[USER] OpenQM third-party reporting tools
Dawn Wolthuis
reporting tools. None that I can find supports OpenQM. Does anyone
know of any that do?
In particular, I'm looking for any that run on linux middle (and back)
tiers and are client-OS independent. Thanks. --dawn
--
Dawn M. Wolthuis
Tincat Group, Inc.
Martin Phillips
We spoke to one such vendor some time ago. They have a system that runs on
UniVerse and uses UniObjects. We suggested that migration to QM should be
easy but they wanted us to fund six months of their time to look at it. This
isn't quite the same attitude that we use to prospective business
opportunities!
Perhaps you should go back to the vendors you spoke to and encourage them to
look at QM.
Martin Phillips, Ladybridge Systems
Dawn Wolthuis
>
> Hi Dawn,
>
> We spoke to one such vendor some time ago. They have a system that runs on
> UniVerse and uses UniObjects. We suggested that migration to QM should be
> easy
Do you think that would also be the case if they use UniObjects for
Java? I know someone was using JNI to access the client/server
libraries in Java at one point, but I don't know if there is anyone
running any live Java-to-OpenQM software. Is anyone here doing that?
> but they wanted us to fund six months of their time to look at it. This
> isn't quite the same attitude that we use to prospective business
> opportunities!
It sounds like an opening move in a negotiation ;-)
> Perhaps you should go back to the vendors you spoke to and encourage them to
> look at QM.
I've talked to two of them, both of the "money talks" camp, but I'm
not feelin' that is going anywhere at this point. The open source
community might be better positioned to be visionary on that front.
Cheers! --dawn
Dawn M. Wolthuis
Bob M.
How are you? Here comes the ad. BlueFinity's mv.NET supports OpenQM
and loads of other MultiValue databases for .NET integration.
BlueFinity is about to release Reporting Services Data Connector (RSDC)
for Microsoft Reporting Services.
Microsoft has spent lots of money developing the product and it now
comes at no cost with SQL Server 2005 SP1, including Reporting Services
as part of the SQL Server Express Edition.
More information can be found at www.bluefinity.com/rsdc.html.
Bob
Dawn Wolthuis
Overloaded. Good to see you are still at jBASE (since I don't think I
know anyone else there anymore).
> Here comes the ad.
I launched a question that was begging for someone to post an AD, so good deal.
> BlueFinity's mv.NET supports OpenQM
> and loads of other MultiValue databases for .NET integration.
> BlueFinity is about to release Reporting Services Data Connector (RSDC)
> for Microsoft Reporting Services.
OK. I happen to have an OpenQM database and wouldn't want my
reporting to be restricted to the way that SQL thinks about data
(single-values and all), but I don't know enough to have any problems
with MS Reporting Services at this point. I know they put a lot of
resources into it and I expect they will have lots of people using it.
But the database it accesses is SQL Server, and possibly other SQL
databases, right?
> Microsoft has spent lots of money developing the product
So we both heard the same news on that. I'm guessing they came up
with some good features too, from what I have seen.
> and it now
> comes at no cost with SQL Server 2005 SP1,
Which would be a database product, right ;-) I have a database. It's
OpenQM. (And another one called UniData). I don't want to buy SQL
Server in order to report from OpenQM, do I? (I'm just helping you
figure out your responses for folks).
> including Reporting Services
> as part of the SQL Server Express Edition.
>
> More information can be found at www.bluefinity.com/rsdc.html.
So far, yours is the best third-party solution for OpenQM reporting
that has come up, so I'll give you that. It isn't my cup of tea, but
it might work for someone else. What a shame to take rich nested data
and push it into a SQL-happy format to get reports out. Heavy sigh.
Thanks, Bob. Cheers! --dawn
Glen Batchelor
Dawn,
I'm leisurely working on the initial XML open source framework, but I hope
to have some stuff ready for the new QPROC when it becomes available. Are
you aware of any XML related database querying standards other than XQuery?
I'm also working on XML shipping, rating, and tracking tools with UPS and
FedEx right now, so I have been putting my old parser through some heavy
trials. We're going live with certified XML UPS shipping in possibly a week,
under D3. I'm at the point now of trying to decide how to implement a DOM in
OpenQM, before I start re-porting my parser code again.
-------------------------------------------
Glen Batchelor
IT Director
All-Spec Industries
phone: (910) 332-0424
fax: (910) 763-5664
e-mail: webm...@allspec.com
-------------------------------------------
www.allspec.com
-------------------------------------------
Martin Phillips
> I'm leisurely working on the initial XML open source framework,
> but I hope to have some stuff ready for the new QPROC when it
> becomes available.
We are probably not too far off being able to publish an interface between
the query processor and third party output components. The new query
processor is working reasonably well but so far nothing has been done to the
optimiser phase which means in particular that it does not yet take
advantage of indices.
I'd like to think that I will find time to continue with this soon but I've
been saying that since before Christmas.
Martin Phillips
Ladybridge Systems
17b Coldstream Lane, Hardingstone, Northampton NN4 6DB
+44-(0)1604-709200
Dave Walker
You go girl...
--
Dave
--
"They who would give up an essential liberty for temporary security, deserve neither liberty or security."
-- Benjamin Franklin
Dawn Wolthuis
I guess that was a little intense. Sheesh, women can get so emotional!
Sorry Bob, and, really, thanks for the tip on the reporting solution,
I owe you one for my reaction. There are books on MS Reporting
Services on the shelf, and I really can appreciate that. SQL Server
also has a lot of nice features these days. I might end up happily
using it someday soon (doubtful, but if it is the best solution for
reporting, well...)
Cheers! --dawn
Tony G
> I owe you one for my reaction. There are books on MS Reporting
> Services on the shelf, and I really can appreciate that. SQL Server
> also has a lot of nice features these days. I might end up happily
> using it someday soon (doubtful, but if it is the best solution for
> reporting, well...)
I would have made similar suggestions but your OP stated quite clearly:
"In particular, I'm looking for any that run on linux middle (and back)
tiers and are client-OS independent. Thanks. --dawn"
For better or worse, when you state up front that you only want to hear
about solutions that run on a specific platform you limit the responses
immensely.
mv.NET works with OpenQM on the back-end but the middle-tier must be
Windows. If you find another Windows-based reporting engine and want to
connect into QM, you could try running mv.NET with WINE, or with VMware
Server (for free) and a Win32 guest over a Linux host. With the current
popularity of application virtualization, this would probably be more
acceptable today with end-users than it was just a year ago. I have
discussed a Mono port with BlueFinity but without cost-justification it
won't happen for a while.
For other tools that more closely match your original requirements for full
reporting tools, have a look at SourceForge. Here is a start:
<https://sourceforge.net/search/?type_of_search=soft&group_id=714&words=rep
orting>
If you find something there that was developed in .NET and you really want
to run it over Linux, have a look at Visual MainWin:
http://www.mainsoft.com/products/index.aspx
HTH
T
Dawn Wolthuis
>
> I would have made similar suggestions but your OP stated quite clearly:
> "In particular, I'm looking for any that run on linux middle (and back)
> tiers and are client-OS independent. Thanks. --dawn"
>
> For better or worse, when you state up front that you only want to hear
> about solutions that run on a specific platform you limit the responses
> immensely.
OK, Tony, so you are saying that if you have any requirements for a
project, then you limit the possible solutions? Hmmm. Good point ?
> mv.NET works with OpenQM on the back-end but the middle-tier must be
> Windows. If you find another Windows-based reporting engine and want to
> connect into QM, you could try running mv.NET with WINE, or with VMware
> Server (for free) and a Win32 guest over a Linux host. With the current
> popularity of application virtualization, this would probably be more
> acceptable today with end-users than it was just a year ago. I have
> discussed a Mono port with BlueFinity but without cost-justification it
> won't happen for a while.
I respect the skills required to live in an MS world and try to figure
out how to get other stuff to play well with it. I also understand
the costs of making an architecture portable from the start. I work a
lot with higher ed and on most campuses you just don't tell an art
professor they cannot use a Mac if that is their preferred platform,
nor an Engineer that a Sun or linux workstation is out of the
question. This type of open systems approach works its way throughout
the organization into IT. Some shops are happy to have MS web
servers, but many are not. There is a cost to diversity, but also a
desire to minimize that cost. Using linux or other *nix hosts for web
servers and other non-lan hosts is one way to try to handle these
costs. Many in higher ed still use Novell for lan servers and have no
MS servers. Reliability in a college is a different beast than other
organizations too. Your student users are using whatever workstation
they bring on campus. Try managing reliability and security in that
environment for a while and you might steer clear of MS and related
viruses and reboots too.
> For other tools that more closely match your original requirements for full
> reporting tools, have a look at SourceForge. Here is a start:
> <https://sourceforge.net/search/?type_of_search=soft&group_id=714&words=rep
> orting>
My own specific interest is in reporting tools that don't think like
SQL, with its single-line of data thinking (it would take a bit to
explain that line, but I think you and most here have a sense of what
our data model gives us on the reporting front).
> If you find something there that was developed in .NET and you really want
> to run it over Linux, have a look at Visual MainWin:
> http://www.mainsoft.com/products/index.aspx
>
> HTH
> T
I always appreciate the perspective, Tony, even if I remain ignorant
on the .NET side of the house only because I just can't keep current
on enough fronts to cover it all. Cheers! --dawn
patric...@yahoo.com
connector with D3 (Tony???). Raining Data has a tutorial demonstrating
connecting to D3
"http://www.rainingdata.com/support/documentation/rsc/30/Tutorial/WebHelp/index.htm".
As you can see in the demo it appears to understand/work with
multivalues. Any of these report designers that actually utilize XML
style datasets (or even the normal microsoft datasets) can now handle
pick nested data.
Outside that, has anybody here had any experience with the Mits
products (www.mitsonline.com). I have recently deployed a Mits
discover here that is essentially a Olap Cube product. All the
datastorage is inside D3 while a small web server running on a windows
machine handles the socket and graphics work. Pretty impressive
product.
They now offer a mitsreporting tool that takes a normal pick report and
essentially stores the output in a external database. The user can
then reference this "report" via a web browser. The neat thing (this
is designed to be placed on existing systems) is that you can take a
sales report and run it for all divisions (instead of running seperate
reports or running one large report and then splitting it up by hand)
to this product. You then define access permissions to your users (Bob
for example can only see division=Sacramento). When he pulls up this
report it is automatically narrowed down to just the Sacramento data.
The end tool understands all the pick constructs (totals, break-on's,
etc) and therefore everything is re-totaled correctly for you. You can
then choose to archive/save these reports as essentially database
snapshots. The end user can also click on column headings and resort
he report, narrow the selection criteria down even more, etc.
I had talked to Mits last year about QM and they seemed interested.
They are attempting to break out of the pick market and I believe they
are struggling with keeping a pick backend or moving to something else
(as they did with Mits Report). QM seemed to interest them as a
possiblity of keeping with a pick backend but I bet in the end they
decided to stick with moving away from MV. I am sure although they
could be pushed to make a qm port (they currenlty support the U2
products and D3).
Even if we build our own new tools from the ground up I thought their
product may be a interesting product offering to look at for ideas. If
there is interest I could perhaps setup a demonstration login to allow
everybody to take a look at the mits discover tool.
Tony G
> Has anybody here actually used the Microsoft Reporting Services
> connector with D3 (Tony???). Raining Data has a tutorial
> demonstrating connecting to D3
>
"http://www.rainingdata.com/support/documentation/rsc/30/Tutorial/WebHelp/i
ndex.htm".
>
> As you can see in the demo it appears to understand/work with
> multivalues. Any of these report designers that actually utilize XML
> style datasets (or even the normal microsoft datasets) can now handle
> pick nested data.
Raining Data sabotaged themselves by opening with a $20k+ price tag and no
commensurate value-add justification. BlueFinity offers a better solution
for a fraction of the cost. You'll rarely see me this black-n-white about
topics but that's just the way it is. I have not used the RD software for
many reasons. At the moment I have one client who does intend to use
Reporting Services with Unidata. The BlueFinity software can be used
across many DBMS products. My development for their Unidata back-end will
be partially done with D3.
> Outside that, has anybody here had any experience with the Mits
> products (www.mitsonline.com). I have recently deployed a Mits
> discover here that is essentially a Olap Cube product. All the
> datastorage is inside D3 while a small web server running on a windows
> machine handles the socket and graphics work. Pretty impressive
> product.
I have a small bit of experience with MITS, even spent a week up in Seattle
in a MITS class. I'd like to do a lot more with it. It's really
impressive.
> I am sure although they could be pushed to make a qm port (they
> currenlty support the U2 products and D3).
As with any tool/service providers in our industry or elsewhere, there
needs to be cost-justification for going through the learning curves, doing
development, preparing for support, and then doing the marketing to attract
sales. Some key messages associated with QM/OpenQM right now are "free"
and "low cost". These messages are intended to attract people with a
specific mind-set. That mind-set is not in-line with many tool providers
who charge for their offerings and expect a prospect base that is capable
and willing to spend money with them. I know there is the notion that a
company that spends less on its DBMS has more money to spend on other
things but this isn't the way the real world works. The less a company
spends on the DBMS the less it expects to spend on everything associated
with it. When the QM market can boast some decent numbers of live end-user
sites and seats, other companies who want to sell to this market will take
note.
> Even if we build our own new tools from the ground up I thought their
> product may be a interesting product offering to look at for ideas.
> If there is interest I could perhaps setup a demonstration login to allow
> everybody to take a look at the mits discover tool.
That sort of proves my point.
T
Dawn Wolthuis
>
> Has anybody here actually used the Microsoft Reporting Services
> connector with D3 (Tony???). Raining Data has a tutorial demonstrating
> connecting to D3
> "http://www.rainingdata.com/support/documentation/rsc/30/Tutorial/WebHelp/index.htm".
>
> As you can see in the demo it appears to understand/work with
> multivalues.
Thanks for the link, Patrick.
I looked around and don't see what you are seeing that looks like
working with multivalues. In particular, I look for a couple of
features that might well be there, but I'm likely not taking enough
time to find it.
1) A query with multiple multivalues (or mv associations) where the
lists are not associated with each other except through the parent.
For example, a person with a (virtual field of) a full multi-line
address block and a multivalued classification of some sort. So, you
would have on a second line of output for this person, the second line
of the address and the second classification code. These are not
related to each other except they are both the second value in a list
associated with the person. In a result set, these would each be
multi-valued attributes.
2) two-valued logic, instead of three-valued. For example, if I ask
if John and Linda have the same birthday and both have "" (MV null),
then I want the answer to be "yes" -- all logical expressions should
evaluate to either true or false, no third value.
3) no fixed lengths. If a field is specified to display in a certain
space, then if the value is longer than that and the report has not
been told to truncate the value, it would naturally wrap the value in
that space
4) The ability for the user to ask questions like requesting a list of
all students who are not majoring in philosophy. I'm sure that MS
reports can do this, but would want to understand how.
> Any of these report designers that actually utilize XML
> style datasets (or even the normal microsoft datasets) can now handle
> pick nested data.
That is great!
> Outside that, has anybody here had any experience with the Mits
> products (www.mitsonline.com). I have recently deployed a Mits
> discover here that is essentially a Olap Cube product. All the
> datastorage is inside D3 while a small web server running on a windows
> machine handles the socket and graphics work. Pretty impressive
> product.
I haven't yet seen it, but I plan to. They said there are no plans to
port to OpenQM.
> They now offer a mitsreporting tool that takes a normal pick report and
> essentially stores the output in a external database. The user can
> then reference this "report" via a web browser. The neat thing (this
> is designed to be placed on existing systems) is that you can take a
> sales report and run it for all divisions (instead of running seperate
> reports or running one large report and then splitting it up by hand)
> to this product. You then define access permissions to your users (Bob
> for example can only see division=Sacramento). When he pulls up this
> report it is automatically narrowed down to just the Sacramento data.
> The end tool understands all the pick constructs (totals, break-on's,
> etc) and therefore everything is re-totaled correctly for you. You can
> then choose to archive/save these reports as essentially database
> snapshots. The end user can also click on column headings and resort
> he report, narrow the selection criteria down even more, etc.
>
> I had talked to Mits last year about QM and they seemed interested.
> They are attempting to break out of the pick market and I believe they
> are struggling with keeping a pick backend or moving to something else
> (as they did with Mits Report).
What are they using as the backend for MITS Report?
> QM seemed to interest them as a
> possiblity of keeping with a pick backend but I bet in the end they
> decided to stick with moving away from MV.
There is a lot more data in the non-MV world than in MV.
> I am sure although they
> could be pushed to make a qm port (they currenlty support the U2
> products and D3).
I gave it a shot.
> Even if we build our own new tools from the ground up I thought their
> product may be a interesting product offering to look at for ideas. If
> there is interest I could perhaps setup a demonstration login to allow
> everybody to take a look at the mits discover tool.
I'm planning to take a look when I'm in Seattle in June (speaking at
the Pick Users Group meeting out there). Thanks. --dawn
patric...@yahoo.com
Dawn Wolthuis wrote:
> On 5/7/06, patr...@spm1.com <patric...@yahoo.com> wrote:
> >
> > Has anybody here actually used the Microsoft Reporting Services
> > connector with D3 (Tony???). Raining Data has a tutorial demonstrating
> > connecting to D3
> > "http://www.rainingdata.com/support/documentation/rsc/30/Tutorial/WebHelp/index.htm".
> >
> > As you can see in the demo it appears to understand/work with
> > multivalues.
>
> Thanks for the link, Patrick.
> I looked around and don't see what you are seeing that looks like
> working with multivalues. In particular, I look for a couple of
> features that might well be there, but I'm likely not taking enough
> time to find it.
I only looked at it quickly but the Orders example did seem to imply it
was working with a multi-value set. I would assume in order for RD to
endorce this they would have to support the multi-values. I am
assuming the backend is all based on pdp.net and in that area I would
assume what blue infinity is offering must be similiar.
It sounds like everything is done in Java (again off a windows server).
Tony G
> I only looked at it quickly but the Orders example did seem to imply
> it was working with a multi-value set. I would assume in order for RD to
> endorce this they would have to support the multi-values. I am
> assuming the backend is all based on pdp.net and in that area I would
> assume what blue infinity is offering must be similiar.
Yes, the whole purpose of these products is to allow a Reporting Services
user to access their MV back-end and return meaningful, structured data
without regard for whether the back end is storing it as multivalues or
joined tables.
You don't "need" the integration products provided by Raining Data or
BlueFinity to use Reportings Services. These companies used their own
products as embedded tools to create an application/utility, and they are
selling the finished product. As any of us in the third-party market would
do, RD and BF have invested a lot of labor into making the final solution
look easy and function transparently - and that's what someone buys if they
want it, not necessarily the underlying tools.
These RS connectivity products serve as examples of what any of us can do
with the tools, to get popular and in-demand products to integrate with MV
as easily as they integrate with relational. I'm hoping that over time
people will start to think less of the individual tools as solutions until
themselves (Why NET? Why Java? Why UO? Why MV? Why QM?...) and think more
about the value-add solutions that they can create using some combination
of available tools.
Tony
TG@ removethisNebula-RnD.com
Dawn Wolthuis
>
>
> Dawn Wolthuis wrote:
> > On 5/7/06, patr...@spm1.com <patric...@yahoo.com> wrote:
> > >
> > > Has anybody here actually used the Microsoft Reporting Services
> > > connector with D3 (Tony???). Raining Data has a tutorial demonstrating
> > > connecting to D3
> > > "http://www.rainingdata.com/support/documentation/rsc/30/Tutorial/WebHelp/index.htm".
> > >
> > > As you can see in the demo it appears to understand/work with
> > > multivalues.
> >
> > Thanks for the link, Patrick.
> > I looked around and don't see what you are seeing that looks like
> > working with multivalues. In particular, I look for a couple of
> > features that might well be there, but I'm likely not taking enough
> > time to find it.
>
> I only looked at it quickly but the Orders example did seem to imply it
> was working with a multi-value set. I would assume in order for RD to
> endorce this they would have to support the multi-values. I am
> assuming the backend is all based on pdp.net and in that area I would
> assume what blue infinity is offering must be similiar.
There is a big, but subtle, difference between a reporting tool that
can have MV databases as a data source, but "thinks like" SQL and an
MV reporting tool. An XML reporting tool that "thinks like" XML would
be close to an MV reporting tool. I do know that Microsoft has worked
on making some of the advances in their data model that would be
appreciated by MV folks, but from what I have read, I cannot yet tell
if they have 2VL reporting tools, handle multiple multi-values as we
would want etc. Maybe they do, but I'm guessing not quite.
If they are still restricted to similar reporting as Crystal Reports
and other SQL-thinking reporting tools, plus their OLAP cubes, that
still doesn't get us as far as the MV Query language in some respects.
Obviously it gets us well beyond the MV Query language for pretty
output, however. I'm kinda demanding in wanting both. Entrinsik
Informer is an example of a tool that "thinks like" MV.
Cheers! --dawn
Bob M.
>OK. I happen to have an OpenQM database and wouldn't want my reporting
>to be restricted to the way that SQL thinks about data (single-values
>and all), but I don't know enough to have any problems with MS
>Reporting Services at this point. I know they put a lot of resources
>into it and I expect they will have lots of people using it.
> But the database it accesses is SQL Server, and possibly other SQL
>databases, right?
No. The mv.NET adapter for Reporting Services allows you to report on
your multi-value data, be it QM, jBASE or whatever. Based on the
information you provide about your data, mv.NET will normalize it
automatically for reporting purposes.
>Which would be a database product, right ;-) I have a database. It's
>OpenQM. (And another one called UniData). I don't want to buy SQL
>Server in order to report from OpenQM, do I? (I'm just helping you
>figure out your responses for folks).
The capabilities of SQL Server Reporting Services include the ability
to report data other than that from SQL Server. The interface is much
more user-friendly than that of Crystal Reports. As of SQL Server 2005
SP1, Reporting Services in part of the free SQL Server Express version.
We are currently evaluating the capabilities of this version in
comparison to those of the "paid" version.
>So far, yours is the best third-party solution for OpenQM reporting
>that has come up, so I'll give you that. It isn't my cup of tea, but
>it might work for someone else. What a shame to take rich nested data
>and push it into a SQL-happy format to get reports out. Heavy sigh.
Of course it its the best and It provides capabilities which are of
interest to a number of multi-value users, and as far as most end-users
are concerned, they don't care how that data is stored, so long as it
looks good when the produce a report.
And of course the BlueFinity technical team helped me with the
responses.
Bob
Dawn Wolthuis
>
> Dawn,
>
> >OK. I happen to have an OpenQM database and wouldn't want my reporting
> >to be restricted to the way that SQL thinks about data (single-values
> >and all), but I don't know enough to have any problems with MS
> >Reporting Services at this point. I know they put a lot of resources
> >into it and I expect they will have lots of people using it.
> > But the database it accesses is SQL Server, and possibly other SQL
> >databases, right?
>
> No. The mv.NET adapter for Reporting Services allows you to report on
> your multi-value data, be it QM, jBASE or whatever. Based on the
> information you provide about your data, mv.NET will normalize it
> automatically for reporting purposes.
If I always wanted to have my data normalized before or while being
reporting on, then I would have used a SQL DBMS in the first place.
> >Which would be a database product, right ;-) I have a database. It's
> >OpenQM. (And another one called UniData). I don't want to buy SQL
> >Server in order to report from OpenQM, do I? (I'm just helping you
> >figure out your responses for folks).
>
> The capabilities of SQL Server Reporting Services include the ability
> to report data other than that from SQL Server.
Can data sources other than SQL-based data sources (ODBC, OLE/DB) be
employed? It sounds like they can, so it would be nice to know what
the other options are.
> The interface is much
> more user-friendly than that of Crystal Reports. As of SQL Server 2005
> SP1, Reporting Services in part of the free SQL Server Express version.
> We are currently evaluating the capabilities of this version in
> comparison to those of the "paid" version.
I knew that there were two database engines (jet and the one under sql
server) that had free downloads. I was not on top of the SQL Server
Express product. I'll have to take a look.
> >So far, yours is the best third-party solution for OpenQM reporting
> >that has come up, so I'll give you that. It isn't my cup of tea, but
> >it might work for someone else. What a shame to take rich nested data
> >and push it into a SQL-happy format to get reports out. Heavy sigh.
>
> Of course it its the best and It provides capabilities which are of
> interest to a number of multi-value users, and as far as most end-users
> are concerned, they don't care how that data is stored, so long as it
> looks good when the produce a report.
No everyone is aware of the advantages of reporting with the
multi-value model, without switching it to a SQL-based model. I know
MS has put dollars into the XML data model as well, however, so
perhaps they have done something interesting in that regard. I'll
have to see if there is a connection to XML data stores, for example
and see whether they retain the data model in that case or continue to
insist on converting it to an SQL mindset.
> And of course the BlueFinity technical team helped me with the
> responses.
Thanks (to you and them). Cheers! --dawn
Tony G
>>> But the database it accesses is SQL Server, and possibly other SQL
>>> databases, right?
>>
>> No. The mv.NET adapter for Reporting Services allows you to report on
>> your multi-value data, be it QM, jBASE or whatever. Based on the
>> information you provide about your data, mv.NET will normalize it
>> automatically for reporting purposes.
>
> If I always wanted to have my data normalized before or while being
> reporting on, then I would have used a SQL DBMS in the first place.
Dawn, this is an internal process. You don't need to normalize the data,
it's not normalized in the MV DBMS and you never see it normalized.
>>> Which would be a database product, right ;-) I have a database.
>>> It's OpenQM. (And another one called UniData). I don't want to
>>> buy SQL Server in order to report from OpenQM, do I? (I'm just
>>> helping you figure out your responses for folks).
>>
>> The capabilities of SQL Server Reporting Services include the ability
>> to report data other than that from SQL Server.
>
> Can data sources other than SQL-based data sources (ODBC, OLE/DB) be
> employed? It sounds like they can, so it would be nice to know what
> the other options are.
This is a raison d'être of Reporting Services, it's a data aggregator and
reporting engine which can use data from any number of data sources.
>>> So far, yours is the best third-party solution for OpenQM reporting
>>> that has come up, so I'll give you that. It isn't my cup of tea,
>>> but it might work for someone else. What a shame to take rich
>>> nested data and push it into a SQL-happy format to get reports out.
>>> Heavy sigh.
>>
>> Of course it its the best and It provides capabilities which are of
>> interest to a number of multi-value users, and as far as most
>> end-users are concerned, they don't care how that data is stored, so
>> long as it looks good when the produce a report.
>
> No everyone is aware of the advantages of reporting with the
> multi-value model, without switching it to a SQL-based model.
Reporting Services allows for much better formatting of reports than any MV
platform. It also does scheduled generation and distribution in different
formats and through different media. You could say "I can email my own
reports", "I can generate PDF", "I can export to Excel"... But this is a
single product that does all of that and much more.
None of this requires switching to SQL. You are using SQL Server as one
tool among many to get a job done, but it doesn't mean you need to switch
or be intimately familiar with its inner workings. IMO, having the name
SQL associated with Reporting Services wasn't a good marketing decision,
but like any of us, Microsoft is promoting the benefits of having their
database embedded within a higher level product/application.
> I know MS has put dollars into the XML data model as well, however, so
> perhaps they have done something interesting in that regard. I'll
> have to see if there is a connection to XML data stores, for example
> and see whether they retain the data model in that case or continue to
> insist on converting it to an SQL mindset.
Dawn, you have to understand that XML is so easy to convert into a
relational datastore that well formed XML and relational tables are
considered synonymous and interchangable to most .NET developers. Import
XML and do a SQL query on it, import nested tables and export as XML with a
schema - it's all the same.
You can use XML data stores with Reporting Services - see my post in CDP
about using XML with MV via ADO.NET and mv.NET. There's only a couple
lines of code required to import XML and convert it to an ADO.NET dataset,
which is what Reporting Services uses internally. (That code is already in
there, you don't need to write code.)
>> And of course the BlueFinity technical team helped me with the
>> responses.
>
> Thanks (to you and them). Cheers! --dawn
HTH
T
Dawn Wolthuis
>
> Dawn Wolthuis wrote:
Apologies for the not-exactly-on-target for OpenQM response. Perhaps
we should move this to cdp, but I'm headed out of town anyway, so last
post for me for a while.
> >>> But the database it accesses is SQL Server, and possibly other SQL
> >>> databases, right?
> >>
> >> No. The mv.NET adapter for Reporting Services allows you to report on
> >> your multi-value data, be it QM, jBASE or whatever. Based on the
> >> information you provide about your data, mv.NET will normalize it
> >> automatically for reporting purposes.
> >
> > If I always wanted to have my data normalized before or while being
> > reporting on, then I would have used a SQL DBMS in the first place.
>
> Dawn, this is an internal process. You don't need to normalize the data,
> it's not normalized in the MV DBMS and you never see it normalized.
I don't think you are catching my point on this one, Tony. I can
normalize or a process can normalize mv data and then we have just as
good an approach as anyone using SQL, but not better. There actually
are some differences between the relational model and the multivalue
model that favor reporting in line with the MV model. Examples
include those I indicated, such as multiple one-to-many's and just
reporting from the perspective of a person, place, event, or thing in
general, if there are multi-valued properties of such entities.
> >>> Which would be a database product, right ;-) I have a database.
> >>> It's OpenQM. (And another one called UniData). I don't want to
> >>> buy SQL Server in order to report from OpenQM, do I? (I'm just
> >>> helping you figure out your responses for folks).
> >>
> >> The capabilities of SQL Server Reporting Services include the ability
> >> to report data other than that from SQL Server.
> >
> > Can data sources other than SQL-based data sources (ODBC, OLE/DB) be
> > employed? It sounds like they can, so it would be nice to know what
> > the other options are.
>
> This is a raison d'être of Reporting Services, it's a data aggregator and
> reporting engine which can use data from any number of data sources.
Take mv.NET out of the picture and tell me if you can specify a data
source and not have SQL queries pass into the data source. Is there a
way to have XQuery queries instead, returning XML documents as results
instead?
> >>> So far, yours is the best third-party solution for OpenQM reporting
> >>> that has come up, so I'll give you that. It isn't my cup of tea,
> >>> but it might work for someone else. What a shame to take rich
> >>> nested data and push it into a SQL-happy format to get reports out.
> >>> Heavy sigh.
> >>
> >> Of course it its the best and It provides capabilities which are of
> >> interest to a number of multi-value users, and as far as most
> >> end-users are concerned, they don't care how that data is stored, so
> >> long as it looks good when the produce a report.
> >
> > No everyone is aware of the advantages of reporting with the
> > multi-value model, without switching it to a SQL-based model.
>
> Reporting Services allows for much better formatting of reports than any MV
> platform.
I am certain that far more dollars have gone into it than anything any
mv vendor or 3rd party has ever spent on reporting tools.
> It also does scheduled generation and distribution in different
> formats and through different media. You could say "I can email my own
> reports", "I can generate PDF", "I can export to Excel"... But this is a
> single product that does all of that and much more.
Informer from Entrinsik does those too (with U2). I'm sure it is far
simpler. But it never has to do the translation from mv to sql in
order to report.
> None of this requires switching to SQL.
Really? There is no SQL changing hands and no SQL "result sets" being
passed in any of this? What does it use instead?
> You are using SQL Server as one
> tool among many to get a job done, but it doesn't mean you need to switch
> or be intimately familiar with its inner workings.
apparently you don't know me ;-) I really would need to know all
intermediate data models in the process from source to report, and
whether there is an extended flavor of SQL that can handle multiple
one-to-manys and other things like mv does.
> IMO, having the name
> SQL associated with Reporting Services wasn't a good marketing decision,
> but like any of us, Microsoft is promoting the benefits of having their
> database embedded within a higher level product/application.
You definitely seem to be telling me that this is not a SQL-based
reporting tool, so I just gotta know what it is then.
> > I know MS has put dollars into the XML data model as well, however, so
> > perhaps they have done something interesting in that regard. I'll
> > have to see if there is a connection to XML data stores, for example
> > and see whether they retain the data model in that case or continue to
> > insist on converting it to an SQL mindset.
>
> Dawn, you have to understand that XML is so easy to convert into a
> relational datastore that well formed XML and relational tables are
> considered synonymous
I know that you can map them, although not necessarily the same going
in both directions (if you apply the switchDataModel function and then
apply it again, do you get what you started with?).
I want to understand how it handles nested lists, di-graphs of data
with trees on the nodes, or however you want to talk about an XML or
MV data model (similar, not the same).
> and interchangable to most .NET developers. Import
> XML and do a SQL query on it, import nested tables and export as XML with a
> schema - it's all the same.
Not to me it isn't. SQL uses a 3VL instead of a two-valued logic that
MV and XQuery use. SQL-92 does not permit nested collection types,
although I'm sure MS SQL does, I just don't know how. SQL DBMS's
don't tend to all implement the most recent SQL standards the same
way. If it always puts data in 1NF before running a report, you lose
some of the beneifts of the richer data model found in MV and XML.
> You can use XML data stores with Reporting Services - see my post in CDP
> about using XML with MV via ADO.NET and mv.NET. There's only a couple
> lines of code required to import XML and convert it to an ADO.NET dataset,
> which is what Reporting Services uses internally. (That code is already in
> there, you don't need to write code.)
OK, I'll look into the data model behind ADO.NET then. While I know
MS wrote new reporting tools, I doubt they are passing requests to
databases with a highly non-standard language or they would be
hard-pressed to use other databases. So, I'm guessing it is SQL, but
perhaps XQuery is an option.
--dawn
Tony G
looking for info on extending their use of their platform of choice. I've
been having exchanges with Dawn off-list and have responded to her non-MV
questions there.
Dawn Wolthuis wrote:
> I don't think you are catching my point on this one, Tony. I can
> normalize or a process can normalize mv data and then we have just as
> good an approach as anyone using SQL, but not better. There actually
> are some differences between the relational model and the multivalue
> model that favor reporting in line with the MV model. Examples
> include those I indicated, such as multiple one-to-many's and just
> reporting from the perspective of a person, place, event, or thing in
> general, if there are multi-valued properties of such entities.
I'm focused on the end-objective to get data on paper, not the mechanics of
data models. The bottom line is that with no normalization or manipulation
whatsoever of your existing database you can generate reports of MV data
using the Microsoft SQL Reporting Services data aggregation and
distribution engine. How tough is that? All of this other rhetoric just
gets in the way of getting the job done.
>> It also does scheduled generation and distribution in different
>> formats and through different media. You could say "I can email my
>> own reports", "I can generate PDF", "I can export to Excel"... But
>> this is a single product that does all of that and much more.
>
> Informer from Entrinsik does those too (with U2). I'm sure it is far
> simpler. But it never has to do the translation from mv to sql in
> order to report.
In short there is no translation which involves the user/developer (more
off-list).
>> None of this requires switching to SQL.
>
> Really? There is no SQL changing hands and no SQL "result sets" being
> passed in any of this? What does it use instead?
Data can be provided in various ways, via MV queries, coding, or a
combination just like we do in BASIC to generate standard LPTR reports, but
no, you don't make SQL queries against your MV database. Enhancements are
being made to mv.NET to allow for more SQL-like queries for people who want
that (one of my clients is more comfortable with SQL, YMMV) but you don't
need to use SQL syntax.
> You definitely seem to be telling me that this is not a SQL-based
> reporting tool, so I just gotta know what it is then.
Ya think? ;)
Let's back up and change the perspective a little. mv.NET is a product to
allow an MV DBMS to be used in an environment that is already sold on .NET.
It's purpose is not solely to sell .NET to MV people, though it's nice if
some MV people catch on and decide they like it. The RSDC from BlueFinity
is not intended solely to sell MSRS to MV people but to give all-Relational
shops one less reason to trash MV.
As an MV person you should try to sell yourself on the MSRS concept first
and then you'll recognize that the RSDC is the best way to get you there.
If you look at MSRS as an end-user solution and it's not what you want then
move on to something else.
As a solution provider and recognized expert for your end-user community,
someone might ask if you can integrate your MV/QM app with other products
that are in common use. mv.NET and RSDC allow you to respond positively
with solutions that are appealing to the people asking those questions, and
that helps you to keep your software in place.
All of these efforts for making the MV model more acceptable (open source,
lower costs, etc) are futile if MV remains isolated from the enterprise.
Being right about the superiority of the MV model has got to take a back
seat to doing business with a large part of the world, which, whether you
like it or not, is using SQL and Microsoft .NET.
The MSRS solution is only one of many. I've also suggested that you can
use open source tools and completely gut out the relational dependencies.
Everything you want is available and most of it for free. I don't know if
any more rhetoric on this topic is really necessary.
T
patric...@yahoo.com
appears to be Jasper Reports
http://jasperreports.sourceforge.net
- Patrick