Slashdot Mirror
Open-source Challenge To Exchange Gains Steam
jbrodkin writes "An open-source, cloud-based e-mail alternative to Microsoft Exchange called Open-Xchange has signed up two new service providers and predicts it will have 40 million users by the end of 2011. Based in Germany, Open-Xchange has tripled its user base from 8 million to 24 million paid seats since 2008, with the help of three dozen service providers including 1&1 Internet, among the world's largest Web hosting companies. Microsoft is still the 800-pound gorilla, with a worldwide install base of 301 million mailboxes in 2010, expected to reach 470 million by 2014. But Open-Xchange is luring numerous service providers who are wary of Microsoft's attempts to compete against its own partners by selling hosted e-mail services directly to its customers."
We have our email hosted on google apps. It's free for 50 employees a year and we couldn't be happier!
Shameless promotion. NetworkWorld has too many of them. Maybe Slashdot should stop posting articles from them?
And I tried to use Open-Xchange once. Not quite that usable, unless they've made big progress.
Why is this better than MS? And can someone still remote wipe all my iStuff remotely? (with/out my permission)
Don't worry about Open-Xchange, OpenChange + SOGo is the real open source alternative:
http://www.openchange.org/index.php/component/content/article/7-news/55-openchange-and-sogo-the-first-interoperable-and-exchange-compatible-groupware-solution
- OpenChange Server is a transparent and native Exchange replacement for Microsoft Outlook users working on top of Samba 4. With OpenChange, you don't need costly MAPI connectors anymore.
- SOGo is a reliable groupware server with a focus on scalability and open standards. Let your Mozilla Thunderbird/Lightning, Apple iCal/iPhone, BlackBerry and now Microsoft Outlook users collaborate using a modern platform.
No per-seat CALS or license fees whatsovever.
Judging from a cursory perusal of the PCI DSS quick reference guide, as long as the business has in place a policy which forbids sending payment card numbers over email in the clear, it should still be able to use a cloud-based email solution. Do you have personal knowledge which contradicts this?
Here's the direct link to go read about it if you don't want to go through the networkworld blogspam article: http://www.open-xchange.com/
The "Server edition" is $1300, and they make you open a blind link to a PDF to figure that out.
Here's a handy feature matrix but noticeably absent is the free "community edition": http://oxpedia.org/index.php?title=OX_Product_Matrix
Also, the activesync thing (oxtender) is completely non-free and only available in the licensed versions.
Prevent linux based DDOS's!
http://linux.denialofservice.org/
I’m just curious as to how much it would be to setup (hardware costs) a local BSD or linux SMTP or pop mail server for say a business of 50 or 100 people?
It needs an immense expertise.
NB: The message above might reflect my opinion right now, but not necessarily tomorrow or next year.
For that few of people almost anything would work. Email isn't processor intensive, spam filtering can be a little but it isn't that much. Just need a lot of disk space per user. For personal email I use a low wattage atom processor, for a business I'd go with any raid system for a little redundancy. Software wise postfix (SMTP), dovecot (POP/IMAP) and amavisd for some spam filtering is a good fit. Plenty of walkthroughs on how to do it if you've never used the software, FreeBSD has these all as default ports to make it easier. I've used this general setup for a regional ISP and it's solid. It doesn't take much for admin and maintenance either. There are optional installs for web interface to control spam filters through like a squirrelmail webmail and other tweaks but those aren't bad.
If the business already has an IT person comfortable with BSD/linux it's a viable option. If not google apps can save a lot of headaches for the cost, or other alternatives like this. Also it depends on if shared calendaring and outlook integration is already part of the business. SMTP Email is easy but if the place is used to outlook calendaring that's another matter.
Find an old malware-ridden Windows machine that someone is dumping, and obtain a Linux or BSD LiveCD from mail order for between $5 and $10.
Total cost - can be between $5 and $10 plus 20 minutes of your time.
For example, I think Slashdot needs to come up with an alternative logo for Microsoft stories. Sure, the old one was really stale - but at least it looked like a Borg. With the new one, it just looks like Gates is wearing a really poorly-designed Bluetooth phone headset.
#DeleteChrome
Are there no decent wizard based setups?
Though i could see that getting everyone able to connect to the server could be frustrating. A large number of Linux distributions come with a mail server on the install disk (are these just never used?). Yast (openSUSE) has wizard that appears to allow you to set a mail server in 3 min if you can figure out what to put in a few fields (A reasonable tutorial should be able to solve this).
For those of you frustrated as I was at the labyrinthine task of actually finding the free version:
http://www.open-xchange.com/en/products/open-xchange-appliance-edition-en/download
I looked into OX a few months ago but actually dismissed it because, despite the "open" in the name, I couldn't tell that it was open source.
First of all, it is technically open source, but the license the community edition uses means it cannot legally be used by businesses.
It is definitely not a free alternative to M$ Exchange.
Each user license costs $52 for this product, an M$ Exchange CAL costs about as much, maybe a few bucks more.
Whoever designed the web access GUI went icon crazy and they are not very meaningful either.
Outlook Web Access is simple, this contraption had me guessing at what buttons do.
I manage an Exchange 2007 environment with roughly 700 users depending on it.
Originally having no experience, I got a test server up and running within a day.
The administrator tools are simple, powerful, and reliable; overall we have not had any serious issues in the past three years.
I also know that if something goes wrong, there is M$ support, service packs, backup software, DB repair tools, forums, etc.
Here is what happens with an open source product:
You install the product and spend the next couple of hours wading through text config files.
When you do manage to get the product to work, the thing does not work as expected.
You spend the next couple of hours cranking up debugging output and wading through source code.
If you are really masochistic you end up compiling your own build after you have found a bug.
Now in some cases going open source is worth the pain, especially when it brings additional functionality and cost savings.
Unfortunately, this open source product has the goal of duplicating functionality at a similar price point.
An additional thing to consider is that most open source products need more maintenance and labor.
This additional labor is highly in demand and is not at all cheap, which might make this an even more expensive solution than the original.
The problem with this view is that it is missing some functionality that people now consider part of email thanks to Microsoft and Outlook/Exchange or Lotus Notes/Domino. If you have never worked in a company that makes use of these features you wouldn't understand - but if any of your coworkers have they will expect them from you and will find your IMAP mail system to inadequate and unacceptable.
First is Calendaring - inviting people to appointments and booking in meeting rooms and shared resources (projectors etc) to those meetings. They even will recommend times when all the attendees and equipment is free. If you change the time it informs everyone and moves in all their calendars. This is not to mention sharing your calendar with others so everyone can keep track of where/what your team is up to. And you can do all of this on your mobile phone (ActiveSync or Blackberry) and have it update your server/client immediately.
Contacts - you can see all the people in your team, department and company. You can share your contacts with your coworkers. When you or they change them your phone updates with the changes immediately. I've seen our director's assistant add contacts to his mailbox via Outlook and he can call them from his phone's contacts within less than a minute when on the road.
Delegation - your assistant/gatekeeper or the person filling in for you when you are on leave can respond to your email and meeting requests on your behalf. It even says Susie Q on Behalf of John Doe etc. You can also have a departmental or a support or an information mailbox that many people check and share responsibility for.
Not to mention that Exchange offers the significant advantages of a large ecosystem of applications, tools and trained professionals that can back it up, maintain it, fix it, merge it, replicate it and all kinds of other things that you will eventually need to do in the life-cycle of an average modern mail system. I am dealing with a merger of two companies at the moment and them both running Exchange is a godsend - I'm glad it isn't an OpenExchange system I am having to merge with...
Can't they hire someone to do it for them? But may be the cost would exceed using Microsoft Exchange.
I'm pretty sure that if you need to have absolute confidentiality, this ain't going to do it. For that you need to pass the info verbally, without records, in an unexpected open space. Better security than that can be had by the maxim: "A secret is a fact known by one person only. If two people know it, it's information. If three people know it, it's rumor."
Help stamp out iliturcy.
SMTP or pop mail server
or?
OpenChange is an actual FLOSS project without a "community edition" and a $$$ "server edition". It integrates with Samba and implements the MAPI protocol.
OpenChange is very promising, but hardly production ready.
SOGo is not a feature per feature match for OX, Scalix, Zimbra or Zarafa. These are all mature projects with a large installed user base. If you are worried about license fees (which usually include paid support), you can always use the free editions of these projects and not use Outlook.
It's not that impressive, really. Zimbra has gone from six million paid mailboxes in 2007, to 40 million in 2009, to 65 million in 2010.
I do believe all these commercial open source projects should work together on the OpenChange project to finally get rid of those Outlook connectors.
http://citadel.org/ Citadel uses a proper database back-end and can handle terabytes of mail for thousands of users.
Excuse me, but please get off my Pennisetum Clandestinum, eh!
OpenChange seems to be a layer for implementing the proprietary exchange protocols used by outlook, so how difficult would it be to make openchange talk to the free versions of these other projects instead of having a plugin on the client?
http://spamdecoy.net - free throwaway anonymous email - avoid spam!
A unix mail server will usually work out of the box for simple use (ie one domain, add users to the os)...
http://spamdecoy.net - free throwaway anonymous email - avoid spam!
I have been working at a major user of Exchange which has recently merged with another major user of Exchange. Mail goes down for periods of time. Mailboxes are 'lost' to parts of the merged bank. Performance has become a joke and senior management are in a state of denial (incidents are being downgraded). Oh, and the process was run by a major outsourcing provider who run the IT services. Yes, it seems that Exchange is not maintainable. The backups are iffy at best and even with the highest level of support, there isn't sufficient expertise to run it.
See my journal, I write things there
IME, that's not Exchanges fault..
This would be funny if it wasn't true in a way. Many non-technical users are treating computers as a system: to them, software and hardware are hard to distinguish. If it doesn't work, you get a new one. Plenty of perfectly good (hardware-wise) PCs have been tossed because Windows has ceased to operate (usually due to malware).
A successful API design takes a mixture of software design and pedagogy.
SOGo already uses open protocols (e.g. CalDAV, CardDAV, GroupDAV, IMAP) for integrating with clients that support open protocols. For example, it works with Apple's iCal or Mozilla Sunbird for calendaring by using this protocol. You only need the OpenChange when using Microsoft's client - anything else should work directly.
I am TheRaven on Soylent News
OpenChange Server is a transparent and native Exchange replacement for Microsoft Outlook users working on top of Samba 4. With OpenChange, you don't need costly MAPI connectors anymore.
Hmm, when i checked a few minutes ago Alpha 13 was the latest Samba 4 release, I certainly hope that no business runs alpha software, so for all practical purposes OpenChange still requires the MAPI connectors:
Exchange 2003 (now 8 years old) was really I/O heavy and wasn't really designed with large mailboxes in mind. Think back to the average mailbox and attachment size in 2003 (what was your HD size 8 years ago for example) and I think that they thought they exceeded what was necessary for a mail system but it is not really workable for a large organisation with modern needs any longer and buckles a bit under modern expectations - especially on older hardware.
2003 did a few things like single-instancing within a mail database which contributed to I/O and required them to limit the size of DBs to ~75-100GB. So in a large organisation you need many many mail databases and managing them all gets a bit overwhelming.
In Exchange 2007 they did pretty much a complete rewrite and removed single-instancing of everything but attachments reducing the I/O by ~70% for the same workload. In Excahnge 2010 they removed even the single instancing of attachments (if you send an email with an attachment to all staff of a 2000 employee company it stores that 2000 times) but were able to improve I/O by 70% again over 2007. It means you need alot more disk space and a mail archiving solution but storage is cheap these days while I/O is not.
The product has gotten much much better and more scalable in the last two versions. Your IT department either needs to do better with it's storage subsystem to provide 2003 with the necessay I/O (FiberChannel or 10 Gig iSCSI SAN with lots and lots of spindles, transaction logs on RAID10) and/or upgrade to a newer version of Exchange.
It should be possible. I know it has been done for z-push (open activesync, originally from Zarafa) and Zimbra, an unrelated product. Implementing OpenChange for these projects might not be in the best interest of the involved companies, since most of their revenue comes from charging for their specific Outlook plug-in.
I remember recieving 3GB uncompressed avi files constantly from co-workers in my mail, honestly.
Change is certain; progress is not obligatory.
It's nice that the source on these hosted applications is open - no sarcasm intended. But since they're hosted, the source is open "read only": as a user, if I change the source to do something different, I can't commit it to the source of the hosted app to change it. Only the hosts can. Unless some host is running instances of the server per hosting customer that can be revised, which I've never heard of.
And then who's going to be the newly featured server admin? That's the really expensive and hard part of having an Exchange server.
--
make install -not war
With these hosted servers, is the administrator they provide any good (skilled, responsive)? Does their customer service quickly resolve issues that can't be immediately processed by the GUI?
--
make install -not war
This is also the case for many of the commercial open source offerings. I know at least Zimbra supports the four protocols you mentioned. These protocols are really great and completely open, but their biggest drawback is increased configuration on the client side - not very user friendly.
Do any of these hosted Exchange replacements allow custom app development to the server's API? Or are they just another black box that just replaces Exchange with some other magic box?`
--
make install -not war
If I switch my 50 user office from Exchange to Zimbra, what will I lose in functions? Will I be able to point Zimbra at a database whose schema I can edit and populate with other apps, whose objects I can CRUD from other apps, including ones I write?
--
make install -not war
Anyone offering hosted Exchange or an alternative "Outlook server" integrated with support for desktop VOIP phones (US48 unlimited minutes, or $0.02:min), at under $35 per month, that has 99.999% annual uptime and good customer service? With an API for integrating my custom apps to its features?
--
make install -not war
Inverse, the company behind SOGo, has also released extensions to Lightning and Thunderbird, that provide a better groupware solution if you want a fat client. User configuration is pretty simple with this installed, but it seems like a red herring - anyone using Exchange is likely to have an IT department pushing out preconfigured client installs.
I am TheRaven on Soylent News
That's one way to get it gaining steam. Call it "cloud based." Because the "Internet" isn't cool anymore. It's got to be "the cloud!" Marketing...
Really? Setting up Sendmail + Dovecot took me a couple of hours last time I did it, and it's something I only do once every few years so I generally need to reread all of the documentation each time to remember how it all works. This is on FreeBSD, with OpenBSD's spamd in front of sendmail. The hardware costs are tiny. Any relatively modern machine should be able to handle a few hundred clients.
Webmail can be a bit more effort, although simple things like SquirrelMail are practically configuration-free (they use the IMAP server for authentication, so just point them at the correct server and you're ready to go).
Or you can just grab an appliance-style distribution. There are quite a few available, just pick one, install it, and you're done.
I am TheRaven on Soylent News
spam filtering can be a little but it isn't that much
OpenBSD's spamd is now in ports for FreeBSD, and it has very low CPU usage. Putting this on the front line can significantly reduce the amount of CPU time spent running your bayesian spam filters.
I am TheRaven on Soylent News
Two weeks ago, I knew next to nothing about mail administration. I do however have enough experience as generic sysadmin. Took me about 3-4 hours reading into documentation for smtp, imap, exim (+addons), then about half an hour of configuration and now our working group (30 people) has a nicely working public facing mail server, all with aliases, mailing lists, synchronisation,...
The law -- and I haven't read it
This does not inspire a high level of confidence in anything which follows.
While it makes sense to have outlook support without requiring a plugin, they don't have native activesync support so you now need a plugin on many types of phone handsets... Last i checked, the funambol plugin was quite sucky...
And sure an iphone will sync using caldav/carddav/imap but it won't do push email and you can't remote wipe the device this way...
http://spamdecoy.net - free throwaway anonymous email - avoid spam!
It is true that OpenChange and SOGo look very promising and I am following the news with quite a bit of interest. One day it will be production ready. That said, Open Xchange is open source to a point. I think Open-Xchange is more crippleware because you have to buy the product in order to get Outlook integration, or at least the last time I looked into it.
Every time groupware/Exchange related topics appear on Slashdot (often as not an "open" replacement solution that isn't quite open, or not quite there, or both), I see a couple of references to the Citadel project:
http://www.citadel.org/
This appears to be a very interesting offering, and I've never understood why it doesn't generate more buzz. Can anyone knowledgeable in this subject explain what is lacking in Citadel to make it a serious contender in this domain? It is compatibility with Outlook/Exchange, missing features, not scalable, or ...?
"I object to doing things that computers can do." -- Olin Shivers, lispers.org
and just what makes your alpha assumption?
i could very well put alpha on all of my apps and know they are rock stable. and some proprietary ??finished?? softwares really don't deserve final version.
alpha very well depends on how you develop
- if you add feature, make it rock stable, add feature... your alpha will just miss some features but still be better than most finalized
- if you add all features and then start polishing... your alpha will really be a piece of shit like you describe
Signature Pro version 1.13.2-3 release 83.5 beta3try7 after-breakfast edition
This is an area I have been following with interest, as a number of clients have asked me about ditching their Exchange servers. There are several "open source" alternatives to Exchange, all with their own drawbacks. The main ones I know of are Scalix, Zimbra, Zarafa, OpenXchange, Citadel, and OpenChange/SOGo, although there are others.
OpenChange looks the most promising in the long term, as I believe it's the only one that promises 100% open source compatibility with Outlook. All the others require some kind of plugin, which generally isn't open source. However, as others have noted, OpenChange is nowhere near production ready.
So far I've been recommending Zarafa to clients, because it's the only one that includes an open source ActiveSync plugin for mobile synchronisation (it's called Z-push). Their support is also fairly good. I haven't tested the other alternatives extensively enough to see how they compare in practical terms though, it would be useful to see a simple objective comparison of them (certainly much more useful than fluff pieces like TFA).
https://alephnull.uk/
With mplayer allegedly about to go 1.0, Duke Nukem Forever supposedly releasing in May, and E17 libraries declaring 1.0, one simply must assume that Samba 4 and Hurd would *finally* release sometime this year too.
XML is like violence. If it doesn't solve the problem, use more.
The thing about Exchange competitors is unfortunately playing nice with Microsoft's monopoly which means outlook is *everywhere*. Microsoft never makes efforts to support open standards, therefore Outlook will never work well with something like citadel (sure, imap is there, but doing anything more than that requires, surprise, a commercial add-on).
The problem for groupware is not that Exchange is so fundamentally awesome beyond hope of competing on a level playing field, it's that it rides on the success of MS office.
XML is like violence. If it doesn't solve the problem, use more.
I have a user overseas who decides to buy a new Blackberry, so I go to the BB enterprise console, find the pin set a password and give the user a call. The device starts synching wirelessly with the users mailbox, (usually). Can you do that with Open Exchange?
Get up!
We just spent $100,000+ on a new management system for our company. One of the reasons was integration with Office. Drag and drop an email and it automatically attaches it to the customer including any attachments in the email. Does this system do that?
The tools available to integrate Office apps into third party software are the real reason Office is the "800 pound gorilla in the room" and will continue to be.
What google offers for free is just amazing. If you have less than 50 employees, or if you're a school, everything is free. Just pay $10 a year for a domain name. Google even offers an easy migration path from MS-Exchange. You also get google docs, google sites, and google calendar, all for free. Google provides all the storage, and most of the admin, also free. Hard to beat the price.
position #1 being the recurring myth that ISPs are subject to common carrier regulations.
The only reason why ISPs are not subject to common carrier regulations is because the FCC is wildly (and opportunistically) misreading the Communications Act, which defines "Telecommunications" as follows:
Sounds like Internet access service, no? In fact there has _never_ been a better example of telecommunications service than Internet access service.
So while you are technically right as a matter of current federal regulation (which can be changed/corrected at any time by the FCC) that Internet access service is not subject to Title II common carrier regulations, all those Slashdot users you so blithely accuse are arguably right as a matter of law. It is only the worst kind of opportunistic FCC legerdemain that keeps the law from being enforced as written.
Title II is far more sound as a basis for net neutrality regulations than whatever crazy legal theory the FCC has been promoting lately to regulate broadband access while simultaneously pretending that it is not a form of "telecommunications". And if they are forced into that position, that is precisely what the FCC will do - read the law the way it was written, objections from all the people who are horrified at the prospect notwithstanding.
One year ago, we were looking for an alternative to implementing Exchange as our central groupware-solution.
Everything we were offered was a huge failure on seemingly obvious accounts.
stability - various connectors would crash
not enterprise ready - storing the mail in the accounts of the users, not on the server
inability to connect smartphones - blackberry stuff etc. wont work
no proper support for commercial virus scanners - (!!!) nuff said
no backup/restore of individual mailboxes or individual mails - "Ummm, I've accidentially deleted that one super-important e-mail, Mr. Admin?"
They're not ready. Exchange isn't just the 800-pound gorilla, MS knows what they're fucking doing and the FOSS-rest does not.
So then, how are your bare metal install trials from backup going? Have you done one yet? You'd better do one so you can give people depressing but accurate downtimes when you have to do it for a server people are waiting for. Maybe it has improved with this new version and if you have gone to that trouble to really learn what you need to run this software you will be able to tell me. Of course you do have a spare MS Exchange licence for the purposes of disaster recovery don't you?
People who say positive things about MS Exchange are typically only end users that know almost nothing about it and don't remember things like the patch that made the MTA an open mail relay by default, or don't know about problems with copying, moving or doing anything with corrupt mailboxes. Has it really improved enough now that MS Exchange admins such as yourself that know enough about other environments to make comparisons such as above actually like the platform?
One thing I do agree with is that an MS Exchange clone will always be chasing the tail. One thing I disagree with is the widely spread Lotus notes/MS Exchange idea of having a tottering tower of interdependent applications and pretending they are all one thing and thus superior to a pile of independent applications. If you only care about email there are many better solutions, only care about a calender there are plenty there etc.
The main reason I no longer use MS Exchange is because my current workplace only cares about email and even a shared calender would not be used much. It's about the worst solution there is if all you need is email simply because of the way it stores and backs up email. Just about every other email server on the planet stores email in such a way that you don't need to muck about with the email server at all to get lost information from backups, search archived mail etc etc. Truly reliable backups and restores were such a pain on MS Exchange that to this day if I had to use it I would alias all the mail to also go to a box with a mature MTA that hasn't lost email since it was in beta.
Google offers everything you mention, and more.
No, you are making the mistake looking for one single application that does everything the pile of applications packaged as MS Exchange does.
If you want something to use commercial virus scanners to scan mail there's MailScanner and a pile of others that will give you a choice of a dozen or more commercial virus scanners as well as a front end to spam filtering.
If you want an integrated system that does everything with a lot of different bits (eg. MS Exchange) then you should consider it that way instead of just looking for a little bit and get disappointed when it doesn't do everything an integrated system does. You looked in the wrong place and asked the wrong people if you asked anyone at all. The words "system integration" are a clue for the next similar project.
That said, a MS Exchange environment is full of so many weird propriety quirks that it isn't even properly compatible with itself over some version changes. Once you are on it any migration, even to a new version, comes with a lot of pain. With the growing habit of idiots using their personal mailboxes as business critical databases you need very compelling reasons to seriously consider migration. MS Exchange is a horrible choice for a new environment but if it's already there you need to keep it alive as well as possible. The point about backing up almost made me laugh because the procedure to restore messages to a different machine in MS Exchange was a horribly time consuming exercise where you actually even had to rename the server to the name of the original (plus have company name and few other things identical). Because of the way the email is stored you can't just take a snapshot of the mailboxes to backup and assume it works. Backups suitable for bare metal restores with everything in every mailbox (the sort of true backup you get with EVERYTHING ELSE) are a pain to do in MS Exchange. What MS Exchange calls a normal backup doesn't get everything you need to get all of the mail back.
Open-source Challenge To Exchange Gains Steam
Wonderful. When did Valve get into Groupware?
Shai Schticks:"You don't make peace with friends, you make peace with enemies"
Tried to build some sogo version and was getting output to console instead of logs. I've said how I was building packages and developer said that I was applying wrong patch set. My problem - my build command does not set which patchset is used. Had to rebuild n times until I noticed that build process uses wrong patch. Somebody updated main patch, but kept old patch version in package build scripts. Quick look at software repo - holy moly that's the same person who kept telling me that I am doing it incorrectly. Reported outdated patch as a bug - response by same shmuck = not a bug. COTS support practices in their best.
Reported security issue in webmail. They kept it unfixed until I reminded about it. When they fixed it, somebody forgot to inform bugtraq about the problem. Security fix is only "fixed something" entry in changelog.
Created translation and submitted it. It took 10 months to get some response. Response asked to update translation, which was never included in sogo.
Asked to fix incorrect display of some character sets in webmail. They "fixed" it. Now webmail does not display any text in those charsets. I say that I can reproduce it in all my Debian Lenny setups. They say that they can't reproduce it and blame iconv although code in question does not call iconv directly.
Thunderbird integration consist of three plugins with hardcoded groupware server address. Those plugins fsckup with Thunderbird user preferences. User sets some pref - restart program and default preference is back. User turns off display of invites - restart program and invites are back. Plugins are incompatible with other plugins I use. Reported compatibility bug with other plugin. reply and fix from other plugin developer - less than 1 week. Reply and fix from sogo - waiting for more than a year. It took n months to finally get plugins for Thunderbird 3. I had to choose between Thunderbird2+groupware+broken message forwarding and Thunderbird3+no groupware+working message forwarding. Guess which part is more important. forwarding that does not lose text or test version of groupware.
Recheck what "reliable" means.
Your rant about patchsets and logs is hardly an argument against the quality of the product itself. What is argued here is that it's the only free software combo that offers native Outlook connectivity.
Idem for the rest: try reporting bugs to the Exchange team and see how long it takes to fix... this does not prevent Exchange from being used by millions of sites worldwide.
Back to SOGo, it's probably hard to get it properly setup at first, but once this is done, the program itself works quite well. Again, the idea here is that you don't need a proprietary license to obtain more functionalities than the webmail interface...
- OX is not crippleware or "commercial open source". Everything except the connectors to proprietary stuff (Outlook, EAS) are open source.
- The UI license is Creative Commons non commercial share alike, it doesn't prevent commercial companies from using the free version of OX at all, it only prevents companies from selling services based on OX for money without sharing with OX. More on the licenses here: http://www.open-xchange.com/node/783
- There is a free connector for Thunderbird / Lightning here: http://www.open-xchange.com/node/949
- All API's of OX are open and free, if you feel like building your own Active Sync or Outlook connector there is a nice, GPL'd Sync Engine that already does a lot of the work, go for it. http://oxpedia.org is you playground
- OX is pretty mature, has 24 million users, provides a rich messaging and collaboration stack built on Java and AJAX with business class functionality, generates a lot of business for our partners and thus finances all the good open source stuff we do - what's so bad about this?
Cheers,
Rafael / CEO of Open-Xchange