Moving to IPv6

Internet protocol version 6 is one of the few new technologies I was told about during my studies that did not become totally mainstream (like say, threads), or disappear completely (most stuff with the terms high-level in it). IPv6 was a bit like fusion, the thing that will work, soon… As the world supply of IP addresses inexorably shrinks, we have slowly been going there. Today I tried to enable it at home, some thoughts and observations

  • Swisscom supports it on the new routers, but you have to activate the feature in your customer centre page, which implies that you have you credential and the right account.
  • Squid version 2.X does not support IPv6, I tried installing Squid version 3.1 on my NAS, but it segfaults at start. The package is probably corrupt.
  • I can ping various sites over IPv6:
    PING6(56=40+8+8 bytes) 2a02:⚿⚿⚿⚿… --> 2a03:2880:2030:2f01:face:b00c::8
    16 bytes from 2a03:2880:2030:2f01:face:b00c::8, icmp_seq=0 hlim=55 time=120.833 ms
    PING6(56=40+8+8 bytes) 2a02:⚿⚿⚿⚿… --> 2a00:1450:4001:c02::5d
    16 bytes from 2a00:1450:4001:c02::5d, icmp_seq=0 hlim=56 time=36.579 ms

    but not this blog.

  • If I disable the proxy, I can connect various ip6 testing sites which gives me more or less green lights.
  • Browsing seemed to work, although it is difficult to assess how much traffic goes over IPv6.

Flattr this!

New Router

Unfortunately My Swisscom isn't able to display any up-to-date data at the moment. Please try again later.

So Swisscom upgraded my ADSL line to the VDSL2 standard, this was not something I asked for, just the carrier upgrading its infrastructure. Despite being a geek, I don’t consume that much bandwidth: the main media I consume on the internet is text, which is not that bandwidth intensive, and the little I consume, I optimise, for instance by running my own proxy server.

So Swisscom shipped me a new router, a Swisscom Piccolo, which is basically a Motorola Netopia 7640-47, with some crippled firmware. Any hope that the new router would be an improvement was quashed when I realised the new one had two network ports where the old one had four, forcing me to add a switch beside the router – of course the ports are still 100Mb/sec ethernet… Second, I checked if the new router would support IP6, as Swisscom supports the 6RD protocol, now enabling such a complicated feature cannot be done simply in the router’s web UI, that would be too simple, instead this must be done over the Swisscom customer support, except, I did not have one, or it was not supported, so I had to ask for confirmation by post.

Meanwhile I saw how the router would behave in normal usage. I asked my NAS to configure itself against the UPN interface of the router. The router crashed. This morning I got an e-mail from my NAS that it could not connect to the internet, basically the router had gone into stupid mode and was injecting some redirect page for every connection. Plenty of SSL certificate warnings.

But what about the customer center UI? I got the letter with the password, and I have access. But it does not see my ADSL connection but the bills are, so Swisscom can bill me for something it does not know about. Impressive. Let’s go old-school, maybe I can call them, of course all their operators are busy. Maybe the chat interface, nope, all their chat-bots are also busy. Swisscom also has an iOS app that let me access my account, maybe this will work better and I’ll be able to see the account later on? Yeah, right. I actually tried to reload it later, but the app crashed and lost the credentials…

Flattr this!

Screen Capture from the document HTML predefined icon-like symbols

Escape Sequences

1994 proposed named escape sequences vs. unicode characters
Original Unicode Symbol
&audio; 1F509 🔉
&cd.rom; 1F4BF 💿
&clock; 1F557 🕗
&diskette; 1F4BE 💾
&display; 1F4BB 💻
&document; 1F4C4 📄
&fax; 1F4E0 📠
&folder; 1F4C1 📁
&home; 1F3E0 🏠
&index; 1F4C7 📇
&keyboard; 2328
&mail; 1F4E7 📧
&; 1F4E5 📥
&mail.out; 1F4E4 📤
&next; 2398
&notebook; 1F4D3 📓
&previous; 2397
&printer; 2399
&sadsmiley; 1F61E 😞
&smiley; 1F604 😄
&telephone; 1F4DE 📞
&text.document; 1F4DD 📝
&trash; 267B

One early proposal to the HTML standard was by Bert Bos to have named entities for often used icons (folders and such). The list of the proposed escape sequences was closely modelled to the icon set of the day: hypercard, gopher, etc.

Some time ago there was discussion on this list about defining a set of standard icons for things like Gopher types, “home” buttons, etc.
The discussion didn’t reach a conclusion. Below is a proposal.
Reactions please! The text and two sets of example icons are also available at:

The idea never took off, but what is ironical, is that in the end, 18 years later, many of these symbols are now part of various segments of unicode, inluding emoji, so the only difference between now and the 1994 proposal is the type of escape sequence: unicode numbers vs. named entities. It is one of the rare cases where something ends up being implemented one level down in the abstraction stack, not in the browser, but in the text rendering system of the operating system (schematically at least, chrome seems to be be doing strange things with emoji).

Edit: here is the W3C proposal

Flattr this!

A Beginner’s Guide to HTML

NCSA Mosaic Logo

One of the basic ideas of the web is that you do not copy data to your machine, instead you keep pointers (urls) to stuff. The underlying assumption is that said pointer will remain valid in the future. When I first learned to do HTML, I clearly remember printing out the page A Beginner’s Guide to HTML written by Marc Andreessen, so I could work offline, at home. I had dialup access, but loading a web page over a modem was just to slow. As internet became faster and more prevalent, I stopped keeping paper versions of documentation, and moved on to more modern HTML features, like the ones supported by Netscape. I eventually lost the paper version.

In response to assorted requests and queries, I have written a simple “Beginner’s Guide” to writing documents in HTML. It’s up for grabs at at the moment; comments are welcome (but no complaints about my coverage or use of the IMG tag that Mosaic supports; it’s important internally).
The guide also points to a rudimentary primer on URL’s that might be of interest to Web beginners (certainly the number of people who have sent me Mosaic bug reports saying “URL ‘’ doesn’t connect to the ftp server”, etc., would seem to indicate that basic knowledge of URL’s is not yet a given on the net).

Finding that particular guide again was not completely trivial, so I’m now putting another mirror online here. This document confirms what I outlined in my post about image formats, that what was considered a reasonable image file in these days is not anymore. We are still struggling with the format of video data, although that subject is already touched upon in 1993. In a sense, HTML-5 is much closer to the spirit of HTML-1, with various teams trying to get something done instead of having a nice formalism.

Flattr this!

Doran Shihan / 動羅師範 2013

Doran 2013 Poster

毎年は動羅 (ドラン) 師範がチューリッヒに合気道のゼミをします。すろごとにとても面白かったです。

Each year, Doran-shihan comes to Zürich for an aikidō seminar, each time it is very interesting.

Chaque année, Doran-shihan vient à Zürich pour donner un stage d’aïkidō. C’est chaque fois très intéressant.

Jedes Jahr kommt Doran-shihan nach Zürich um ein Aikidō Seminar zu geben, es ist jedes mal sehr interessant.

While interesting, Doran’s seminars tended to be challenging for me, he tends to do complex techniques which look very fluid when he demonstrates them, but which I have trouble reproducing. This year there were less people at the seminar, I think there was some other high-ranking teacher giving one somewhere else in Europe, so people had to make hard choices. While bad news for the organisers and the teacher, I found this helped make the lessons clearer and more serene, and while some technique still confused the heck out of me, I found them generally easier to grasp.

I really liked the fact that the techniques are both elegant and very practical, with attacks that are quite different from the usual: knife, leg kicks, you sometimes need to do something different from 正面打ち (shōmen-uchi) 一教 (ikkyō) .

Flattr this!

Orc-Idée 2013

Grande Kermesse
13-14 Avril
Son Saloon ⬩ Ses Salles de Jeu ⬩ Son croque mort

Orc-idée est probablement la convention à laquelle je vais le plus régulièrement. C’est pour moi une occasion de revoir du monde, et c’est presque devenu une tradition, de faire une partie de Rêve de Dragon.

L’organisation aura été un peu chaotique, j’ai eu très tôt quatre joueurs inscrits, mais l’un d’entre eux c’est désisté, l’autre n’est simplement pas venu. Je n’aime pas tellement jouer avec seulement deux joueurs, mais heureusement trois joueuses se sont spontanément jointes à la partie. L’exercice m’a surtout forcé à faire ma partie bien plus vite, vu que les joueuses en question avaient une autre partie enchaînée juste derrière et qu’il m’a fallu un peu de temps pour les trouver, bref une partie de 4 heures jouée en 2:30 pour elles, 3:00 pour ceux qui sont restés jusqu’à la conclusion, le tout avec une seule personne connaissant l’univers et le système. C’était la première fois que je faisais jouer un nouveau scénario, La plaine d’Urh. Je pense que je ne m’en suis pas trop mal sorti : les participants avaient l’air d’être contents.

Par contre cette danse des joueurs confirme mon impression: autant avoir des parties inscrites et annoncées est une bonne idée, autant les inscriptions en ligne des joueurs ne fonctionnent pas bien, il y a régulièrement des gens qui s’inscrivent mais qui n’on ne voit jamais à la table. Résultat : des parties qui apparaissent pleines en ligne, des joueurs qui ne viennent pas, et en fin de compte des maîtres de jeu qui ne peuvent démarrer faute du bon nombre de joueurs. Bref, une situation plus chaotique que si les inscriptions étaient faite sur place. J’ai aussi l’impression que cette forme d’inscriptions n’encourage pas à l’exploration, on se retrouve dès lors avec des groupes de joueurs qui se connaissent, cela donne un jeu plus fluide, mais aussi moins de surprises.

Flattr this!

WP Retina 2×

User interface for the Retina 2× plugin

Following my article on high resolution graphics for my blog, I have looked around for some plugin that would handle high DPI graphics for this blog. I have found WP retina 2×, which use Apple retina approach to the problem: have a second image at twice the resolution with the string "@2x" appended in the name before the extension, if the display is detected to be in high resolution, some javascript tries to rewrite the urls for the various images. The plugin does not generate the 2× graphics automatically for existing images, but provides a tool to do so: for each image in the image library, one can automatically generate the versions at twice the resolution.

This is particularly useful for the thumbnails, in this blog they are typically 250 pixels wide, when the thumbnail was generated from a higher resolution image, then the plugin can generate a 500 pixel wide thumbnail and substitute that on high DPI displays. Of course, this does not work for hand-optimised low palette count PNG files, which I often used.

I generated substitute images for some the thumbnails in this blog, but for many the available image just has to few pixels, or was added to the blog using an older media mechanism that the plugin does not recognise, but still, it works, and from what I saw the images do look better. Still I only do periodic checks so if you see issues, please let me know.

Flattr this!

High Resolution Graphics


There is a demonstration Pixel Chromebook at work, so I tried looking at this blog using it. The display is really nice, with text appearing crystal sharp – from what I have seen of retina display Macbook, it’s the same. The difference seems even more dramatic as the resolution transition on mobile phones.

While the blog rendered fine (the browser is Chrome after all), the low resolution of the bitmap was showing, much more dramatically than on the mobile phone. Some observations:

  • SVG images look very sharp
  • Emoji character look very sharp, except when they are not included in the font used by Chrome.
  • Bitmap images within the text appear much more blurred than the background images, probably because they are within the centre of focus.

While there are tricks to selectively serve different image resolutions to different clients, this is a hassle, because it means twice the graphical resources and more work. SVG images are fine for logos and symbols, but what about photos? I found an interesting article about high-resolution graphics, the quick take-away is that for photographs, having a photo at twice the resolution (i.e. four times the pixels) but with a much more aggressive JPEG compression (so that the file is four times smaller), you end up with similar visual quality on classic display, and improved resolution on high-resolution displays. This makes sense, instead of algorithmically a value for each group of four pixels, you send all four of them, but give much more slack to the compression algorithm, in the worst case, it will average the group of four pixels, i.e. do the same as down-sampling does.

I think the example in the article with the logo make little sense, sending such data in JPEG format makes little sense, he would be better off with either an SVG version, or a PNG. I would need to check how sending high-resolution, low-colour count PNG works out.

Flattr this!