Re: [Omail-devel] omail-webmail: what next?
Brought to you by:
swix
From: Carter S. <der...@de...> - 2004-03-18 07:02:35
|
Olivier, Yes, it does seem that omail-webmail is in need of a change of something. I'd just like to point out that one of the special features that omail-webmail has that other webmail packages don't seem to have is the sheer number of translations. Here are a few ideas that I've had which may or may not revitalize the project: - Implement omail-webmail using Template Toolkit (bugzilla uses it) to allow easier separation of code and UI. This allows a non-programmer to modify the UI without having to change the logic that drives the app. - Rewrite omail-webmail using the perl MailTool library. This would allow omail-webmail to access imap, pop3, maildir, mbox, MH, and Outlook Express (dbx) folders all at the same time or per user. I also question whether or not webmail-omail has a future with products such as squirrelmail being available, so I am curious to know what other users of webmail-omail think. Carter On Wed, 2004-03-17 at 14:00, Olivier Mueller wrote: > Hello, bonjour! > > Since a few months I'm wondering what to do with the > omail-webmail project: should it be continued or > stopped? This is what I'd like to ask you (the =~ 200 > people one these lists). > > The main concern I had 4 years ago, when I started > omail-webmail (forked from the Neomail project) was speed > & performance (& fun :) : it was then the case with direct > Maildir access compared to complex access via imap. > Openwebmail continued on this way for "mbox-based" accounts > (www.openwebmail.org), with many nice features, while > omail-webmail stayed a bit behind: there are many > features missing, especially folders support and > decoding of base64-encoded mails to name just two of them. > > But omail-webmail is now old and IMHO not secure > enough for "serious" hosting (it even came on bugtraq last year: > http://www.securityfocus.com/archive/1/334056 and there > are probably other security issues): I belive that > running a suidperl based CGI is not the best solution > nowadays. Other problem: the installation is still quite > complex (check the ML archive :). > > > Today it's easier to setup qmail + vmailmgr + courier-imap > and an imap-based webmail like squirrelmail than the > same with omail-webmail, there are even automated tools to > do that (cf freshmeat). And it's safer for the server too: > no more suidperl, and PHP is quite standard now too. > I remember that it wasn't the case 3-4 years ago, with > IMP/horde setup and uw-imapd libraries, etc. > > So the question is: what next? :-) Are there still > people/servers using omail-webmail? > > > >From a personal point of view, I think I'd rather concentrate > on other open source projects (omail-admin, ospam: spam > fighing add-on for qmail+vmailmgr, and phpmyadmin) than > spending more time on omail-webmail, trying to reinvent the > wheel and implement weird things (MIME support, attachments > decoding, folders support etc) which are already nicely > programmed elsewhere (squirrel, IMP, etc.). Or are > there people around willing to continue de developement > of omail-webmail? > > > All my customers are still using omail-webmail, so > in any case I will have to find a solution anyway :). Now > I'm thinking about an automated transition to > squirrelmail, with scripts to (server-wide) convert > omail's pseudo folders to courier-imap compatible IMAP > folders, and if possible do the same with the addressbook. > Maybe you even already coded something like that? > > > Feedback welcome! It was fun working with you :-) > > Regards, > Olivier |