I don't mean to be cold water, but it is an impossibility to consider this approach.

This is a product that has been under development since 1984...20 million lines of code, some of which pre-dated WINDOWS, would not really make for a worthy open source project even if IBM could. There's a lot of IP in there that simply can't go "open".

Further, despite the doom and gloom FUD, IBM's revenues from Domino and Notes are substantial and *growing*, and it would make for a very difficult business discussion to say "oh, yeah, we'll just make it up on support".

I have been working with a team to make more of the application layer open around Notes/Domino, and I think with the new 8.5 Xpages approach we have a new opportunity to recruit partners and developers into the ecosystem. Am considering several new packaging approaches for Domino and Designer as a result.

I don't want to stifle enthusiasm but there is a heck of a lot of consideration before any company would open source any existing intellectual property.... it doesn't happen very often. Ian's initiative to try to do things different is much appreciated, but the arguments are emotional rather than substantive.

It's a neat idea to dream, but I can't see IBM doing this...
As they say on Little Britain USA, "Computer sez No!"

Cannot see it happening nor do I see it as being desirable.

Is there really a large community of C/C++ engineers out there interested in extending this product?

Good point Jeff. I have demoted the idea.

Even if there weren't the obstacles mentioned above, IMHO the principal reaction to such a move -- aided and abetted by Microsoft marketing, no doubt -- would be:

"OMG! IBM is abandoning Notes and Domino!"

The Law of Unintended Consequences strikes again.

That's one code base I never want to touch.

It is simply a bad idea.

@Ray Bilyk - Excellent reference! :-D

@Dan - I think this is a key point. Even if IBM were willing or interested, an open source project is only going to succeed with developers who want to work with the code and make it better. Even with extensive experience of the C API and a fairly good understanding of many of the complex data structures and such, I have no desire to go in and muck with the internals of the code base. I can imagine few in the Domino community who would, and fewer outside the Domino community.

I demoted the idea, but what I do want is that the documentaion of the C API and the C++ API is brought to the newest level. There are samples that will not work on a release higher than R5!

Take the sample MakeView. In a view there is a complex attribute $ViewFormat that is build from several blocks. In many releases this block is extended with new sub-blocks. And there is no documentation on the format of these sub-blocks. Only by hard labour and many red-screens I was able to generate a view from scratch in R6 and R7.

Another sample is the ODS format. Again no updates for newer structures beyond R5. So you cannot use ODSReadMemory nor ODSWriteMemory.

On a corporate level I agree with Ed @1.
But on a personal level I agree with Richard @6. We do not need to give MS any more ammunition that they have already.
Sorry, but I demoted it.

Even if this could happen, the codebase would have to be documented.

Aye, there's the rub ;o)

[1] The scale of such an endeavour is indeed huge however I believe that the Open Source community at large is mature enough to accept the challenge. The IP issue will indeed create some "Black Holes" in the initial codebase - these could be filled with alternate Open Source components or indeed licensed proprietary "plugins", again a huge task - but doable.

8.5 Xpages is of course great & eagerly anticipated ... but .. is it a Game Changer?

[2] Ray, Great quote.
[4] Get the product out there and they will come.
[6] MS would certainly try to spin the message that way .. the initiative should launch with maximum IBM market effort - use the opportunity to attack MS - not always defend.
[10] In my belief the initiative would draw a lot of talent into the area.

I demoted it, but feel that I ought to provide you with some of my reasoning:
1) Domino is more accessible now than ever. Especially to the SMB market in Lotus Foundations. Without looking for it (I am lazy), IBM has made rumblings of putting Domino in the cloud (hosted). They have very aggressive pricing in the non-profit and education markets.
2) Expecting that C developers will fall all over themselves to add to the code is tough for me to fathom. Especially since they have no understanding of why it is as it is (institutional knowledge).
3) I feel better knowing that IBM holds sacred the building blocks of the security for the software.
4) There are too many organizations (or decision makers) that view Open Source in a negative way. Once they discover that IBM is no longer developing the software, I feel that the migrations away from the platform will increase.

Demoted, see { Link } as to why.

Just a bad idea, plain and simple.

Demoted.

Sorry guys, just a flat out terrible idea.

I agree, I don't know how this helps Lotus. What would be better is a marketing dept at IBM that could sell the IBM product set and not send in 20 - 30 people for a sales call. IBM needs to also figure out how to sell this to small companies and maybe give away some product/services to make them see the power of Lotus vs. everything else.

[15] Good points, well made.

1. I agree it is more accessible, however,a) it is not aggressively and effectively marketed into the target space b) we lack a compelling message for startups to commit to the platform and c) the pricing model is too complex and obscure.

2. For this very reason IBM would need to maintain a strong and ongoing leadership role in the development community.

3. I believe that over time the pros and cons for security in Open Source offerings tend to balance out, this seens to be the experience from RHEL.

4. Again this is a 50/50 call some organisations mandate the use and some are strongly against. I would not see IBM leave the development of the product but rather enlarge the development team by embracing the Open Source community.

Further reason to demote: I don't believe for a minute that an open source project could manage the level of QA required to keep customer confidence. I.e., cross-platform and multi-version feature, performance and scalability testing, plus extensive regression testing and forward and backward compatability testing. This requires skilled and dedicated QA staff, and an envioronment with a large network of enterprise-scale servers running tests for all the different features of Domino, on many different versions of Domino, on all the supported OSes on all the supported platforms for the particular Domino versions, with all the different client platforms and versions, and with a variety of different localized versions for both client and server. Even running a statistically valid sampling of the above for a small but well-defined subset of Domino's functionality is a huge undertaking. I don't believe there could possibly be enough motivated volunteers with access to the necessary resources, the available time to do this job well, and the skills and discipline to get the job done adequately.

While it is probably impractical for IBM to open the source code up to everyone, I hope they will consider the proposal I made during the Lotusphere 2008 Q&A session to provide limited source code licenses for all Business Partners. There have been several times when we ran into intermittent Domino failures and had to go round and round with IBM Software Support for months, first to convince them that the problem was real and then to get a series of custom debugging hotfixes to isolate the root cause. If we could have added our own debugging code and recompiled that would have shortened the troubleshooting process, and it would save money for IBM as well.

I can't really speak to the statistics but the impression I get is that Exchange absolutely dominates the collab/calendar/email market, the only company i am aware of that actually uses Lotus anymore is National Australia Bank. Open Sourcing the lotus suite would put Lotus into a much more tenable position against Exchange whilst as we've seen with OpenSolaris it is still possible to retain a degree of control of the actual brand & product whilst benefiting from the user led adotion rather than the top down (RFP approach so popular at IBM

Looks like this hit The Register and is now being discussed on Slashdot.
The Register article: { Link }
Slashdot Discussion: { Link }

Perhaps a hybrid closed/open source project may work. Create a scaled down version of Notes for use with Symphony and make that open source. That way you can try out new ideas for functionality/design etc at a faster pace than you can now. This would give the Notes Design team access to a broad base of ideas and suggestions and allow them to solicit feedback in near real time. The winning ideas could then be melded into the closed source version over time. Similar to the Novell SLED/OpenSuse model. OpenSuse proceeds at a faster pace then SLED, but SLED provides more stability and business usability than OpenSuse. Another idea is to take the QA process for Notes and open it to more people - think Beta process on steroids. IBM could have their internal QA and the move it to an external QA process before releasing it to GA. Let's face it there are too many configurations out there for any one company to simulate (even IBM). Change the release scheme so that X.Y i.e(8.0,8.5) are stable supported releases and X.Y.Z i.e.(8.5.1) are quick fix releases that IBM doesn't provide support for. I have to tell you that the only thing I hate more than waiting for a fix - is applying a fix that breaks something else.
Let's face it - Notes has a bad rep. I keep running into people who used it in the past and loath it. A guy in my office complex was just converted to Notes - and I can hear him complaining down the hall way. Granted they are only on version 7 (I'm told it was a multi-year rollout and his office was the last one), but it doesn't work like Outlook which is what they were using. We need to get the UI up to snuff and we need marketing to let people and companies know that Lotus is alive and kicking!

I'm a huge fan of open-source. If Lotus Notes/Domino didn't exist, we'd be exclusively running Postfix and PHP and JSP based code.

Why we don't is simple- Applications.

It could be released as open-source, but the risks of "holes" as other put it in the code, in the expertise, in the ramp-up time, loss of "knowledge", all are scary.

IBM has a great Foundations product. However, it needs another point release to fix such things as Domino's HTTP (80 and 443) not being available on the parts with Domino included. It makes having your "default" web site being Notes almost impossible if you are using a SSL key on 443. However, the Foundations admin GUI overall is very good. Installing the Domino package was also easy.

We already have the Express offering. It's only missing one thing - a supported free server O/S. I'd love to see IBM officially certify CentOS and Ubuntu. CentOS is the same thing as RedHat ESx anyway. Ubuntu has support, too. Then the actual purchase price is just the user CALs. I've yet to see a commercial targeting the <100 user company showing them they can have almost all the features at a price that is FAR less than an MS SBS Server w/Outlook/Exchange/SharePoint.

Basically, the product is already fine. From my perspective, keep the pricing and offering simple, just make it available with options that SMBs would like to see. Basically, there should also be small SMB bundles where they get a Foundations like install where the Ubuntu or CentOS server is added directly to "iron", distributed as a VM, or even as an appliance.

I think it's aleady as good as it's going to get: moving to open-source Eclipse framework allows IBM to focus on improving a very complex product but still have the benefits of open source in some parts of the platform.

One additional reason to demote is that Notes includes a lot of security, that other products do not have. Security, that originally came to being from the CIA and DoD (example, the hieroglyphics during login introduced in R4 {now key rings}).

If that algorithm were to be open source, then anyone could make a copy-cat phishing version and the end-user might not know they were being scammed. Thus, IBM loses a major competitive advantage and the nefarious types score a huge victory for fraud and identity theft.

Notes/Domino is remarkably successful and making huge strides forward on the web side with 8.5. I don't see any reason for IBM to change the business model and give away the IP to hungry little developers. Notes/Domino is the Megalith (2001:Space Odyssey) of software platforms. Because the vast majority cannot fathom it, it becomes a religious curiousity. That's why we have to deal with the question of "is Notes/Domino on the way out" 25 times in 25 years and make such ethereal suggestions.

I'm just happy that it made it through the last takeover relatively unscathed. Agree with Ed on this one.

I remembered "opensource" iNotesX.nsf.
If it ever happens and Domino goes OS I doubt we will see more than "... wUzBFAA== ... haiku ... B64#XAAAAE ..."