CaelC

I got an answer about this in the FMP7 area of the forum. Has anyone else heard this and can confirm? Apparently CDML is gone in favor of XML/XSLT server side translations when FMS Advance comes out? I have an email in to Blueworld to see if they plan on supporting FMP7 with Lasso. Has anyone heard anything on this? Thanks! Cael. Version: v7.x Platform: Mac OS X Panther

I definitely feel bummed about FM7 costs for web hosting. I too run FMU on a remote CoLo server and don't need Server to run the web sites I do. I wanted to upgrade my CoLo server to come up to date with Lasso and FMP but the cost of software alone is going to make the cost of the XServe look like chump change. I'm also clear now that CDML is gone and I for one won't really miss it except than now I'll need to really learn XML. I'm unclear on whether or not BlueWorld will be supporting FM7 with Lasso or not. I have an email in to them. I don't see how I'll be able to afford the upgrade to run my web sites. This is going to be a lost sale for Apple and Blueworld too as I won't be able to afford the whole package. I had $1000 approved in my budget for FMPU and there's no way the CFO is going to tack another $1500 onto the price tag of this migration! Bummer. Cael.

I can't figure out how to duplicate a table inside a DB file. A minor annoyance but is there a way to do it? I don't want to have a to recreate all my basic fields in every table (date_modified, date_created, uid, etc. etc.) It would save a little time to be able to duplicate a template table as many times as needed and then start adding the unique fields to each table. Can this be done in 7? Cael. Version: v7.x Platform: Mac OS X Panther

Thanks for the info! You've probably saved me hours of reading and digging. Regarding Find Unique ... In the little bit I've dabbled in SQL queries, there isa very easy command that finds and returns all unique iterations inside a column. So for example ... If you find unique inside a big database, under a field holding the 2 digit abbreviation for State (e.g. - CA, OH, FL, etc.), the result set would contain 1 each of the 50 abbreviations (assuming you only have US states in there and no weird deviations). This is easy to get at with FileMaker's Sub Summary reports, I know that. But I haven't been able to figure out how to easily get this kind of information for Lasso/CDML based web deployments where I don't want to run any scripts in the back end database. So specifically, it would be for providing a list to the web of unique iterations of data in a certain field, across all records. Then a user could click on one of those to see all records that match that iteration (e.g. all people in CA by performing a search on the DB for CA). Then, they could drill down to detail on a single CA record. Does this make sense? Am I missing something really obvious? Your help is appreciated! Cael.

I'm unclear on a couple of things after reading a bunch of the information about FMP7 on FMI's web site: 1) Is CDML dead? It seems like it isn't mentioned and XML, XSLT, XHTML, etc is mentioned a lot. Is this just in relation to "instant web publishing" and I just missed the section on "custom web publishing"? 2) Will Lasso work with FMP7? 3) Do we need to spend $2500 on FMP Server Advance to do web implementations (instead of $1000 on FMU)? And at that, does it really only support 100 remote users? Or, is this all in relation to hosting directly into a DB via TCP/IP? 4) I haven't read all the security briefs yet, but is record level security in there? 5) Is there a "Find Unique" criteria in 7? In SQL it's really nice to find only the unique iterations of data in a column, it would be great to do that in FMP. I think that's it for now. I can't wait to start using it. Any help would be greatly appreciated! Cael. Version: v7.x Platform: Mac OS X Panther

I was using a G4-400. When you try to pull up pages on the same machine they are hosted on = painfully slow. When accessing from a remote machine, it was 10X to 20X faster. I haven't tried this in the last couple of years since I have a dedicated CoLo machine using Lasso now. Cael.

I think they were at MacWorld SF in 2000. I talked to the resident Guru and told him I develop in FMP and asked him to tell me why I might want to use Helix. His response funnily enough was something like, If you use FileMaker, why would you want to use Helix? I expected him to try to win me over or something but he didn't. When I pressed him he gave me the full demo ... about 20 minutes worth. It seems like it was "Object Oriented" if that is the right term. You drag and drop fields on the screen and then define what they are to do. I remember it didn't have a calculation field per se but that you defined some sort of routines that accomplished the same thing but in a round about fashion. I think when you relate files to each other, you draw a line from one to the other and then define what that "line" does. It was almost like you were building a flow chart and definining what the objects on the flow chart were to do and how they were to react to one another. He claimed it was the first relational database system for Mac and based on his timeline that he provide, I would concur. Overall, it seemed like a foreign and cumbersome way to work if you understand databases ... maybe for people that don't understand basic database concepts could create in it since they wouldn't have preconceived notions of how it should work. I left the booth wondering why I had bothered pressing him into a demo and thinking I should of taken his initial advice "why use Helix when you use FMP"? I found there web site but didn't bother reading any of it: http://www.helixtech.com Cael.

Yes, Mac OS is perfect for doing all that you mention there but there is some sort of circular reference that isn't resolved properly when you are testing on the same machine you are serving from. Testing from secondary machines is no problem. Not sure where it stands with OSX. Cael.

Thanks. Do you just run a script with an Execute SQL step in there? What does the SQL step have to do? Does it replace all data in the remote records, or just changes to certain fields? Thanks, the help is appreciated! Cael.

Regarding issue 1): You will be entering line items through a portal in Invoices and a the point of creation of the related line item record you will be automatically marking it with the same UID as the invoice has ... based on the defined relationship controlling the portal. I don't think you'll be creating an invoice from the line items file ... you'll be creating line items from the invoice file. Does that make sense? Cael.

I agree with LiveOak and would add that whe you have a separate field for group name, if you want to, you can always calculate a field that tests to see which fields are empty and only disply the one that is filled out. In simple english: if group name is blank, then display firstname and surname, otherwise display group name. Something like that on layouts for generating labels, letters, view only etc. will keep things nice and clean. Also, in a small database such as yours one file is best for this sort of thing. When you get into very large systems with many, many possible contacts for a single organization or lots and lots of possible phone numbers, types, etc. it is often best to "normalize" the database in several different files. An example of this is shipping address information, where there can be zero, 1 or many, many shipping addresses for one primary contact, that is best normalized to a separate database. If it's designed properly it can scale nicely and offer far fewer problems than a single file solution. That's probably more than you wanted to know, but thought I should clarify that sometimes it is good to have multiple files with contact information and sometimes it is not. Cael.

I have a remote MS-SQL DB I am trying to "hook up to". I have an SQL person helping me as I am clueless here. We got a DSN assigned and I have scripted a series of imports from ODBC data source to bring data down from the web server via TCP-IP (I'm still having some trouble bringing in data from his stored procedures and/or join fields however). Apparently in Access, it is also easy to have the local DB act as an interface for the remote data source, even allowing UPDATING of data on the remote server. For example...I change a field value locally and it automatically changes on the server to match. 1. Is this possible in FileMaker? 2. Is it rocket science or can a competent FileMaker person do it without knowledge of SQL??? Please help! Thanks, Cael.

I'm about 3 months late with my reply here ... but ... You can get satisfactory test results out of the G4 if you are using it as a true "server" and access the files from a "client" machine. If you try to access them via IE from the same machine you are serving web companion from it will be painfully slow. You'll see speed increase at least a dozen times when set up using a "client" machine to access. This is more true to what will happen in your final web installation. I'm not sure why it's so slow when testing on the same machine but it is unbearably slow.

My understanding, of what I've seen of PHP, is that it is much, much more powerful than CDML. I think this is especially true in it's text pre-processing capabilities. For example, you could perform a find and have PHP pre-process the results and highlight each recurrence of the word you were searching on. I think documentation for PHP is much, much better than CDML also. I would say that the resources out there for PHP are probably 25 to 50 times more than there are for CDML. Also, I think it is super easy to scale your back-end from FMP to MySQL, PostGres SQL or even MS-SQL or Oracle when you use PHP middleware rather than CDML which is proprietary and only works with FMP as far as I know. This is based on the research I have done, please correct me if I'm wrong. Thanks!

Hi, I don't know much about XML, just trying to figure out whether or not I should be learning it. My question is basically this: In order to serve FMP data as XML to allow client side sorting, searching, etc... Do I have to use XSL Transformations server side to output HTML files for compatibility purposes? I'm serving general data to the web ... retail application so compatibility is of paramount importance. I'd like to be at least V.4 browser compatible and up. Also, If I do have to do it that way, why not just serve using Lasso, the load on the server isn't really reduced if I have to parse XML using XSLT on the server side is it? Plus when you factor in DTD verifications (if needed), seems like it will take longer than just using Lasso. I'm sure I'm missing something here but it seems like it's only good for intranet deployment where the V.5 browser can parse the XML without having to download helper apps (and even at that, I'm not totally clear if V.5 is totally compatible in both browsers). Can anyone shed some light? Thanks a lot!