[Mondrian] preferred communication mechanism?

m h sesquile at gmail.com
Tue Sep 30 16:57:04 EDT 2008

On Tue, Sep 30, 2008 at 1:22 PM, Julian Hyde <jhyde at pentaho.com> wrote:
> You're using the right mechanism.
> The other mechanism, of course, is to buy a Pentaho subscription. Then you
> have access to someone whose primary goal is to make you happy by answering
> your questions. (My goal is to make you happy, but in different ways.)

Good to know.

> Subscription costs money, but not a lot. If you value your time at, say,
> $100 per hour, then a subscription pays for itself very fast; if system
> downtime costs $5000 per hour, then it pays for itself very fast indeed.

There's one way to do it.  I guess I'm concerned with whether Mondrian
supports what I want.  I'm not about to pay for support without even
knowing that.  (Or even getting basic aggregation working).

> For the record, these are my communications preferences. I prefer the
> mailing list mondrian at pentaho.org for developers, discussions about possible
> bugs/features, and for advanced users. The forums are better for general
> help and for newbie questions.

I apologize if I'm posting in the wrong area, it wasn't at all clear
to me that this was the developer mailing list.  I just assumed it was
the Mondrian mailing list, and that there was a Mondrian forum as
well.  Then there's the sourceforge stuff as well....

> I see emails on the mailing list within an hour or two, and respond
> immediately or within a couple of days depending on the importance of the
> issue. I check the forums at best once a day, at worst every 3 or 4 days.
> The volume on the developer list is pretty good right now because people are
> using it sensibly. If the community at large sees my above comments and
> starts to use the developer list for other issues, then we will have to do
> something about the developer list.
> I do not use IRC at all. Sorry to those who like it as a communications
> medium, but my bandwidth is maxed out so I can't take on a low latency
> channel.

Understood.  IRC can be very distracting at times.  I guess I'm just
sort of used to it for the projects I use/contribute to.

> Your questions are of a general help nature - although some are not out of
> place on the developers list.
> I confess that I skip a lot of questions in the forum and the mailing list
> that I know the answer to, or could find the answer to. It pains me to do
> that, but I have to prioritize, and that means focusing on things that only
> I can do. To give you an idea of my priorities, here are some examples of
> posts I have made in the last week:

This is somewhat disappointing.  I realize you think I'm just throwing
out questions without doing it the "smart way".  If I could easily
find the answer I wouldn't be asking, believe me.  (Note I'm now
tracing the code through eclipse and still asking questions.)

Perhaps having an FAQ with these questions/answers?  A wiki?

> * http://forums.pentaho.org/showthread.php?p=195198
> * http://forums.pentaho.org/showthread.php?t=64775
> * http://forums.pentaho.org/showthread.php?t=64643
> * http://julianhyde.blogspot.com/2008/09/is-oracle-about-to-embrace-mpp.html

I'm fully aware of the challenges of "commercial" open source and
developer priorities.

> I'm sorry you are having problems getting aggregations to work. I'm afraid I
> tend to avoid questions about aggregations, just like installation
> questions. They take a lot of cycles to fix and rarely generate any insight
> that would help us make the product better. I wish that were not so.

This is too bad, because I'm interested in using (and contributing
to/developing for) mondrian for my work, where I already have a
homegrown solution that supports aggregation and db partitioning.
Mondrian might have some features that prove useful, but if I can't
get agg to work, it's a no go.

I'm surprised that agg doesn't appear to be important to you.  It
seems that any decent sized data set is going to benefit from it.

If agg has issues, that's fine.  Like I said, I've implemented agg (in
python) for my solution.  I'm willing to code and contribute to get
stuff working.  But if I'm on my own, then I'll just go back to my
homegrown stuff.



More information about the Mondrian mailing list