Hi all,
I’m evaluating this tool to use with PostgreSQL and I have to say that I like a lot.
Actually is the #1 candidate to buy.
However, I have found some things I’d like to see in the product…
Here is the list:
-
I’d like to define primary key fields with a checkbox directly from attributes list, instead of having to go in primary key tab. As there’s always only one primary key for
a table, the primary key can be generated using standard naming convention rules from the fieds with the primary key attribute checked. -
In workspace I’dlike to see a hint when I pass over identity/relation/view indicating the name of the object (and maybe comment also?). Can be greatly useful when you set “fit to screen” to have a big picture of the model but object are shown very little. Also right clicking on entity shortcuts and choosing Find object is not of great usefulness when object are little because is not too much visible… can be maybe found a way to make it more visible? Finally, in some concurrent products I’ve seen a nice feature: when you click on a reference the related fields in both tables are highlighted… maybe is more spectacular than useful but…
-
check constraint rules: imho its usefulness is limited by the fact that in SQL page you have to put correct field name also, so you cannot reuse the same rule for many different field names with same check: could be interesting having something like a placeholder: for example %fieldname% in (‘A’, ‘B’, ‘C’) that is then replaced with “real” field name. This cannot solve situation when you check more than one field in same instruction, but in my experience 90% of constraints are on single field only.
-
Somehow related with previous point: if I specify a domain with a check constraint rule inside, and then I use this domain to define a field, the check constraint is generated in script without a name. Is not possible in that case to give him a name using standard field check contraint naming convention rule? Or one option for named/unnamed check constraint… (I never use unnamed check constraints!)
-
Domains: could be a good idea to have there null/not null also? (doubful on this)
-
When generating DDL script, having an option to keep all check constraint external from table generation instead of embedded inside create table like now
-
In DDL script generation, make optional quoting of identifiers (actually are all always quoted). This is particularly important with PostgreSQL that is case sensitive for names!
-
Make all these DDL options (+ the existing ones, like adding create or replace for view/procedures) valid not only in DDL but also in sql preview tabs
-
I like a lot to see in the attributes form, the combobox populated with all names of other fields so that I can switch to another field quickly (and also new buttons up/down added in beta are great!) but personally I’d prefer to have fields ordered not alpha but in generation sequence… or some option to choose ordering
-
In procedure definition, I’d like to have a grid to define params instead of a free text memo space like now (more clean IMHO) with following columns: param name, data type (not only standard types, because in postgres you can define a param like table.field%type, or table%rowtype), kind (in, out, inout, variadic), default value
-
I’d like to prepare some “standard” function bodies (particularly useful for triggers!) and then in sql tabsheet of function choose one of these and have it “injected” inside… (ask before: to avoid activation by mistake on a 2,500 rows procedure!!)
-
In function sql tabsheet I’d like to have a button to invoke an external editor, and can choose between one or more “registered” external editors… (like powerdesigner does )
-
(specific for postgres) Maybe is too much have some flag to identify trigger functions from “standard” functions, and have only the former listed in function combobox in table’s trigger definition?
Ok, enough for now… I’ve something more about naming conventions, but I want to carefully study how they works right now before saying something…
regards…
