SugarCRM and Open Source blog

I’ve been reading a few blog posts about what the commentators (mainly Americans) are calling OSS 4.0. For the uninitiated, OSS 4.0 is ultimately about the death of Open Source. Apparently we’re all so greedy and obsessed by making a fast buck that all Open Source will either be bought over by large corporations such as Oracle or the owners of Open Source projects will realise the error of their ways, commercialise their Open Source projects, abandon their “hippy” (the usual derogatory term applied to anyone that dares to defy market driven capitalism) ideals and “grow up".

As Ha-Joon Chang argues persuasively in his excellent book “23 Things They Don’t Tell You About Capitalism”, human beings are naturally more disposed to collaborate than they are to compete. Without collaboration, society doesn’t and can’t work. Without competition, only capitalism doesn’t work. The two are not mutually inclusive.

The main thrust of many of the articles is that apparently it’s impossible to put food on the table by being involved in Open Source projects.

I’ll cite two companies as examples of why that view is wrong.

SalesAgility

OK, we’re not the size of Oracle and don’t have aspirations to be so. We’ve been involved in the Open Source version of SugarCRM for five years. During that time we have resisted overtures to become a SugarCRM partner and sell their proprietary versions. Because we do not see the future as being governed by restrictive licensing covenants.

We have ploughed our own furrows, building value for our customers on top of the excellent Community Edition of SugarCRM. We run one of the major SugarCRM community projects called Advanced OpenSales which builds Product and Quotation modules for the Community Edition of SugarCRM (shortly to include Invoicing and Contracts Modules).

Advanced OpenSales is a great example of why and how Open Source works. The original project, which was called Open Sales, ran out of steam in 2008 and development stopped. But it was an important project for the Community Edition. So, we picked it up and have moved it forward considerably and have a roadmap for the future. We have gathered excellent collaborators along the way and the modules are built collaboratively by us, supported by developers in America, Ukraine and Germany.

The Advanced OpenSales modules are superior to the SugarCRM developed ones contained in the paid-for versions. They include parameter driven quotation template design and multi-tax applied at line item. Functions that have been absent from the proprietary version for years.

And all the code is given free and gratis to the Community under the Affero GPL. We don’t earn a penny from it …. actually, we do.

Without Advanced OpenSales, many of our customers would use the paid-for versions of Sugar. So, we give them a compelling reason to use Open Source. Internally, we probably do more development work around quotes and products than we do on any other parts of SugarCRM. The phrase “we all float higher on a rising tide” is germane here.

We’ve never been busier. We’re growing. We’re putting lunch on the table. We engage internationally. Our markets are horizontally and vertically broad. We know no boundaries. And yet, we’re end to end Open Source.

IBM

The thrust of Lou Gerstner’s (chief executive officer of IBM from April, 1993 until 2002) excellent book “Who Says Elephants Can’t Dance?” is that in order to thrive and survive, IBM had to do what their customers wanted them to do. They had to be a customer-centric services business. In applying this transformative philosophy, IBM has become a company that earns more from Open Source than any other company on the planet. And IBM is one of the largest contributors to Open Source with 177 projects that it either manages or contributes to listed on their website.

IBM earns service revenue from its Open Source activities. And service and service revenue is the game changer that some commentators either don’t get or don’t want to get.

I case you hadn’t noticed, IBM is one of the most consistently profitable corporations around, putting a considerable amount of bread on the table.

So, next time you read about OSS 4.0, smile and move on. The “hippies” have taken over the asylum.

I’ve blogged about this before.

There is a gap in the provisions of GPL versions 1, 2 & 3 that allows individuals and organisations to take GPL’d code, enhance it substantially, derive significant revenue from the enhanced code and yet, not have to give so much as a comma’s worth of code back to the project.

The GPL states that if you amend GPL’d code and distribute it then the amendments to the code are also governed by GPL and should be available to the community as a whole to benefit from.

The loophole is (was) in the definition of “distribute". When the original GPL was written, most code was distributed physically. You bought a bit of software and it arrived on a CD (remember those days?). Code distributed this way is covered by GPL. Additionally, code that is distributed electronically (you buy some software and are sent a .zip or other executable) is also covered by GPL.

If however code is distributed in a software as a service mode, then GPL does not cover this. So, company A (let’s call them Google) acquires open source code from Project B which is covered by GPL2, (let’s call it Linux). Company A substantially improves Project B. Company A deploys Project B code in tens of thousands of it’s own servers and derives billions of dollars of revenue from applications that depend on Project B.

And company A does not have to give so much as a single comma of the enhanced code back to Project B.

It’s a parasitical existence. The parasite could not exist without a host to feed on. The parasite thrives on the host. The host does not benefit from the parasite and in some instances may die.

Affero GPL closes this loophole. It requires that any adaptation of software covered by a AFGPL license must prominently offer its source code for download to users who interact with it over a network.

So how does this affect SugarCRM?

From SugarCRM Release 6.0 onwards, all SugarCRM open source code will be release under the Affero GPL 3 license. This is a shrewd move on SugarCRM’s part and may give some much needed impetus to AFGPL.

In particular it will kill stone dead those business models predicated on building additional functionality on top of Sugar code and competing directly with Sugar - something that happens quite prominently at the moment.

It also positions SugarCRM in the vanguard of Open Source thought leaders and that, compared to where they were a couple of years ago when they were viewed with deep suspicion by the open source community, is quite a turnaround.

We recently undertook a migration from NetSuite (which just has to be the worst designed application out there) to SugarCRM Community Edition.

Why people fall into the trap of trusting their data to organisations like NetSuite is completely beyond us. CRM data is business critical. Yet when you sign up to hosting with NetSuite and other similar providers, you virtually surrender ownership of that data. If you want your data back, you can only get it in .csv format, with incomplete records and no indexes. This makes the data almost useless for the non-technical user with the task of migrating to an alternative application and problematic for the IT professional charged with rebuilding a coherent database from partial records. If you want access to more complete records, then get your cheque book out because it’s going to cost many thousands of pounds to get access to the tools that will enable you to do so.

This is not common to NetSuite. We did a SalesForce migration to SugarCRM Community Edition last year and encountered similar issues.

An absolutely fundamental question that users of cloud services have to ask is “If I want to migrate to another application, can I get a copy of my database in a format that can easily be migrated?". If the answer is either “No” or “Yes, but ….", then run a mile because you’re going to be trapped in a data cul-de-sac with the only way out being a hard technical slog which may end up costing several times your annual license fee.

SugarCRM on the other hand were a dream to deal with when a client of theirs wanted to migrate from the paid for “Express Edition” to the Open Source “Community Edition” and retained SalesAgility as the migration and hosting partner.

We started a Case on the SugarCRM Support Portal saying that we wished to migrate their customer. Within 24 hours we had a MySQL dump of the database and a copy of the relevant application files that were need to rebuild the complete application. It only took us a couple of hours to re-instantiate the SugarCRM instance and our client was up and running with no loss of service!!

What a great attitude. I posted a thread of congratulations for fantastic customer support on the SugarCRM support forums. The reply, from SugarCRM’s Director of System Services was informative:

Thank you for the kind words. We appreciate it, and glad you’ve noticed we’re “not like the other guys”

We hadn’t just noticed, we’re absolutely delighted. It’s great working with fantastic software backed by an ethical supplier. Anyone thinking of using cloud services really ought to sit up and take note.

Please accept my apologies if we seem to have skipped a generation and ommited any reference or critique of SugarCRM 5.5. This is entirely intentional. As a well respected collaborator said in a recent email to me - “Did SugarCRM cut any new code in the last six months?". The probable answer is “No!". SugarCRM 5.5 was a huge (with huge being underlined and in bold) bug fix release and not before time.

If there is one difference that the appointment of Larry Augustin as CEO has made, it is that the race for functionality has been replaced by a race for stability. It’s all very well having a product that does everything that competing products do, but if it only does them half as well or breaks in the process, it might as well not do them at all. Larry Augustin has realised this and cleaned out the stables. In 5.5 there were more than 1,000 bug fixes … we should know … we submitted many of the bug reports and participated in the testing program.

However, this is not about 5.5, it’s about what’s coming down the line and some of the more radical changes that are being made.

In SugarCRM 5.0.x, Sugar released a brand new email module using code from the Yahoo email project. Not to put too fine a point on it, it was not good. It was buggy, dysfunctional, complex, unintuitive. At SalesAgility we advised all our clients to avoid it and continue to use their traditional desktop applications.

In SugarCRM 5.5.1, the email module is the really big change. It has been radically re-engineered with a stripped down user interface, less functionality in the places that don’t matter and more in the places that do. It won’t please everyone as support for POP3 has gone (IMAP only). But as we’ve seen before with SugarCRM, pleasing all the people all the time ain’t possible.

In the RC (Release Candidate) version that we’ve been testing, there are one or two niggly bugs (Search in particular) … but that’s to be expected in pre-production releases. However, our first impressions are to give it an overwhelming thumbs up.

If this translates into a stable release, the strong likelihood is that the next upgrade that we do for our clients will miss 5.5 altogether and go straight for 5.5.1. And, for the first time we can start to recommend that clients use the SugarCRM email module in preference to traditional desktop applications. Now that’s real progress!!

However, as always, until we get the production release and a couple of patches, we’ll be keeping clients on good old 5.2(k), dependable, stable and we know where the bugs are and how to fix them.