Slashdot Mirror
Can We Finally Ditch Exchange?
"With new releases on the way, like Mandrake 9.0 and the new Lycoris can we who try to use Free Software in business environments hope for any change? Do the commercial Linux distros have any plans to implement a free replacement for Exchange, including a Win32 client-side bridge? If not, why not? Do you feel it is too cost prohibitive to imitate Bynari in this case, or is it a decision more along the lines of 'we'd rather you used Evolution and Mandrake/Lycoris/Whatever, rather than OutLook and Win32'? If it's the latter I'd be severely disappointed, and I don't think I'm alone. Any discussion on this topic would be appreciated; but what I'd really love is a community push to get this done. Perhaps a running Web-A-Thon to raise the money to simply purchase the technology from Bynari? I personally think it would be a great move towards grabbing market share from some of the other distributions, some of which have the technology but choose to keep it closed, as well as from the Great Dragon. What do you think?"
Alot of companies/admins are waiting for an Exchange replacement. I for one have considered dropping exchange for a flat out mail server that runs in a *nix environment but it always comes back to the scheduling of exchange. With all of the people out there writing code, it still amazes me that nothing has surfaced. Why not take time off f the useless mp3 player/id3 reader/all of the other crap and contribute to a worthwhile project?
My sig of choice is Marlboro
This is not trying to be a troll, but it seems there is always one more "clincher" in the movement away from MS products. IE / Office / Outlook / Photoshop you name it, but now it is Exchange. OSS always makes a replacement, but it is only 98% there in terms of functionality in most cases. As soon as we get Exchange out of the way, there will still be something else left to take its place to prevent adoption.
Buying a Dell computer is equivalent to dropping the soap in a prison shower.
as for number 2, companies just need to buy a copy of redhat server, or whatever. As long as they buy a version of linux, they can get instant support. This excuse in general doesn't work anymore now that we have so many linux companies ready to sign deals.
Only dead fish swim with the stream...
If that were true, it would have already happened. There are several good calendar/messaging systems avaliable in the open source market however exchange remains as the corporate standard.
The reason for this is simple. Exchange uses the Outlook client; the Outlook client comes with Office; Office is the de-facto standard software for almost any corporation that uses computers.
At first I thought if there was an open source system that was compatible with Outlook that would do the trick, however HP offered a system that did just that, and even it didn't make a dent in Exchange's market.
second society
I agree.
However creating something , even Opensource that does what Notes and Domino do is quite a task. Do they do it well ? not hardly.. but its effective and a great many places are as entrenched with domino as others are with exchange..
you need to make your solution protocol compatible.. you need to make your solution make the transition as painless as possible.. and then provide all the functionality that was had before.. but in new and better ways.. its the only way to get it accepted. Many many great software packages go unused because they came along after inferior products were entrenched and didnt provide a solution for seamless painless cross over.
If you want to kill Domino (and god knows i do too).. then dont only create a replacement.. create a bridging application to get the corporation from the ugly wasteland that is Domino to your utopia... that my friend is where the true battle lies.
Nobody ever got fired for buying Microsoft.
>
You mean they can hire MCSE's to rebuild their server after every e-mail virus attack.
You know Notes/Domino runs on many different not Windows OSes, including Linux.
Note to IBM: make a native Linux client for Notes, so we can stop having to use the Domino webmail interface.
- An OSS backend that replaces Exchange for calendaring at least. We can use IMAP and LDAP and such for the other functionality because at least there exist standard protocols, but there is no competetive backend for calendaring applications. It would be really nice to have an OSS all-in-one that does IMAP/LDAP/calendaring, but first things first.
- An OSS front-end like Evolution, but one that works on Windows as well as Unixes.
At our organization (a small university), we are desperately looking for a comprehensive calendaring solution -- one that supports teams, conflict resolutions, notifications, palm synching, and can be used either at home or at the office. I am the OSS advocate in our IT department, but I just can't find a suitable OSS solution. In fact, the only solution I can think of that even remotely fits the bill is Outlook and Exchange.The unfortunate part is that we're using Netscape 4.x here, mainly because of its mail client. (We're using IMAP and LDAP on our backend and NS 4.x Messenger is still pretty good, even though the browser sucks.) Netscape 7.x / Mozilla 1.x is nearly there, but not quite. If there was a calendar solution that worked with Mozilla/NS7 that had those features and had a OSS server, it would be like a dream come true. As it stands, I may have to roll out a small deployment of Outlook and Exchange just to solve this problem (which has come down from the president BTW, so it can't just be ignored until a suitable OSS solution comes along). Now suddenly we're mix-mashing between NS 4.x over IMAP with Outlook and Exchange. You can see what is going to happen with that nice IMAP/LDAP solution in a year.
I think what we really need is a standard protocol, de facto or otherwise, for network calendaring. There is iCal, but from what little I know about it, it's just not comprehensive enough. (Does it deal with network transport?)
Jason.
First of all, if you have ever migrated a corporate messaging system (email, calendar, shared documents, etc), you wouldn't make light of such a task.
Second, I should be more specific; it needs to work with Outlook. The reason for this is that given a choice, users will work with software that is already installed vs. installing something new; and this includes the users in the IT department. If you don't beleive me, look at browser usage statistics; why install Netscape when you already have IE installed? Why install AIM when you have MSN right there on your new XP box?
Make it work with outlook and you immediately have a client base. Make a tool to ease the migration and you're golden.
second society
If we truely want to provide an alterative to Exchange as someone who works in an entirely exchange based environment, here is my analysis of what my PHB's would have to see.
Server Side:
1. The replacement must support Outlook as a client, people actually like Outlook as an integrated client.
2. The Replacement must work with the Sendto functions of Microsoft Office
3. The Replacement must be able to scale to 10's of thousands of users, in geographically diverse locations.
4. Must Support Multipule languages
5. Must be easily scannable for Virus protection, and must be able to deny delivery of messages that fit certain criteria
6. Easy rules based scripting of mail events stored on the server as part of the user's mail box.
7. Must support enterprise calendaring/scheduling.
8. Must inter-operate with Exchange during migration
9. Must support server and OS of choice at the company(You know what that means)
10. Must offer web mail capabilities equal too or better than OWA(this includes the ability to secure the web mail client via SecureID)
11. Must support massive data stores, on the order of 500GB-1TB(yes exchange can do this)
12. Must Integrate with our directory services, like exchange 2000 integrates with AD.
13 In short it has to do all the things that exchange can do, and more, and better.
Client Side:
1. Must have a client which supports all the functions of the server side. In short its gotta work like Outlook.
2. Must Support OS, and hardware of choice.
3. Easy Rules based scripting interface to server and client side rules(Think Outlook rules wizard)
4. Must be dead simple for users to use, users don't learn they want everything to work just like it always has, even if you give them a new application to do it. When we moved from Banyan Beyond Mail to Outlook when we went from a banyan network to an NT one it was a nightmare for all of the administrative assistants as their workflow was massively changed.
So there you have it....rebuild exchange as an OSS roject and get back to us...this is not meant as Troll, this is a real world example of how a corporation is going to look at such a thing.
Power Corrupts,Absolute Power Corrupts Absolutely, leaving one person(group)in charge is absolutely corrupt.
What about RFC 2447? The iCalendar protocol looks to have been developed jointly by Netscape, MS and Lotus. Exchange may support this, and even if it doesn't, this would be a good place to start.
As for the client-side, I think that I fully-featured web mail system can easily replace Outlook on the corporate desktop. They may all have Office, but they've got browsers too!
Next, they like to have someone to sue
Are you kidding - who is going to sue microsoft if Exchange borks? or doesnt perform? or doesnt do what they promised?
They'll be lucky if M$ fixes their issue in the next version --- and they'll be charged for the privilage.
All this hooey about "Companies like someone to sue" falls apart when talking about MS. No one is going to sue MS. Now, OTOH, if someone bought support for the OS and an Open Groupware/Scheduler app from an GNU/Linux company, say redhat/suse/connectiva/mandrake who would their legal department be able to strong arm? M$ or the afore mentioned Good Guys?
Further, the argument about "Who are they going to sue" falls apart when presented with the source. Who would sue vs. paying one of the above to implement their feature?
Open Source, and the super value proposition (synergy* between companies paying developers to work on a codebase and keep the application Open) has yet to be realized, when it is, you'll see alot of big companies break down and say "why do i want to pay to rent (license) something when I can pay to the commons and have EXACTLY the features I want.
*excuse the corpspeak.
...because the problem isn't interesting enough.
Open source shows a strong prediliction for solving interesting problems well ahead of boring ones. For instance, we had useful, powerful distributed databases, cryptography, new languages and C compilers long before we had a functional word processor and spreadsheet combo. Quite simply, we have already solved mail distribution and address-book sharing on their own, and have relatively little interest in peeling apart a proprietary MS standard for same which is liable to change next week. This is also the reason why OpenOffice is great for everything except reading and writing Word documents.
This flows into my new theory about how Microsoft intends to go about attacking Linux: A deluge of boring, repetitious, pointless APIs and interfaces for problems that were long ago solved but now must be addressed using these new, uselessly variant interfaces simply because that's what everyone else has to do. (Think dotnet.) A hacker's familiarity with extant interfaces is his or her number-one resource, and is therefore that which he or she will part with least readily --- even at the expense of the compatibility or useability of the code they're writing.
Microsoft's strategy is reminiscent, in some ways, of an ancient Incan technique for pacifying politically difficult villages and towns. By forcibly migrating the entire settlement to some distant part of the empire, the usefulness of the skill-sets of these hunter-gatherers was greatly reduced, making them dependent on the (massively centralized) government for handouts, and therefore suddenly rather polite in their relations with the regime.
In the same way that a hunter-gatherer depends on his knowledge of the land, a geek depends on knowledge of the problem and solution spaces. Furthermore, most OSS projects are extremely long-term endeavours; think GCC, think Emacs, think the Linux kernel(*). OSS developers work by building things slowly and correctly with a minimal expenditure of precious manpower; Microsoft works by using more coders, more money, insane work hours and a blase attitude toward standards (even difficult, complicated, important standards) so that they may get to market early , recoup such expenditures, and get to work on the next total (and totally incompatible) revision of their product, which people will use simply because of the upgrade path that MS will kludge together with exactly the same bloodyminded application of superior capital.
Simply put, we need stability more than they do, because they have more time and money. We write things right the first time, whereas they have the luxury of making as many mistakes as they need to in order to grab market share. But more importantly, we need the projects of the past to have been written right the first time; we need a working libc, kernel, and so forth, otherwise OSS simply doesn't happen. Microsoft has no such prerequisites to its growth, as, in a pinch, *it can simply replace its foundations by fiat*. Their hunter-gatherers can, metaphorically speaking, simply create (with a certain expenditure of time and effort) the landscape best suited to their requirements. Thus they can march along beside us, setting the pace, forcing a speedup, replacing good APIs with new because every step into new territory costs them less than it costs us, dissociates us from our well-known and powerful (if somewhat lacking) APIs and encourages our work to depend on their own work, which will then be changed, etc, rendering ours much less useful.
Ultimately, the strategy is designed to encourage hackers to go take up billiards or chess or something with a potential of being useful to remember or think about or use five minutes hence. The ultimate goal of cycling APIs is to induce *indifference*, as we face a choice between working harder on minutia or walking away, hands in the air.
(*)Note that, of these projects, two are sufficiently low-level to be immune to all but the most radical shifts in design; this is again indicative of what OSS excels at.
- undoware.ca
It may be possible to replace Outlook as a client with something like Lotus Notes or the like, but it had better be at least as good as a client (Notes is not.) Outlook is an excellent client. So is Evolution, but Evolution is still about 20% shy of serviceability - it needs to be 20% ahead to justify a migration (and the feature of virtual folders could be half that battle right there) - but it doesn't run in Win32 environments yet and there's no indication it will. Mozilla calendar still lacks a lot of finish, including basic sync conduits for Palm/Pocket PC and, of course, enterprise calendaring.
I've got non-computer savvy users who blow me away with how far they push the functionality of Exchange and the calendar/meeting functions. It's been an incredible boon for us to have this system in place.
On the flip side it's horribly complicated, unreliable, resource intensive, and when it breaks it breaks BAD. But even with all those negative things going against it, there's NOTHING else we can use to replace it. There is no competition for our dollar in this area, commercial or free.
And as far as Microsoft support... try getting them to help you fix your broken Exchange 5.5 installation sometime. We don't call Microsoft for anything--we don't believe they could be of any real help. As with any software that the user has to modify after installation, there's not much a phone tech guy can do to help.
Maxim: People cannot follow directions.
Increases in truth directly with the length of time spent explaining them
Sysadmins have no power if the boss says "We are an Outlook shop and nothing else". My boss actually said that to me. He could care less about the black magic behind the scenes, but he wanted a unified client interface.
;-)
I am sure this applies to many sysadmins, and in reality, the only people with the power to swith over users software are sysadmins of one
This is a decision for IT heads or even presidents of companies to make depending on the size. If it makes economic sense to swith over they will, but not because it "isn't microsoft". If we actually want to create a viable alternative, we need to entice corporate decision makers with dollar signs, not with rhetoric.
Bye!
Open Source is a good thing for the simple reason that the app doesn't chain you to the OS. For instance, Sun Java will not run on the next versions of Mandrake and Red Hat, because of ABI changes brought on by gcc-3.2. If Sun's Java was actually opensourced (rather than their half-assed attempt), it would be a simple matter to rebuild it for new distros.
- They dont have STABLE standards support (IMAP, POP3, iCal, etc...)
;p
I use IMAP and POP3 through GWIA for 700+ users, off one box. It's been up 60 days, and that's because we moved offices two months ago.
- They still havent integrated GW's user/password database into Novell's famed eDirectory/NDS database.
Maybe not, but I manage them using the same utility. Nobody has anything better, really. And because of the way the post office works, you have to communicate with a specific server agent, not just any server in the tree, so intergrating passwords wouldn't really help any, unless you have no tape backups.
- Very little administrative control over the mailboxes.
What complete bullshit. In NWAdmin, I can control every option of the GroupWise client, I can set it remotely, and I can grey out the option so the user can't change it. What the can't you do? You want to add rules or specific proxy access, just go in to their box with the client, and do it.
- Poor backup solution (you MUST shutdown the email system to get a reliable backup). No, the GWTSA's dont cut it (based on my personal experiences, and statements from senior techs at Novell)...
Not based on my experience with Backup Exec 9.0. Even if you don't use the GWTSA's, you just make everyone access the post office over IP, instead of file access, and backup the directory. The files locked by the agent can be rebuilt from the files that will never be locked.
- Novell has POOR support for automated administration and report generation out of GroupWise - GWCheck just does not cut it...
Hmmmm... I've never cared about getting a report, really. Besides, GWCheck is for repairing the system, not reporting. But since I don't know what kind of reports you'd like, I'll leave this one alone.
Groupwise is *great*. No, I don't work for Novell. Yes, I do administer a 2000 user enterprise system that runs Groupwise 5.5. We don't even need a dedicated e-mail guy, even for all 2000 users. And it doesn't even take up a big chunk of my time. I have 15 domains, 22 post offices, two internet gateway agents, and WebAccess set up. No issues, anywhere.
*ever*
I think you're doing something wrong.
The Outlook/Exchange user experience, especially in large corporations, is the best email, calendar, and collaboration experience available today by orders of magnitude over anything other product combination out there.
The Exchange store is practically bulletproof (when it goes down, it goes down HARD, but it hardly ever goes down), and the integrated Active Directory user administration makes account administration relatively easy.
The only real administration headache I have heard about (and this is a biggie) is that backups are extremely difficult. Also, when you're running Exchange, you are completely locked into Microsoft and it's practically impossible to get off that treadmill.
I haven't tried Oracle's solution, but I haven't heard anything (good or bad) about it from anyone else either.
I WISH someone would come up with an alternative, because I have to run an Exchange server in my home office (yes, I have RedHat Linux here, too) to get the user experience and functionality that I need. I get a real kick out of using Evolution to access my Exchange email, though. Excellent work, guys!
But it doesn't exist today, and that's not going to change anytime soon. I don't even see anyone taking on this problem, or I would jump in and help them.
But beware! If Microsoft puts a Home version of Exchange on their Home Media Server with, say, 5 email accounts on it, everyone will be running Outlook and Exchange at home, too, and sharing their calendars with each other like Apple's new iCal!
I see many people here aren't real fond of Lotus, but what would happen if IBM really decided to shoot the whole works on Linux and release Lotus as open source on Linux/Unix platforms? Here are some points for and against it - I suppose only IBM really knows how they would balance out.
For:
1) Would be a huge boost to their Linux effort, and might convince a lot of companies to get a Linux solution from IBM.
2) As open source, people could finally start to address all the things they don't care for in Lotus. Perhaps a license could be arrived at which retained for IBM exclusive rights to distribute binaries for non-open platforms, and include on those platforms innovations submitted by the open developers. For open platforms such as Linux and BSD, full availability.
3) As a free and open solution, Lotus might begin to do some serious damage to that end of Microsoft's business, and at the same time focus more IT departments interest on IBM.
4) Support contracts could still be offered, and in large scale operations would probably still be bought. Even with an open Lotus, IBM is still the logical supporter for the programs.
Against:
1) IBM wouldn't get any direct license fee income from Lotus on Linux.
2) Legal issues with releasing the code could be considerable.
"I object to doing things that computers can do." -- Olin Shivers, lispers.org
Not a troll, just the truth. First, there isnt and never has been a functional replacement for the Exchange/Outlook combination. Wanna know why?
Because its DAMMED HARD to build. If it were a no/brainer, you would have a dozen choices by now. The fact is that all these OSS wannabe coders bitching about Exchange are in no position technically or financially to replace it, no matter how much bitching about it they do.
I was in the Exchange group, back in 98, and I can tell you that as bad as Exchange is, it is still better than anything else approaching its functionality, and all these people know it. Dont like Exchange? Go fuck with simple SMTP, and shutup. Hell, I remember when Lotus Notes installs had to cross their fingers every time they sent an attachment. Groupware? Kiss my ass. People dont work that way.
The reason that Exchange/Outlook are #1, is because they do the most things that people want done, more OFTEN and naturally than its competition. Is it perfect? Hell no, but it is BETTER than everything else out there.
Microsoft wont say it, but Exchange's reputation is more the fault of brain-dead MCSE's than anything else. It is Microsoft's fault that they made the product approachable to any idiot that could get their hands on the software, but the fact remains that Exchange is the best Messeging/Calendar product on the market, and if anyone things its easy to beat, SHUT UP and BUILD IT!!!
This AskSlashdot is trying to demand code from the Open Source community. That is rude in my opinion. If he wants a copy of Exchange and Outlook, I expect him to put the effort into it. Learn to code and then start a project. Demanding that people who code on things for enjoyment start working on something else just because you need a free alternative to a costly product, that is arrogant as well as rude. From the sounds of it, he wouldn't even like to contribute to the project only use it.
Aside from the managerial politics mentioned by the others, as well as the technical issue of moving between e-mail servers, there's also the Dick Factor. Specifically, the users will see the IT department as a bunch of dicks for forcing them to change from something they like. If they know specifically who ordered the change, and don't get an extraordinarily good reason for it (virtually impossible), they will see that SysAdmin as the biggest dick of the group, and some of those people have political power, which can result in a SysAdmin being a dick to the unemployment people.
You can never go home again... but I guess you can shop there.
Why is it when the open source people make changes that break application software, it's automatically the fault of the closed-source software writers? Jesus Christ, just don't make changes that break software!
This is why it takes MS a month to release security patches. The patch itself could be out in a few hours (they have a 24-hour security team), but they spend the rest of that time testing it on every configuration possible to make sure it doesn't break other software.
Just more pro-Open Source FUD.
Aw, fuck it. Let's go bowling. - The Big Lebowski
Next, they like to have someone to sue.
BS. That's standard open-source "why we aren't accepted" fodder. How about "Exchange has proven itself in the industry, so without a credible competitor it's the natural choice"? I see the other classic "Oh, they just like Outlook because it's included with Office" argument throughout the threads as well: Classic, delusional, self-consoling nonsense. Exchange is a superb groupware solution, with no credible peers, and Outlook is the best, bar none, PIM software on the Windows platform. I would love if there was something better and some good competition, but the laughable nonsense seen here on Slashdot doesn't encourage it.
Corporations care much less about software politics than most of us do. The open source solution has the benefit of getting out of proprietary formats, but I don't think that's very high up on the list of priorities of the people making the decisions.
The second sentence is redundant: Corporations do care if "proprietary formats" have specific advantages or disadvantages, but they don't care when it's just politics (which often it is, particularly with the open source community). If you can say "Well product X offers Z ISO standard functionality, which means that we can use the highly rated Y client for 1/4th the price of Q", then you've offered an advantage. If, on the other hand, you stomp your feet and have a tantrum about how Microsoft added extensions to the extension fields of Kerberos, well then you're just politicking. Every one of these nuances has end results, and if you can't quantify them for the suits, then don't expect to get respect. Sadly, the introverted geeks club (of which I am happily a member) are incredibly incapable of appreciating or estimating the costs of the various factors, but instead approach technical issues with a moral righteousness that turns most people off.
You've touched on the truth, though: The cost of Exchange, even the super Exchange server with thousands of CALs, is trivial chump change to most corporations, and the cost of a couple of sysadmins to administer the mail server far eclipses it in a single year. Tiny shops that are willing to deal with more hiccups and put more elbow grease into it may care about cutting some corners, but no real corporation is going to.
> This is why it takes MS a month to release
> security patches. The patch itself could be
> out in a few hours (they have a 24-hour
> security team), but they spend the rest of
> that time testing it on every configuration
> possible to make sure it doesn't break other
> software.
Bull!! Should we count how many times Microsoft's service packs broke applications or even worse broke another serivce pack or re-exposed a previously patched problem ?
If a fix for a problem in one application requires such a long time to test, then perhaps it is wise to re-examine its design. One should always understand that tightly integrating functionalities of anyone application into the core of anyone operating system will always have some kind of undesired side-effect as exhibited time and again by the many security problems that keep hounding MS.
Anyways, what do I care. I am not a Windows user.
But to address your question, and the brilliant observations of the iarchitect.com web site: please do keep in mind that Notes was developed before Windows 95, and long before Microsoft released the Common Interface guidelines, at a time when there was no agreement on how GUI interfaces should look or work. Every application at that time (including the various Microsoft apps) had its own look and feel. The designers of Notes built a platform-independent GUI from scratch without any guidelines to use and before most of what passes for "UI research" today was published (I am no big fan of the Microsoft CUI standards myself).
As for iarchitect.com, the first thing they rip on are the fat double-click buttons and tabbed desktop. Whereas when I was supporting Notes, the first thing that non-geekaziod users would ask me after they had used Notes for a while was how they could add the fat buttons and tabs to their other applications! So I am not quite so convinced that was a bad choice...
I also find it funny that the denizens of this site post violent diatribes concerning Microsoft's stifling monoculture and lack of innovation, but when faced with something a little different rip on how it "doesn't follow standards"!
sPh
I too am gluing together an Open Source solution for email/webmail and the Calendaring is always the biggest pain....
Why?
The majority (not all) of geeks look at calendars, project plans, palm pilots as useless. In my case, I don't even wear a watch.
The only way to get this done is for someone to write a check and/or fund a grant. Once the project gets rolling, it will quickly catch up to Outlook and perhaps even merge a few OSS projects like Mozilla and Squirrelmail.
I hate to agree with the previous posters, but this is BORING, non-challenging work. Many have gone down this path, only to fall asleep and find something better to hack on.
Unfortunately, this is one case where Microsoft actually excels over Opensource. They have enough money to pay programmers to do mundane work and complete bloatware with a pretty front ends.
Has the Open Source Community mets it's match?
Is creating calendaring code that interoperates with Outlook beyond our reach?
It is the proprietary software makers problem that he didn't release the source code.
This is what I'm talking about. It sounds like they "have" to release their code for it to work -- once again, not their problem. Windows 3.1 apps still run on my WinXP machine w/o a problem. I've never had my hands on the code. Perhaps there is a better solution than releasing the code... hmmm??
Aw, fuck it. Let's go bowling. - The Big Lebowski
I guess you have *never* heard of the many occasions when 'service packs' from M$ have broken competing products (the most glaring being the NT SP that broke lotus notes on NT 4 server)? I'll give them the benefit of the doubt and just assume they never tested it with Lotus Notes, but that kind of blows the whole 'they test it everywhere with everything' argument, doesn't it? Or perhaps you believe they did it on purpose? Hmmmmm. . .now THAT is a good argument for Open Source, isn't it?