Saiku, Open Source and sustainability

Recently Tom Barber (Magicaltrout) made a post here talking about the troubles of funding an open source product such as Saiku.

In talking to some colleagues there’s an angle here which perhaps some have not considered – which is that of the community version of the Pentaho BI server.  Should Saiku disappear where does that leave the BI Server?

Well; In a very bad state I would suggest. Essentially then you’re a BI server with no olap client.  So what is included in CE then? Well nothing other than reporting.  (And not even adhoc reporting as of 5.0)  Hmm; Thats strangely like another recently acquired “open source” BI server out there where the community version is rarely seen in production.

So; I just hope that those in the Pentaho community (Including those at the corporation itself) realise the benefit that Saiku brings to the BI suite

And before anyone points out dashboards in CE ( which are actually industry leading) – They’re not part of the default CE product and that is a great shame – A surprising number of people don’t even know about the whole Ctool stack.

I personally like some of the models I’ve seen recently where the community collaboratively fund specific features. Maybe that could work for Saiku!

Advertisements

8 thoughts on “Saiku, Open Source and sustainability

  1. I like Saiku! Love the tool, the project and the people behind it. For sure, I don’t know what to say about writing open source software purely from a commercial viewpoint, I never been in that position with Kettle. For me, the choice for an open source license simply indicates to all that you have some intention to collaborate with others (and that includes other companies) on the level of the source code. The license simply clarifies the rules. However, the license doesn’t say anything about a business model, where and how you want to monetize the codebase and so on.
    In other words, I think Tom might be slightly mixing up “free as in speech” with “free as in beer” in the mentioned blog post.
    Personally I think one option is more collaboration with the community, increase adoption (being happy when other companies adopts Saiku regardless of payment), make code contributions flexible, and so on… on the source code side.
    There seems to also be a real interest to monetize Saiku so I would tweak the business plan, re-organize sales, support, marketing, events, training and so on as professional as I could and give it another shot. I’ve been told doing a lot of marketing to fuel the sales funnel seems to be the key there. As someone who failed to do exactly that with Kettle I have nothing but sympathy for the hard problem that Meteorite needs to solve there.

    BUT! Concepts like “making money from open source software” are nonsensical. You make money selling expertise, services, integration, support, training, builds, patches, QA, helpdesk,… not code. Unless it’s in the form of closed source add-ons. In the case of Saiku, closed source visualization plugins, data connectivity, collaboration, security, a visual designer for Mondrian perhaps, …

    Then your idea for Pentaho forking over money to help develop a competitor to analyzer… It seems to be contrary to the “everyone is welcome” open source platform that we’re building. What about Pivot4j (http://www.pivot4j.org/) Should Pentaho sponsor them as well?

    Finally, as an early monetary contributor to Saiku I think your crowd funding idea might have some merits and official CTools support will be offered by Pentaho in the near future.

    Just some random thoughts!

    Cheers,
    Matt

  2. Many thanks for the comments Matt – I wasn’t deliberately suggesting that Pentaho sponsor Saiku (although I had considered it) and I can indeed see the conflict there. The key point is where is CE without Saiku. So what do we (all of us) risk if we lose Saiku? I’m suggesting that if we did lose saiku the community and therefore the corporation would lose a LOT.

    I also presumed someone would mention pivot4j, hehe! Not had much chance to check it out yet but looks promising.

    Now it would be putting words into Toms/Pauls mouth, but part of the “monetising” is simply paying Pauls salary. AFAIK he just wants to develop and push Saiku forward – whilst keeping it open. Making that work is indeed tricky. And their great work being the first guys to use olap4j, and the first to use Mondrian4 is helping the core product enormously too.

  3. I thought about a bit and I actually think there’s a lot of useful technology in BA Server CE. Even without any extras installed it’s a nice base for others to work with. Let’s not take security, integration, plugin frameworks, perspectives and so on for granted. Let’s remember that the server was not an open source part at JasperSoft. We’re actually going to great lengths to continuously provide better integration for our community through REST, OSGi and what not. Just brushing that aside would be a bit over the top I think.
    In that respect you can invert the question: would Saiku have all the functionality provided by CE if it wasn’t availability? Or perhaps we don’t ask the question since it is in fact available to all.

  4. OK, wow, i thought the jasper server was open, but all it had was reporting. I had assumed that the open server had security etc, I took that as a given. So it’s even less functional in CE than i thought.

    So I’m not bashing CE at all, the reverse. CE is indeed totally usable in production. But a significant part of that is Saiku in my experience of CE installs. I for one would be very unhappy with anything that reduces the size and quality of the CE stack. The fact there is a fully functional CE stack validates the EE proposition in my view.

  5. Hi guys, you know that I have my weak spot for that guilty pleasure of ego searching, so I wouldn’t miss a post with ‘pivot4j’ keyword in it 🙂
    Anyway, I’m grateful that I’ve been able to maintain Pivot4J without much need for financial support, as I only do it on my spare time as a hobby.
    But sometimes I wonder what better features I could have written for it if I had some kind of sponsor for it. Now I’ve launched a startup company which has litter to do with BI, so I won’t be able to accept it even if I had such a chance.
    However, it might be different for guys with Saiku, and I believe nobody can blame them as long as they keep their project under open source license.
    If they want to put more time and effort into their open source effort than what can be maintained without a sponsorship. I only can say that I respect their dedication and want to wish them good luck.
    As for my project, I hope I could find enough time to spare for it to keep it going long enough so that I could see more people find their Pentaho CE useful with Pivot4J.
    I believe Pivot4J has come a long way since it’s first began two years ago, and can be a viable alternatives to Saiku for many people. But I’ll be the last person to wish them disappear. Actually, there presence itself has been a good motivator for me to keep working on my project.
    So I hope it stays that way for long time, and I believe that’s how open source projects evolves while competiting with each other.
    Wish them luck!
    Xavier

  6. I know this is very old, but an interesting letter came from Tom this morning…

    I fear Xavier’s prediction might happen. Then what?? I know our budget couldn’t afford BI products based on the licensing models out there. What’s a “user”?

    What to do… other than contribute, I guess. But that doesn’t mean that we would ever be able to afford the version of Pentaho where the new Saiku (dahboards, for example) would be exposed because, as mentioned, since 5.x this is not in CE.

    Seems time for a reset.

    ==== The letter I received from Tom Barber on 11/17/2015 ====
    Thanks for taking the time to read this email. We are running a crowdfunding campaign for Saiku to allow us to build and release Saiku Reporting as an open source tool that will become part of Saiku Analytics Community Edition. We have 20 days to raise our goal of £15000 which will allow us to build and deploy Saiku Reporting and allow the end users the ability to create formatted adhoc reports with the same easy to use interface as Saiku Analytics.
    For those of you who are unaware of Saiku Reporting, it was written by the community a number of years ago and allowed Saiku users to create “Pixel Perfect” reports from within the web browser. Sadly the development was discontinued and the code base was never completely stable. We want to be able to provide Saiku Reporting via Saiku Standalone and within the Pentaho 5 & 6 BI server as plugins. To do this we need help from existing users of Saiku CE to sponsor the work required to stabilise the code base and extend it to provide the experience users have been asking for.
    As this is a Kickstarter campaign we don’t get a penny unless we hit our goal, so if you are interested in using Saiku Reporting, please consider funding the project and allowing us to build this new Saiku module. All our code will be released under the existing Apache license and will be freely available via github. As a thank you we are offering a number of discounted Saiku Enterprise licenses to major sponsors, so that organisations who want to take advantage of EE features and support can do so for a year and help to develop the Community version.
    If you have any questions about the product, then please feel free to reply to this email and I’ll happily discuss them with you.
    If you would like to help fund the project, click here to visit the site.
    Thanks for your time, and continued support.
    Tom Barber
    Saiku Analytics Founder

    • Hi Bill,

      Dan pointed me in your direction, I hope you dont mind me taking the time to reply.

      I have always maintained, personally, I don’t like the open core model I think it sucks, I am an open source guy and maintaining 2 code bases for our code to try and get people to pay money to cover development costs has never been what I wanted to do, but……

      PAT/Saiku 1 was a hobby it was an interesting way to relearn to program, Saiku 2 we asked people to sponsor development, or just pay some money as a business to cover our development costs, there have only ever been 2-3 regular committers for a large and complex codebase, and it is a full time job just keeping the CE stuff running, but donations never amassed close to enough to cover our costs (I live in the countryside and drive a very old car, we’re not high maintenance).

      So Saiku 3 went open core, not out of choice but because it was the only way we could survive whilst developing code, I can give you the contact details of a few guys at Amazon and Samsung who have used Saiku for a number of years and never sponsored a dime, if we could cover development costs there would be no EE at all, but that isn’t the case. So we try and price it cheap, we price per user so small departments can benefit without the outlay that Pentaho EE or Tableau or whatever would cost. That said, I’ve never turned down a contribution, and I know a lot of companies wrote dashboard frameworks for Saiku before we wrote ours, if any had wanted to donate the code for the CE product I would never have blocked it, but Saiku used by consultants and developers to facilitate their end goal, and that stuff is never passed down.

      So in summary, I really wish someone would come up with a way for us little guys without VC investment to make enough cash to cover our costs just doing Open Source development, but sadly with the mindset of current companies, this isn’t the case.

      Tom

      • Sorry, I should also point out, that running the kickstarter is a way of us actually paying for the work that we would other do, at our own cost, and we would apache license and again people would use without offering any time or money to the project, and so if we can cover the costs before the work begins, at least we’re not on the street at Christmas! 🙂

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s