiler les deux depuis les sources - en me faisant un
> > noeud au cerveau pour que les deux utilisent les mêmes bibliothèques -
> > pour éviter les problèmes.  
> 
> Déjà en buster c'est une ancienne version qui n'est même pas la stable 
> actuelle.
> Donc oui, le seul moyen d'utiliser FreeCad sous debian en profitant des 
> dernières avancées c'est de compiler depuis les sources.
> 
> Effectivement à une certaine époque j’ai dû aussi compiler des dépendances 
> (pivy, soqt, coin…), mais vu que c’est sur mon système ainsi qu’une longue 
> liste de paquets dépendants, je ne sais plus très bien ce qui est en paquet 
> debian ou pas et ce qui est vraiment nécessaire, je compile sans me poser de 
> question.
> 
> Mon prochain objectif sera de faire le tri dans tout ça et apprendre à 
> construire un paquet debian de freecad 0.19 pour avoir une routine de 
> compilation/installation maîtrisée. Sauf si j’apprends que la version de 
> développement de freeecad (0.19 en ce moment) est dispo en paquet debian, 
> d’où ma question...
> 
> --
> Benoit
> 

Re: The .xsession-errors problem

[rhkramer]

On Tuesday, October 27, 2020 11:05:31 AM Teemu Likonen wrote:
> * 2020-10-26 20:04:55+03, Reco wrote:
> > On Mon, Oct 26, 2020 at 06:35:45PM +0200, Teemu Likonen wrote:
> >>   - Do you configure some rotating system, perhaps with logrotate(8)?
> >>   
> >> (Why doesn't Debian have this automatically?)

I simply (like someone else mentioned on the list) truncate the file 
periodically (actually, once every minute) using a crontab with a task like 
this:

  *  *  *   *   * echo "Cleared on $(date) by $USER cron" > 
/home//.xsession-errors

Back in some day, the .xsession-errors was growing by, iirc, multiple messages 
every second.  (This might have been in Debian Wheezy, or a Debian before 
Wheezy, or even in a version of Mandrake / Mandriva.)

Re: What's going on with snapd?

[Brian Vaughan]

On 10/27/2020 6:54 AM, Andrei POPESCU wrote:

What release are you running (stable, testing, unstable, etc.)? The
package in stable will only receive security updates.


I'm running unstable. But if I've got the dates right, some of them were 
reported before Buster became stable.


I suppose my main concern here is that most of the listed bugs seem to 
have gotten no response at all, for a few years, so a framework that (I 
was told) is considered a reliable tool for daily use at Canonical seems 
to be badly broken on Debian, so there's some break  in communication, 
and I'm wondering if there's anything to be done about it.


The immediate reason I ran into an issue was that I wanted to try out 
Ember, a client for WorldForge, an open source MMO; it's in the Debian 
repository as a snap package -- which strikes me as an odd choice, and I 
gather there was some controversy about it. Anyway, it seems like the 
issues I ran into were all issues with snapd, which seems like a tool 
intended for more extensive use and a broader concern.



Please provide some example for the bugs you're interested in.

The ones I ran into myself:

#880174 snapd: swrast error (nvidia-legacy-340xx) while launching UI 
installed-snaps

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=880174

#911906 snapd: Boot hangs due to snappy/snapd: `A start job is running 
for Snappy daemon (../1min31s)`

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=911906

#931776 snapd: Installed snaps do not appear in desktop launcher in 
Debian buster.

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=931776

#947325 snapd: strict confinement is not enabled
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=947325

Of those, #931776 looked the most like a problem I understood; it looks 
to me as if someone simply copied over a configuration script meant for 
headless servers and didn't check if it would work with GUI desktops, 
which it doesn't, because the script's not executed in that case. But 
startup configuration dotfiles are a maddening labyrinth.

Re: Mounting a USB device

[Cindy Sue Causey]

On 10/27/20, Mick Ab  wrote:
> If a filesystem in /etc/fstab has a noauto entry, can that filesystem only
> be mounted manually using the mount command or
> is there any chance that it will be automatically mounted by
> usbmount ?
>
> The filesystem is used in a USB port.


Is there a "sub" anything on that filesystem that would be... sorry,
words are failing.. that would be quietly "sub-mounted"? I've had a
partition be mounted when a directory was "mount -B" mounted
*somehow*, but I can't come up with a scenario. I looked at past
/etc/fstab entries (still saved as comments), but they're not
triggering any related memories.

It has only happened once that I noticed. A full partition was mounted
to my great shock. At some point not long after, I realized that a
directory within that partition was some form of "mount --bind". I
just figured that was surely how it happened because the partition had
to be mounted first to reach that directory.

Yes, I know. Everything I look at for my usage case, it shouldn't have
happened backwards. The directory mount should have failed and exited
as e.g. "can't find" the way mount did just now.

Thought it was a handy shortcut then forgot about it. It was kind of
like using "-p" flag when manually "mkdir" creating embedded
parent-child directories.

But, again, am now not able to duplicate where it had happened for me.
ONCE (that I know of).

Cindy :)
-- 
Cindy-Sue Causey
Talking Rock, Pickens County, Georgia, USA

* runs with birdseed *

Re: Mounting a USB device

[Nicolas George]

ghe2001 (12020-10-27):
> On my system (Buster) USB things are called /dev/sd.  From
> the CLI, I type "sudo mount /dev/sd" to see what's already there,
> then plug in the USB device, and  hit  again to see what's new. 

dme is faster.

Regards,

-- 
  Nicolas George


signature.asc
Description: PGP signature

Re: Mounting a USB device

[Brian]

On Tue 27 Oct 2020 at 15:51:06 +, Mick Ab wrote:

> If a filesystem in /etc/fstab has a noauto entry, can that filesystem only
> be mounted manually using the mount command or
> is there any chance that it will be automatically mounted by
> usbmount ?
> 
> The filesystem is used in a USB port.

I too used usbmount in the distant past; it is now consigned to the
refuse bin of history. In my /etc/fstab I have

LABEL=ARCHIVE-10 /media/ARCHIVE-10 vfat 
ro,gid=1000,fmask=0117,dmask=0007,noatime,noauto,user,x-systemd.automount,x-systemd.idle-timeout=5,x-systemd.device-timeout=1
 0 0

The USB stick is mounted when it is accessed and unmounted after 5
seconds of non-use. Works brilliantly.

-- 
Brian.

Re: La Comisión Europea aprueba la nueva Estrategia de Software de Código Abierto 2020-2023

[Nicholas Geovanis]

On Tue, Oct 27, 2020, 1:03 PM mick crane  wrote:

> On 2020-10-27 14:35, Kenneth Parker wrote:
>
> > Good article, about the European Commission embracing a "new Open
> > Source
> > Software Strategy".  Being the European Union is awesome, simply
> > because of
> > how big it is, of course.  I like their principle, "Think Open".
>
> If the European Union say they want to "embrace" you I would run.
>

It's just a poor translation of the cognate French verb. They're really
just saying they want to kiss you 

mick
>

Re: Mounting a USB device

[Kenneth Parker]

On Tue, Oct 27, 2020, 11:51 AM Mick Ab  wrote:

> If a filesystem in /etc/fstab has a noauto entry, can that filesystem only
> be mounted manually using the mount command or
> is there any chance that it will be automatically mounted by
> usbmount ?
>
> The filesystem is used in a USB port.
>

I have a dislike of Gnome, because it seems to mount  *every*  Filesystem I
have, even ones that I consider sensitive.

But it doesn't occur until the GUI comes up.  (I set the SystemD Default
Target to multi-user and only type "systemctl start graphical.target" after
I finish my "Apt Ritual").

Not sure what Gnome Package does this.  Any Gnome Experts here?

Kenneth Parker

>

Re: getmail, getmail6, testing, unstable, python-is-python3

[Reco]

Hi.

On Tue, Oct 27, 2020 at 01:20:17PM -0400, songbird wrote:
> Reco wrote:
> ...
> > Because python. Be it python 2 or python 3 - it's (in)famous for one
> > thing - lack of backwards compatibility. Sooner or later they replace
> > python 3.8 (current sid) with, say, python 3.9 - and the things will
> > break again.
> > Because they did on 3.7->3.8 transition, did before that on 3.6->3.7
> > transition and this list goes long way back.
> 
>   i've not had any problems with python3 here, but i 
> don't do strange things with python3.

Because it's not your burden ;)
Every time a major change like python transition happens in Debian -
packages start breaking. It's up for the maintainer to fix such
breakages (obviously), and once in a while a maintainer goes MIA.
What happens next is packages get deleted from testing first, sid second.

Luckily the bullseye package freeze should be happening Soon™, so it's
unlikely we'll see python 3.9 (and inevitable collateral damage it
brings) in bullseye.


> oh, let me rephrase that, during a recent round i had a problem with a
> program, but that wasn't anything i did so it was somewhere in the
> stack of modules that program used.  does that mean it was
> incompatible?

I cannot tell you that without going into specifics (python version,
modules used, all usual stuff). Usually it's pretty clear from Python
backtrace.


> no, it just meant there was a bug introduced which
> eventually got fixed.

It means you got lucky that time. I would not count on such luck next
time.

Reco

Re: aptitude safe-upgrade vs apt-get upgrade.

[David Wright]

On Tue 27 Oct 2020 at 15:05:36 (+0200), Andrei POPESCU wrote:
> On Lu, 26 oct 20, 09:55:00, John Hasler wrote:
> > Andrei writes:
> > > dpkg does its own dependency checking, in addition to APT (the
> > > software, not the command), and will prevent any inconsistencies
> > > unless you use one of the --force switches.
> > 
> > What it does not do is resolve dependencies.  Apt recursively resolves
> > dependencies, installing them as required. It also detects conflicts
> > and offers to resolve them as well as breaking loops.
> 
> dpkg can only work with the set of .deb files that were passed on the 
> command line.
> 
> If all dependencies are included (or already installed), fine, otherwise 
> it will bail out as it doesn't (by design) have the capability to search 
> for them in repositories and download them (if this is what you mean by 
> resolving).
> 
> I believe someone demonstrated quite recently on list that dpkg has some 
> limits in the number and/or combination of packages it can deal with at 
> once, so APT might have to pass them in smaller chunks and/or specific 
> order (in case of Pre-Depends: maybe?).
>  
> > Dpkg is safe but can be rather frustrating.
> 
> As far as I'm concerned it does just fine what it was designed to do.

Years ago, I used to do mega installs with dpkg, copying the contents
of /var/cache/apt/archives/ from one machine onto a caddy, and then
installing them all on another.

I'm probably the person who recently demonstrated dpkg's (in)ability
to cleanly install 1558 packages simultaneously without help:

https://lists.debian.org/debian-user/2020/07/msg00032.html
(its References are all dead links because they fall in the previous month.)

This followed a demonstration of apt-get's reversibility:
specifying 271 "top-level" packages that resulted in 1558 being
installed altogether, and all 1558 being cleanly purged again.

Cheers,
David.

Re: La Comisión Europea aprueba la nueva Estrategia de Software de Código Abierto 2020-2023

[mick crane]

On 2020-10-27 14:35, Kenneth Parker wrote:

Good article, about the European Commission embracing a "new Open 
Source
Software Strategy".  Being the European Union is awesome, simply 
because of

how big it is, of course.  I like their principle, "Think Open".


If the European Union say they want to "embrace" you I would run.

mick

--
Key ID4BFEBB31

Re: Mounting a USB device

[David Wright]

On Tue 27 Oct 2020 at 15:51:06 (+), Mick Ab wrote:
> If a filesystem in /etc/fstab has a noauto entry, can that filesystem only
> be mounted manually using the mount command or
> is there any chance that it will be automatically mounted by
> usbmount ?
> 
> The filesystem is used in a USB port.

According to   man fstab   noauto protects against automatic mounting
by   mount -a   which is what's done at boot time. Your automounter
can't be using   mount -a   as that could mount totally unrelated
filesystems, so there's every likelihood that noauto is ignored.

Cheers,
David.

Re: The European Commission approved the new Open Source Software Strategy 2020-2023 of the Commission

[steve]

Le 27-10-2020, à 18:50:50 +0100, NoSpam a écrit :



Le 27/10/2020 à 18:45, steve a écrit :


https://ec.europa.eu/info/departments/informatics/open-source-software-strategy_fr



sauf que

https://www.zdnet.fr/blogs/l-esprit-libre/la-commission-europeenne-et-les-logiciels-libres-ambition-molle-pour-l-april-39912003.htm


Autant mettre le lien direct

https://april.org/strategie-logiciel-libre-de-la-commission-europeenne-un-esprit-ouvert-qui-manque-de-force

Re: [testing] noyau 5.9

[Gaëtan Perrier]

Le lundi 26 octobre 2020 à 22:10 +0100, F. Dubois a écrit :
> Le 26/10/2020 à 21:03, Kohler Gerard a écrit :
> > a savoir :
> > incompatibilité actuelle entre le kernel 5.9 et les drivers proprio 
> > Nvidia,
> > il faut attendre la mise à jour de Nvidia
> > 
> Ou passer les drivers en experimental, sid noyau 5.9 tout fonctionne.
> 

Non car je suis en legacy-390xx ...


signature.asc
Description: This is a digitally signed message part

Re: The European Commission approved the new Open Source Software Strategy 2020-2023 of the Commission

[NoSpam]

Le 27/10/2020 à 18:45, steve a écrit :


https://ec.europa.eu/info/departments/informatics/open-source-software-strategy_fr 



sauf que

https://www.zdnet.fr/blogs/l-esprit-libre/la-commission-europeenne-et-les-logiciels-libres-ambition-molle-pour-l-april-39912003.htm

Re: La Comisión Europea aprueba la nueva Estrategia de Software de Código Abierto 2020-2023

[Nicholas Geovanis]

Just curious: Since Britain is shortly leaving the EU, has PM Boris' regime
made any similar commitments? As part of the divorce settlement? 

On Tue, Oct 27, 2020, 9:36 AM Kenneth Parker  wrote:

>
>
> On Tue, Oct 27, 2020, 6:29 AM Weaver  wrote:
>
>> On 27-10-2020 20:01, Iker Bilbao wrote:
>> >
>> https://www.casadomo.com/2020/10/27/comision-europea-aprueba-nueva-estrategia-software-codigo-abierto-2020-2023
>>
>> Bravo!
>> --
>
>
> Thank you Weaver.  I had ignored the original email, due to it being
> "English Challenged", but you inspired me to click the link and let Google
> translate it.
>
> Good article, about the European Commission embracing a "new Open Source
> Software Strategy".  Being the European Union is awesome, simply because of
> how big it is, of course.  I like their principle, "Think Open".
>
> Kenneth Parker
>

The European Commission approved the new Open Source Software Strategy 2020-2023 of the Commission

[steve]

https://ec.europa.eu/info/departments/informatics/open-source-software-strategy_fr

Re: getmail, getmail6, testing, unstable, python-is-python3

[songbird]

Reco wrote:
...
> Because python. Be it python 2 or python 3 - it's (in)famous for one
> thing - lack of backwards compatibility. Sooner or later they replace
> python 3.8 (current sid) with, say, python 3.9 - and the things will
> break again.
> Because they did on 3.7->3.8 transition, did before that on 3.6->3.7
> transition and this list goes long way back.

  i've not had any problems with python3 here, but i 
don't do strange things with python3.  oh, let me
rephrase that, during a recent round i had a problem
with a program, but that wasn't anything i did so it
was somewhere in the stack of modules that program 
used.  does that mean it was incompatible?  no, it 
just meant there was a bug introduced which eventually 
got fixed.


  songbird

CIBERSEGURIDAD: La gran vencedora durante la pandemia

[Francisco Gamboa Ornelas]

LIVE TRAINING
ESPECIALISTA DE SEGURIDAD INFORMÁTICA CORPORATIVA 2021
- Amenazas avanzadas, nuevas herramientas y analítica forense - 

Fecha de inicio: 05 de Noviembre 2020


Debido al aumento de ataques informáticos, la demanda de expertos en seguridad 
informática creció 250% durante la pandemia. 

¡PROGRAMA TOTAL DE CIBERSEGURIDAD PARA EMPRESAS DEL HABLA HISPANA!


DESCARGAR FOLLETO COMPLETO

>>> Recibir toda la información vía WhatsApp aquí



¿Mayores informes? Comuníquese con alguno de nuestros ejecutivos:

CDMX / 55 2450 6187
MTY, N.L. / 81 2974 7731
GDL, JAL. / 33 2005 0994



















Este boletín informativo tiene como objetivo crear valor en usted y en su 
Organización. Si desea dejar de recibir este tipo de información conteste con 
la palabra BAJASEGURIDAD774.
O en su defecto haga click en el siguiente enlace: unsubscribe from this list

Re: Problema con bluetooth en Buster

[Camaleón]

El 2020-10-27 a las 16:35 +0100, Josu Lazkano escribió:

> El mar., 27 oct. 2020 a las 16:08, Camaleón () escribió:

> > > [5.176848] brcmfmac mmc0:0001:1: firmware: failed to load 
> > > brcm/brcmfmac43241b4-sdio.clm_blob (-2)
> > > [5.176859] firmware_class: See https://wiki.debian.org/Firmware for
> > information about missing firmware
> > > [5.176866] brcmfmac: brcmf_c_process_clm_blob: no clm_blob available
> > (err=-2), device may have limited channels available
> > > [5.177153] brcmfmac: brcmf_c_preinit_dcmds: Firmware: BCM4324/5 wl0:
> > Jul 17 2013 07:36:07 version 6.10.197.71 (r412987) FWID 01-882d2634
> > >
> > > En Google no aparece por ningun lado, es un poco raro.
> >
> > Creo que lo tienes en este paquete:
> >
> > https://packages.debian.org/buster/firmware-brcm80211
> >
> > Contents:
> > * Broadcom BCM43241 rev 4 firmware (brcm/brcmfmac43241b4-sdio.bin)
> >
> 
> En ese paquete tengo el fichero "brcmfmac43241b4-sdio.bin", pero no el
> "brcmfmac43241b4-sdio.clm_blob".
> Cambia la extensión.

Entonces ya tienes instalado ese paquete y te sigue apareciendo el 
mensaje ¿no?

Si no tiene ningún efecto secundario, más allá del error, seguramente 
ya lo habrán corregido en una nueva versión del kernel.

Me parece que se trata de esto:

brcmfmac: Use request_firmware_direct for the clm_blob
https://patchwork.kernel.org/project/linux-wireless/patch/20190107113401.6824-1-hdego...@redhat.com/

Saludos,

-- 
Camaleón 

Re: Replacement Email Client

[John Hasler]

Greg Wooledge wrote:
> [1]I used to read slashdot regularly, and on slashdot, the front page
> had a bunch of news stories and a poll.  The poll was written as a
> vanilla HTML form.  If you participated in the poll, it would send you
> to a new instance of the home page, because a form *must* load a new
> page.  Doing that would lose my place, showing a new set of stories,
> even if I hadn't finished reading the ones on the previous instance.

In Firefox hold down  while clicking the button and the new page
will open in a new tab.
-- 
John Hasler 
jhas...@newsguy.com
Elmwood, WI USA

Re: Replacement Email Client

[Joe]

On Tue, 27 Oct 2020 07:43:43 -0400
Greg Wooledge  wrote:

>
> 
> [1]I used to read slashdot regularly, and on slashdot, the front page
> had a bunch of news stories and a poll.  The poll was written as a
> vanilla HTML form.  If you participated in the poll, it would send you
> to a new instance of the home page, because a form *must* load a new
> page.  Doing that would lose my place, showing a new set of stories,
> even if I hadn't finished reading the ones on the previous instance.
>
It doesn't have to be like that. Nearly all of my web applications just
use the one page, though of course it does have to be reloaded after a
submit. Anything I want to be persistent, I need to arrange through
hidden controls, appearing as parameters in the reloaded page. If
someone is showing you a large number of random entries on a page, then
of course it may be too much trouble to do this, but it is certainly
possible.

I do use the very occasional smidgen of JS to replace things that have
been left out of HTML, such as making a radio button group invoke a
submit when changed, but on the whole I believe client-side scripting
to be the work of the devil. 

-- 
Joe

Re: OT: Crear paquetes deb

[Gonzalo Rivero]

El mar, 27-10-2020 a las 10:43 -0500, Ismael L. Donis Garcia escribió:
> Ante todo disculpen el off topic.
> 
> He realizado unos pequeños programas con Lazarus, y quisiera
> empaquetarlos 
> para que puedan ser instalados tanto en debian como en ubuntu, pero
> no tengo 
> la más mínima idea de como realizar esta tarea.
> 
> Alguien me podría dar una pequeña ayuda en este tema?
> 
> Como link o donde buscar tutoriales o cosas por el estilo.
> 
http://www.debian.org/doc
en particular: 
https://www.debian.org/doc/devel-manuals#packaging-tutorial

Re: Replacement Email Client

[Joe]

On Tue, 27 Oct 2020 07:43:43 -0400
Greg Wooledge  wrote:

> On Mon, Oct 26, 2020 at 03:16:21PM -0700, Patrick Bartek wrote:
> > The ones I respond to are known to me and are legit --
> > organizations, businesses, government agencies, etc. -- that I do
> > business with.  To respond, I must switch to a web browser, login to
> > my email account (like gmail), find that particular email, and
> > enter the requested data either by keyboard, drop-down menu,
> > buttons, etc., then SUBMIT it.  This happens all within the email.
> > I never get forwarded to another web site. I always stay on the web
> > mail page.  
> 
> If you don't get a new page, then it was not a vanilla HTML form
> submission.  Those *always* give you a new page, as defined by the
> form's action field.  That's why people stopped using them.[1]
> 
> What you're describing sounds more like a Javascript button made to
> look like a form submission.  Those can do *anything*.
> 
> [1]I used to read slashdot regularly, and on slashdot, the front page
> had a bunch of news stories and a poll.  The poll was written as a
> vanilla HTML form.  If you participated in the poll, it would send you
> to a new instance of the home page, because a form *must* load a new
> page.  Doing that would lose my place, showing a new set of stories,
> even if I hadn't finished reading the ones on the previous instance.
> 
> At some point I wished fervently for the option to middle-click the
> form submit button to open the form's action page in a new tab/window.
> That never happened, obviously because web browser developers do not
> have the same priorities that I have.  They've proven that many times.
> 
> Eventually it became a moot point, because I stopped reading slashdot,
> and because web page designers have stopped using HTML forms.  They're
> just too limiting.  It's all custom Javascript stuff now.
> 

Mounting a USB device

[Mick Ab]

If a filesystem in /etc/fstab has a noauto entry, can that filesystem only
be mounted manually using the mount command or
is there any chance that it will be automatically mounted by
usbmount ?

The filesystem is used in a USB port.

OT: Crear paquetes deb

[Ismael L. Donis Garcia]

Ante todo disculpen el off topic.

He realizado unos pequeños programas con Lazarus, y quisiera empaquetarlos 
para que puedan ser instalados tanto en debian como en ubuntu, pero no tengo 
la más mínima idea de como realizar esta tarea.


Alguien me podría dar una pequeña ayuda en este tema?

Como link o donde buscar tutoriales o cosas por el estilo.

Saludos
--
Ismael 

Re: Problema con bluetooth en Buster

[Josu Lazkano]

El mar., 27 oct. 2020 a las 16:08, Camaleón () escribió:

> El 2020-10-27 a las 15:46 +0100, Josu Lazkano escribió:
>
> > El mar., 29 sept. 2020 a las 13:20, Camaleón ()
> > escribió:
> >
> > > El 2020-09-29 a las 10:05 +0200, Josu Lazkano escribió:
> > >
> > > > He instalado Debian Buster un mini PC (
> > > > https://www.ecs.com.tw/ECSWebSite/Product/Product_LIVA/ES/LIVA),
> pero
> > > me da
> > > > problemas el bluetooth:
> > > >
> > > > [4.859123] Bluetooth: hci0: BCM: chip id 84
> > > > [4.859634] Bluetooth: hci0: BCM: features 0x0f
> > > > [4.860790] Bluetooth: hci0: BCM4324B3
> > > > [4.860797] Bluetooth: hci0: BCM4324B3 (002.004.006) build 
> > > > [4.876567] bluetooth hci0: firmware: failed to load
> > > brcm/BCM4324B3.hcd
>
> > > Revisa este bug de Ubuntu, y verifica que hayas cargado el firmware
> > > correcto y en la ubicación adecuada:
> > >
> > > Broadcom Bluetooth brcm/BCM.hcd firmware not found
> > > https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1743336
>
> (...)
>
> > La mayoria de los errores de firmware han desaparecido, pero todavía me
> > falta el fichero "brcmfmac43241b4-sdio.clm_blob" que no lo encuentro por
> > ningun lado:
> >
> > [5.155659] Bluetooth: hci0: BCM: chip id 84
> > [5.156196] Bluetooth: hci0: BCM: features 0x0f
> > [5.157324] Bluetooth: hci0: BCM4324B3
> > [5.157329] Bluetooth: hci0: BCM4324B3 (002.004.006) build 
> > [5.158058] bluetooth hci0: firmware: direct-loading firmware
> brcm/BCM4324B3.hcd
> > [5.176800] brcmfmac: brcmf_fw_alloc_request: using
> brcm/brcmfmac43241b4-sdio for chip BCM4324/5
> > [5.176848] brcmfmac mmc0:0001:1: firmware: failed to load
> brcm/brcmfmac43241b4-sdio.clm_blob (-2)
> > [5.176859] firmware_class: See https://wiki.debian.org/Firmware for
> information about missing firmware
> > [5.176866] brcmfmac: brcmf_c_process_clm_blob: no clm_blob available
> (err=-2), device may have limited channels available
> > [5.177153] brcmfmac: brcmf_c_preinit_dcmds: Firmware: BCM4324/5 wl0:
> Jul 17 2013 07:36:07 version 6.10.197.71 (r412987) FWID 01-882d2634
> >
> > En Google no aparece por ningun lado, es un poco raro.
>
> Creo que lo tienes en este paquete:
>
> https://packages.debian.org/buster/firmware-brcm80211
>
> Contents:
> * Broadcom BCM43241 rev 4 firmware (brcm/brcmfmac43241b4-sdio.bin)
>
> Saludos
>
> --
> Camaleón
>
>
Gracias de nuevo,

En ese paquete tengo el fichero "brcmfmac43241b4-sdio.bin", pero no el
"brcmfmac43241b4-sdio.clm_blob".
Cambia la extensión.

Un saludo.

-- 
Josu Lazkano

Re: getmail, getmail6, testing, unstable, python-is-python3

[Cindy Sue Causey]

On 10/27/20, Reco  wrote:
> On Tue, Oct 27, 2020 at 10:45:23AM -0400, songbird wrote:
>> Reco wrote:
>> ...
>> > Indeed. Switch back to fetchmail, because the less you're depending on
>> > python and the software that uses it - the better ;)
>>
>>   i never used fetchmail so that wouldn't be a
>> "switching back" and since i already have my setup
>> working as i want for getmail6 why would i bother
>> doing more work for no gain?
>
> Because python. Be it python 2 or python 3 - it's (in)famous for one
> thing - lack of backwards compatibility. Sooner or later they replace
> python 3.8 (current sid) with, say, python 3.9 - and the things will
> break again.
> Because they did on 3.7->3.8 transition, did before that on 3.6->3.7
> transition and this list goes long way back.


Hi.. Am posting this because it just came through my inbox and feels
like it's appropriate. It's from the Debian-Python list (YESTERDAY)
and says, "Re: help with FTBFS bug due to python3.8/3.9 confusion":

https://lists.debian.org/debian-python/2020/10/msg00066.html

The original post is a little lengthy and might provide some
commiseration that's of interest. Being a professional lurker, I'm
fully in agreement that this is an ongoing deal the last few years.

Between this and the GTK2 v. GTK3 deal, Developers do get frustrated
when it seems like they're constantly chasing breakage issues. When
messages pop about "abandonware" type issues, about the first thing I
peek at (out of curiosity) is whether Python or GTK is part of the
packaging. :)

Cindy :)
-- 
Cindy-Sue Causey
Talking Rock, Pickens County, Georgia, USA

* runs with birdseed *

Re: Problema con bluetooth en Buster

[Camaleón]

El 2020-10-27 a las 15:46 +0100, Josu Lazkano escribió:

> El mar., 29 sept. 2020 a las 13:20, Camaleón ()
> escribió:
> 
> > El 2020-09-29 a las 10:05 +0200, Josu Lazkano escribió:
> >
> > > He instalado Debian Buster un mini PC (
> > > https://www.ecs.com.tw/ECSWebSite/Product/Product_LIVA/ES/LIVA), pero
> > me da
> > > problemas el bluetooth:
> > >
> > > [4.859123] Bluetooth: hci0: BCM: chip id 84
> > > [4.859634] Bluetooth: hci0: BCM: features 0x0f
> > > [4.860790] Bluetooth: hci0: BCM4324B3
> > > [4.860797] Bluetooth: hci0: BCM4324B3 (002.004.006) build 
> > > [4.876567] bluetooth hci0: firmware: failed to load
> > brcm/BCM4324B3.hcd

> > Revisa este bug de Ubuntu, y verifica que hayas cargado el firmware
> > correcto y en la ubicación adecuada:
> >
> > Broadcom Bluetooth brcm/BCM.hcd firmware not found
> > https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1743336
 
(...)

> La mayoria de los errores de firmware han desaparecido, pero todavía me
> falta el fichero "brcmfmac43241b4-sdio.clm_blob" que no lo encuentro por
> ningun lado:
> 
> [5.155659] Bluetooth: hci0: BCM: chip id 84
> [5.156196] Bluetooth: hci0: BCM: features 0x0f
> [5.157324] Bluetooth: hci0: BCM4324B3
> [5.157329] Bluetooth: hci0: BCM4324B3 (002.004.006) build 
> [5.158058] bluetooth hci0: firmware: direct-loading firmware 
> brcm/BCM4324B3.hcd
> [5.176800] brcmfmac: brcmf_fw_alloc_request: using 
> brcm/brcmfmac43241b4-sdio for chip BCM4324/5
> [5.176848] brcmfmac mmc0:0001:1: firmware: failed to load 
> brcm/brcmfmac43241b4-sdio.clm_blob (-2)
> [5.176859] firmware_class: See https://wiki.debian.org/Firmware for 
> information about missing firmware
> [5.176866] brcmfmac: brcmf_c_process_clm_blob: no clm_blob available 
> (err=-2), device may have limited channels available
> [5.177153] brcmfmac: brcmf_c_preinit_dcmds: Firmware: BCM4324/5 wl0: Jul 
> 17 2013 07:36:07 version 6.10.197.71 (r412987) FWID 01-882d2634
> 
> En Google no aparece por ningun lado, es un poco raro.

Creo que lo tienes en este paquete:

https://packages.debian.org/buster/firmware-brcm80211

Contents:
* Broadcom BCM43241 rev 4 firmware (brcm/brcmfmac43241b4-sdio.bin)

Saludos

-- 
Camaleón 

Re: The .xsession-errors problem

[Teemu Likonen]

* 2020-10-26 20:04:55+03, Reco wrote:

> On Mon, Oct 26, 2020 at 06:35:45PM +0200, Teemu Likonen wrote:
>>   - Do you configure some rotating system, perhaps with logrotate(8)?
>> (Why doesn't Debian have this automatically?)
>
> For Debian, it may work. For RHEL, for instance, such logrotate policy
> would be denied by SELinux.
> That, and inviting running-as-root logrotate to cleanup user files opens
> all kinds of trouble.

I'm using KDE Plasma desktop and my .xsession-errors grows quite fast.
I'll probably write some rotation system for the file. So far the
simplest seems to be adding /etc/logrotate.d/my-xession-errors with
contents like below. Logrotate uses the same owner and permissions as
the original file. Nothing else is needed.

/home/*/.xsession-errors {
rotate 3
monthly
compress
notifempty
}

Logrotate could be used as normal user. For my personal system it's
probably too complicated for such a small thing but there could be
configuration file like this:

# Xsession-logrotate.conf
~/.xsession-errors {
rotate 3
monthly
compress
notifempty
}

Xsession script could run a command like this:

/usr/sbin/logrotate \
--state "$HOME/.local/var/lib/logrotate/status" \
/etc/X11/Xsession-logrotate.conf

On the other hand all this is probably too complicated because the
.xsession-errors file is not that interesting. People have small
additions in their /etc/X11/Xsession script: delete the file every time
or move the old file to ".old". That would be good enough, I think.

-- 
/// Teemu Likonen - .-.. https://www.iki.fi/tlikonen/
// OpenPGP: 4E1055DC84E9DFF613D78557719D69D324539450


signature.asc
Description: PGP signature

Re: getmail, getmail6, testing, unstable, python-is-python3

[Reco]

On Tue, Oct 27, 2020 at 10:45:23AM -0400, songbird wrote:
> Reco wrote:
> ...
> > Indeed. Switch back to fetchmail, because the less you're depending on
> > python and the software that uses it - the better ;)
> 
>   i never used fetchmail so that wouldn't be a
> "switching back" and since i already have my setup 
> working as i want for getmail6 why would i bother 
> doing more work for no gain?

Because python. Be it python 2 or python 3 - it's (in)famous for one
thing - lack of backwards compatibility. Sooner or later they replace
python 3.8 (current sid) with, say, python 3.9 - and the things will
break again.
Because they did on 3.7->3.8 transition, did before that on 3.6->3.7
transition and this list goes long way back.

Reco

Re: Problema con bluetooth en Buster

[Josu Lazkano]

El mar., 29 sept. 2020 a las 13:20, Camaleón ()
escribió:

> El 2020-09-29 a las 10:05 +0200, Josu Lazkano escribió:
>
> > He instalado Debian Buster un mini PC (
> > https://www.ecs.com.tw/ECSWebSite/Product/Product_LIVA/ES/LIVA), pero
> me da
> > problemas el bluetooth:
> >
> > [4.859123] Bluetooth: hci0: BCM: chip id 84
> > [4.859634] Bluetooth: hci0: BCM: features 0x0f
> > [4.860790] Bluetooth: hci0: BCM4324B3
> > [4.860797] Bluetooth: hci0: BCM4324B3 (002.004.006) build 
> > [4.876567] bluetooth hci0: firmware: failed to load
> brcm/BCM4324B3.hcd
> > (-2)
>
> (...)
>
> > Creo que he metido bien el firmware, ¿me podéis ayudar con esto?
>
> Revisa este bug de Ubuntu, y verifica que hayas cargado el firmware
> correcto y en la ubicación adecuada:
>
> Broadcom Bluetooth brcm/BCM.hcd firmware not found
> https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1743336
>
> Saludos,
>
> --
> Camaleón
>
>
Gracias Camaleón,

He seguido estos pasos:

apt-get install firmware-brcm80211
cd /lib/firmware/brcm/
wget https://github.com/Asus-T100/firmware/raw/master/brcm/BCM4324B3.hcd
wget
https://github.com/Asus-T100/firmware/raw/master/brcm/brcmfmac43241b4-sdio.txt

La mayoria de los errores de firmware han desaparecido, pero todavía me
falta el fichero "brcmfmac43241b4-sdio.clm_blob" que no lo encuentro por
ningun lado:

[5.155659] Bluetooth: hci0: BCM: chip id 84
[5.156196] Bluetooth: hci0: BCM: features 0x0f
[5.157324] Bluetooth: hci0: BCM4324B3
[5.157329] Bluetooth: hci0: BCM4324B3 (002.004.006) build 
[5.158058] bluetooth hci0: firmware: direct-loading firmware
brcm/BCM4324B3.hcd
[5.176800] brcmfmac: brcmf_fw_alloc_request: using
brcm/brcmfmac43241b4-sdio for chip BCM4324/5
[5.176848] brcmfmac mmc0:0001:1: firmware: failed to load
brcm/brcmfmac43241b4-sdio.clm_blob (-2)
[5.176859] firmware_class: See https://wiki.debian.org/Firmware for
information about missing firmware
[5.176866] brcmfmac: brcmf_c_process_clm_blob: no clm_blob available
(err=-2), device may have limited channels available
[5.177153] brcmfmac: brcmf_c_preinit_dcmds: Firmware: BCM4324/5 wl0:
Jul 17 2013 07:36:07 version 6.10.197.71 (r412987) FWID 01-882d2634

En Google no aparece por ningun lado, es un poco raro.

Sigo buscando.

Un saludo.

-- 
Josu Lazkano

Re: getmail, getmail6, testing, unstable, python-is-python3

[Reco]

Hi.

On Tue, Oct 27, 2020 at 10:19:01AM -0400, Celejar wrote:
> On Tue, 27 Oct 2020 17:02:22 +0300
> Reco  wrote:
> 
> > Hi.
> > 
> > On Tue, Oct 27, 2020 at 01:31:19PM +, mick crane wrote:
> > > >   this was just a quick heads-up for those who are stuck
> > > > on getmail like i am (and quite happy with it).  :)
> > > > 
> > > 
> > > as far as getmail goes maintainer thinks is an unnecessary panic.
> > > 
> > > ""getmail goes out of official support by my distro" may be a theoretical
> > > problem, or a philosophical one, but it it certainly is not a significant
> > > practical problem.
> > 
> > Indeed. Switch back to fetchmail, because the less you're depending on
> > python and the software that uses it - the better ;)
> 
> Here's the getmail author's opinion of why getmail is preferable to
> fetchmail:
> 
> http://pyropus.ca/software/getmail/faq.html#faq-about-why
> 
> Doubtless opinionated, and certainly dated, but would you or anyone
> else here care to comment?

It boils down to two things:

1) Configuration of fetchmail is teh hard.

If I have to choose between hard-to-configure software and will-cease-to
function software - I always go with the first variety. YMMV.


2) Fetchmail is insecure, getmail is bulletproof.

As [1] and [2] show us - it's true somewhat. fetchmail has 5 times more
known vulnerabilities than getmail.
Problem with such numbers approach is - last reported CVE for fetchmail
is dated 2012, and for getmail it's 2014. I.e. both can be considered
secure enough in this regard.

CVE-2020-5239 - [1] - corresponds to some *person* (let's put it this
way) who apparently thought that putting outdated fetchmail in docker
along with the unspecified Agile/Scrumm-level quality "fetchmail script"
will make things secure by some magic.
A morale of the story here - running a random docker image is comparable
to running a random binary downloaded from the Internet as far as
security concerned.

And my favorite:
"getmail users have not had to worry about any of these security holes
or design and implementation errors".
Instead getmail users have to worry about [3]. It's not php-level mess -
[4], but venerable nevertheless.

Reco

[1] https://cve.mitre.org/cgi-bin/cvekey.cgi?keyword=getmail
[2] https://cve.mitre.org/cgi-bin/cvekey.cgi?keyword=fetchmail
[3] https://cve.mitre.org/cgi-bin/cvekey.cgi?keyword=python
[4] https://cve.mitre.org/cgi-bin/cvekey.cgi?keyword=php

Re: getmail, getmail6, testing, unstable, python-is-python3

[songbird]

Reco wrote:
...
> Indeed. Switch back to fetchmail, because the less you're depending on
> python and the software that uses it - the better ;)

  i never used fetchmail so that wouldn't be a
"switching back" and since i already have my setup 
working as i want for getmail6 why would i bother 
doing more work for no gain?

  i do use python3 so python won't be gone from this
machine entirely anyways.


  songbird

Re: La Comisión Europea aprueba la nueva Estrategia de Software de Código Abierto 2020-2023

[Kenneth Parker]

On Tue, Oct 27, 2020, 6:29 AM Weaver  wrote:

> On 27-10-2020 20:01, Iker Bilbao wrote:
> >
> https://www.casadomo.com/2020/10/27/comision-europea-aprueba-nueva-estrategia-software-codigo-abierto-2020-2023
>
> Bravo!
> --


Thank you Weaver.  I had ignored the original email, due to it being
"English Challenged", but you inspired me to click the link and let Google
translate it.

Good article, about the European Commission embracing a "new Open Source
Software Strategy".  Being the European Union is awesome, simply because of
how big it is, of course.  I like their principle, "Think Open".

Kenneth Parker

Re: La Comisión Europea aprueba la nueva Estrategia de Software de Código Abierto 2020-2023

[Stefan Monnier]

>> > https://www.casadomo.com/2020/10/27/comision-europea-aprueba-nueva-estrategia-software-codigo-abierto-2020-2023
> Source (in English):
> https://ec.europa.eu/info/departments/informatics/open-source-software-strategy_en

Nice!

Regarding:
> 2. The Commission shall ensure a level playing field for open source
>software and demonstrate an active and fair consideration of using
>open source software – taking account of the total cost of
>ownership of the solution.

I'm afraid this won't take into account the indirect benefits of the
cost:

There's *spending* $10M on a software provided by a company
headquartered in a country outside the EU, and then there's *investing*
$10M into a software developed within the EU.

While it's not directly related to the difference between proprietary
and Free Software, with Free Software, you can almost always arrange for
the development to be done locally, whereas with proprietary software
it's usually out of the customer's hands.


Stefan

Re: getmail, getmail6, testing, unstable, python-is-python3

[Celejar]

On Tue, 27 Oct 2020 17:02:22 +0300
Reco  wrote:

>   Hi.
> 
> On Tue, Oct 27, 2020 at 01:31:19PM +, mick crane wrote:
> > >   this was just a quick heads-up for those who are stuck
> > > on getmail like i am (and quite happy with it).  :)
> > > 
> > 
> > as far as getmail goes maintainer thinks is an unnecessary panic.
> > 
> > ""getmail goes out of official support by my distro" may be a theoretical
> > problem, or a philosophical one, but it it certainly is not a significant
> > practical problem.
> 
> Indeed. Switch back to fetchmail, because the less you're depending on
> python and the software that uses it - the better ;)

Here's the getmail author's opinion of why getmail is preferable to
fetchmail:

http://pyropus.ca/software/getmail/faq.html#faq-about-why

Doubtless opinionated, and certainly dated, but would you or anyone
else here care to comment?

Celejar

Re: La Comisión Europea aprueba la nueva Estrategia de Software de Código Abierto 2020-2023

[Andrei POPESCU]

On Ma, 27 oct 20, 03:29:13, Weaver wrote:
> On 27-10-2020 20:01, Iker Bilbao wrote:
> > https://www.casadomo.com/2020/10/27/comision-europea-aprueba-nueva-estrategia-software-codigo-abierto-2020-2023
> 
> Bravo!

Source (in English):
https://ec.europa.eu/info/departments/informatics/open-source-software-strategy_en

Kind regards,
Andrei
-- 
http://wiki.debian.org/FAQsFromDebianUser


signature.asc
Description: PGP signature

Re: getmail, getmail6, testing, unstable, python-is-python3

[Reco]

Hi.

On Tue, Oct 27, 2020 at 01:31:19PM +, mick crane wrote:
> >   this was just a quick heads-up for those who are stuck
> > on getmail like i am (and quite happy with it).  :)
> > 
> 
> as far as getmail goes maintainer thinks is an unnecessary panic.
> 
> ""getmail goes out of official support by my distro" may be a theoretical
> problem, or a philosophical one, but it it certainly is not a significant
> practical problem.

Indeed. Switch back to fetchmail, because the less you're depending on
python and the software that uses it - the better ;)

> Keep a Python 2 binary around, download getmail, untar it,
> and run it right out of the directory - that's literally all that's required."
> 
> what exactly the python 2 binary is I haven't had need to look as yet.

/usr/bin/python, currently provided by "python-minimal" package.

Reco

Re: What's going on with snapd?

[Andrei POPESCU]

On Lu, 26 oct 20, 15:24:24, Brian Vaughan wrote:
> Recently I wanted to try out an open source game that was being distributed
> as a snap package, so I tried installing snapd. It apparently installed
> successfully, and I could apparently install snap packages, but I couldn't
> execute them. Long story short, I eventually made my way to bugs.debian.org
> and found that most of the issues I'd seen with snapd, had bugs filed
> against them.
> 
> https://bugs.debian.org/cgi-bin/pkgreport.cgi?dist=unstable;package=snapd
> 
> However, as far as I can tell, most of these bugs have been filed two to
> three years ago, and no developer has responded to them. Meanwhile, the
> package has been updated a few times since then.

What release are you running (stable, testing, unstable, etc.)? The 
package in stable will only receive security updates.
 
> snapd is a Canonical project, and from what I've heard from an acquaintance
> who works with Canonical, they use snapd internally quite a lot. So it
> surprises me that these bugs have gotten no attention. Do the messages about
> bug reports just not get through? Is it just a Debian problem?

Please provide some example for the bugs you're interested in.

Kind regards,
Andrei
-- 
http://wiki.debian.org/FAQsFromDebianUser


signature.asc
Description: PGP signature

Re: getmail, getmail6, testing, unstable, python-is-python3

[songbird]

mick crane wrote:
...
> as far as getmail goes maintainer thinks is an unnecessary panic.
>
> ""getmail goes out of official support by my distro" may be a 
> theoretical
> problem, or a philosophical one, but it it certainly is not a 
> significant
> practical problem.  Keep a Python 2 binary around, download getmail, 
> untar it,
> and run it right out of the directory - that's literally all that's 
> required."
>
> what exactly the python 2 binary is I haven't had need to look as yet.

  note that is his opinion, the debian python and getmail 
packaging workers have decided otherwise and now the
original getmail package is no longer compatible in
testing with other packages so it is removed.  i tried to
use my own version but there was a problem with the
python install missing some module so i gave up and then
i found the python6 package and that took care of the
whole problem.

  the newer getmail6 package is based upon getmail, but
it is compatible with python3.  that is all i care about, 
it works, i don't have to mess around with yet another
dangling bit of stuff to keep after (that's why i use a
distribution and not roll my own after all - i have 
enough of that already).


  songbird

Re: The .xsession-errors problem

[Andrei POPESCU]

On Ma, 27 oct 20, 07:55:00, Greg Wooledge wrote:
> On Mon, Oct 26, 2020 at 11:07:37PM +, Tixy wrote:
> > On Mon, 2020-10-26 at 18:35 +0200, Teemu Likonen wrote:
> > > It seems that ~/.xsession-errors file can still grow to infinity in
> > > size. Sometimes it grows really fast. This is nothing new: we have all
> > > seen it and talked about it. What do you do to maintain this file?
> > 
> > Don't do anything here. The file is created fresh at each boot and is
> > 30 lines long [...]
> 
> Something that you're doing, or something that was done for you, is
> clearing that file.  Your case is not the default.  By default, that
> file is never cleared, and just keeps growing.  Most people prune it
> manually whenever they notice it getting bigger than they like, which
> is usually somewhere between "once a year" and "never".
 
On my system the file is rotated (renamed to .xsession-errors.old), on 
every login as far as I can tell.

Didn't find (yet) what is doing this (using lightdm, LXDE and minimal 
Xorg).

Kind regards,
Andrei
-- 
http://wiki.debian.org/FAQsFromDebianUser


signature.asc
Description: PGP signature

Re: getmail, getmail6, testing, unstable, python-is-python3

[mick crane]

On 2020-10-27 13:08, songbird wrote:

well, yesterday finally became the day when i tried to
see if i could get the old getmail to be left alone, but
nope, none of the options worked.

  the good news, is that getmail6 is in unstable and was
a drop in replacement for getmail.  there's a bit of an
odd message coming out of it but that's not an error and
everything is ending up where it should.

  during all of this fiddling i was wondering if the
python2 to python3 transition was complete enough that
i could run all my programs that i use and it all works
just fine.  so i installed python-is-python3 and the
system rebooted and none of things i use the most have
failed.  that's all i need from a system (browswer, e-mail,
usenet, gimp, go, libreoffice).  i've never done much
with python2 and the game i wrote was python3 so that's
all to the good.

  this was just a quick heads-up for those who are stuck
on getmail like i am (and quite happy with it).  :)



as far as getmail goes maintainer thinks is an unnecessary panic.

""getmail goes out of official support by my distro" may be a 
theoretical
problem, or a philosophical one, but it it certainly is not a 
significant
practical problem.  Keep a Python 2 binary around, download getmail, 
untar it,
and run it right out of the directory - that's literally all that's 
required."


what exactly the python 2 binary is I haven't had need to look as yet.

mick

--
Key ID4BFEBB31

getmail, getmail6, testing, unstable, python-is-python3

[songbird]

  well, yesterday finally became the day when i tried to
see if i could get the old getmail to be left alone, but
nope, none of the options worked.

  the good news, is that getmail6 is in unstable and was
a drop in replacement for getmail.  there's a bit of an
odd message coming out of it but that's not an error and
everything is ending up where it should.

  during all of this fiddling i was wondering if the 
python2 to python3 transition was complete enough that
i could run all my programs that i use and it all works
just fine.  so i installed python-is-python3 and the
system rebooted and none of things i use the most have
failed.  that's all i need from a system (browswer, e-mail,
usenet, gimp, go, libreoffice).  i've never done much 
with python2 and the game i wrote was python3 so that's
all to the good.

  this was just a quick heads-up for those who are stuck 
on getmail like i am (and quite happy with it).  :)


  songbird

Re: aptitude safe-upgrade vs apt-get upgrade.

[Andrei POPESCU]

On Lu, 26 oct 20, 09:55:00, John Hasler wrote:
> Andrei writes:
> > dpkg does its own dependency checking, in addition to APT (the
> > software, not the command), and will prevent any inconsistencies
> > unless you use one of the --force switches.
> 
> What it does not do is resolve dependencies.  Apt recursively resolves
> dependencies, installing them as required. It also detects conflicts
> and offers to resolve them as well as breaking loops.

dpkg can only work with the set of .deb files that were passed on the 
command line.

If all dependencies are included (or already installed), fine, otherwise 
it will bail out as it doesn't (by design) have the capability to search 
for them in repositories and download them (if this is what you mean by 
resolving).

I believe someone demonstrated quite recently on list that dpkg has some 
limits in the number and/or combination of packages it can deal with at 
once, so APT might have to pass them in smaller chunks and/or specific 
order (in case of Pre-Depends: maybe?).
 
> Dpkg is safe but can be rather frustrating.

As far as I'm concerned it does just fine what it was designed to do.

Kind regards,
Andrei
-- 
http://wiki.debian.org/FAQsFromDebianUser


signature.asc
Description: PGP signature

Re: Replacement Email Client

[Eric S Fraga]

On Sunday, 25 Oct 2020 at 10:36, Patrick Bartek wrote:
> On Sun, 25 Oct 2020 07:19:28 +0200
> Teemu Likonen  wrote:
>> GNU Emacs mail clients "Gnus" and "Notmuch Emacs" automatically render
>> HTML mail nicely as plain text. User can can also open HTML and other
>> MIME parts in external viewer like web browser.
>> 
>
> I can already view the text of HTML emails with Claws-Mail.  But I need
> to view the entire email: images, graphics, etc. and be able to interact
> with all the links, etc. and not just view them.  Want to get away from
> having to login to the mail account with a browser to do so.  So, EMACS
> won't work for me.

The emacs gnus HTML support is not just for viewing: you can interact as
well.  I do receive HTML emails, unfortunately, and can process most if
not all of them within Emacs.  But I've never received one with an
embedded form to submit (I wouldn't respond to such in any case).

-- 
Eric S Fraga via Emacs 28.0.50 & org 9.4 on Debian bullseye/sid

Re: PC wont boot to GUI

[Greg Wooledge]

On Tue, Oct 27, 2020 at 01:18:42PM +1100, David wrote:
> You need to install the firmware.
> You do that by running this command as superuser:
> 
> # apt install firmware-amd-graphics

You may have to add "contrib" and "non-free" to your sources.list first,
and then run "apt update".

See  if
you need instructions for editing the file.

Re: The .xsession-errors problem

[Greg Wooledge]

On Mon, Oct 26, 2020 at 11:07:37PM +, Tixy wrote:
> On Mon, 2020-10-26 at 18:35 +0200, Teemu Likonen wrote:
> > It seems that ~/.xsession-errors file can still grow to infinity in
> > size. Sometimes it grows really fast. This is nothing new: we have all
> > seen it and talked about it. What do you do to maintain this file?
> 
> Don't do anything here. The file is created fresh at each boot and is
> 30 lines long [...]

Something that you're doing, or something that was done for you, is
clearing that file.  Your case is not the default.  By default, that
file is never cleared, and just keeps growing.  Most people prune it
manually whenever they notice it getting bigger than they like, which
is usually somewhere between "once a year" and "never".

Re: The .xsession-errors problem

[Greg Wooledge]

On Tue, Oct 27, 2020 at 11:28:12AM +1100, David wrote:
> On Tue, 27 Oct 2020 at 10:56, David Wright  wrote:
> >   fuser -v "$j"
> >   [ $? -ne 0 ] && gzip "$j" && mv -i "$j.gz" "$HOME/.monitors/xsession/"
> 
> > (Script improvements always appreciated.)

> https://www.shellcheck.net says:
> """
> Line 2:
> [ $? -ne 0 ] && echo hi
>   ^-- SC2181: Check exit code directly with e.g. 'if mycmd;', not
> indirectly with $?.
> """

In other words, instead of writing:

fuser -v thing
[ $? != 0 ] && other thing && third thing

You can write it this way:

if ! fuser -v thing; then
  other thing &&
  third thing
fi

The $? form isn't technically *wrong*, but it's awkward, and doesn't
let you have an "else" action.

Re: Replacement Email Client

[Greg Wooledge]

On Tue, Oct 27, 2020 at 09:11:42AM +0100, to...@tuxteam.de wrote:
> We still don't know whether
> 
> - those forms are plain old HTML forms of yore with a "classical"
>   SUBMIT action
> - or they are some AJAX-y abomination in which a piece of Javascript
>   plays ping-pong with the server
> 
> In the first case I'm pretty sure you could teach your Claws/Dillo combo
> to cope with it. In the second case... I'd switch providers (I'm going
> to do this with my gas provider for a similar reason).

Or just ignore the survey.

Re: Replacement Email Client

[Greg Wooledge]

On Mon, Oct 26, 2020 at 03:16:21PM -0700, Patrick Bartek wrote:
> The ones I respond to are known to me and are legit --
> organizations, businesses, government agencies, etc. -- that I do
> business with.  To respond, I must switch to a web browser, login to
> my email account (like gmail), find that particular email, and
> enter the requested data either by keyboard, drop-down menu, buttons,
> etc., then SUBMIT it.  This happens all within the email. I never get
> forwarded to another web site. I always stay on the web mail page.

If you don't get a new page, then it was not a vanilla HTML form
submission.  Those *always* give you a new page, as defined by the
form's action field.  That's why people stopped using them.[1]

What you're describing sounds more like a Javascript button made to look
like a form submission.  Those can do *anything*.

[1]I used to read slashdot regularly, and on slashdot, the front page
had a bunch of news stories and a poll.  The poll was written as a
vanilla HTML form.  If you participated in the poll, it would send you
to a new instance of the home page, because a form *must* load a new
page.  Doing that would lose my place, showing a new set of stories,
even if I hadn't finished reading the ones on the previous instance.

At some point I wished fervently for the option to middle-click the
form submit button to open the form's action page in a new tab/window.
That never happened, obviously because web browser developers do not
have the same priorities that I have.  They've proven that many times.

Eventually it became a moot point, because I stopped reading slashdot,
and because web page designers have stopped using HTML forms.  They're
just too limiting.  It's all custom Javascript stuff now.

Re: La Comisión Europea aprueba la nueva Estrategia de Software de Código Abierto 2020-2023

[Weaver]

On 27-10-2020 20:01, Iker Bilbao wrote:
> https://www.casadomo.com/2020/10/27/comision-europea-aprueba-nueva-estrategia-software-codigo-abierto-2020-2023

Bravo!
-- 
`One day the great European War will come out of some damned foolish
thing in the Balkans' (1888).   
  -- Otto von Bismarck

La Comisión Europea aprueba la nueva Estrategia de Software de Código Abierto 2020-2023

[Iker Bilbao]

https://www.casadomo.com/2020/10/27/comision-europea-aprueba-nueva-estr
ategia-software-codigo-abierto-2020-2023

Re: Pourquoi n'y a-t-il pas FreeCad 0.19 en testing ou sid ?

[benoit]

Sent with ProtonMail Secure Email.

‐‐‐ Original Message ‐‐‐
Le mardi 27 octobre 2020 09:41, BERTRAND Joël  a 
écrit :

>
> FreeCAD est un logiciel un peu tordu à compiler. J'utilise cet outil en
> rolling release et il y a des dépendances qui ne sont pas dans Debian
> avec quelques incompatibilité entre OCC et OCE, sans compter d'autres
> bizarreries avec vtk et consorts qui dépendent aussi des deux larrons
> précédents.
>
> Pour utiliser Kicad (5.99) et FreeCAD (0.19) sur la même machine, je me
> suis résolu à compiler les deux depuis les sources - en me faisant un
> noeud au cerveau pour que les deux utilisent les mêmes bibliothèques -
> pour éviter les problèmes.

Déjà en buster c'est une ancienne version qui n'est même pas la stable actuelle.
Donc oui, le seul moyen d'utiliser FreeCad sous debian en profitant des 
dernières avancées c'est de compiler depuis les sources.

Effectivement à une certaine époque j’ai dû aussi compiler des dépendances 
(pivy, soqt, coin…), mais vu que c’est sur mon système ainsi qu’une longue 
liste de paquets dépendants, je ne sais plus très bien ce qui est en paquet 
debian ou pas et ce qui est vraiment nécessaire, je compile sans me poser de 
question.

Mon prochain objectif sera de faire le tri dans tout ça et apprendre à 
construire un paquet debian de freecad 0.19 pour avoir une routine de 
compilation/installation maîtrisée. Sauf si j’apprends que la version de 
développement de freeecad (0.19 en ce moment) est dispo en paquet debian, d’où 
ma question...

--
Benoit

Re: Replacement Email Client

[Curt]

On 2020-10-27, Carl Fink  wrote:
>
> Meaning no offense, I doubt it. I have been using email since before Gopher,
> and I have literally never received an HTML  email with an embedded form
> (that I opened, at least). I find it hard to believe that you get many 
> of them.
>

>From what I've read html forms can be embedded in an email (and probably
have), but the chances of the recipient being able to submit them with
success from within his or her email client are problematic (sporadic
support is apparently the consecrated expression).

It seems like a corner case (and a very tiny corner at that). Maybe the
OP could show us the html code in question and lay all vaporous
speculation to rest (although it probably ain't worth the trouble ((but
then again we're doubtless soon to be confined here once again, and any
remedy for boredom might do, as the bars and bistros and brasseries may
be closed until further notice)).

Re: What's going on with snapd?

[didier gaumet]

Hello,

Did you follow these steps?
 https://snapcraft.io/docs/installing-snap-on-debian

Re: Pourquoi n'y a-t-il pas FreeCad 0.19 en testing ou sid ?

[BERTRAND Joël]

benoit a écrit :
> Bonjour,
> 
> La version stable de FreeCad est la : 0.18.4 et est empaquetée dans
> cette version aussi bien en testing qu'en sid.
> 
> Pourquoi la 0.19 qui est la version de développement n'est-elle pas
> disponible en testing ou en sid ?

Bonjour,

FreeCAD est un logiciel un peu tordu à compiler. J'utilise cet outil en
rolling release et il y a des dépendances qui ne sont pas dans Debian
avec quelques incompatibilité entre OCC et OCE, sans compter d'autres
bizarreries avec vtk et consorts qui dépendent aussi des deux larrons
précédents.

Pour utiliser Kicad (5.99) et FreeCAD (0.19) sur la même machine, je me
suis résolu à compiler les deux depuis les sources - en me faisant un
noeud au cerveau pour que les deux utilisent les mêmes bibliothèques -
pour éviter les problèmes.

Bien cordialement,

JKB

Pourquoi n'y a-t-il pas FreeCad 0.19 en testing ou sid ?

[benoit]

Bonjour,

La version stable de FreeCad est la : 0.18.4 et est empaquetée dans cette 
version aussi bien en testing qu'en sid.

Pourquoi la 0.19 qui est la version de développement n'est-elle pas disponible 
en testing ou en sid ?

--
Benoit

Sent with [ProtonMail](https://protonmail.com) Secure Email.

Re: Replacement Email Client

[tomas]

On Mon, Oct 26, 2020 at 03:42:24PM -0700, Patrick Bartek wrote:

[...]

> Okay.  Here's a trivial one to keep it simple:
> 
> Recently I went to my bank in person. A few days later I get an HTML
> email [...]

I think we won't advance unless you look into that HTML.

(If you post it here: make sure to obliterate whatever might seem
confidential).

We still don't know whether

- those forms are plain old HTML forms of yore with a "classical"
  SUBMIT action
- or they are some AJAX-y abomination in which a piece of Javascript
  plays ping-pong with the server

In the first case I'm pretty sure you could teach your Claws/Dillo combo
to cope with it. In the second case... I'd switch providers (I'm going
to do this with my gas provider for a similar reason).

Cheers
 - t


signature.asc
Description: Digital signature

Pulp pour gérer un dépôt Debian privé

[Olivier]

Bonjour,

Je viens de découvrir [1].

Qui a déjà utilisé ce logiciel pour gérer un dépôt privé Debian ?
Quel retour d'expérience ?

[1] https://opensource.com/article/20/10/pulp-debian

Slts