Re: The Great UTF-8 SWITCHEROO

On June 24, 2005 at 06:17, Jeff Breidenbach wrote:

> It's been brought to my attention that the Great UTF8 Switcheroo
> on June 19th may have had some side effects. Some lists are 
> showing  some corruption on index pages. Not a complete disaster,
> but fairly annoying. 

Not exactly corruption.  What your are seeing is the "raw" version
of the subject text.

> For example, on brygforum, things look reasonably ok after June 19th,
> but before then subject lines are misdecoded and show up prepended
> with strings like  =?iso-8859-1?. 

You are right that the transition changes is the triggering factor.

The technical reason is that mhonarc assumes that all non-ASCII
encoded data gets decoded when a message is first read when TEXTENCODE
is enabled.  Therefore, a separate routine is used when converting
resource variables (like $SUBJECT$).

Ideally, TEXTENCODE is enabled when an archive is initially created.
I did not consider the implications when TEXTENCODE is enabled
for existing archives.

It should be technically possible to write a script to "update" an
existing mhonarc database file so all non-ASCII encoded information
is decoded and converted.  Drop me a note if you are interested.

--ewh

_______________________________________________
Discussion list for The Mail Archive
[email protected]
http://jab.org/cgi-bin/mailman/listinfo/gossip