Hi

I use BaseElements a database analysis/forensics tool, to evaluate databases for errors, and then fix the errors. A new client is using SeedCodeComplete. Examination of their (customised) database, revealed a substantial number of code errors ( x 328), so a decision was made to purchase the latest version, as a fresh and presumptively more economic starting point.

An analysis of a fresh copy of SeedCodeComplete reported a total of 528 errors ( plus 5 warnings), as profiled in the attachment.

In short, there are 246 script step errors in 25 scripts, and 242 calculation errors, plus a small number of errors in TO/Objects/FileRefs/layouts.

Fixing all these is going to occupy quite a lot of time. All such fixes will be overwritten by any downstream update to SeedCodeComplete. Even at 10 minutes per error (optimistic) across 528 errors, that's 88 hours. Granted some errors may have more impact than others, but establishing the difference probably takes even more time than fixing them.

Thoughts?

BaseElements report done ( and repeated) on a fresh copy of SeedCodeComplete 3.11 and licensed.

Mac OSX 10..5.8 , FMPA 11.0v4 , BaseElements 3.0.11

Hi Q,

We love BaseElements (especially since the incorporated fmSerachResults ), but just because it returns errors doesn't mean that performance is going to be affected, and we certainly don't think they need to be fixed.

A few examples.

Almost all of the errors reported are related to the plugin required for QuickBooks integration: if you don't have the plugin installed BE throws script, calc, and script step errors. The others are from a couple optional routines that use the scriptmaster plugin, unmapped fields for Zulu integration options, and scripts where we have unused branches waiting for new script routines to be added by our customers.

We DO have an orphan TO in the data file, the old CalendarRows reference is in there , and probably shouldn't be. Deleting that should take a lot less than 88 hours, and thanks so much for bringing that to our attention!

Hope that is good news.
-Jason

Hi Jason.

can I draw your attention to this:


http://www.seedcode.com/ccp51/cgi-bin/c ... complete11

<<
Features
SeedCode Complete includes...
Simple, Well Documented Code
Easy to modify
Native FileMaker 11, no plugins required.
>>

I suppose that might not be seen as inconsistent with:

<< Invoicing
Simple Invoices & Payments
Invoices linked to contacts, projects, and calendar appointments.
Multiple payments per invoice.
Simple inventory system tracks quantity of items on hand.
Persistent list of unpaid invoices.
Easily email invoices as PDFs.
Optional QuickBooks integration. >>

So having looked at the offending scripts, and documented the called by's, all the QB script folder scripts could be deleted, and the corresponding layout objects also deleted.
THe few non QB scripts can be corrected ad hoc.

Hi Q,

Right, the plug-in functions are optional, and are not required for Complete. If you want use the additional features with the plug-ins, then the code is there and quite valuable.

We do have routines in the plug-in/ninja-cal related scripts to make sure things are short-circuited, so none of these missing references should ever be called or affect performance. Our intention was that they would just be ignored if not used. However, it is an open solution, so removing them is certainly doable.

You may want to keep the layouts and just remove the scripts. The layouts themselves shouldn't trip any reference problems, and that way if you wanted to add the scripts back in at some point, the layout references would be preserved.

hth,
Jason

The quickbooks plugin is windows only, the client is mac only, so the quickbooks inclusion is not optional, and is redundant.

Am I missing an alternative?

Hi Q,

Our opinion is that this code is harmless if not being used, and there is a potential downside to removing it if you wanted to use it at some later point.

If you're sure that you will never need it, then you can carefully remove it.

Cheers!
Jason