I've just checked some updates into svn, and updated hep.fettig.net. Changes:
* "Sites" are now referred to as "Connections"
* You can see the status of your connections, and try to reconnect if the last attempt failed
* You can change your password
* Yarn now limits the number of simultaneous connections it will try to make at once, which should keep Hep responsive during restarts, or when somebody imports a big OPML file
* It's possible to add users without restarting, so hep.fettig.net should only have to be restarted for upgrades
* You can log out of the web interface
Some of the new web pages are a little ugly, but I'll be cleaning them up soon.
Hep is now feature-complete for the 0.7 release, which I'm targeting for Wednesday. If you have an account on hep.fettig.net, please take it for a spin, let me know what you think of the changes, and report bugs!
I've been getting some strange behavior related to the MIME
encapsulation in messages. I occasionally open a message to find the
MIME encapsulation broken, i.e. the raw text of the message appears
rather than just the part that Mail.app usually displays by default.
In the process of investigating this, I extracted the raw text of the
message and discovered that the "boundary" parameter in the
Content-Type header did not match the boundaries in the message
itself. This explained why Mail.app was having a problem parsing the
message.
What was strange was that when I went to report this, the message had
*fixed itself*. I don't know if this was a temporary issue related to
restarting or perhaps an on-the-fly bug fix, but it was darned spooky!
:-)
If it happens again, I'll try to capture the sucker before it changes
and forward the raw message text to you.
> I've just checked some updates into svn, and updated hep.fettig.net.
> Changes:
> * "Sites" are now referred to as "Connections"
> * You can see the status of your connections, and try to reconnect if
> the last attempt failed
> * You can change your password
> * Yarn now limits the number of simultaneous connections it will try to
> make at once, which should keep Hep responsive during restarts, or when
> somebody imports a big OPML file
> * It's possible to add users without restarting, so hep.fettig.net
> should only have to be restarted for upgrades
> * You can log out of the web interface
> Some of the new web pages are a little ugly, but I'll be cleaning them
> up soon.
> Hep is now feature-complete for the 0.7 release, which I'm targeting for
> Wednesday. If you have an account on hep.fettig.net, please take it
> for a spin, let me know what you think of the changes, and report bugs!
Matt Huyck wrote:
> I've been getting some strange behavior related to the MIME
> encapsulation in messages. I occasionally open a message to find the
> MIME encapsulation broken, i.e. the raw text of the message appears
> rather than just the part that Mail.app usually displays by default.
> In the process of investigating this, I extracted the raw text of the
> message and discovered that the "boundary" parameter in the
> Content-Type header did not match the boundaries in the message
> itself. This explained why Mail.app was having a problem parsing the
> message.
I was seeing this myself in Mail.app a while ago, so I know what you're describing. I haven't seen it in a while, though, so I thought it had been fixed in the last batch of IMAP bugfixes.
The boundary issue you describe certainly seems like a bug, but the weird thing is that I've never seen this problem in the other clients I test on (Thunderbird and Outlook Express). I think if there was really a MIME boundary issue it would show up in those other clients as well. Hmm. I'll have to spend some more time testing on Mail.app, and see if I can get it to choke on a message.
> What was strange was that when I went to report this, the message had
> *fixed itself*. I don't know if this was a temporary issue related to
> restarting or perhaps an on-the-fly bug fix, but it was darned spooky!
> :-)
> If it happens again, I'll try to capture the sucker before it changes
> and forward the raw message text to you.
It's possible that this is just a Mail.app problem. I thought I had
just caught it again, but then it resolved itself (again) before I
could figure out what was causing it.
It's good to know that I'm not hallucinating the broken messages. Or,
at least, that we are sharing the hallucination! :-)
