Not a Number Values in the Database

Our application has a table, which is populated via ADO.Net from C# with data originating from a C++ COM call. Today I encountered an entry that is C++ code for an undefined value: -1.#IND stored in the database. However, I could only discover what was stored in the table by Casting the value to a varchar -- simply selecting returned an error.

Is this expected behavior or a bug? It does not seem correct that SQL Server should store a value that cannot be displayed. In essence, either the value should not be allowed in the table because it violated the domain or SQL Server ought to have a way to display it with a Select *.

As fas as our application is concerned, we will be masking these values -- initially by ignoring them in the queries and eventually the loading program will convert to null.

What is the defined datatype for the field?|||

The column in question is a float column.

I would argue that the datatype should not be that important. The server should not allow a value to be stored which cannot be displayed.

|||

IND could be (COULD BE) referencing the unicode index control character. It looks like you had some sort of odd data transformation from the COM-C++-C#-SQL conversion process, especially considering it's a floating point variable which can lose precision through casting, it didn't surprise me to see that that was the data type in question.

Have you considered testing that single row by running it through your conversion process again by itself? Is the original float data for the column and row something... unusual?

|||

Hmm, very interesting. Try casting it to a numeric of some sort (particularly really large precision after the decimal)... That might give the clue. What datatype are you using in your C# application? And what prints out when you select from the column in Management Studio?

I can't imagine there is any value that is illegal for a float, but stranger things have happened.

|||

The specific data in the database was purged, due to some maintenance, so I'll have to see if it returns.

However, I made a simple example:

static void Main(string[] args)

{

double d = Double.NegativeInfinity;

using( SqlConnection conn = new SqlConnection( "server=(local);initial catalog=Northwind;integrated security=true" ) )

{

SqlCommand cmd = new SqlCommand( "Insert Foo(zero, one) Values( Datepart(mi, getdate() ),@.val )", conn );

cmd.Parameters.Add( "@.val", d );

conn.Open();

int rows = cmd.ExecuteNonQuery();

Console.WriteLine( rows );

}

Console.WriteLine( d );

}

with a target table:

Create Table Foo(

zero int not null,

one float null

)

Using any of Double.NegativeInfinity, Double.PositiveInfinity, or Double.NaN gives an odd result. The code succeeds and reports that a record has been added. However, the command Select * from Foo will not return records with these values. The values cannot be displayed by Casting to a numeric type, but output is produced by Casting to a varchar.

Using Select Count(*) from Foo will report the correct number of records.

|||

If you worked with MDX queries (analysis server), it is very common issue.

When you store the INDEFINITE/INFINTE value into your variable, compiler uses the special values to identify the INDEFINITE/INFINTE number/value,

1. 1.#IND (positive indefinite) or -1.#IND (negative indefinite)

2. 1.#INF (positive infinite)or -1.#INF (negative infinite)

Both are used to identify the compiler the variables hold the INDEFINITE/INFINTE, usally when you try to divide by zero it will be throw an error. But here you are forcefully stroring the infinite value. It is the compiler specific data. It can’t trusted across the compilers. So when you execute the code on your .NET framework will work fine. When try to pass this value into SQL Server it doesn’t understand, there is no specification available to identify the indefinite value in database.

Logically you can have nothing in your column (NULL/ZER0), but you are not allowed to store indefinite value in your column. Indefinite values are not comparable, these are imaginary values. But database store the masseurable values.

Solution:

Before passing the value to your database, validate the data, if it is INDEFINITE/INFINTE reset the value to NULL or other identifiable value for your future logics...

|||

I know you are a moderator, but as a general rule the thread creator should close threads. Being a moderator does not automatically make your information infallible.

As a matter of fact, on this issue you did not really add any additional information to the thread other than that you have experienced the issue as well.

I would now contend that this is, in fact, a bug in the SQL Server implementation. Either these values -- NaN, PositiveInfinity, and NegativeInfinity -- should be excluded from the domain of the Float datatype and rejected on Insert; or they are included in the domain and should be handled in a logical and consistent manner -- certainly they should have a display representation and they should have defined behavior in functions.

The current implementation is just a landmine waiting to go off at inopportune times.

|||

I marked the thread as 'answered'. As a 'general rule', we all take responsibitly to close a thread when there doesn't seem to be anymore 'interest' and/or it appears to be adequately 'answered'. And anyone can choose to 're-open' that thread by 'unmarking' the marked 'answer'. It's just a 'housekeeping' matter to help focus limited resources (the time and energy of volunteers) to the active and unanswered threads.

It appeared that Mani provided an adequate explanition for the behavior, and it seemed to me to be doubtful if there could be a 'solution' posted for your question. It appears that there to be a 'anomoly' involved in this situation -concerning a C++ interaction with the datastore, but I haven't been able to locate a declared 'bug' about the issue. (Doens't mean it's not there, just that I haven't devoted enough time to find it if it exists.) Mani's response satisfied the need to provide information to subsequent readers of the thread.

My mistake was in not encouraging you to post a 'potential' bug. I'll correct that now.

Please go to: http://connect.microsoft.com/sqlserver and post your observations and concerns. With everyone's keen interest and help to provide feedback and critique, the SQL Server will only continue to get better.

If you want this thread to stay 'open' to see if additional comments are forthcoming, that is perfectly fine.

As a postscript, let me add that there is often an issue of 'the' answer. We work on the concept of helping the questioner find a way to solve his/her problem. At some point, one has to say, ok, maybe there is a 'better' answer, but what is here now works, so lets move on.

|||

I have found more information on the issue:

First to clarify, the issue is not restricted to C++ interaction. The code I show above is C# and can be used to demonstrate the behavior when NaN or Infinity is stored in the database. Furthermore, I was surprised to find out that the data can be read out with a query from a C# program. My code and example were run against Sql Server 2000.

Looking on Microsoft Connections, I found an interesting item. It appears that MS has chosen to no longer support NaN and Infinity in the database in SQL Server 2005. In fact, a request had been made to restore the 2000 functionality. There is, at least, one organization which is using SQL Server to store engineering data and who has the need to occassionally store Infinities in the database. Evidently they are not performing any ad hoc queries or calculation within the database.

I plan to post a recommendation that MS embrace full support for the IEEE floating point numbers rather than simply disallowing them. It seems like the best, long term, behavior for the database and would be keeping in line with industry trends. Ideally, some form of patch would be made available for SQL 2000 to more gracefully handle NaN or Infinity when it is entered in the database.

|||Yes, this is a behavior change from SQL Server 2000. We now validate float values and Unicode data when you pass them via RPC calls or Bulk Copy APIs. So for float even though values like NaN and Infinity is allowed in the IEEE 754 spec (floating point implementation) we do not allow it now. Upon upgrade to SQL Server 2005 from older versions you need to modify the data in your tables to use them within TSQL. TSQL does not support any way to query these type of data or manipulate it explicitly - older SQL Server versions used to just allow them to be stored but you can only manipulate it on the client-side.|||

Is there a strong reason to not implement full support for the IEEE 754 spec? It specifies the rules for all operations involving the extra values. Some external representation of the values would be needed, but that has been solved for C# already.

It is a little surprising that it is not easier to support them than to prevent them. I assume that much of the Server software is written in a high level language that supports IEEE floats.

With competing products already supporting IEEE floats, the change seem inevitable.

Thanks for your input.

|||

I agree with your assessment. It is really unfortunate that we do not support it yet. Did you vote on the connect bug at https://connect.microsoft.com/SQLServer/feedback/ViewFeedback.aspx?FeedbackID=239674? I would also encourage anyone reading this thread to vote on this bug. The issue today is that several built-ins in SQL Server (those that work with floating point data) and other facilities cannot handle these values. I will follow-up with the owner of this bug and see if we can do anything about it in SQL Server 2008.

sql

Normalize this data

How would you normalize this data into a useable database?
Customer Name City Contact Address
Contact First Name Contact Phone Number Contact Fax Number
Billing Address Country Contact Email
Customer Number Contact Title User Name
Contact Last Name Shipping Address Password
Postal Code Contact Cell Phone User Role PrivilegePut it in a table?
<billsahiker@.yahoo.com> wrote in message
news:1192905303.503854.262450@.i13g2000prf.googlegroups.com...
> How would you normalize this data into a useable database?
> Customer Name City Contact Address
> Contact First Name Contact Phone Number Contact Fax Number
> Billing Address Country Contact Email
> Customer Number Contact Title User Name
> Contact Last Name Shipping Address Password
> Postal Code Contact Cell Phone User Role Privilege
>|||On Oct 20, 1:18 pm, "Jay" <s...@.nospam.org> wrote:
> Put it in a table?
> <billsahi...@.yahoo.com> wrote in message
> news:1192905303.503854.262450@.i13g2000prf.googlegroups.com...
>
> > How would you normalize this data into a useable database?
> > Customer Name City Contact Address
> > Contact First Name Contact Phone Number Contact Fax Number
> > Billing Address Country Contact Email
> > Customer Number Contact Title User Name
> > Contact Last Name Shipping Address Password
> > Postal Code Contact Cell Phone User Role Privilege- Hide quoted text -
> - Show quoted text -
Two or more tables, and indicate primary and foreign keys. This was a
test item and that is all the information given.|||<billsahiker@.yahoo.com> wrote in message
news:1192908678.033282.164490@.k35g2000prh.googlegroups.com...
> On Oct 20, 1:18 pm, "Jay" <s...@.nospam.org> wrote:
>> Put it in a table?
>> <billsahi...@.yahoo.com> wrote in message
>> news:1192905303.503854.262450@.i13g2000prf.googlegroups.com...
>>
>> > How would you normalize this data into a useable database?
>> > Customer Name City Contact Address
>> > Contact First Name Contact Phone Number Contact Fax Number
>> > Billing Address Country Contact Email
>> > Customer Number Contact Title User Name
>> > Contact Last Name Shipping Address Password
>> > Postal Code Contact Cell Phone User Role Privilege- Hide quoted text -
>> - Show quoted text -
> Two or more tables, and indicate primary and foreign keys. This was a
> test item and that is all the information given.
>
How about doing your own homework...
--
David Portas|||Why do you assume it is school work? It was a job application test I
took yesterday.
> How about doing your own homework...
> --
> David Portas- Hide quoted text -
> - Show quoted text -|||Ah, because you told the truth, I will help. And there is little difference
between an employment test and homework. You should have said so in the
first place.
Basically, it looks like they got together and looked for ways to screw
applicants up. They probably thought they were being cute mixing up and
eleminating necessary attributes too.
The first thing that jumps out is that you're storing people and addresses,
so that's two tables:
People is 1-to-many to Addresses (more on addresses later).
Second (and it's glaring): "Customer Name" vs. "Contact First Name" &
"Contact Last Name"
These columns should be: "Prefix", "First Name", "Middle Name",
"Last Name" & "Suffix".
(Mr. John Adam Smith, Sr)
Third Customers have "User Role Privilege", "User Name" and Password,
Contacts do not. Therefore, some people have attributes that others do not.
This suggests storing Customers and Contacts seperatly.
--
If you take the list of attributes they provided and arrange it, more things
become apparent:
Customer Number
User Name
Password
User Role Privilege
Customer Name
Billing Address
City
Postal Code
Country
Shipping Address
Contact Title
Contact First Name
Contact Last Name
Contact Address
Contact Phone Number
Contact Cell Phone
Contact Fax Number
Contact Email
Comments:
"Billing Address" and "Shipping Address" are clearly attributes of the
Customer. However, "City", "Postal Code" (Zip Code) & "Country" are vague.
Are they for Billing, Shipping, or Contact. In fact, you really need those
for each address. Same for attributes that appear in Contact, you should
have those for Customers too. In general, they really messed with the
attribute list.
Also, what "State" is the address in?
People (not including addresses):
There are two approaches here: store the "Contact" as attributes in
Customer, or create three tables: Customer, Contact & People, where "People"
is the parent table. You'll probably have to define your own PK for people
too and store it as a FK in Customer alone with the Customer Only info.
Addresses:
Create a seperate address table with all approporiate attributes. Then
you need a way to link people to addresses. You can do this by adding a
column to one of the people tables, or creating a many-to-many relationship
between people/customer/contact tables and addresses.
Since you have three distinct address types, you might want an AddressType
column in addresses constrained either by a lookup table, or a check
constraint. There are pluses and minuses for each. I prefer lookup tables,
but I'm a 3NF+ bigot.
Well, that covers the basics. Can you normalize the tables now?
<billsahiker@.yahoo.com> wrote in message
news:1192905303.503854.262450@.i13g2000prf.googlegroups.com...
> How would you normalize this data into a useable database?
> Customer Name City Contact Address
> Contact First Name Contact Phone Number Contact Fax Number
> Billing Address Country Contact Email
> Customer Number Contact Title User Name
> Contact Last Name Shipping Address Password
> Postal Code Contact Cell Phone User Role Privilege
>

Normalize data?

I have two columes F1 and F2
Both are NUMBER but were orginally imported from a TXT file.
Is there any way to ENSURE, that there are no spaces and such inside the
fields to ensure proper matching? i.e. if I am trying to match F2 with F2
and one records is "1" and the second is "1 " (space added to the end) it
won't match, how do i remove spaces like this?
DennisIf both columns are numeric datatypes then they won't have leading or
trailing spaces. I don't know what you mean by "Both are NUMBER". There
is no built-in datatype called "NUMBER". Maybe you have a user-defined
type in which case you could just change the column to be a proper
NUMERIC or INTEGER for example.
David Portas
SQL Server MVP
--|||db wrote:
> I have two columes F1 and F2
> Both are NUMBER but were orginally imported from a TXT file.
> Is there any way to ENSURE, that there are no spaces and such inside
> the fields to ensure proper matching? i.e. if I am trying to match
> F2 with F2 and one records is "1" and the second is "1 " (space added
> to the end) it won't match, how do i remove spaces like this?
> Dennis
If your NUMBER columns are actually a character datatype, you should
trim your values before you insert into the table. You can also issue a
trim on the columns now to remove any leading white space.
David Gugick
Imceda Software
www.imceda.comsql

Non-browsable custom properties

Does anyone happen to know if it is possible to set a custom data flow component property to be non-browsable? I have a number of custom component properties, and would prefer that they only be updateable through my custom UI as opposed to via the property grid on the SSIS designer,

thanks

Sure, use the

[Browsable(false)]

attribute on the property.

Kirk Haselden
Author "SQL Server Integration Services"

|||This would apply to a task only. It doesn't work for a pipeline component, since properties are not defined as regular properties on a class, so unfortunately cannot be attributed as normal properties are. I cannot see a way to apply such attributes/settings to custom component properties. I would expect it to be on IDTSCustomProperty90 if anywhere. Browsable and read-only would be usefull properties to have there.

Non-additive measure - Actual and Average

I am working on a project at a manufacturing client. The measure I'm having trouble with is LeadTime (the number of days from manufacturing schedule to completion). It is used in a calculation to determine over/under inventory levels.

I need it to use the actual LeadTime for the calc at the SKU level and the average LeadTime at all of the aggregate levels.

Thanks for the help,

Dave

Hello! I am not sure about what you mean with SKU level but if you have two dates in the fact table, ManufacturingScheduleDate and ManufacturingCompletionDate, you can make a named calculation in the data source view(Analysis Services 2005) or a calculation in the ETL-process with SSIS.

Use the TSQL function DATEDIFF() for that and the difference between the MScheduleDate and the MCompletionDate.

I am not sure about your problem with the averages and what you are doing averages of?

You should be able to solve this with the MDX AVG() function.

HTH

Thomas Ivarsson

|||

SKU is the lowest level of data...the individual product being manufactured and sold. The LeadTime does not need to be calculated it is already determined for each product.Here is a table that I hope helps:

LeadTime

Product

Actual

Average

1001

3

4

1002

4

4

1003

5

4

Sub-total Sum

12

Sub-total Avg

4

When I set the measure LeadTime to Sum the calculation that uses LeadTime is correct at the product level, but wrong at the aggregate level.If I set the measure to Average then the calc is wrong at the product and right at the aggregate level.

I need to show the cube data with LeadTime as actual at the product level and average at the aggregate level.

Dave

|||

Hello! Can you also indicate the expected values in the table for Actual and Average Lead time?

From what I have seen with other clients is that you have a manufacturing order id for each production of a product.

It is not a part of your table but can it be a part of the problem?

Regards

Thomas Ivarsson

|||

Here is a little description of the table from my last post.Three product ID’s (1001, 1002, 1003) with actual LeadTime for each product in days (3, 4, 5 – respectively).

The LeadTime measure is used with other measures (On-hand, Daily demand, Safety stock, etc.) in an inventory over/under calculation.But, LeadTime is the only non-additive measure.Because of that the calculation is correct at the Product ID level (3, 4, 5), but wrong at the aggregate level (12).If I set the measure to Average LeadTime at the Product level is wrong (4, 4, 4), but the aggregate is correct (4).I’m looking for a solution that gives me the correct calculation at all levels

This project is only dealing with inventory levels.Orders are not a part of the calculation except as a part of the Daily Demand figure.

No. of User Accounts in SQL SERVER 2000 and SQL SERVER 2005 ..plz

I want to know the maximum number of users that can be created in SQL SERVER 2005 and the previous version...?

Yorker:

I am going to assume that you will hit the limitations of the "uid" field. This field is a SMALLINT datatype; however, it appears that groups begin @. 16384 and that this field must be non-negative. Therefore, it appears that a user database MIGHT be limited to 16384 distinct users. However, on a serverwide database you can have many databases therefore leading to at least N x 16384 distinct users. There does not look to be any "variable size" limitation on the LOGIN side because this is based now on SIDs whereas in much older versions of sql server it was formerly based on "SUID"s -- The cardinality of the number of SIDs appears to be enormous.

Would someone please check me on this?

Dave

|||The maximum number of logins in 2000 was 35,365, in 2005 it was increased to 65,000 (I think).

However, I would HIGHLY recommend using windows authentication and group membership for rights, instead of 35,000+ SQL logins.|||any evidence to prove your numbers... ? Gentleman ?|||See this: http://msdn2.microsoft.com/en-us/library/ms187376.aspx

This says in 2005 "you can create more than 32,767 users". I was slightly off with my number in 2000. I thought I read somewhere else about the 64k limit in 2005, but I might be mistaken.

I still it is a nighmare to manage 32k logins.|||From http://msdn2.microsoft.com/en-us/library/ms187376.aspx, it infers that SQL2K can only have 32767 SQL users, but SQL2K5 can have more. How much more, I don't know off-hand.

But this doesn't say how many logins there could be (although I would guess it would be the same). http://msdn2.microsoft.com/en-us/library/ms174355.aspx doesn't give many clues.

I would strongly recommend using Active Directory, where the limit doesn't actually apply in the same way. Plus, do you really want to have your DBAs managing this type of thing?

Rob|||But... it turns out that http://support.microsoft.com/default.aspx/kb/303879 quite clearly states that SQL2000 can only have 16K users.

But I guess the types used in sys.sysusers will cater for more.|||

I think that is refering to "database logins" and not "server logins".

The other one does say "This is because, in SQL Server 2005, you can create more than 32,767 users, groups, and roles, and 32,767 data types." This kind of implies, users/groups/roles use the same id counter, so combined you can't have more than 32,767.

MS added more than 16k of users for a reason, although I find it hard to believe someone requested that to be added because they ran out of logins. I have a hard enough time managing 100 SQL logins (excluding AD Groups), I can't imagine the headache I would have with more than that. :)

No. Of Subscribers in Transactional Replication...

Hi,

I am Using Transactional Replication with Updatable Subscriptions. Is there any limitations in the number of subscribers (Servers) used for this type of replication?

I have a scenario of configuring this replication in 60 Subscribers(Servers). And the replication should be in the Continuous running mode. Will Transactional Replication with Updatable Subscriptions work in this scenario? Or is it meant to work for less than 10 subscribers?

Please reply asap.

Thanks in advance.

Regards,

Swapna.B.

Hi Swapna,

This question has been answered in the following posting.

http://forums.microsoft.com/MSDN/ShowPost.aspx?PostID=754018&SiteID=1

60 subscribers should be fine for transaction replication. However, you should go over the "Considerations for Using Updatable Subscriptions" in following article.

http://msdn2.microsoft.com/en-us/library/ms151718.aspx

Gary

|||

Hi Gary,

Thanks a lot.

Regards,

Swapna.B.

No. Of Records inserted into a table/database

How do I findout number of records inserted/sec into a table or database.
What is the version of SQL Server?
"Balaji" <Balaji@.discussions.microsoft.com> wrote in message
news:406D4E62-A0D3-4CB9-9585-127A5A61D67B@.microsoft.com...
> How do I findout number of records inserted/sec into a table or database.
|||I want to findout this information both in SQL Server 2000 and SQL Server 2005
"Uri Dimant" wrote:

> What is the version of SQL Server?
> "Balaji" <Balaji@.discussions.microsoft.com> wrote in message
> news:406D4E62-A0D3-4CB9-9585-127A5A61D67B@.microsoft.com...
>
>
|||If you want only INSERTED rows, you can create in both versions trigger on
table for INSERT and calculate them.
In SQL Server 2005 take a look at an OUTPUT clause
"Balaji" <Balaji@.discussions.microsoft.com> wrote in message
news:6FBE1B30-CC16-4BCB-9C6C-4E6D3852BC42@.microsoft.com...[vbcol=seagreen]
>I want to findout this information both in SQL Server 2000 and SQL Server
>2005
> "Uri Dimant" wrote:
|||In almost all cases, you have to clock it yourself.
Linchi
"Balaji" wrote:

> How do I findout number of records inserted/sec into a table or database.

No. Of Records inserted into a table/database

How do I findout number of records inserted/sec into a table or database.What is the version of SQL Server?
"Balaji" <Balaji@.discussions.microsoft.com> wrote in message
news:406D4E62-A0D3-4CB9-9585-127A5A61D67B@.microsoft.com...
> How do I findout number of records inserted/sec into a table or database.|||I want to findout this information both in SQL Server 2000 and SQL Server 20
05
"Uri Dimant" wrote:

> What is the version of SQL Server?
> "Balaji" <Balaji@.discussions.microsoft.com> wrote in message
> news:406D4E62-A0D3-4CB9-9585-127A5A61D67B@.microsoft.com...
>
>|||If you want only INSERTED rows, you can create in both versions trigger on
table for INSERT and calculate them.
In SQL Server 2005 take a look at an OUTPUT clause
"Balaji" <Balaji@.discussions.microsoft.com> wrote in message
news:6FBE1B30-CC16-4BCB-9C6C-4E6D3852BC42@.microsoft.com...[vbcol=seagreen]
>I want to findout this information both in SQL Server 2000 and SQL Server
>2005
> "Uri Dimant" wrote:
>|||In almost all cases, you have to clock it yourself.
Linchi
"Balaji" wrote:

> How do I findout number of records inserted/sec into a table or database.

No. Of Records inserted into a table/database

How do I findout number of records inserted/sec into a table or database.What is the version of SQL Server?
"Balaji" <Balaji@.discussions.microsoft.com> wrote in message
news:406D4E62-A0D3-4CB9-9585-127A5A61D67B@.microsoft.com...
> How do I findout number of records inserted/sec into a table or database.|||I want to findout this information both in SQL Server 2000 and SQL Server 2005
"Uri Dimant" wrote:
> What is the version of SQL Server?
> "Balaji" <Balaji@.discussions.microsoft.com> wrote in message
> news:406D4E62-A0D3-4CB9-9585-127A5A61D67B@.microsoft.com...
> > How do I findout number of records inserted/sec into a table or database.
>
>|||If you want only INSERTED rows, you can create in both versions trigger on
table for INSERT and calculate them.
In SQL Server 2005 take a look at an OUTPUT clause
"Balaji" <Balaji@.discussions.microsoft.com> wrote in message
news:6FBE1B30-CC16-4BCB-9C6C-4E6D3852BC42@.microsoft.com...
>I want to findout this information both in SQL Server 2000 and SQL Server
>2005
> "Uri Dimant" wrote:
>> What is the version of SQL Server?
>> "Balaji" <Balaji@.discussions.microsoft.com> wrote in message
>> news:406D4E62-A0D3-4CB9-9585-127A5A61D67B@.microsoft.com...
>> > How do I findout number of records inserted/sec into a table or
>> > database.
>>|||In almost all cases, you have to clock it yourself.
Linchi
"Balaji" wrote:
> How do I findout number of records inserted/sec into a table or database.

No Values in Dropdowns

I'm having a problems with SQL 2005 reporting services. We have deveoped a number of reports that operate fine if the developer runs them but, when a user accesses them, the dropdown fields are empty. Also, when selecting a date from the date picker, the value reverts back to the default set in the report.

Any ideas?

Thanks

Brian

I should add this is in the report viewer webpage. The use can see all forlders and execute the reports. We place data from an "environments database in the header and those fields are showing up fine. It appears that there is a problem with users modifying the report parameters at run time...... Driving me crazy....

Brian

|||Problem solved. My user had ie version 5 sp4. The minimum is v 5.5 sp2. Preferably v 6 sp1.