Page 1 of 1

Any Quality Open Source Olap ??

Posted: Mon Apr 09, 2012 10:03 pm
by John Hobson
Now that Jedox seem to have left Palo OS to wither away are there any other OS Olap products out there that can offer a credible alternative to TM1 for planning?

Re: Any Quality Open Source Olap ??

Posted: Mon Apr 09, 2012 10:24 pm
by Alan Kirk
John Hobson wrote:Now that Jedox seem to have left Palo OS to wither away are there any other OS Olap products out there that can offer a credible alternative to TM1 for planning?
Don't know of any open source off the top of my head. (Too much work for too little reward, I suspect, which is probably also why unpaid Palo was effectively dumped by Jedox.) Two commercial alternatives that I can think of are Infor and of course Ben Hills' Keystone.

Re: Any Quality Open Source Olap ??

Posted: Tue Apr 10, 2012 6:34 am
by John Hobson
Is Infor what i knew as MIS Alea a while back?

Re: Any Quality Open Source Olap ??

Posted: Tue Apr 10, 2012 6:51 am
by Alan Kirk
John Hobson wrote:Is Infor what i knew as MIS Alea a while back?
It's an ancestor; Infor acquired (indirectly) MIS a few years back. However I think it might be stretching the point to say that it's the same thing, any more than it is to say that there's more than a DNA connection between TM1 10.1 and the last keystrokes that left Manny's fingers all those years ago. It still has Grampa's eyes and hair colour, but the other features have doubtless changed quite a bit down the generations.

Re: Any Quality Open Source Olap ??

Posted: Tue Apr 10, 2012 12:15 pm
by tomok
The open source model only works long term when you have a community of developers working together. With Palo, you have that community but they're not developing Palo, they're developing with Palo. The programmers at Jedox are the only ones doing any enhancement of the tool. It's kind of hard to justify paying someone to develop something you can't be paid for. I guess they weren't making enough money just charging for consulting and support. Can't say I blame them for going commercial.

Re: Any Quality Open Source Olap ??

Posted: Sun Mar 10, 2013 11:03 am
by Jamo
Taking a close look at Jedox 4 looks an improved interface, I am interested to know how well the excel to Web publishing goes.
The Web Excel however looks ok and I think this approach to building "Forms" looks like its flexible and can look very attractive.

Insight to me is just miles away, you simply cannot get enough information on one sheet, you are unable to reduce the size of the crosstab (font etc) plus the charts are minimal, surely for dashboards we need
some micro charting ability to see trends.

If someone has some limitations of Jedox be keen to hear, I'll be investing some time and effort soon to see if I can port my TM1 models over.

I hear it has not ability to "hold" for spreading, but that this is being worked on now.

Regards
Jamo

Re: Any Quality Open Source Olap ??

Posted: Sun Mar 10, 2013 12:20 pm
by John Hobson
Jedox have recently announced a change of policy as regards the OS and Premium versions of their product

With V5 due for release next month the two versions will move back into synch. For those of us who were disappointed by the build quality of the existing OS version this is great news.

The difference (I think) is going to be that the GPU processing, ETL and WEB components will only be available in Pro, so it ay be worth waiting a few weeks to see what comes out of this new initiative.

As a free OS tool Jedox 5 should prove pretty attractive to a lot of the smaller organisations (and maybe a few big ones too) as a risk free way of proving concepts. It is very easy to port models to and from TM1 and Jedox Palo.

AFAIK the existing OS version does not allow holds, and spreading is not as effective as in TM1 on calculated elements taht are enterd at N:. I don't know about Jedox V4

I had to fight a bit to get Manny Perez to accept that it was valid in TM1 to spread in a situation where for example you enter % Margin at N: and back calculate consolidations a C: levels but want to spread down from C: onto the entered N: levels.

Example :

Margin % N Product 1: 40% (entered)
Margin % N Product 1: 50% (entered)

Sales £ N Product 1: 100 (entered)
Sales £ N Product 1: 100 (entered)

Profit £ N Product 1: 40 (calculated)
Profit £ N Product 1: 50 (calculated)

Sales C: 200 (consolidated)
Profit C: 90 (consolidated)
Margin % C: 45% (calculated)

Spreading on that Margin % C: 45% is logically valid as its N: level components are all enterable.

As soon as Manny understood this he changed TM1 pretty quickly.

I haven't had the same success with Jedox so you still get illogical results when spreading on numbers like that conslidated % (It just splits is like a normal number)

Hope this helps

Re: Any Quality Open Source Olap ??

Posted: Sun Mar 10, 2013 4:09 pm
by Jamo
Hi John,

Thanks for the reply,
I am pretty committed to testing Jedox V4 with my complex budgeting and planing model to evaluate the functionality, I'll post a comprehensive report once its all done as its a pretty massive job to fully evaluate and it seems others may be wanting to look a this option now its starting to mature nicely.

Its the overall business case that I am questioning now in terms of the licensing TOC etc

My hope is that as a younger more agile company who are focused on the one product, maybe we can see some inovation based on feedback from us developers and users.

I think its worth my investment, Just downloaded the suite to test now so in the next month I will se how it manages my TM1 modules....

Regards
Jamo

Re: Any Quality Open Source Olap ??

Posted: Wed May 08, 2013 8:09 am
by John Hobson
Unfortunately Jedox appear to have hobbled the OS version of V5 to such an extent that it won't even load my models any more.
They have taken out significant functionality (like use of attributes)

Be very careful of using this product if you think you may need an upgrade path, as changes like that, which simply break models could give you a very nasty shock!

Ah well - TM1 is looking good again LOL.

Re: Any Quality Open Source Olap ??

Posted: Wed May 08, 2013 9:25 am
by lotsaram
John Hobson wrote:Unfortunately Jedox appear to have hobbled the OS version of V5 to such an extent that it won't even load my models any more.
They have taken out significant functionality (like use of attributes)
Was always going to happen, Jedox's concept of "open source" seemed to be always in name only and more of a marketing ploy than anything else.

Re: Any Quality Open Source Olap ??

Posted: Wed May 08, 2013 9:51 pm
by Martin Ryan
There's no sight like hindsight.

Re: Any Quality Open Source Olap ??

Posted: Wed May 08, 2013 9:56 pm
by John Hobson
Ain't that the truth :roll:

Re: Any Quality Open Source Olap ??

Posted: Mon May 25, 2015 8:29 am
by John Hobson
Any recent developments in Opens Source OLAP I ought to know about?

Re: Any Quality Open Source Olap ??

Posted: Sat Jul 18, 2015 12:49 am
by Alan Kirk
John Hobson, almost 2 months ago, wrote:Any recent developments in Open Source OLAP I ought to know about?
I think the breeze flowing around the long-abandoned livery stable, the small dust clouds near the decrepit hitching rails, and the tumbleweeds rolling past the weathered woodwork of the former saloon are in themselves an eloquent answer to this.

I'm wondering if open source anything has in fact reached its zenith and is at the start of a long, slow decline. It's easy to be generous with your time and write cool code when you're getting a respectable salary for doing similar work during the day. But with IT work also devolving to "a lower overall cost structure" (and we all know what that's a euphemism for), it's harder to justify spending your hours doing much work for no return. (Or no return aside from a limited value of fame.) I honestly can't imagine why anyone would want to put in the time to develop a new open source OLAP project these days.

Re: Any Quality Open Source Olap ??

Posted: Mon Jul 20, 2015 12:04 pm
by jim wood
I think the whole "get rich quick" phone development stories haven't helped either. Historically getting free code out there used to be a good way to get known in the coding community. These days you just have to write an app.

Re: Any Quality Open Source Olap ??

Posted: Tue Jul 21, 2015 3:06 am
by ninow
Hello Colleagues,
I played with Palo in the early days up till 3.2 but never trusted it to be supported long term as open source.
I do not know of good open source but I have had a long standing relationship with PARISTechnologies who make PowerOLAP.
I have implemented PowerOLAP with serious clients using big models and at all times have had excellent support from PARISTech.
They do offer a 'personal edition' that has a limitation on number of cubes, dimensions and members and very inexpensive for the smaller business. (This makes it 'almost' free).
I have done a lot of development with TM1 and Hyperion so I do understand and know the products.
This may be of interest, I would be pleased to know of anyone's experiences if they try it.
Cheers
Ray

Re: Any Quality Open Source Olap ??

Posted: Wed Jun 21, 2023 9:14 am
by Elessar
Hello everybody!

10 years have passed, so I want to revive this topic.
Is there any open-source OLAP with planning capabilities now? Jedox has become commercial, leaving the last open-source 5.1 version in 2015.

Re: Any Quality Open Source Olap ??

Posted: Wed Jun 21, 2023 10:54 pm
by Adam
The only thing that comes to mind for open source OLAP is this recent blog post: https://community.ibm.com/community/use ... source-tm1

Re: Any Quality Open Source Olap ??

Posted: Thu Jun 29, 2023 2:01 pm
by lotsaram
Well there's tinyolap https://tinyolap.com/
But this is single user in-process. Although the project bio talks about xP&A use cases I can't think of any real planning use case which is viable without multi-user and a security model.

Re: Any Quality Open Source Olap ??

Posted: Fri Jun 30, 2023 7:36 am
by Elessar
lotsaram wrote: Thu Jun 29, 2023 2:01 pm Well there's tinyolap https://tinyolap.com/
But this is single user in-process. Although the project bio talks about xP&A use cases I can't think of any real planning use case which is viable without multi-user and a security model.
Wow many thanks!