Slashdot Mirror
Drop-In Replacement For Exchange Now Open Source
Fjan11 writes "Over 150 man-years of work were added to the Open Source community today when Zarafa decided to put their successful Exchange server replacement under GPLv3. This is not just the typical mail-server-that-works-with-Outlook, it is the whole package — including 100% MAPI, web access, tasks, iCal and Activesync. (The native syncing works great with my iPhone!) Binaries and source are available for all major Linux distros."
I've not looked at this software, but Exchange is one hell of a piece of machinery. Say what you want about MS, but I've seen an Exchange server with terabytes of email, gigabytes per day, keeping up fine. It's a pain in the ass sometimes to be sure, but I wouldn't trust my production network to this today anyway.
..I'll DEFINATELY be installing this for our company's mail server. I currently have Zimbra setup, which is very nice, but the bosses don't like it because it doesn't integrate into Outlook very well (iCalendar, contacts, etc), without the outlook connector that you have to pay for. No hate on Zimbra though...I absolutely love it's capabilities and ease-of-use, but it's a deal-breaker w/the management types if won't support the 'advanced' features in Outlook.
The nice thing about GPL software is that it's easy to go in and change arbitrary limitations like that.
Give me Classic Slashdot or give me death!
Actually you want to provide additional capacities so that going back to Exchange is a true downgrade.
"It's the height of ridiculousness to say for those 9 lines you get hundreds of millions."
...Zarafa decided to put their successful Exchange server replacement under GPLv3. This is not just the typical mail-server-that-works-with-Outlook, it is the whole package including 100% MAPI, web access, tasks, iCal and Activesync...
While I hail this development, I wonder what "successful" means in this story. Here are questions I might want answered:
Was it "successful" at sales? If so where are the figures? I would not really praise them that much if the original goal - to make money, could not be reached making these fellas to opensource everything...much like what Netscape did years ago.
Was it "successful" at actually replacing Exchange with no [significant] trouble for Systems Administrators? I need to know. How come it is not that known in IT circles? What's going on?
Over the last few months, I've been forced to use Exchange/Outlook a lot, and for the life of me I don't get the big deal.
I don't think it is a very big deal. I've supported Exchange servers in companies of various size, and it's really not doing anything astoundingly complicated, and what it does it doesn't do all that well. But still it does a bunch of things that other solutions haven't done an even worse job at, and does them all together. Things like "I can send a meeting invitation to my boss and his assistant can check his mail, accept his invitation, and reply on his behalf without actually logging in as him."
I know, it doesn't sound like that sort of thing would be all that important, and it's not even clear all the time that it makes a lot of sense, but there are companies that run on this sort of procedure. So there are a bunch of random things like shared calendars and push-email to phones that people don't want to live without, and unless you can provide a seamless replacement, you're stuck with Exchange.
I, for one, am eager to see a suitable Exchange alternative. I have a real love/hate relationship with Exchange. There are some options out there, but none of the options I've tried have worked out.
Sure, but there's a lot of difference between offering a similar feature set, and being a drop-in replacement that is compatible with all the crufty MS protocols.
Not until Q4 2008
From the features pdf
http://download.zarafa.com/zarafa/en/Featureslist620.pdf
"Integration with the Blackberry Enterprise Server to get email, calendar items, contacts and tasks real-time on your Blackberry. Available Q4 2008"
"In the end, there is simply no weapon more devastating than the truth, delivered in just the right way." - tnk1
Yet another open source exchange replacement that didn't open source everything required to interact with outlook.
Without that, whats the point?
Google for the debate on debian-legal about whether it complies with the DFSG. Anyway, the crux of the matter is that authors can embed unmodifiable sections in their code, and you are not allowed to alter that code even if you will not be giving copies of it away. The theory is that you're distributing the output of the program, which is part of the program itself - or some nonsense like that. This goes against decades of precedent for the idea of usage versus distribution.
For example, if you VNC to a machine on my home LAN, you could potentially run Quickbooks. It would be executed on my machine and exporting its display to yours, but no one would ever consider this to be distribution. However, if I were running an AGPL'ed equivalent of Quickbooks on my home web server and you accessed it, the authors of the AGPL would claim that I distributed a copy of that application to you. That's their legal theory behind restricting my usage of it.
Another poster said I was spreading FUD. Yeah, I am, and with good reason. I fear that some project I depend on may adopt the AGPL. I'm uncertain that I'd be able to use it given the additional restrictions that it piles on top of the GPL, to the point that I actually doubt it.
Dewey, what part of this looks like authorities should be involved?
why enterprises run Windows?
Does anybody have links to success stories of large(-ish) corporations converting to Zarafa?
thegodmovie.com - watch it
I've tried other Exchange alternatives, and with some of them, even if the directions look pretty easy, it takes a ton of tinkering to get the thing running.
Much like the actual product.
As a former MAPI programmer (don't worry - I've largely recovered) I have to point out that this is utterly irrelevant.
The only compelling reason to use an Exchange compatible server is to support Outlook. The issue is that Zarafa charges for the Outlook connector. This is not a new business model, people, and truth be told its been a fairly common paradigm of 'Exchange-killers' for quite awhile now. Nothing is killed until the connector is free. Full stop.
So why doesn't anyone offer a free connector? Because it is ridiculous amount of work to build and it is something corporations are willing to pay for. It's not that replicating the server functionality is difficult, it's that Microsoft twisted and violated open standards into something utterly unholy known as Exchange to ensure that nobody but Microsoft could communicate with it. MAPI is Microsoft's obfuscation of traditional messaging protocols and is infamously poorly documented.
I wrote about this issue for Redmond magazine about 2 years ago and nothing's changed. The connector is still the kicker and, regardless of how nifty the back-end is, until an open-source Outlook connector appears Exchange will remain one of MS's top 5 products.
Nothing but PR to see here. Move along...
Well, the reality is that senior executives have always had personal assistants (used to be secretaries) who really opened all their mail, sorted it, and typed responses to the mundane, and took dictation for the serious ones.
The executives typically have 100% trust in their admin's and this feature is absolutely necessary to the proper functioning of a senior management team. It may seem like a security risk, but in the cases that I am aware of, both users are aware of their status, and it rally operates like it did in the pen and paper days.
Suppose you were an idiot and suppose you were a member of Congress
Having others able to act on your behalf without using your login simplifies the process of proving that you did/didn't actually do something. The information about who logged in and did what on whose behalf can easily be logged. If, on the other hand, you have a system where your login has to be used to act on your behalf then the logs can only show your username no matter who actually used your account.
"Welcome to our world. We are the wasted youth. And we are the future too." Yes, I know these are stupid lyrics.
FUD? Well, I'll let others be the judge of that but the difference between GPLv3 and Affero GPLv3 are HUGE:
"Notwithstanding any other provision of this License, if you modify the Program, your modified version must prominently offer all users interacting with it remotely through a computer network an opportunity to receive the Corresponding Source (...)"
This is the FSF on steroids, it's the anti-ASP license and it's also unsuitable for any software you want to improve internally and not share if it in any way interacts with externals. It's still not an EULA as creating derivates is one of the copyright holder's exclusive rights (except fair use) but it's definately stretching copyright law to the absolute limit in order to force the release of code.
Live today, because you never know what tomorrow brings
You have to have an account specifically granted the privilege to do what he is describing, and you can place restrictions on what they can do. In his example, the secretary has the authority to say "my boss will attend this meeting", and that authority translates to his electronic calendar, the same way it would in real life. Sure his secretary could have a breakdown and screw up his schedule, but she could do that anyways just by not doing her job, and how many office environments have you seen where the secretary didn't frequently have physical access to her boss's machine while it was logged in?
Jesus H. Fucking Christ. 99.9999999% of all companies just want to buy a tool that works. They don't want to build the fucking thing. They don't even want to fix it. That's why they buy the support license. This whole 'we can customize the code if we want' is a huge stinking load of specious crap. Companies of any size BUY their software because they don't want to customize software they don't have to. Like office software. Customizing a huge billing system is one thing (if you are a big enough company to warrant doing that), but why would an insurance company, or a local widget maker, or a medical clinic want to become an email server programming company???? Get a grip. They'll go out and buy exchange or lotus notes or whatever because they just want the frickin tool. And if it is buggy so what? It works for the most part and they don't have hire programmers or keep programmers around to fix bugs that said programmers introduced when they screwed around with the source code. It's cheaper to pay for the licence for a year than to pay for an unneeded programmer for a year.
-- I ignore anonymous replies to my comments and postings.
If others can be delegated permissions to act on your behalf in specifically designated manners without logging in to your account, then, if the system logs who did what under what account, there will be accountability.
OTOH, if others can't act on your behalf without logging in as you, and you have a business need them to act on your behalf, you have no choice but to give them your access credentials (dongle, password, whatever) and then there really is no accountability, and no control over the manner in which they can act on your behalf.
So, rather than destroying accountability, supporting delegation enhances accountability (and security).
it's around 2,400-- but then you also require 25 outlook licenses.
once again, the price of the software is negligable compared to the cost of 25 employee's salaries...
go ahead, waste a week of each one's time teaching them whats different about the new program.
every day http://en.wikipedia.org/wiki/Special:Random
Well, project maintainers tend to be very lazy when it comes to things like Freshmeat, which is why I maintain something like 120 project records and semi-regularly update 50+ others that I've subscribed to. (I should be paid full-time by Sourceforge or someone for the level of work I put in. Yeah, right.) If project maintainers were, oh, a little more forthcoming and not hiding releases, we'd all be a damn sight better off. Google should be a good source, but the problem there is that it's so hard to search for anything where all your keywords are common words. There's no good, juicy, unambiguous keyword to use, and search engines just aren't any good at semantics. They're only useful for syntax-based searches. The FSF's software page is excellent for projects the FSF knows about and is associated with in some way (even if just as an umbrella or as a webpage host), but you'd not get much done if you only used software they linked to. A pity, but they can't do everything, though they do try to do exactly that. Sourceforge's search engine seems to turn up everything Sourceforge hosts, and there are a million and one Sourceforge-like sites these days. Getting information, therefore, depends massively on volunteers trawling every imaginable report, rumour and hint of Open Source and indexing it somewhere. And there just aren't anything like enough volunteers to make anything close to a dent in what's out there. Which is a real problem, as projects that nobody knows about WILL die. Even when something IS known about, if updates aren't announced in a meaningful way, it will also die. Likewise, if people don't contribute, the project will die. Or if the maintainer doesn't release early and often, the project will die. And even after all that, if it's not in any of the major distributions, there won't be a sustainable userbase or a large enough supply of bug reports and the project will die. In this case, I think most of the above apply.
It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
there are a lot more copmpanies out there that barely know what a PC is. Software houses are a tiny minority compared to retailers, to name one. Think how many back office staff exist to serve them compare to the number of programmers. Then go to the next class of business, repeat a thousand times.
I'd say 0.0000001% is a bit of an exaggeration, 0.1% is more like it.
it's around 2,400-- but then you also require 25 outlook licenses.
once again, the price of the software is negligable compared to the cost of 25 employee's salaries...
go ahead, waste a week of each one's time teaching them whats different about the new program.
This is server software we are talking about here. The end users don't change their software (that's the entire point). So there is no cost for retraining end users.
You would obviously have to train the server administration staff, but even if you did put in a "Genuine Microsoft" Exchange server, you would probably still have to do this.
Besides, even if the front end did change, a week of training is a LOT. As it would be replacement software, the concepts are the same, it's only which button you push to do it that changes. So if you can't train them in a matter of hours, if not minutes, you really do need new employees.
Where I work, we use a non-MS stand alone calendaring solution. Our end user training takes a couple of hours.
How long do you think it would take to train users to use the new version of MS Office?
Ever stop to think
That's not to say that all protocols developed by open processes are wonderful, but on average they seem to be better.
The MS protocols are crufty because the direction of the design of them is left up to middle managers.
That should explain it fully to anybody curious.
That being said, though, I would guess that the greatest contribution isn't actually the program itself, but rather the fact that it lays open the protocols involved, so that other MAPI servers could be written. Maybe it could even be implemented as another protocol for Dovecot?
"...they can finally have something that works!"
Hear, hear. And maybe they can finally have something that doesn't try to break protocol standards, introduce a non-interchangeable mail archival format, artificially create a need to have ten times as many server licenses as necessary... in short, businesses would do well to uh, swap it for something else.
You can't send a takedown notice to an already printed newspaper.
``Jesus H. Fucking Christ. 99.9999999% of all companies just want to buy a tool that works.''
You are right about that (well, maybe not about the percentage, but the general point), but that doesn't contradict anything I said. They _would_ rather buy something that Just Works. They _would_ rather spend money on software licenses than spend more money on developing their own software. And all this makes perfect sense.
``They don't want to build the fucking thing. They don't even want to fix it. That's why they buy the support license. This whole 'we can customize the code if we want' is a huge stinking load of specious crap.''
Now, I don't know where that comes from. Who said anything about having to "build the fucking thing"? Did you miss the part where Zarafa is called a "drop-in replacement for Exchange" and "binaries ... are available"? You don't have to build anything, and you don't have to fix anything. You seem to have confused what you _can_ do with what you _must_ do. Having to fix something is a Bad Thing. Being able to fix something is a Good Thing.
The comparison is like this:
On the one hand, you have Microsoft Exchange. There are various versions, each with their own bugs and limitations. If you need those bugs and limitations removed, you might be able to buy a different version, you might have to wait for a newer version to become available, or you might be simply out of luck. You pay for the software itself, and for client licenses - the more people use the software, the more you pay.
On the other hand, there is Zarafa. There will be various versions, each with their own bugs and limitations. If you need those bugs and limitations removed, you might be able to obtain a different version, you could wait for a newer version to become available (possibly with a patch from some company in the same position as you), or you could remove the bugs and limitations yourself. You can get everything for free.
Now, you tell me which seems to be the more attractive option.
Finally, I would like to point out that open source software tends to get easy installation procedures and low maintenance cost once Linux distributions start packaging it. Also, if running a particular piece of software is too tedious for you, you can always get it hosted by someone else. There is Exchange server hosting, and I imagine there will be Zarafa server hosting, as well.
Please correct me if I got my facts wrong.
And you think that exchange is admin friendly?