Jump to content


  • Content Count

  • Joined

  • Last visited

Community Reputation

0 Neutral

About Dali

  • Rank
  1. Hack would be the right way to call it, but unfortunately it remains the only way until filemaker decides to allow their users to make these changes programatically, such as changing data reference through a script. Many developers out there who implement the data separation model would benefit from this. Of course the low level hack would not be done on the "master" file so there's no need to worry about corruption if done properly. Hence, the reason I'm looking for a documentation. Thanks, Dali
  2. Hi there, Are there any docs available on the filemaker (fp7) file structure? I'm basically looking to edit (in binary mode) an fp7 file. I need to modify the string in External Data Source box for about 100 UI files. UI files are all the same the only difference is that each points to a different DATA file on the FM server. This happens on an ongoing basis, every time we issue an update. I'm designing a C++ app that handles this, just having a little trouble figuring out where/how the file reference string is stored inside the binary file. I was able to locate and modify the string to so
  3. Hi, If I add an SQL view to the relationship graph in FM a pop-up comes up with: The table "xyz" does not provide a pre-defined unique key... If however I chose the underlying table instead of the view then filemaker properly identifies the primary key and the pop-up is not shown. The SQL table of course has the PK. Question: is there a way to construct the view so FM will recognize the PK of the underlying table? Perhaps indexed views should be used? If I simply choose the field for the unique-key in FM then I'm afraid filemaker will create its own indexes, performance will
  4. Thanks Jonathan, I was not aware of that. FYI, ODBC Admin 32-bit can be found at c:\windows\syswow64\odbcad32.exe
  5. We're running filemaker 11 on Windows Server 2008 x64 and would need to access data on MS SQL 2008 through ESS. All machines are x64 but it seems that Filemaker ODBC drivers are 32-bit only. Does filemaker support x64 ODBC drivers? Are there any 3rd party drivers available? Thanks!
  6. Yes, filemaker application does that but it also means that if I need to check 100 scripts and make sure the import order hasn't changed in each script (due to changes in the database) I would need to first find all 100 scripts, then manually click on each one, then click on import order, then manually inspect each. Way too much room for error. That is why a nice automation application that would compare things automatically would be great. If only I could get my hands on where that is stored inside fm file. Sure it will take some time to get it working but it will absolutely save huge amount
  7. Hi all, this might be a bit of a stretch but I'm looking for a way to access a filemaker database file and performs queries or even make modifications like perhaps changing a database reference file or even more important query/modify import/export order of import/export scripts. I know this is possible since a tool like FMDiff does exactly that. Question is where to get this sort of information? Of course, I'd be creating a separate application in .NET/Java to access FM files. Thanks, Dali
  8. Thank you for your replies, I use command tools to validate against xsd schema and handle xslt 2.0 transformations so no hiccups there. Could you give little more detail regarding your comment. Not really sure what a "pedestrian" validation is. What's the best way for testing data importing/migration? "However, you could do your own "pedestrian" validation, and generate records only if all tests passed. If there are no records, nothing will be imported - and you will get an error message.
  9. Hi everyone, Is it possible or are there any tools that can validate data imported into filemaker from an xml file For example say I have a Contact table with fields: CONTACT: company fname lname After I export this into xml the xml file would contain nodes company, fname and lname. Sometime later my database schema changes and fields fname and lname are renamed to first_name and last_name Now if I import the original xml file with nodes fname and lname the import will still work without any errors thrown from filemaker and only company will be imported sinc
  10. We use the separation model with two files, Interface file and Data file. Basically, we have one db application over Terminal Services where every client database has these two files. Hot fixes and updates are therefore a real hassle. Any changes need to be made on every file - very tedious and potentially error prone. There is nothing we can do about the DATA file but the only difference in each of the interface files is the External Server Group Name (from Active Directory) and a File reference. You can see now the benefit of having a single interface file. Based on user login the DATA fil
  11. unfortunately, I don't have access to FM 9 on a Mac, we're a windows shop.
  12. This is driving me nuts, DDR report (XML/HTML) crashes when scripts are included in the report. No problems running DDR on other files. Tried file recovery, optimize/compact/file, I even tried deleting all scripts in the file but whenever I run DDR with scripts selected on this particular file FM crashes. I tried it on a different machine and same thing. Are there any logs I can look into? Our main file has some 1500 scripts so migration to a fresh empty file is not an option. I tried running DDR on the same file with FM 8.5 Advanced and it worked fine. Running FM 9 Advance
  13. Your understanding is correct. I'm currently working on a project that uses FM as a front end for MS SQL. In the entire project the only thing FM has access to are views. In FM world a view appears just like another table, in fact you can't even tell the difference. But you need to make sure to deny permissions to actual tables and grant at least select permissions to views. Dali
  14. I'm not familiar with Oracle but in MS SQL you can filter the rows using a view. For example CREATE VIEW Children AS SELECT * FROM People WHERE Age < 10 When filemaker looks at the Children "table" only people younger than 10 will be transmitted between the Oracle and FM. Performance is also better using views since views are already compiled and can also be indexed. Cheers, Dali
  15. Hi, How can I use SQL INSERT/INSTEAD-OF INSERT triggers with FM9. Filemaker 9 accesses SQL Server 2005 database via System DSN. Only views are visible to FM9. I get an ODBC error when trying to create a record in FM when the underlying view has an INSERT or INSTEAD OF INSERT trigger. Looking at ODBC tracing I see that FM is trying to execute: "INSERT INTO tablename(column) OUTPUT INSERTED.ID VALUES('Joe')" which is illegal and therefore results in error: "The target table 'tablename' of the DML statement cannot have any enabled triggers if the statement contains an OUTPUT claus
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.