This morning, we are officially publishing the User Data Manifesto 2.0.

Today, most of users’ data are not stored on their computer’s hard drive any more, but rather online on a service provider’s server somewhere in a data center.

While most computing happened on local machines, in the late years a new kind of “computing“ has emerged in daily use. Marketers have called this “Cloud computing”—but do not mistake, as there is no cloud, it’s just some one else’s computer.

Most popular online services nowadays are gratis, but that does not mean that they come at any cost. Instead of paying with money, people are paying allegiance to service providers. In the land of “Minitel2.0”, Google and Facebook are like feudal lords of the Internet and we are their mere subjects. The exploitation of user data and of personally identifiable information is leading to numerous privacy invasion, some of which were only revealed thanks to Edward Snowden’s leak from the NSA.

Online Communities 2, XKCD

If you’re looking to protect your privacy or if you want to know how your rights can be affected when using these online services, you usually don’t have many options but to look for adjusting increasingly complex privacy settings or you need to be a full time lawyer to read the lenghty terms of service.

The User Data Manifesto aims at defining the basic rights that users should have on their own data when using online services. Recognising these rights is an important first step towards a free society in the digital age, along with Free Software.

Indeed, users should have:

1. Control over user data access

   User data should be under the ultimate control of the user. Users should be able to decide whom to grant direct access to their data and with which permissions and licenses such access should be granted.

   Data generated or associated with user data (e.g. metadata) should also be made available to that user and put under their control just like the user data itself.

2. Knowledge of how user data is stored

   When user data is uploaded to a specific service provider, users should be informed about the geographic location that specific service provider stores the data in, how long, in which jurisdiction that specific service provider operates and which laws apply.

   This point is not relevant when users are able to store their own data on devices in their vicinity and under their direct control (e.g. servers) or when they rely on systems without centralised control (e.g. peer-to-peer).

3. Freedom to choose a platform

   Users should always be able to extract their data from the service at any time without experiencing any vendor lock-in.

Read the full text of the manifesto here

I’m very happy that, with Frank and Jan, we are today announcing the release of version 2.0 of the manifesto during the ownCloud conf keynote.

Organisations and activists defending digital rights are joining in this effort to support online services that respect users’ rights. I am proud to be part of that effort and I hope it’s the start of a constructive debate and, hopefully, a humble contribution to our society.

I look forward to your feedback on the manifesto, which I view as a starting point rather than an end in itself.

Thanks to ownCloud, Inc for inviting me over to the ownCloud conf in Berlin.

The EUCJ has just published another decision regarding data protection that got me puzzled (but I’m not the only one!).

This one is primarily concerned with the interpretation of exceptions to the 1995 directive, but it also has interesting things to say regarding the infamous so-called right to be forgotten decision where legitimate interests in personal data processing were involved.

The facts are simple: someone puts a camera to monitor the entrance of his house. One day, people break in, but they are later identified thanks to the camera. Then, these suspects challenge the legality of the camera system on the grounds that they were not notified of the processing of their personal data.

Article 3 of the 1995 directive provides:

2  This Directive shall not apply to the processing of personal data: […]

– by a natural person in the course of a purely personal or household activity.’

But for the Court, (emphasis is mine)

33  To the extent that video surveillance such as that at issue in the main proceedings covers, even partially, a public space and is accordingly directed outwards from the private setting of the person processing the data in that manner, it cannot be regarded as an activity which is a purely ‘personal or household’ activity

This is a strange reasoning in my opinion, as it seems to make no distinction between purely personal activities and purely household activities–they are now combined under the criteria of the “private setting.”

So here’s how this applies to us: thanks to Neil, we already have a solution!

How does this relate to the so-called right to be forgotten?

The Court notes that:

34  At the same time, the application of Directive 95/46 makes it possible, where appropriate, to take into account — in accordance, in particular, with Articles 7(f), 11(2), and 13(1)(d) and (g) of that directive — legitimate interests pursued by the controller, such as the protection of the property, health and life of his family and himself, as in the case in the main proceedings.

I wish the Court followed the same approach in the so-called Right to be forgotten decision. But instead, the legitimate interest of the public to access published information has not been taken into account.

This is my first post.

I’ve been thinking about using ikiwiki for some time now. I don’t exactly know what I’ll do here, but I suspect I’m moving away from wordpress-powered http://blog.hugoroy.eu. The main reason is that I’m fed up of using the wordpress admin interface when all I want is to publish something I’ve written on my machine.

And git history.

I know I haven’t started writing yet, I’m already worried about technical issues. How can I handle writing in different languages here? I’ve seen the po plugin but it requires me to set a “Master language” for the whole system. Rather, I want to be able to set different master languages for each post.¹ Because sometimes I start writing in French, some other times I start writing in English. Most of the time I don’t translate myselft. So translation isn’t really what I’m looking for.

If you’ve got any suggestion on how to solve this, you’re very welcome. Until then, I’ll just stick to tagging.

C’est aujourd’hui que paraît le roman de Suzanne, Meurtre à Sciences Po ! Je ne peux que vous inviter à aller le lire, car il est très divertissant et qu’il dépeint avec un humour pointé de sarcasme, les individualités parfois originales qu’on trouve rue Saint-Guillaume ! Évidemment comme son titre l’indique, il s’agit d’un roman policier. Vous me direz si on reconnaît l’influence d’Agatha Christie 😉

Et non, ce n’est pas moi sur la gauche, mais Maxime ! Je suis sur la droite mais il faut croire que je gâchais toute l’harmonie visuelle de la photographie ☺

I've had some problems with the ?timeline lately. I'm trying to see if there's a way to "reboot" it and get clean of the bits that should have been left out… This will probably flood the RSS a bit, sorry for the inconvenience!

Mozilla is currently promoting the new Firefox 29 (Go get it!). Now, they're asking us on Twitter: What do you want for the Web? So I clicked on their link and here's what I got.

A broken Flash-YouTube video

I haven't been able to play YouTube videos for weeks now. Sometimes, it works though. I have no idea what's going on…

Dear Mozilla, next time you publish a video on your website, I don't want Flash and I don't want YouTube. I want HTML5 video (in an open standard format, i.e. free of patent restrictions) and I don't want you to promote a platform with crappy terms of service.

I was reading an article by Lorrie Cranor in the MIT Technology Review on how it’s difficult even for her to protect her privacy online.

However in this article, I was disappointed to see Ghostery mentioned. Ghostery is an browser extension supposed to help users against tracking and surveillance on the web. The main problem is that Ghostery is not released as Free Software[^akaos]

[^akaos]: a.k.a Open Source. Both these terms designate the same set of programs.

Earlier on Twitter I quickly posted my frustration about this. People who promote web privacy should stop promoting Ghostery, as it’s proprietary. What’s their business model exactly? ;-)

Ghostery has been playing on the ambiguity for too long. This hypocrisy must stop. See these tweets from years ago…

It seems Secret is the new thing. So I had a look at their terms of service. Here are some extracts:

TL;DR: They’re not good.

Oh, and here's the Privacy policy.

In case you thought you were “anonymous” when using Secret, think again:

We may share information about you as follows or as otherwise described in this privacy policy:

• In response to a request for information if we believe disclosure is in accordance with any applicable law, regulation or legal process, or as otherwise required by any applicable law, rule or regulation;

Wall Street Journal: The encryption flaw that punctured the heart of the Internet this week underscores a weakness in Internet security: A good chunk of it is managed by four European coders and a former military consultant in Maryland.

According to Brett Simmons, That pretty much wraps it up for C.

The whole heartbleed bugs also reminds me that OpenSSL is also an example of bad idea when it comes to licensing issues.

The heartbleed vulnerability is not only a catastrophic security issue, it also spans other interesting topics.

XKCD has an amazingly simple explanation of how the vulnerability works.

Somebody working at Mozilla put together a timeline of facts surrounding Brendan Eich’s resignation.

And the real tragedy here is that Mozilla would have sorted this out satisfactorily if it hadn’t been sensationalized by the media and turned into an internet witch hunt. Anyone who wrote a news story, posted to their blog, or tweeted about Brendan without understanding paragraph (i)(c) of the Community Participation Guidelines was part of the mob that brought Brendan down.

For more than 15 years, Brendan fought for openness and freedom on the web, and led many of the people who built that open and free web. This week, in a senseless, vicious convulsion, the web turned on him.

Meanwhile, Mozilla published an FAQ.

Q: Was Brendan Eich forced out by employee pressure?

A: No. While these tweets calling for Brendan’s resignation were widely reported in the media, they came from only a tiny number of people: less than 10 of Mozilla’s employee pool of 1,000. None of the employees in question were in Brendan’s reporting chain or knew Brendan personally.

In contrast, support for Brendan’s leadership was expressed from a much larger group of employees, including those who felt disappointed by Brendan’s support of Proposition 8 but nonetheless felt he would be a good leader for Mozilla. Communication from these employees has not been covered in the media.

Which echoes something written in the timeline mentioned above:

11) On March 27th, a small number of Mozillians tweeted variants of “I am an employee of @mozilla and I’m asking @brendaneich to step down as CEO”. These tweets were reported by the tech press, and my perception is that this was the start of the media firestorm. Most (or perhaps all) of the Mozillians who tweeted this were employed by the Mozilla Foundation, not the Mozilla Corporation which means that they report to the executive director of the foundation and not to the CEO. As foundation employees, they did not share the same org chart as Brendan.

This is why pieces like this trouble me:

Both writers seem concerned that Eich's resignation is a defeat for freedom of expression. If anything, it is a victory – the ouster of a founder and CEO by his own people, at a foundation based on open and equal expression, should be the new textbook example of the system working exactly as it should.

I hope this episode is now closed and that everybody learns a lesson from this.

(Especially, the guys at Rarebit who, after publishing an article “5 reasons why Brendan Eich should step down” now write “I want to say how absolutely sad to hear that Brendan Eich stepped down.” No comments.)

The IRS says that Bitcoin is property, not a currency. Bloomberg: “It’s challenging if you have to think about capital gains before you buy a cup of coffee.” No kidding!

It's interesting, as I was discussing the relationship between property and value yesterday night with Basti.

Hôtel de Ville de Paris, le 26 mars 2014, avec le drapeau chinois

Je me baladais hier avec Basti (@skddc) qui était à Paris pour la première fois. Lorsque soudain, j'aperçus des drapeaux chinois mêlés à ceux de la république et aux armoiries de Paris ! Malheureusement, ce n'est pas aussi drôle que dans le film de Jean Yanne et son adaptation à l'opéra, Carmeng.

Since I first set up Firefox Sync, things have changed. Mozilla thought that they needed to completely change the user experience of setting it up in firefox, thus discarding the previous firefox sync server for a totally new system of “Firefox Accounts”.

That sounds nice, however at the moment it’s nowhere as easy to set up if you want to self-host it instead of relying on Mozilla’s services.

You have to start 3 services:

• Sync 1.5
• and the Firefox Accounts Server
  • https://github.com/mozilla/fxa-auth-server/
  • https://github.com/mozilla/fxa-content-server/

Unfortunately, the READMEs are not as good as before. Sometimes, they ask you to change settings, but they don’t always tell you in which file you should modify it; or it also happens that the file they mention does not exist (e.g. the “config.json”).

I’ll have to give it another try… I hope that for next time, the documentation will have improved.

I get a CORS issue, but then… what?

I’m having a splendid Sunday at my desk, working on some moot cases for school.

Sometimes to get going, I need some good music that fits the mood. If you’re like me, you’re probably listening to some ambient or minimal music.

I’m on Trentemøller’s 2006 Last Resort right now and it feels great. I have no idea where I got that from, I just don’t remember. But anyway, thanks to the person who gave me this!

BBC: Microsoft admits reading Hotmail inbox of blogger:

Microsoft is caught up in a privacy storm after it admitted it read the Hotmail inbox of a blogger while pursuing a software leak investigation.

While the search was technically legal, [Microsoft's deputy general counsel] added Microsoft would consult outside counsel in the future.

So if it’s not just legal, but “technically legal”: what does that mean?

Yes, it means the way companies like Microsoft handle privacy is wrong. Yet another example.

I just came back from the cinema, where I watched Spike Jonze’s Her. This movie has got me thinking.

One thing I notice which was funny is how Theodore’s job kind of makes him fit the same role that her, the AI, is doing for him. Let me explain a little bit. Theodore, the main character, works at beautiful-handwritten-letters.com a service where people ask him to write beautiful letters to their wife for their 50th wedding anniversary, or to their son for his diploma, etc. You get it. By writing letters for other people, expressing some of their most personal emotions for them or even, instead of them, he participates to this society where the human self dissolves.

I guess what I’m saying isn’t making sense if you haven’t seen the movie. So go see it! There aren’t movies like this every year!

Here some little known, yet awesome apps or tools that I use. Thanks to the people working on these (I’m glad to have met some of them, and they’re awesome too)!

Transportr

Transportr is an Android app to help you use public transports systems. It’s simply the best one I’ve seen, and it supports a lot of systems (city-wide like Berlin or Paris and even long-distance).

Feedbin

Feedbin is an RSS web reader. It provides a pleasing reading experience and you can easily browse through items and share links. If you’re looking to host it yourself, have a look at the sources.

ikiwiki

ikiwiki powers this blog, hosted by branchable. If you like git and markdow, and editing your texts with your favourite text editor, this is for you.

Known

Known (formerly “idno”) is more “socially aware” than ikiwiki. It runs with PHP and it’s basically your easy-to-run indieweb space. If you use it with http://brid.gy you will enjoy a nice integration with twitter and other silos (see an example of my own).

YunoHost

YunoHost is custom debian distribution aiming at making self-hosting easy. It provides a nice web interface for administration of your self-hosted server and for users of the web server. If you have basic linux administration skills, this will be very helpful.

Pinboard

Pinboard a simple and efficient bookmarking app that also archives the content of marked pages (if you pay for it). Unfortunately, Pinboard is not released as free software. But you can export your bookmarks.

Sharesome

Sharesome lets you easily share files on the web. It has a pleasant interface that works well on all devices I have tested so far. It’s also available as a web app. The neat feature is that you can choose where to host your data (for instance, with remotestorage; you can get an account at https://5apps.com).

Terms of Service; Didn’t Read

Some shameless self-promo with ToSDR, the app that tells you what happens to your rights online by rating and summarising Terms of service and privacy policies. You can also get it directly in your web browser or as a web app.

If you’re looking for a curated list of awesome web services that are free of charge and based on free software and open data, look no further than Jan’s Libre projects.

Hier, j'étais au loop avec Okhin pour refaire le chiffrement de mon serveur mail, que je viens de réinstaller (Kolab 3.4, Debian 8).

Voici quelques liens qui nous ont été utiles :

• How to create a self-signed SSL Certificate

  Attention cependant, en suivant ces instructions on crée un certificat avec le flag CA false alors que pour utiliser un certificat autosigné sur Android, il faut absolument que le flag CA soit true (voir ci-dessous).

  Finalement, voici la commande qui m'a permis de faire le certificat, dans /etc/ssl/private/:

  openssl req -x509 -new -key totosh.ampoliros.net.key -out totosh.ampoliros.net.csr -days 730
  

  Cette commande remplace l'étape 2. L'étape 4 n'est donc pas nécessaire semble-t-il.

  Ensuite, je copie totosh.ampoliros.net.csr vers /etc/ssl/certs/totosh.ampoliros.net.crt puis je configure mes services pour utiliser ce certificat ainsi que la clé privée générée à l'étape 1.

• Je veux HTTPS

  Utilisé pour configurer Apache2.

• Exemples de configuration Postfix de Benjamin Sonntag

  Sans oublier de générer dh2048.pem:

  openssl dhparam -out /etc/ssl/private/dh2048.pem -outform PEM -2 2048
  

• Securing all Kolab Services

  Utilisé pour configurer Cyrus IMAPD.

Importer un certificat autosigné dans Android

Normalement, avec la commande openssl donnée plus haut, on génère un certificat autosigné acceptable pour Android.

On peut par exemple utiliser l'application CAdroid pour importer le certificat.

Si on est root sur l'Android, on peut aussi copier le certificat directement dans le système en passant par adb. C'est la solution que j'ai choisie.

Installing CAcert certificates on Android as 'system' credentials without lockscreen - instructions

(En passant: j'ai un bug sur CyanogenMod qui m'enmpêche pour le moment de modifier les paramètres de sécurité. Je suis donc passé par adb pour autoriser l'installation de sources autres que Google avec la commande: adb shell settings put secure install_non_market_apps 1.)

Sometimes, you want to get numbered paragraphs throughout a document.

I'm looking for a way to achieve this with pandoc, in a way that I can write once, and publish in several formats.

I added this in pandoc's LaTeX template:

$if(numberparagraphs)$
%% numerate paragraphs with a label
\newcounter{paranumero}
\newcommand{\para}[1]{%
\noindent\refstepcounter{paranumero}\llap{\small\bfseries\theparanumero\label{#1}.~}}
$endif$

That way, I can write

\para{shortId} A paragraph...

and later in the document, I can refer to that paragraph\vref{shortId}

That's good, but it only works in LaTeX. I'd like to write something similar in a Markdown document and be able to also produce HTML, or even ODT or OOXML.

I'll experiment a bit, but if you have ideas of your own... tell me :-)

TODO: allow to write \para without any value and assign it a predictable \label... for instance the first letter of the first five words of the paragraph.

I've been using LaTeX since about 2009.

Here's a collection of tips for Xe/La/TeX with an inclination towards use in the French language, within a legal context or in combination with ?pandoc.

L^AT_EX

X_ƎT_EX

<span class="latex">L<sup>A</sup>T<sub>E</sub>X</span>

<span class="xetex">X<sub>&#398;</sub>T<sub>E</sub>X</span>


.latex sub {
  vertical-align: -0.1ex;
  margin-left: -0.1667em;
  margin-right: -0.025em;
}

.xetex sub {
  vertical-align: -0.1ex;
  margin-left: -0.1667em;
  margin-right: -0.125em;
}

.latex sub, .latex sup, .xetex sub {
  font-size: 0.9em;
  text-transform:uppercase;

}
.latex sup {
  font-size: 0.85em;
  vertical-align: -0.2em;
  margin-left: -0.26em;
  margin-right: -0.05em;
}

Edit: You should have a look at msmtp-queue-scheduler to solve this :-)

Thanks Nicolas!

I've been looking for a solution to this problem:

Sometimes, I just sent an email, and after 10 seconds I realise: “Oh, damn, I forgot to add something!” or “Oh, nooooes, I sent it to the wrong person!” Whatever. This is annoying.

Making mistakes is human. Email clients are for humans, therefore they should be able to cope with our mistakes and help fix them. That's why software should allow us to “undo” or “cancel” our actions.

Gmail does this right, so why can't we do it in Mutt too?

I wrapped my head around this a little bit. And I'm no programmer, so after trying to add some stuff here and there, I finally decided to have it with this very, very, very dirty hack. You've been warned. So here's how I do it:

• I use the msmtpq script which allows me to queue emails when I'm offline, so that msmtp can take care of sending queued email when I'm back online.

• I have tried to put a delay of 30 seconds any time before calling msmtp or I tried to force msmtpq to queue all outgoing email for at least 30 seconds. But I couldn't make it work. So instead,

• I added sleep 30 && at the beginning of the msmtpq script!

• Result: now when I send an email, I have 30 seconds to realise when I made a mistake. Then, I just need to activate plane mode on my laptop, and fix whatever needs fixing from here.

  One issue with this workflow is that I can't just edit the outgoing email, I also have to make sure I update the openPGP signature. So I'd probably just delete the email and start again from Mutt.

I warned you, this is very ugly :-)

But as they say: the best way to get an answer on the net is not to ask a question, but to give the wrong answer!

So if anyone wants to implement this feature in msmtpq, that would be great. Here's how it could work ideally, in the .muttrc:

set sendmail="msmtpq --wait 30" 
set sendmail_wait=-1  #send in the background

then msmtpq could queue the email for 30 seconds before testing the connection and feeding the email to msmtp. The Queue management could then allow the user to pause all outgoing emails, cancel the email containing the mistake, then sending remaining emails from the queue.

The msmtpq script is available at: http://sourceforge.net/p/msmtp/code/ci/master/tree/scripts/msmtpq/

Here’s a list of options I modified in my firefox’s about:config.

• middlemouse.contentLoadURL: false

  I use the mouse middle click to copy and paste stuff. So when I accidentally press that mouse middle click while reading a web page, I don’t want Firefox to load whatever’s in my buffer as a URL.

• browser.urlbar.trimURLs: false

  Mozilla decided to look more like Google Chrome and started to hide relevant parts of the URL like http://. This feature has helped me exactly 0 time, while on the other hand it was annoying many times when copy/pasting from the URL bar would not give me the entire URL (and the results were not consistent).

• Disabling unsecure ciphers

  • security.ssl3.ecdh_ecdsa_rc4_128_sha
  • security.ssl3.ecdh_rsa_rc4_128_sha
  • security.ssl3.ecdhe_ecdsa_rc4_128_sha
  • security.ssl3.ecdhe_rsa_rc4_128_sha
  • security.ssl3.rsa_rc4_128_md5 [this one seems required for Youtube's https to work]
  • security.ssl3.rsa_rc4_128_sha

  You can now check if your browser uses secure SSL/TLS cipher.

• Replace Google with DuckDuckGo on Firefox

  • browser.search.defaultenginename: DuckDuckGo
  • browser.search.selectedEngine: DuckDuckGo

Other tips for Firefox:

1. Install the linux-xps13-archlinux kernel (now in the AUR)

   Note: Since I wrote this, it’s possible that the patched kernel now has more features than only touchpad support.

2. Install xf86-input-synaptics and, from AUR, touchegg and touchegg-gce-git (this last one is to be able to configure gestures with the graphic interface).

3. Edit /etc/X11/xorg.conf.d/50-synaptics.conf

   Section "InputClass"
           Identifier "touchpad catchall"
           Driver "synaptics"
           MatchIsTouchpad "on"
           Option "TapButton1" "1"
           Option "TapButton2" "0"
           Option "TapButton3" "0"
           Option "ClickFinger2" "0"
           Option "ClickFinger3" "0"
   
   
   # This option is recommend on all Linux systems using evdev, but cannot be
   # enabled by default. See the following link for details:
   # http://who-t.blogspot.com/2010/11/how-to-meta:ignore-configuration-errors.html
           MatchDevicePath "/dev/input/event*"
   
   
   EndSection
   

4. Configure your gestures with Touchègg

   Here's my ~/.config/touchegg/touchegg.conf:

5. Add to your session (using gnome-session-properties for instance):

   • touchegg

The real improvement is that I can use three-finger tapping to simulate the middle-click mouse button which is used for quick pasting or for opening links in a new tab.

As far as "pinching" is concerned, it does not work reliably at all for me.

Once you have Mutt up and running with the basic tools that will allow you to fetch, index, view, edit and send email, there are a couple of things that you probably need in your .muttrc, no matter what.

• In Mutt, even if you are on the pager that lets you read a selected email, the <up> and <down> keys will not help you navigate inside the email; instead they will bring you to the previous or to the next email in the index.

  That’s rather weird and unexpected. One of the first thing I did was trying to scroll inside an email, and Mutt suddenly browsed dozens of emails from the index (even marking them as read…). That’s quite unhelpful. To solve this, add:

  bind pager <up> previous-line #scroll inside the message rather than the index
  bind pager <down> next-line
  

  That way, you can scroll, or use the arrow keys to read an email inside the pager.

  When the pager is not opened, the up and down keys will behave as usual in the index.

• To make Mutt faster

  set sleep_time = 0 # be fast
  

  Don’t worry, this does not affect anything, it will just be faster. Here’s the doc about sleep_time:

  Specifies time, in seconds, to pause while displaying certain informational messages, while moving from folder to folder and after expunging messages from the current folder. The default is to pause one second, so a value of zero for this option suppresses the pause.

• When you display email in the pager, it can look ugly when the lines are too long and that Mutt, by default, is not very nice about line-wrapping because it can cut words in the middle. You probably want to add:

  set smart_wrap
  

• Changing between mailboxes can be a pain by default in Mutt. You need to tell Mutt where to find your several mailboxes, so you’ll be able to switch between them more easily (for instance by pressing y or with the sidebar). Add:

  mailboxes $Mail
  

  I have seen on some Mutt configs that y sometimes does not toggle the list of mailboxes. In that case, you can add something like:

  macro index y "<change-folder>?<toggle-mailboxes>" "show incoming mailboxes list"
  

The way it works is:

• You’re reading an email on Mutt and you think: Oh, I should really remember that because I need to do X.
• You pass this email to emacs’ org-mode containing:
  • the subject, the date and the from
  • a link to that email which is mutt:Message-ID
• You add some information if you wish to, and you save this in your org-mode todo file.

Now, you’re in your org-mode todo list, you can work with it as usual. Now you have this bit, and you need to remember what it’s about exactly. You can click the link in org-mode, and it will open Mutt and show you the right email in Mutt!

• More info on Zack’s blog
• Get org-mutt directly from his git repo

Make mutt-open work with mutt-sidebar

I used mutt-sidebar and I couldn’t get ./mutt-open to work correctly. I fixed this by changing:

HIDE_SIDEBAR_CMD="B"     # set to empty string if sidebar is not used

You need to update B with whatever you use to toggle the sidebar. I use CTRL-B so I changed it to HIDE_SIDEBAR_CMD="\CB". That’s it!

This is the due to the fact that mutt-open has something called mutt_keys that’s used to display the right email in mutt. However in its current state mutt_keys decides to toggle the sidebar before it actually shows the email. Hence if the toggle sidebar key is not well defined in mutt-open, mutt will hang up there. So I changed this to:

mutt_keys="/=i$msgid\n\n$HIDE_SIDEBAR_CMD"

Make mutt-open work with mutt-kz

Since mutt-kz comes with a sidebar as well, the previous paragraph applies. But you might need further changes to the original mutt-open script. Here’s how mine works:

mutt_keys="<vfolder-from-query>id:$msgid\n\n<entire-thread>"

This should work with any carddav server, but if you use Kolab's carddav server here's some extra tip!

The problem you want to fix is: it's impossible to remember everyone's email address. This problem is solved by most email programs because they are linked to a contact list already. However, for those of us using Mutt, there's no full contact integration so you need to rely on something else.

One obvious solution is to rely on a mail indexer to search and find addresses in emails from the past. If you use mu, here's how Karsten does it.

However, that's not really helping if you have contact information from multiple sources (e.g. typing on your mobile the email address of somebody you just met AFK). This is where a contact server is handy.

If you use Google's contacts, you can use goobook it works well but it's quite slow IMHO. And obviously, the problem is that you have given up your whole contact list to Google.

Find your Kolab addressbook

With Kolab 3.1 comes a CardDav/CalDav/webDav server! Version 3.1 was just released today. So let's use that instead.

• When I set up Kolab 3.1 before the official release, I got a packaging bug in CentOS, but it's easy to fix.

• The *Dav server is located at https://kolab.example.org/iRony. Now, you need to find how to link to a specific addressbook. I tried to have a look at the Roundcube interface, for a folder id or something, but I could not find any that was working.

• Just connect a webDav client (in Nautilus, File > Connect to a server) to the iRony folder, and then just navigate to find the addressbook identifier (look in the address bar!)

Hopefully, this will soon not be needed any longer. There will be a "Show address book URL" setting directly in Roundcube's contacts menu.

Sync your CardDav addressbook with pyCardDAV

Now install pyCardDav which just landed on Debian last month:

# apt-get install pycarddav

• Sync pycarddav after you entered the Kolab addressbook resource in the config file with pycardsyncer

  I advise you run this with --debug to make sure that it does not get stuck in case you have some illegal characters inside one of your vCards. If it gets stuck, then you can just go back to your webDav client and edit the file that's causing trouble.

• If all goes well, you should be able to search for contacts inside your local copy:

  % pc_query hugo
  searching for hugo...
  Name: Hugo Roy
  TEL (CELL): +...
  EMAIL (INTERNET\, WORK): hugo at fsfe dot org

Lookup directly from Mutt

I just added this to my ~/.muttrc:

set query_command="pc_query -m '%s'"
bind editor <Tab> complete-query

That way, in Mutt, just type ‘Q’ to search for a contact. Or you can also press ‘m’ to start a new message, start typing in the ‘To:’ field the name of your contact and just press [Tab] to have autocompletion!

Hey, did you know that on Mutt, ‘CTRL’+K will import public openPGP keys attached to emails (with the right MIME type)?

That makes signing and sharing keys even easier ☺

Thanks gollo for the tip!

I began using Mutt in March 2013. (I was using Gnome's Evolution before, but its searches were really slow and one day I really freaked out when Evolution did something weird which led me to believe that I couldn't see my emails since 2010. It turns out everything was fine, but Evolution hang up on me).

Beginning on Mutt is “not easy.” I went to Mutterwares to have experienced user show me how they use it. You might be interested in some of the information collected there.

First tips for Mutt beginners

Once you have Mutt up and running with the basic tools that will allow you to fetch, index, view, edit and send email, there are a couple of things that you probably need in your .muttrc, no matter what.

• In Mutt, even if you are on the pager that lets you read a selected email, the <up> and <down> keys will not help you navigate inside the email; instead they will bring you to the previous or to the next email in the index.

  That’s rather weird and unexpected. One of the first thing I did was trying to scroll inside an email, and Mutt suddenly browsed dozens of emails from the index (even marking them as read…). That’s quite unhelpful. To solve this, add:

  bind pager <up> previous-line #scroll inside the message rather than the index
  bind pager <down> next-line
  

  That way, you can scroll, or use the arrow keys to read an email inside the pager.

  When the pager is not opened, the up and down keys will behave as usual in the index.

• To make Mutt faster

  set sleep_time = 0 # be fast
  

  Don’t worry, this does not affect anything, it will just be faster. Here’s the doc about sleep_time:

  Specifies time, in seconds, to pause while displaying certain informational messages, while moving from folder to folder and after expunging messages from the current folder. The default is to pause one second, so a value of zero for this option suppresses the pause.

• When you display email in the pager, it can look ugly when the lines are too long and that Mutt, by default, is not very nice about line-wrapping because it can cut words in the middle. You probably want to add:

  set smart_wrap
  

• Changing between mailboxes can be a pain by default in Mutt. You need to tell Mutt where to find your several mailboxes, so you’ll be able to switch between them more easily (for instance by pressing y or with the sidebar). Add:

  mailboxes $Mail
  

  I have seen on some Mutt configs that y sometimes does not toggle the list of mailboxes. In that case, you can add something like:

  macro index y "<change-folder>?<toggle-mailboxes>" "show incoming mailboxes list"
  

Here is a list of mutt tips. My ~/.mutt is public.

I’m a new ikiwiki user. I’ve been interested in using this, because of the git version control, and I liked the idea that I could just clone the entire ikiwiki on my laptop and thus edit it offline with my favourite editor.

So, as every new user, I make some mistakes which could have been pretty easily avoided. Here’s one quite significant.

In order to generate the ?timeline I use the aggregate plugin. That way I can pull content from lots of sources through feeds, and put it into one page and one feed. This timeline is like a big feed of things I do everywhere on the web. In a way, it’s a PESOS approach in the IndieWeb.

One important technical detail is that the directive displays information on the page itself, such as when was the last time the source feed was checked. This means that each time the feeds are checked, the page that contains the directive has to rebuild in order to update the information.

Obviously, it was a very bad idea for the timeline, because that meant that each time the sources were checked, the whole timeline has to rebuild whole over again. And it took CPU time.

The people running http://branchable.com fixed it for me. Thanks to them! I’ll make sure to use it as a reminder and have separate files for that in the future!

If you have other tips to make ikiwiki more efficient, please comment!

It’s quite annoying when you click somebody's name or email on a web page and that Firefox cannot figure out how to rely on your system preferences to send an email.

For instance, my system is set up so that when I click on an email address or mailto: link anywhere, gnome-terminal opens up with Mutt ready to send an email. For some reason, Firefox tries to figure out all by itself which program I should use.

I have tried to make Firefox use gnome-terminal with Mutt, but it didn’t work. However, it's possible to add your own webmail there (for some reason, I could choose between Yahoo Mail, Gmail and Mykolab.com but I can't remember how I did that). Now that I have my own Kolab instance with roundcube, I decided to add my own webmail there. Unfortunately, it seems there's no way from the graphical interface, so I went straight to: ~/.mozilla/firefox/iceweaselprofile.hugo/mimeTypes.rdf.

It's quite a big file FWIW (885 lines here). I did not have a look at the details, but hopefully just adding this helps (for the second block, make sure to merge with existing mailto handlers):

<RDF:Description RDF:about="urn:handler:web:https://kolab.example.org/roundcubemail/?_task=mail&amp;_action=compose&amp;_to=%s"
       NC:prettyName="Kolab Groupware"
       NC:uriTemplate="https://kolab.example.org/roundcubemail/?_task=mail&amp;_action=compose&amp;_to=%s" />

<RDF:Description RDF:about="urn:scheme:handler:mailto"
       NC:alwaysAsk="true">
    <NC:possibleApplication
    RDF:resource="urn:handler:web:https://kolab.example.org/roundcubemail/?_task=mail&amp;_action=compose&amp;_to=%s"/>
</RDF:Description>

(Just replace https://kolab.example.org/roundcubemail/ with your own Roundcube location. And of course, if you don't use Kolab replace "Kolab Groupware" by whatever.)

When you get used to a text editor that you like, you start to get annoyed by all the times you have to enter some text and you're using a shitty editor. This is so true on the web when you try to write something in more than 140 characters.

Matthias showed me a nice firefox extension: “it's all text” (you can also install it from debian repos). For each text box, you just click a small “edit” button and it will fire up vim[^1] and let you edit the text there, then send it back to the web interface.

[^1]: Actually, I had to configure it to start GVim instead of vim. I suppose there's a way to launch a terminal and then vim…

Reply to an email, easily delete the parts below line

When I reply to emails, Mutt fires up vim. Then I can quickly navigate until I reach the parts I want to reply to specifically. Sometimes it’s just a portion of the whole email, so I want to leave out the rest. Matthias shared a nice tip that allowed him to do that quickly.

Unfortunately, it wasn’t working properly for me. I figured out it’s because in markdown mode, I use folding. By default, my own signature is folded. So I modified the rule a little bit, and here it goes, just add in your ~/.vimrc:

So when you're below the part you’re replying to, just press ,dd and it will delete the rest of the email between the line and your signature.

It only works if your signature is well delimited by --  and if folding is on. If you do not use folding, then use the original config provided by Matthias:

noremap ,dd :.;/^-- $/dO-- 

I began using vim in January 2013. (I was mostly using gedit before, but it was so slow at times that it couldn’t keep up with me typing on the keyboard.) I suppose my use of vim isn’t the most widespread. I write code occasionally only, I mainly write emails and documents, so my set-up reflects this and the tips I share here are not the most interesting for coders.

Here is a full list of vim tips. My beginner’s ~/.vim is public.

Most of the hardware devices are, even as you read this, available as off-the-shelf items, just waiting to be plugged into each other in order to put an end to the record business as we know it.

Frank Zappa with Peter Occhiogrosso, The Real Frank Zappa Book (p. 336, Chapter 18, “Failure”), A Touchstone Book, published by Simon & Schuster, 1988-1989. Also see, “A proposal for a system to replace phonograph record merchandising”.

– Bah, la vérité, la vérité… Qui peut dire ce qui est vrai et ce qui ne l'est pas, commandant ! Pour nous, l'essentiel, ce n'est pas la vérité, c'est la justice, non ?

Pierre Lemaitre, Alex (2011)

Il est juste que ce qui est juste soit suivi ; il est nécessaire que ce qui est fort soit suivi. La justice sans la force est impuissante, la force sans la justice est tyrannique. La justice sans force est contredite, parce qu'il y a toujours des méchants ; la force sans la justice est accusée. Il faut donc mettre ensemble la justice et la force et pour cela faire en sorte que ce qui est juste soit fort, ou que ce qui est fort soit juste.

La justice est sujette à dispute, la force est très reconnaissable et sans dispute. Ainsi on n'a pu donner la force à la justice, parce que la force a contredit la justice et a dit qu'elle était injuste, et a dit que c'était elle qui était juste. Et ainsi ne pouvant faire que ce qui est juste fût fort, on a fait que ce qui est fort fût juste…

Ne pouvant faire qu'il soit forcé d'obéir à la justice, on a fait qu'il soit juste d'obéir à la force. Ne pouvant fortifier la justice, on a justifié la force, afin que le juste et le fort fussent ensemble, et que la paix fût, qui est le souverain bien.

Pascal, Pensées (à vérifier)

Information is not knowledge. Knowledge is not wisdom. Wisdom is not truth. Truth is not beauty. Beauty is not love. Love is not music. Music is the best…

Frank Zappa, “Packard Goose”, Joe’s Garage Act III (1979)

„Nein,“ sagte der Geistliche, „man muß nicht alles für wahr halten, man muß es nur für notwendig halten.“ „Trübselige Meinung,“ sagte K. „Die Lüge wird zur Weltordnung gemacht.“

Non, dit le prêtre, on n’a pas à tenir tout pour vrai, on a seulement à le tenir pour nécessaire.

— Triste opinion, dit K. ; c’est le mensonge érigé en loi de l’univers.

Franz Kafka, Le Procès, Dans la Cathédrale (1925)

It's easier to move without a piano.

Michiel B. de Jong in Berlin, Kreuzberg (2012)

God, grant me the serenity to accept the things I cannot change,
courage to change the things I can,
and wisdom to know the difference.

Que Dieu
m'accorde la sérénité
d'accepter les choses que je ne peux pas changer,
le courage de transformer celles qui s'y prêtent
et la sagesse
de savoir toujours les distinguer.

Billy Pilgrim, in Slaughterhouse-Five, or The Children's Crusade: A Duty-Dance with Death (1969)

With great power comes great responsibility.

Uncle Ben in Spider-Man (2002)

Toute Société dans laquelle la garantie des Droits n’est pas assurée, ni la séparation des Pouvoirs déterminée, n’a point de Constitution.

Article ⅩⅥ de la Déclaration des droits de l'homme et du citoyen du 26 août 1789 (inspiré de Montesquieu)

Violence is the last refuge of the incompetent.

La violence est le dernier refuge de l'incompétence

Salvador Hardin, in Isaac Asimov's Foundation (1951)

Without deviation from the norm, progress is not possible.

Frank Zappa (1940-1993) video

Dans la pratique — les révolutionnaires se trompent toujours, parce qu’ils croient toujours la vérité trop simple, ont trop confiance en eux-mêmes et s’imaginent qu’ils ont trouvé et déterminé le terme du progrès humain ; tandis que le propre du progrès, c’est de n’avoir pas de terme, de n’atteindre ceux qu’on lui propose qu’en les transformant, de ne résoudre les problèmes qu’en en changeant les données.

In practical matters the revolutionary spirits always make mistakes, because they always believe truth to be too simple, because they are too confident of themselves, and imagine that they have found and fixed the end and aim of human progress. Whereas real progress is to have no end; is to reach those ends which one has put before one’s self, only to transform them; to solve problems only by changing their data.

Jean-Marie Guyau, Esquisse d’une morale sans obligation ni sanction, « Le risque métaphysique dans la spéculation », 1885