Mark Williams

My initial thoughts were: On start up fetch the data from the key id field into Query1 for all required rows. This should be very fast even for a large database. When data is needed for the tree it is fetched into Query2. I think this query would have to also contain the key id field so there is a duplication of data, but that would be relatively minor. Set no of virtual nodes in the tree to number of rows in Query1 When the tree needs to populate a node it looks to Query1 for the document id. With that id you would then use FindKey on Query2 to see if it already contains the data. If not, request the data from the database and add it to Query 2. Then pass the relevant RecNos from Query2 back to the tree. Possibly store the recNo from Query2 as node data so that cut out the small overhead of having to go back to Query1 each time a node requests data. I think this is pretty much in align with your thoughts. I am not too concerned with fast scrolling, I would just show a holding string ("Data populating...") or some such. I am more concerned with handling what happens when scrolling stops for a time. I previously implemented a thread which fired once the user stopped scrolling for a short period of time. It analysed scroll direction and populated above or below accordingly grabbing a few tree pages more than was necessary. This worked ok, but not great. More requests were being fired to the server than was really necessary and occasionally when the scrolling stopped nothing happened. These are just fine tuning and debugging issues. Before I plunge in and try to implement something like the above (which will be quite a bit of work), I was just looking for a steer from VirtualTree users who have encountered this issue as to whether this is a sensible approach or whether there is a better one, In particular, I need to know what are the best events to handle in order to implement this. My thoughts are OnScroll, but I have a feeling that there may be better suited events from the many that VirtualTree exposes. Many thanks for your input. As always highly appreciated.

When using Array DML to insert new records is there a way of obtaining the data for any autoinc key fields for the inserted rows so that the key field of each newly inserted record can be updated in the relevant TFDQquery If using ApplyUpdates you merely set the AutoInc field of UpdateOptions and away you go. Can't see any obviously simple way of doing this with Array DML.

It certainly handles data from local memory, but I am pretty sure it was also designed for displaying data from databases.

Thanks for the tip. I've had a quick look on their site. I'm not sure how useful it would be to download their demo product without the source code. There are so many users of VirtualTree and, whilst Peter's observations are never to be taken lightly, I can't believe that my issue, which perhaps might be better summarised as "How best to use TVirtualTreeView virtually with a database", has not been addressed previously. But I cannot find any demos or help online. I suspect that I may be looking to the wrong events and properties for handling this. If so, I can't work out which are the right ones. OnInitNode may be, but it fires for each node and I need to query to get data for all visible nodes. I don't think my original approach was a million miles off, but it didn't perform as well as I hoped. It may be just a case of taking a long hard look at my code and seeing how it can be improved. However, before I do that I would really like some feedback from other virtualTreeview users who have already addressed this problem.

I suppose so. I have used both approaches. If it were just three different forms I would be inclined to use panels. If more forms, a page control may make it visually simpler to manage at design time, with the common search box in a separate panel below the page control.

Afraid to say "yes". It's used in a document database where the documents can be a random assortment and often poorly titled. Trying to sift through the documents by categories often doesn't work well and the only way to find what you want may be by way of a laborious scroll through the table (even item by item), tedious as it may sound. Your reply also seems to query the point of the virtual paradigm. Why have the ability to host millions of nodes if there is never any point in hosting them? I could load 250/500/1000 items a time and have a button "Get Next 1000", but I don't wish to do it that way. I would prefer (and believe it is also better from a usability stand point) to have a tree that purports to provide access to all the relevant documents simply by scrolling and allows the user to get to the point they want to be by scrolling through the tree quickly. Having to do so in chunks is in my view clunky and obstructive. VirtualTreeView, as I have always understood it is the remedy to such clunkiness: if only I could figure out how!

The forms in the screenshot seem to be pretty simple with a common search box and buttons. Why don't you use a single form (AutoSize=true) with a series of aligned panels. The controls common to all form types on a client aligned panel and the optional controls on top aligned panels. You can then make the optional panels visible/invisible as required in onShow event. And after posting this suggestion I saw that you have lots of these forms. A host of panels would become rather cumbersome!

The column definition is ALTER TABLE public.bundles Add COLUMN pages bytea; The query statement INSERT INTO bundles (pages) VALUES(:P)

I am using a TFDQuery component to load and edit records from a table using CachedUpdates. Records can be deleted from the table as well as edited and appended. The BatchMove component in dmDelete mode deletes all records in the query and not just those that have been flagged for deletion. Is there any way of using the BatchMove component so that it only deletes records where the updateStatus is usDeleted?

I was configuring the Batchmove as follows: FBatchMove := TFDBatchMove.Create(nil); FReader := TFDBatchMoveDataSetReader.Create(FBatchMove); FWriter := TFDBatchMoveSQLWriter.Create(FBatchMove); try FReader.DataSet:=FDQueryPG; FWriter.Connection:=FDConnectionPG; FWriter.TableName:='dummy'; FBatchMove.Mode := dmDelete; FBatchMove.Execute; I assumed (obviously incorrectly) that the batchmove component would only delete the records with an update status of usDeleted. I assume there is a way of configuring it for specific ideas, however, I have had so muh trouble with BatchMove that I have completely moved away from it now and using Array DML instead which is so much less hassle.

I am trying to work with integer arrays in a Postgres integer[] field. I have read somewhere that with Postgres you need to use NestedDataSets to read the data rather than TArrayField and that seems to be correct. I can successfully write back to the dataset using TArrayField in an Insert Query. I can read from the integer[] field using TDataSet and nested fields. However, I cannot work out how to write to the TDataSetField. I have tried the following: Edit; TT:=TDataSetField(Fields[2]); for i:=0 to high(arr) do begin TT.NestedDataSet.Append; TT.NestedDataSet.Fields.Fields[0].Value:=arr[i]; end; The error I get on trying to post data edited as above is I tried to use "AsInteger" instead of value, but got the same error.

It's too slow for large updates.

Is there anyway of indicating that a post has been answered in this forum?

I am having difficulties with a parameterized query designed to be used repeatedly. It works first time but never a second. The query is fairly complicated so I prepare it on app initialization for repeated use, but it works just once and returns an empty dataset on every other occasion (even with the same parameter). If I reset the query text (FDQuery.SQL.Text := FDQuery.SQL.Text) it works as expected. That rather defeats the purpose of preparing the query in advance. Is it the case that parameters cannot be used in repeated queries?

So simple! Thanks.

I have been using ClearDetails to remove all records from TFDQuery components. However, I have experienced a problem today where a call to ClearDetails for a reusable query that is designed to return a single record is not removing that record. The description for this function states: Pretty much what I was expecting it to do, but (in this one case at least) it is not. I have switched to EmptyDataSet which does the job, but should I be updating all places where I use ClearDetails to update to EmptyDataSet?

That's good to know. Thanks. Have you any experience of Batchmove and reverse inrements?