Filesystem size limits and the complication of when errors are detected

One of the practical things that further complicates limiting thesize of things in filesystems isthe issue of when people find out about this, or rather how thisinteracts with another very desirable property in modern filesystems.

For practical usability, you want people (by which I mean programs)to be notified about all 'out of space' errors synchronously, whenthey submit their write IOs, or more exactly when they get back theresults of submitting one (although this is effectively the samething unless you have an asynchronous write


A small irritation with Go's crypto/tls package

I generally like Go's TLS support in the crypto/tls package , to the small extent that I'veused it; things pretty much work in straightforward ways as I'dexpect and it's easy to use. But I do have one issue with it, andthat is that it doesn't supply a .String() method for any of theTLS related constants and types it provides.

The obvious problem is that this leads to duplicated work for everyonewho wants to report or log


Ubuntu LTS is (probably) still the best Linux for us and many people

I write a certain amount of unhappy things about Ubuntu here. This is not because I hate Ubuntu, contrary towhat it may appear like; I don't write about things that I hate,because I try to think about them as little as possible (cf spam ). Ubuntu is a tool for us, and I actuallythink it is a good tool, which is part of why we use it and keepusing it. So today I'm going to write about the attractions ofUbuntu, specifically


Remembering that Django template code is not quite your Django Python code

Recently I was doing some work with our Django web application where I wanted to do something specialfor one particular field of a form. As is usual in Django forms,we iterate through the form fields with the ' {% for %} ' templatetag .Stripped down, this is something like:

{% for field in form %}  {{ field.label_tag }} 
{{ field }}{% endfor %}

I had to make one field generate extra HTML, which


My little irritation with Firefox's current handling of 'Do-Not-Track'

The Do-Not-Track proposed HTTP feature was either a noble or naive attempt by variouspeople to get websites not to track you if you asked them not to.It worked about as well as you'd expect , whichis to say not at all in practice. Allegedly, for a long time havingyour browser send a DNT header made it easier to fingerprint youbecause so few people did it that you stood out all the more.

(This may no longer be the case, for reasons we'


Some magical weirdness in Django's HTML form classes

In Django, HTML forms are defined in what seems to be the conventionalapproach; each form is a Python class and fields are variables definedin the class definition. This looks like this, to crib a skeleton from our Django web application :

class RequestForm(forms.Form):  login = forms.CharField([...])  name  = forms.CharField([...])  email = forms.EmailField([...])  aup_agree = forms.BooleanField([

The Ubuntu package roulette

Today I got to re-learn a valuable lesson, which is that justbecause something is packaged in Ubuntu doesn't mean that it actuallyworks . Oh, it's probably not totally broken, but there's absolutelyno guarantee that the package will be fully functional or won'tcontain problems that cause cron to email you errors at least oncea day because of an issue that's been known since 2015 .

I know the technical reasons for this, which is that Ubuntu pretty


Googlebot is both quite fast and very determined to crawl your pages

I recently added support to DWiki (the engine behind WanderingThoughts ) to let me more or less automatically generate'topic' index pages, such as the one on my Prometheus entries . As you can see on that page , the presentation I'm using has links toentries and links to the index page for the days they were postedon. I'm not sure that the link to the day is particularly usefulbut I feel the page looks better that way, rather than just havinga big list of


Magic is fine if it's all magic: why I've switched to use-package in GNU Emacs

A while back I tweeted :

I revised my .emacs to use-package for all of the extra packages thatI'm using (not yet for stock Emacs things, like c-mode settings). Itwas mostly smooth and the result is cleaner and easier to follow thanit used to be, although there is more magic.

Since I deal with my .emacs only occasionally, I think clean and easyto follow is more useful than 'low magic'. In practice everything inmy .emacs is


If you don't test it in your automated testers, it's broken

A while back I tweeted :

The Go developers keep breaking Go's self-tests when you have gccgoinstalled. Since they don't install it on almost any of theirautomated builders, I don't think they really care and so I'm donedealing with it myself. It's time to uninstall gcc-go.

I could file yet another Go bug report about this breakage due togccgo that could as well have been filed by an automated machine,but I've lost all my