[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-devel
Subject:    Re: kio is a mess...
From:       Rik Hemsley <rik () rikkus ! demon ! co ! uk>
Date:       1999-05-24 20:37:27
[Download RAW message or body]


On 23-May-99 Alex Zepeda wrote:
> On Sun, 23 May 1999, Rik Hemsley wrote:
> 
>> You are correct. I learnt Qt through the API docs, and didn't have to bother
>> with the tutorial. Sadly the KDE API docs aren't that good (yet :)
> 
> Well you've got commit access :)

I think I've finally got the idea of how these things work now too :) I seem to
have forgotten (?) somewhere along the line that of course they all work with
URLs. At least I know now.

I will actually have a go at documenting this a little more. Even just the
example I got will be a good start.

I don't know how much more I can do though - your code is a little, er, weird
looking for me :)

> Actually, I'm not quite sure how long it'll stay connected, I should
> probably have it noop to keep the server itself at bay.  I'm not quite
> sure how long KDE trys to keep the slave alive either.

I think the POP timeout is 10mins (?) Doing it every five should work well.

>> By 'index' do you mean the output from LIST or that from UIDL ? I
>> could really do with both if you don't mind. The implementation should
>> be the same.
> 
> I'm thinking from LIST, but I suppose another "command" could be done to
> get the unique identifier (or a list thereof).
> /me goes and checks the imap rfc to see if this can be done with imap4

Yes, it's important that most of the POP RFC can be done, as things like UIDL
are (almost) necessary to allow more than just a big drag down of all messages.

> Did I mention, I hate imap4?

I'm with you on this ! I was so happy when you said you were working on an
ioslave, as it was going to be the thing I left 'til last :)

>> BTW my implementation can also handle the case where UIDL is not
>> available, as the message IDs are unique to the mailbox/folder and all
>> transactions are done using URLs as reference, so there's no
>> confusion.
> 
> I'm still not quite sure, what do UIDs/UIDLs gain you?  It appears you
> can't request message <UID>, can you?

No, you can't ask for a message by UIDL, but...

You need them if you want to avoid downloading a message twice. You can keep a
local index of those you already have. This way, if your connection drops, you
can reconnect, do a UIDL again, and only treat those you know you haven't
downloaded as new.

I'll try and give a hand with the pop slave if I can. At the moment I'm
fighting with Empath, because it's so close to being completely usable and I'm
desperate to start using it as my usual client. Also, although many people have
expressed interest in working on it, no-one has actually started on it, so I
think I need to get it working well to get more interest.

Cheers,
Rik


--
KDE - Colour outside the lines  : http://www.kde.org
[[without]] - software for KDE  : http://without.netpedia.net

[prev in list] [next in list] [prev in thread] [next in thread]