bb.donnay-software.com

I have a customer loan account program where the collectors enter collection comments into a xbpmle, using a fixed width font, and the results are then split into individual lines and loaded into the comments dbf file. A DCBrowse is used to via the comments. This is based on the original wordwrapper clipper program approach x-Memo approach. I continued using this approach instead of using fpt memo files initially due to the problems caused by corruption of memo file. We have files with over 99 million lines of comments. We use ADS now and are looking at using the fpt memo files.

So I have 2 questions.

1. For ADS users using extremely large memo files, say 3-6GB in size, have you ever had corruption and lost data and how do you handle it.
2. The comments are not one comment per account, they are 1 comment today, one or more comments the next day, comments for logging of changes, etc. So you would still have the same comments file with account#, date, time, comment type and instead of the single/multiple line of text, you would have one record with a memo file link for each comment. The question I have is how do you accumulate all the individual comments into a single scrollable window for user reviewing.

Current screen: The * indicates a new comments. Some are single lines and some are multiple lines. What is the best way to show this. They may have hundreds or thousands of lines of comments over many different threads that need to be show in a single scrolling type dialog, preferably being updated as new items are added. Additions would be generally at the end prior to the reminder notes that always have an internal date of 12/31/9999. I don't think it would be good to build an array everytime you start looking at this screen as it would be creating a delay for the user. Currently, since it is a table, there is no delay.

Back in the Clipper days there was a product called Word Wrapper. Basically, it used one record of a database for one line of text and then tied it all together with a common key. They had their own display and edit program that would delete all the records and rewrite them. This was a rather simple concept and worked well. In the GUI world, however, text lines can vary in width due to proportional spacing. I'm not sure how I would want to approach this, but if FPT's are not acceptable, then I see no other solution.

I originally used word wrapper. Xmemo, from Clayton Jones, was its Xbase++ replacement. I used that in a somewhat modified version. We are currently not using proportionately space fonts because of this. However, we are looking to using FPT's, but need to know how to incorporate them into a similar dialog.

With the single line items, you just browse the table. But with the FPT memo approach, you have a collection of of xbpmle's. What do you use to effectively put them into a similar scrollable display of those items. The attached screen was from our windows GUI dialog. It is just an DCBROWSE. How would we do the same if the source was not a database of 61 character text fields and instead was a table of 0-many memos. I don't know what would be the display mechanism in Xbase++/Express++ as I have never seen an example of this.

It is like the webboard, but putting all the replies for a thread into a single scrollable window.

Cliff,

we use a single dbf file to store these as out system audit log.


{ 'DATE ','D', 8, 0 },;
{ 'TIME ','C', 8, 0 },;
{ 'ID ','C', 7, 0 },;
{ 'TYPE ','C', 1, 0 },;
{ 'PRINTED ','L', 1, 0 },;
{ 'USER ','C', 8, 0 },;
{ 'STATION ','C', 2, 0 },;
{ 'LOGTYPE ','C', 30, 0 },;
{ 'PRINTLINE ','C', 200, 0 },;
{ 'LOT ','C', 8, 0 },;
{ 'BUYER ','C', 8, 0 },;
{ 'SELLER ','C', 8, 0 },;
{ 'IDNUM ','C', 24, 0 },;

we split the transaction into multple lines of 200 chars each. only the 1st record geta the record number of that record in the id, then all the following records for the transaction are gathered into additional lines that go into the same file., these files are parsed out be something that predates the abacus..


They can be quickly assembled and entered into just about any format you can choose...

I've got some routines i can send youi if you want to go down this roasd. it automatically gathers and scatters comments from mle fields into segments that can be stored in a normal dbf. ours is encrypted, but is easy
to remove the encryption.. et me know if you want to see then.. i can zip 'em up and send /e;l to y9ou off list..

Here's a screen shot... all the key info is on the left and all the comments are on the right...
the origirnal scatter & gather routines cane frm a precusor of clayton's workss p.m. me off list for more code related to this..

I would appreciate the opportunity to look at the code. So to change from one log items to the next, do they need to navigate the items on the left, or do they do it on the right. Say if at the first item on the log entry and they press up on the left side, does it move to the previous log entry. Or is that only provided for on the left. Also, on the right, can they see portions of two logs entries, say the bottom x lines on 1 log entry and then the first x lines of the next log entry.

navigation is from the left...
one item at a time is on the right...

Here is a sample also from the ACT program that would be nice to be able to emulate. They also have to so they can configure it to be integrated entirely on the left or with an additional preview window on the right, like yours, with a shortened version on the right.

It is something like the multi row sample in the browse samples, but the number of lines would not be fixed number of lines for each line. I would need to be variable based on the size of the memo comment.

Cliff Wiernik wrote:It is something like the multi row sample in the browse samples, but the number of lines would not be fixed number of lines for each line. I would need to be variable based on the size of the memo comment.

you'd need to ck w/ roger for the multi line comments..
i don't have a clue...

Are you saying that you would like a browse with a multiline column to show comments?