Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - Barry_Pearce

Pages: [1] 2 3 ... 5
1
[Solved]
Indeed the problem was the ODBC driver.

Updating to EA 13 didnt help - not surprising.
The MariaDB ODBC driver is not similar (enough) to the MYSQL ODBC driver.
Replaced with the MySQL ODBC driver (still connecting to the MariaDB database) and all has recovered without requiring any additional patching. Now safely transferred to an EAP file.
All good.

2
Hi,

Im trying to recover an old model of ours from a MySQL database.
A mysqldump was take prior to a hardware change and Im trying to get it up on the new system as it has some important information on it still.

We now have mariadb. A new database schema was created, with a different name (due to conflict) - although I believe this shouldnt cause an issue.
We have the latest Maridb ODBC connector and configuration a User data source works and tests OK. DBeaver can also use the same ODBC datasource and shows full (appropriate for permissions) access across the database server.
The client is on Windows 10. With EA 12.0

Im seeing EA refuse to connect with invalid data source check configuration.

Does this error also occur if the database is an old format?
I assume I need to upgrade the database to the latest EA standards (2 vers back) will this likely solve the invalid data source issue or is there another problem?

Cheers
Barry

3
Hi m8,

yeah. I see. I did try this - but the limitation is that you are forced to add a datatype to add the product - I hadnt realised that....

The reason I didnt do this is you dont declare 'types' in python.
Yeah there are strings, ints etc but you just bind these values to an 'identifier' and the 'identifer' is now related to that type.

i.e.

class X:
   x = 23
   x = "hello"

When the class finishes being defined x == the string hello. If in a function I now say X.x = 34.56 its now a floating point number!

Henceforth implementation of types for python is pretty ridiculous - you never say int x;

ive now defined a type called type and this has indeed caused the templates I needed to occur - I am now transfering over the data :)

And I pretty much am rolling my own CGT as there is little of use in the majority of the others anyway..(for python that is)


Cheers!
Barry

4
OK. I have got some preliminary Python code generation templates together .

Due to EA limitations these replace the Delphi templates...if you would like to test/help collaborate please drop me a private message or reply here!!!

5
Aha. We have exactly the same issue!!!

6
Suggestions and Requests / Re: template classes
« on: June 24, 2004, 11:28:33 am »
Yes its still a major problem in 729. We will be chasing EA in the coming days....

This is a major issue when deriving classes from STL (assuming you can get STL in to start with).

7
Suggestions and Requests / Re: c code generation option
« on: July 29, 2004, 06:57:48 am »
? curious. As of 2000 their code generators didnt...When did this change?

What on earth does it produce?!

8
Suggestions and Requests / Re: New Template ParameterNotes
« on: July 13, 2004, 01:08:29 pm »
Wellthats an interesting one as the throw specification is actually done via a tag rather than any direct support....which leaves documentation in a real limbo...

9
Suggestions and Requests / Re: New Template ParameterNotes
« on: July 08, 2004, 03:34:31 pm »
Incidentally ive recently logged an issue regarding lack of decent return documentation for an operation...which is the last bit of missing info about the method.

10
Suggestions and Requests / Re: New Template ParameterNotes
« on: July 07, 2004, 03:49:44 pm »
Yeah that stuff already appears from the current CGT as you say...but the problem is the nice semantic information about in/out/in-out (and possibly fixed) have been lost and are not making it into the code.

11
Suggestions and Requests / Re: New Template ParameterNotes
« on: July 07, 2004, 08:44:09 am »
Bit fickle eh? Anyway...developers shoud be looking in the model for documentation not the source code ;)

Mind you it would be nice.

12
Suggestions and Requests / Re: C Reverse Engineer
« on: July 29, 2004, 06:57:01 am »
And what would constitute an object?!
How do you model functions?

IMO this is a case of thinking a good OO tool can do things for non-OO projects....and I cant see how the tool could usefully produce anything...

Id rather the guys spent their time fixing c++ problems... :)

13
Suggestions and Requests / Re: C# Indentation when generating code
« on: July 06, 2004, 05:46:58 am »
Reboot not required,

However, if you are working on a multi-user system the changes are not automatically propogated - you will be OK but other users wont - reload currrent model wont help your colleagues as it doesnt reload the CGT (fault already logged with sparx) - you have to exit the application and reload it.

Hope this helps.

14
Suggestions and Requests / Re: C# Indentation when generating code
« on: July 05, 2004, 02:31:01 pm »
Indeed, our standards also dictate ANSI style bracing and not K&R style.

We have changed the C++ codegen templates heavily to get the code generating to be correct and well formatted C++.

Basically you only need to add a couple of bits to the CGT to do what you want....

ClassBody - need to be altered for default constructor/destructor generation stuff.

OperationBody - Add \n before the { and that should be all you need to do :)

This may be clobbred by a trim or something...we have modified a lot - so if this doesnt work give me a shout.

15
Suggestions and Requests / Re: Generalised (non-UML) drawing
« on: June 30, 2004, 06:12:04 am »
?!?!
I hope this never happens - sparx focus should be on engineering solutions not drawing tools - if you dont like microsofts prices then try Dia for diagrams - its open source and free, and for graphics editing try GIMP. :)
Hope this helps.

Pages: [1] 2 3 ... 5