"It all looks different this morning!"
I've been using chronicle since I started this blog in January of this
year, but the time has come to replace it.
I thought about modifying the code (it's distinctly cleaner than
much of the Perl I come across) but in the end I wrote my own from
scratch. So this blog is now being generated by aikakirja, named
after the Finnish word for a chronicle.
-
My main concern was that the chronicle version in Debian/stable
(4.6) has to rebuild the whole blog each time one adds or changes a
page or a comment, and this takes some time. The latest version 5.x
apparently does a bit better by storing page text in a database, but
I wanted to rebuild only those pages which had been changed (by
adding a comment, or a tag, etc.). The obvious way to do this was to
store modification times and some content in a database. (And so
I've been remembering SQL, which I haven't had to use for years.)
-
chronicle uses textile for comment markup (if it's installed), which
isn't UTF8-clean, and for no apparent reason quite a few web
browsers change entered text into non-ASCII smart quotes. And my
commenters ought to be able to use foreign characters if they want
to. (Not to mention, none of them has put in a textile markup
sequence anyway, possibly because it's not advertised.) So that's
been removed. Would markdown be interesting to commenters?
-
chronicle doesn't support emacs' org-mode as an input format. Fixed.
(emacs documentation writers mostly don't think you should be able
to do things on the command line, because you should be doing them
from within emacs, so that took some finding.)
-
The cross-references feature doesn't work the way I expect it to, or
with the template I like. Fixed. (And you can insert
cross-references to posts that aren't in the blog yet, and they'll
appear once those later posts are published.)
-
The ordering of comments is based on file modification date, not any
internally-stored date – not even the one in the filename. Fixed.
-
The font size used for the various entries in the tag cloud is based
on absolute numbers of posts including that tag, so tags tend to get
bigger as the overall number of posts increases. Fixed.
-
There's now an RSS feed for all comments everywhere on the blog
(/comments.rss) in case you want to keep up with
everything rather than watching individual posts. It's not so
high-volume that this should present problems.
-
URLs should be unchanged from the chronicle version of the site.
There are doubtless still bugs, most of them likely to be connected to
insufficient paranoia about rebuilding pages that might have been
changed by action elsewhere. If anything breaks, let me know. Source
code will be released once it's been tidied up a bit.
A merry Christmas to all my readers.
Comments on this post are now closed. If you have particular grounds for adding a late comment, comment on a more recent post quoting the URL of this one.