Gary Mugford

Vincent, I could see the issue being paging IF THE PROCESS FIELD in the child table wasn't RENUMBERED!?!?! And, while BDE did something to live with the data, NexusDB, an honest to goodness Database Server wanted NOTHING to do with displaying that relationship, crashing out with an AV as soon as the relationship was hit. 

I've already decided to change 30+ years of relationship and create a GROUP designation and connect it to a database with the sub-parts, while removing it from the child table. The particular part is handled as a difference kind of business model, so trying to push those sub-parts into the old model was and is a lesson in frustration. By creating a new workflow model, I can accommodate the customer that has requested the client work to a certain way that LOOKS like the normal model, but is not. 

Never great when you have to write alternative workflow streams, but modern businesses adapt. And so must I. Thanks for your thoughtful question. GM

Although the BDE display chopping off 200 child records AND renumbering a field is bad, turns out I forgot to include some OTHER critical info. I actually export the Master and Child table data to NexusDB, recreate the index based on the PartNumber and then run an analysis report. It hits the problematic part and crashes with a Memory Access Violation error. Just touching it bombs the program 11 minutes into the run and after the transfer took 37 minutes. I TRIED to test for the PartNumber and if it was the problem part, continued in the interior loop. Not a solution. Neither was testing for the PartNumber BEFORE the problem part and then doing a double NEXT;. That touches the part, so I should have expected the issue. The CURRENT WORKING SOLUTION, is that I transfer the data, including the problem part and it's monstrous child data. I get to the part BEFORE that part in the internal loop, but then do a findkey for the PartNumber AFTER the partnumber that causes the problem. It's ... working ... but it's TERRIBLE programming and I've been told to expect another part with a massive child presence this month. I DO turn 65 in July and retirement appeals more and more every day. I will pursue this as a NexusDB issue. There is no question that something is malformed in the child data in BDE, but the data is not corrupted. And when it gets moved to NexusDB, the tolerance BDE has for the situation is NOT shared in NexusDB.

While I haven't figured out whether this is an issue ONLY for Woll to Woll database components or for any of my BDE equivalents from other vendors/open source yet, I do have one of those chill to the bone issues I feel very uncomfortable about.

I have a main table connected to a child table through the one-to-one connection of a PartNumber. Using wwDBGrid connected to the corresponding tables, I show the child table as having 62 records with a field called Process (AND YES, I know that's a horrible name but this is legacy 1988 database design) which counts from 1 to 62 as it is shown. The field is an integer field and the numbers are IN the database, not calculated. The issue that I have JUST discovered is that there are 262 records in the child database and the 62 that are showing in the application are the LAST 62 of the child table records. That noted, I have two PartNumbers with 80 records in the child database, and all 80 show in each case. So the PROBLEM PartNumber is singular at this point, although I understand my snapshot of the data is far enough out of date that another PartNumber with a HUGE (300+) number of child record entries now exists.

I do a LOT of analysing the data for various needs where I loop through the main table and then loop through the child table. If the inner loop is only going 62 records at runtime I've got a problem. The really odd part is the 'renumbering' of the Process field in the wwDBgrid to show numbers 1-62 instead of 201-262. So, I have TWO issues, the reduction of showing only 62 records AND the renumbering. 

ONE more oddity, in a totally different part with only 15 child records, somebody managed to enter in the data with a SPACE before the name of the PartNumber in the PartNumber field, then later re-entered the child records with the proper PartNumber in the key field. That still leaves those space-PartNumber records atop the child database when you look at it directly via Database Desktop. The data is NOT corrupted (proven multiple times through Rebuilder), it's just that somebody managed to enter in via a method I haven't discovered, all those Space-PartNumbers. And the extra 200 child parts for my main problem. 

This is BDE, Paradox, Delphi 7, Woll to Woll InfoPower components. I'll eventually ferret out how the data got INTO the child table. I've left a back door  somewhere. But the display and renumbering issue is beyond frightening for me.

ANY IDEAS???

On 5/4/2021 at 10:26 AM, Dany Marmur said:

I don't think i have any SQL in my .pas (a lot in dfm's though).

But...

 

SQL.Text :=

'Select ...' + #13!0 +

'From table t' + #13#10 +...

 

MMX and GExperts have functions, i think, for converting the SQL typed into an editor to pascal strings and vice-versa (?).

Use only when lacking a proper DAC.

Dany,

  Towards the end of MMX development before it went open source, I was an active 'suggester' of features. I might have been the first to suggest a tool breaking apart the With statement. I do use MMX for precisely the task you are talking about. Too bad active development in MMX no longer includes Delphi 7, but I'm such a small niche I can hardly expect to be catered to. 

  I've read the preferences here regarding SQL code keeping and I accept that my style is ... not that of the community. But the efforts of all of you are appreciated. I'll probably REMAIN a stick in the mud. Given my templates and leading /// notation atop query construction and execution, I find looking for SQL code in my .PAS files to be fairly easy. It's less fidgety than searching from DFMs. At least to me, that's the case. I'm a LITTLE intrigued by the DIMOMA SQL resource creator. It has the charm of having a default test SQL version that creates code usable by my report generator and then having a V2 that has query variables in it to use in Run Time work. I'm going to at least look at it. But the single DBMS is where I might run afoul of the licence since I want to convert Paradox to NexusDB and will need both in action for some time. Won't know whether what I do is actually within the licence parameters until I give it a test. 

  Thanks all. GM

Fr0st.Brutal,

I admit, most of the world does exactly as you suggest, although there might be a few who use the BeginUpdate and EndUpdate lines. For me, line-by-line SQL is more readable than the format you suggest. When I have errors in my SQL, and I DO have errors, it's better for the line number identification when I try the code directly into the SQL.lines of the BDE SQL query object. Since I am still a learner here in my dotage, directing me to the direct line of the error is of measurable value. Thank you for your suggestion, though. GM

Dany,

  Last century, well back into last century, I devised an audit trail feature for my applications. All changes in the data were noted and documented in such a way that WHAT the changes were, were easily accessible and thus it mitigated what I called the "NotMe" problem (this plays off a regular feature in Bil Keane comic strips). The NotMe problem was the almost herd-like autonomic response to problems. The History Notes simply eliminated the issue since it was clear who, when and what. This was back in the early 80's and was a feature very valued by my customers. It didn't allow for rollback, but it did leave a change trail to be manually reversed. And it eliminated SOOOOO much angst in the office(s) over what happened. And that SINGLE feature probably kept me employed for as long as I have been, despite no academic background (well, ALMOST no school training) in computer programming and a stick in the mud adherence to not change with the times and keep current with language and database development. However, as I pass officially into senior citizen-hood, I've changed my mind and decided to try and be more au courant. 

  And with people such as yourself, who hold out helping hands, I MIGHT just stretch out my usefulness and put off the switch over to writing books JUST a little longer. Thanks again.

Dany,

  The SQL in BDE for Paradox is a sub-set of, IIRC, SQL-97. The important limitation is no sub-clauses. In fact, there's even a missing Parameters setting that I see in most Parameterization examples. It simply throws an error and leaving it out doesn't seem to affect things. (Params.ParamCheck := true; specifically).

  The template I'm using as I add the Params code to my code looks like this:

//Params.ParamCheck := true; // DOES NOT WORK IN D7!! Might in XE7??
Params.ParamByName('eDF').asDataType := ftDateTime;
Params.ParamByName('eDF').asDate  := eDateFrom.asDate; // value can also be used
Params.ParamByName('eDT').asDataType := ftDateTime;
Params.ParamByName('eDT').asDate  := eDateTo.asDate; // value can also be used

  I have a composite component that is built of a panel, a menu button and two date components from the indispensable ESB library. I call it DateRanger and the menu button in the middle provides quick resetting of the two dates to Last, Current and Next; Week, Month, Year and Quarter, as well as some occasionally asked for ranges like YearToDate, Last365Days and AllDates. Since I use this panel in a great many forms, I can afford a template to include, as examples, the use of the data from eDateFrom and eDateTo. There is nothing to do but change the the query variables to :qvZZ as theorized by François, the DataType ftString and then the value from the proposed TEdit. Then delete the next two lines. 

  At SOME point in time, I will end up working with NexusDB seriously. At that point, a fulll panoply of SQL language features will be made available. THEN, some of your code becomes available for me to experiment with. 

  IN REVIEW: Changing my code to use Parameters in the On The Fly creation is a good idea. Nobody has raised the specter of problems with query variable reuse, although I will be shying away from changing DataTypes whilst doing that. And I can continue to use On The Fly style.  The likelihood that anybody would want to disassemble my code as some start to cracking it for nefarious purposes is slim, maybe exceedingly slim, and hard to reach since the server it sits on is not accessible via the Internet. Strictly sneaker net. At least for most of it. 

  Thanks all for the help. GM

29 minutes ago, FPiette said:

@Gary MugfordI was talking about the string constant "ZZ" and not the Checkbox named xZZ. I tried to say that yoyr SQL request built with Delphi code is OK and DO NOT suffer of SQL injection risk. I was trying to say that if instead of the constant "ZZ" you want to use a value from a TEdit, then suddenly your code becomes at SQL injection risk and to avoid it you have to use a parametrized SQL request.

And as I hope I clarified, there was no string constant involved. Your warning of using a string value from a TEdit IS IN FACT a good thing to know, since it legitimizes the work in switching to Parameters, since in OTHER cases, I have done EXACTLY what you are warning me about. So, thanks for pointing out the risks my habits have exposed me to. GM

First of all, thanks all for clearing up some of my misunderstandings. Let me clarify my code better than I did in just plopping it into the code brackets.

ZZ is NOT a variable, but a checkbox object CALLED xZZ. And because it might be true or false, the SQL code changes somewhat considerably for such a short piece of code. That is in BDE and thus doesn't get to use things like cast. I assume NexusDB won't have the issue once I transform the megalines of code, using a closer to current IDE like XE7 ... which is what, eight generations old at this point?? In fact, this procedure dates back to 1998. 

FPiette: "If - as an example - the constant "ZZ" must be changed by something coming from the user interface, then use a parameter. You'll have to make the same condition where you'll give a value to the parameter to avoid giving a parameter if the parameter doesn't exists because it is not part of the query because of the condition evaluated to FALSE. Using the same condition around parameter value assignment will solve your problem."

Because of the ZZ confusion, some of what you say is based on a misapprehension my writing caused. I THINK you were trying to answer my first issue about reusing parameter NAMES. Especially if different dataTypes. As such, I'm NOT sure what you are saying and what I'm asking are the same problem. If you could clarify I would appreciate it.  I'm VERY much trying to use parameters, but it takes time to fix LOTS and LOTS of OLD code with on-the-fly generation that has become my habit. Thanks.

Dany, the contained code won't work in BDE. I have been programming simple small internal server apps that began as peer-to-peer apps using Lantastic which began as standalone apps using just Paradox, not even Delphi. Been going almost strong for close to four decades now. I'll probably be retired when my successor takes a look at my code, tosses it almost immediately and starts programming a 21st century interface for the IoT that I have nothing to do with. Thanks for clearing up my confusion about using parameters as a competing concept vs On the Fly SQL Code creation. I really thought the latter was actually code injection. Old dog, new tricks. 

Stano, always appreciate anybody spending their time on my behalf. Thanks. 

I have come to understand that SQL code injection is a bad, bad, BAD thing for we programmers to do. BUT I've come up with two questions regarding switching to parameterization in my code. I'm using Delphi 7 and BDE as my bog standard and have XE7 so if there are differences in your answer(s) between the two, I'd appreciate the extra paragraph. 

My first question is the regular use of the same 'variable' name in the SQL lines in the query object. IF I include:

DateExpected between :eDF and :eDT

Do I run any risks using EXACTLY the same variables, albeit maintaining dataType, possibly applied to a different field in a different query, so that I can standardize on the lines of codes dealing with the Parameters that I use in the code before issuing the .Active := true or ExecSQL lines? I am TRYING to remain consistent that eDF is a DateTime variable, but what if I do something stupid like use :A for the variable and make it a string variable in one query and an integer variable in a second query somewhere in proximity in the code to the first??

Next, I have a question where I see no way AROUND using code injection. Say I have a filter I want to apply to the query in the query construction stage. Is that not a case where I NEED injection?? Simple example case in point:

procedure TForm1.Btn1Click(Sender: TObject);
begin
  with s do begin
    active := false;
    Sql.Clear;
    sql.BeginUpdate;
    sql.add('SELECT distinct');
    sql.add('  s.PartNum, s.RAF');
    sql.add('FROM');
    if xZZ.checked
      then sql.add('  "SHIPPING.DB" s, "CATALOG.DB" c')
      else sql.add('  "SHIPPING.DB" s');
    sql.add('WHERE');
    sql.add('  (s.DateDidShip > "' + dateToStr(date-366) +'")'); /// CHANGE to a parameter
    if xZZ.checked then begin
      SQL.add(' and (s.PartNum=c.PartNum) ');
      SQL.add(' and (c.Revision <> "ZZ")');
      end;
    sql.EndUpdate;
    active := true; 
    end; // end sql query for s
  pip.Print;
  ppFooterBand1.visible := false;
  exporttoExcel(pip,'c:\data\Active.XLS', false, true, false);
  ShowMessage('Done! Saved to: c:\data\Active.XLS');
  close;
end;

I see no way to turn this into a set of strings I can call from the original query object. I have a condition (a checkbox on the form) that is used twice, which changes the FROM clause and then the WHERE clause. I CAN and will change the DateDidShip comparison to a parameter. But the other two IF sections?? And remember, this is D7 and BDE. I know this probably is handled easily in newer versions and different databases. But I'm dancing with the same old partner as I have forever. This example might be more easily handled with two SQL.text strings using the dating parameter. But what if there are a half-dozen or even more, switches/filters that are applied to the SQL object's SQL text?? There has to be a complexity line SOMEWHERE, where constructing as above does actually make sense.

Any and all guidance appreciated. Updating software is not a currently realistic option. Thanks, GM

5 hours ago, c0d3r said:

Here is another Delphi IDE color scheme  I used to use:

 

I could use that. Orange and Brick Red are playmates for green out in the wild.  

10 hours ago, Stano said:

He probably has corrupted color vision.

Honestly, I've NEVER heard of that particular impairment for THAT colour. We have had a fellow with Red-Green Blindness. Good guy, so I created a toolbar that would be toggled in as a replacement for the normal toolbar and was strictly words. Had to be two layers, but he didn't mind losing the grid rows. He could see what he was doing and that was a good thing. I'm not sure I can work in Yellow. But if he's going to stay, not much of a choice.

As I mentioned, not everybody has my taste in colour. My house is decorated in green and brown. This is a logical continuation. And yes, It DOES work better with a dark-themed IDE outside of the editor.  So I completely understand Lars. As for Mike and Stano and any others, maybe try a darker green. Really, it IS easier on your eyes. Limit the variation between the various sections ... the light green ones are areas I don't get too all that often being a basic stick in the mud procedural programmer. So, there is less striations in real time use for me. Again, depending on your complexity level necessitating more of the various coding sections, well complexity has a price. There MIGHT be a compromise hue of Green between mine and black, which is what you are going to default to with Dark Theme. 

I have NOT had to involve myself with Styles but it is rearing it's potentially pretty little head now that I've experienced a fellow who decided he wanted a YELLOW BACKGROUND as the Win 10 default. Not gold, not amber, not honey, not parchment, YELLOW. And he's asked if he could personalize his interface. Ahhhhhh, no, not right now. But maybe in a month I'll revisit it ... hoping a month is 14 days more than his career at that company. But if he sticks, then I'll be looking strongly at Almediadev. YELLOW!

I use the Delphi IDE Theme Editor to apply the base colour scheme, which was green-apache. Then I went about making the system greener as far as backgrounds, choosing in most cases to make the background a shade of green enough different that your eye could spot the difference while still maintaining the whole GREEN is good for the eyes sort of thing. I also made the default font 14pt because, as a senior citizen, BIGGER is better. As mentioned, Consolas is the choice of fonts. I like the ones and the slashed zeroes. Once you are finished playing with the editor, save your theme to a name of your choosing and then apply your master work. As mentioned, I like the background, but there are days where I go to a second theme that is darker green in the background with the slightly lighter background of this theme being swapped for it. Look at the enabled breakpoint line to get what I'm saying. 

Again, I understand that this is not everybody's cup of tea. Some will have a preference for bold and non-bold. Yes, some will actually want italics in the comments just for familiarity's sake. But RRuz's work providing a quick library of starting points, allowing you to build up a set of themes for specific work ... I can see having a theme for debugging. 

I forgot that I DID have to do some cleaning up IN Delphi. The comment block was that pinkish red ON A BLACK BACKGROUND. I THINK that was the only post-IDE Theme Editor change. The marked block was money green on navy blue. Yeah, those are the differences. Sorry about that. GM

I've worked on black backgrounds it seems like forever, until I remembered reading a study about how dark grass green is psychologically best for human eyes. We like grass. I remember the underside of our ball caps were always dark green to help cut down on glare. So, I gave a dark green background a try, turning this and that different colours to make them pop out. I do so little work with assembler blocks and IF defined blocks that I was able to repurpose Pinkish Red for comments, losing the italics in the process. Non-Bold white for the majority of the code, bold white for keywords, etc. I found that I preferred it to the black background. 

The one thing I don't think I could go for at all, is a lightish background that isn't pale, pale grey or pale, pale parchment yellow. White just feels like a flashlight to the eyeballs at this stage in my life. Others are certainly allowed to differ.

But give that Grass Green background a try and see if your eyestrain isn't soothed just a tad bit more than normal. 

And for fonts, Consolas does it for me. 

Dalija,

  I aspire to being a better coder. I once asked you for some help and you couldn't have responded any better. I got my help and even more. While a lot of what you do is beyond my capability level, some of it does sink in. The sign of a good teacher. So, there's a middling chance I might actually become competent in this area after decades of being a very linear thinker and programmer. So thanks again, in advance.

  Hope this purchase provides some small recompense.

  Be safe. GM

I am still using Paradox and Delphi 7 and up until just before the pandemic hit, Win10 was making me think of retirement. I WAS slowly, GLACIALLY REALLY, transitioning from D7/BDE to XE7/NexusDB, but The Boss kept asking for new features seemingly whenever he detected I was getting close to having time to devote to the updating. And Win10 was making our 50 user operation over three buildings unhappy, crashing about three times a day. THEN, my hardware guy finally gave in and gave my data it's own server. And overnight, the crashes went away. I've had exactly three instances of needing ReBuilder in the last nine months. With that hanging over my head, I've given into The Boss' featuritis itch and been doing THAT work instead of the changeover to NexusDB. I WILL say, however that NexusDB is the best solution for switching from BDE if you are a table object user. At least in my opinion. I've been using Paradox since version 1.11i for DOS, with it's Pair of Ducks icon, and it's hard to give up a product that has been so very, very stable (until Win10) over the years.

But we don't live in the 20th century any more. As much as I would like to get to my retirement without actually doing the full switchover, the fact is that NexusDB is simply a better product. I have a working example of our main app that uses 268 table objects against 119 Paradox databases/NexusDB duplicates. And the Nexus side simply whiz's past the Paradox side of the app. The Aussies know their business.

I hope you have success in your endeavour, GM

I recommend using Amazon Glacier to do your backups. I watch my pennies and storing my backups in Glacier is the coldest of cold backups. If something goes south (again ... yeah, again), then I can get my files out of storage and have my HW guy put them on the new machine. It's not a quick process, but my HW guy doesn't always have the time to jump to my aid immediately. So I don't mind the slowest recovery of my stuff in return for the cheapest price. Since we are talking about disaster, having something immediately at hand (less than six hours) isn't worth the king's ransom that Amazon demands.

My daily routine on my production machine is to have all of my boot drive, data drive and apps drive disk imaged to another drive (external) to the system. I keep a month's worth of nightly backups. The data drive is backed up over my home network to my impending production machine and to an old Windows XP machine I keep around to test my sales register software because half the stores are still using XP. If it works, why fix it??? Thus I have XP, 7 and 10 copes of my data that are simple single copies. Once a week on the weekend, I have the Glacier software prepare a new update to Glacier and upload it while I watch whatever sports are on the TV. And with all that, I still use and ABSOLUTELY NEED, AJC's Active Backup for versioning. The AJC data is on the data drive and thus gets it's share of backups.

I have a small sliver of a percentage of the software the respondents of this thread have in terms of personal production. I would think you folks have much more to lose than a guy like me with a few clients and the projects they require me upkeep and add to. I DO have more than 40 years in the programming game though. And I could not bear to part with it permanently.

Dummzeuch,

Thanks for getting back to me. The reproducible answer would be, by necessity, a small project and the answer would inevitably be, have all database connections start with being off (perhaps through the expert in question) and then start them at run-time according to need. That way, when working with ReportBuilder on a single form at any given moment, you can ignore the fact that RB ACTIVATES the datasource while working on it within the IDE. IF you forget to deactivate the datasource before compiling, then come runtime testing, it will hit for example, table.emptytable, the program will throw and error, citing that the table is busy, Again, if you are starting over and (in good programming form) starting up the databases at run time in the code, this is the obvious situation the expert is for. 

MY Problem is that this is a 30 year old program that wasn't designed all that well at the beginning and only got worse as kludges and new features were added willy-nilly. I have, over hundreds of forms, and the two datamodules, several hundreds of database connections (I think a shade more than 400 at last count). Some are inactive. MOST aren't. They come up ready to go without me writing any lines of code. The expert DOES say it will only affect the active files in the editor, which in my case includes the BIG central form, with around 80 active databases on it .... yes, I can read the comments to be written already and acknowledge, in advance, their correctness ... as well as two data modules with about 200 connections, mostly active in the dLookUp module and mostly in active in the dData module. The expert, despite neither datamodule being in the editor at the time, indicated it would reset all connections in them, as well as the big central form and the one form I was working on. I needed it to be on the form I was working on ONLY. 

I think my narrow niche need for work on this is beyond making a request official. It's not worth other programmer's valuable time. Sigh. I was hoping the change would be a minor effort. Since it's not your expert, I understand what I am asking for. So, I will have to change things around just a little bit. I've lowered the top edge of the IDE and overtop of the button that I click to compile, I have placed a graphic telling me to DEACTIVATE the CBOX.DB. Crude, but it seems to be working.

Again, thanks for all you do for the community. GM

First of all, kudos to all the folks who participate in making GExperts. It is all appreciated. And in this particular case, I guess Robert Wachtel gets a big share. Thanks.

Now, I THINK the reason for the expert is that people get busy working on a form and at some point, the compiled application won't do something particular, like say emptying a table, because it's still active in the IDE. I'm running head on into that nightmare right now.

The Component Property Replace expert would save my mental fugues if it was JUST a little more particular about what it chooses to switch and what it doesn't. I have a BIG project, and it's an OLD project done in D7. There are a couple of data modules, one BIG central calling form and a nightmarish number of sub-forms. All of this grew organically and is out of control. A replacement is underway, but it's nowhere near usable yet. So, I'm stuck updating the old app.

I'm working on a form that centres around a central database ON THE SUB-FORM that a number of reports run off of. I change things in any of the form's report objects and it auto-starts the database in question which I later need to deactivate before compiling. Something I do about HALF the time. Getting old and stupid. I would very much like this expert to replace my brain and do what I forget ... set the database.active to false JUST ON THIS FORM before compiling. Which the expert will be happy to do. But it also will reset the data modules and the big form, the latter one I expected from the fact it is open in the IDE, the former two, the data modules, somewhat surprisingly.

There is a setting that allows you to limit the resetting to just the forms open in the editor. I sort of have to have the main form up all the time. I COULD try going without it just for this focused (or lack of focused) programming frenzy with the sub-form in question, but it would cause other issues. So, I make the following suggestion: Please have the expert ONLY apply to the SINGLE ACTIVE File TAB in the IDE's editor. Alternatively, the open files in the editor (which should exclude the data modules in MY working example because they don't have an actual tab in the editor). That would let me make the decision to temporarily shut down the main form to concentrate on the sub-form. The best of all worlds would be a radio box that allows:

All files in project

All open files in editor

Just open file TAB in the editor

Maybe reverse the three based on left-side/Top-side DEFAULT bias for picking since it's the small end of the working group numbers.

Lastly, and this is probably the fly in my ointment ... I use Castalia as my editor. That MIGHT be why the data modules get included in my run of the expert. Or would, if I turned it loose to actually protect me against the ONE thing, while causing me to go back and but in the code to turn on databases on Form Creation for all the affected switch offs. A daunting task and why I am not actually USING the expert currently.

Second lastly, a different approach, which I think is impossible ... letting me cite Form and component as a direct target. I'm actually doing this for ONE annoyance currently. Say I could put in TWorkingForm.TTable and have THAT set to false on compile FIXES ALL MY PROBLEMS and it doesn't create the possibility of mistakes spreading if I pick the wrong choice in my hypothetical radio box above.

Regardless of whether this spurs anybody into their own programming frenzy, just want to repeat how much appreciation GExpers has from me. Thanks.

Actually, I'm unfortunately going to have to agree with Mr. Heffernan. The end goal is a working umbrella process, it's true. BUT the end goal is also to have one cohesive program to run against Delphi Parser with a capped licence of 1M lines of code. So, I'm trying to pare down everything, no duplication, no sub-folders. Just code to update via Parser so as to change Database backbones. The idea of calling exe's would BE a solution if there was no next step. But there is. At that, they have the umbrella already ... the Windows desktop, in the form of LOTS of icons. Well, not lots. Each user gets icons for sub-applications that are pertinent to that user. 

My problem right now is that I have apparently wasted my time collecting all the code into one great-big folder and 74 sub-folders with subfolders and subfolders etc. That action was premature. I was hoping to speed things along, but the opposite has happened. The issue is that in ADDING units that weren't in the settings as search path entries LIKE my own home-grown utility libraries, they hard-coded PART of the paths in the DPR's uses unit. Now, I'm DISCARDING the DPR, but there's something in the DNA of the program that when I bring it up to rename the form, planning to save this copy of the form PAS file into a different name, something goes astray every time. It MIGHT have something to do with AJC Active Backup maybe getting confused given the file name no longer tracks back to the last version in it's database. Don't think so, but I'm throwing mud against the wall looking for a hidden message. Not everything, Maybe one out of four or five, but then the time with that one takes FOREVER to get done. Also, any mini-app with sub-forms takes two or three times the time of a single form, but that's to be expected. 

As I asked in the OP, I was looking for pitfalls. And I've found them. For future readers similarly run afoul of their own design, I hope reading this in advance of trying what I'm doing will help, although the better advice is ... Don't Screw It Up In The First Place.

4 hours ago, Fr0sT.Brutal said:

Ok, now it's more clear. I see no reasons why it shouldn't work if you'll have one DPR and thousands of forms. Just rename them in namespace manner and use in umbrella DPR, no auto-create is needed. Just do Application.CreateForm in your umbrella main form's menu onClick handlers. However, you'll have to ensure your under-umbrella apps contain nothing in their DPR but default App.Init/CreateForm/Run.

Frost,

  The renaming is not quite as easy as I thought it might be. I wanted to go in and edit the DFM and the PAS files and replace FrmMain with FrmMainActivityReportToolset in the copy I had made of FMain.Pas renamed to FMainActivityReportToolset.Pas/DFM. Then, I used the search and replace to find frmMain with the longer monicker. But I'm getting a lot of errors when calling up the form from the umbrella. The DPR file in the sub-folder might be involved somehow. So I'm going to work on that theory for the next few hours and then cutaway to watch the Clippers Raptors game. Back at it come midnight.

  Thanks for suggestion. GM

4 hours ago, Rollo62 said:

Wow, Ok you have a bigger task to do then.

I am lucky not to need to keep on old Delphi, I try always move further to the latest Rx version.

Anyway, your project sounds like a task for big GIT scenario with clever CI.

Good luck.

Rollo,

  GIT and whatever CI is are beyond my skillset. I'm just a slogger working through it, one app at a time. Problem is, it's one app every six hours at the rate I'm going. And year end is 80 days away. Seventy if I want to try things out first. Which doesn't leave much time for basketball on the TV and things like eating, sleeping and going to see the doctors. Sigh.

  Deadline pressure is good, if it doesn't kill you first. [G]

  GM

AE,

  Actually, there's a ground-up rewrite using a wholly customizable user interface that I was in the middle of when the breaking of BDE by the Microsoft Windows Update Fall 2017 made continuing to use the app, which dates back to DOS days as a Paradox app and went Windows in the late nineties with Delphi 1, untenable. So, it's past 30 years old and nearer 35 than not. What I think of Microsoft is not allowed to be printed here, I think.

  In order to survive until I can get the new software up and running, the plan became to get the app and ALL the satellite apps together and use Delphi Parser (a SHOUTout here for the amazing customer support) to do the backbone switch for me. There's also a gazillion instances of components from companies no longer in business. They have to be replaced along the way too. And I was shocked and dismayed to find I hadn't produced XE7 for my OWN components (merely extensions of existing components with more properties set). When I wrote ComponentKing to find the mish-mashes between D7 and XE7, it found the expected ones missing in XE7 like the long-cherished, but forgotten ABC components from Australia. At least I know what has to be replaced and have a Rosetta Stone of what goes where (technically, what gets replaced by what and WHAT I have to do post swap). And ComponentKing did give me a list of units to un-use via Parser or Replace. So, I've got a decent start. I never used things that trigger spasms from unicode. Replace will do the string/ansistring swap and the assorted cousins.

  BUT the whole shebang has to be ready for Parser to do the backbone swap and for me to have the app working lickety split. I've never had a single day of parallel testing and the longest beta test on record was one day. I'm expected to be functionally perfect (acknowledging I might not catch EVERYTHING ... until day two).

  I would follow your advice in other circumstances. But I agree with it one hundred percent.

  Thanks, GM

Lars, Rollo and Frost,

  Appreciate your comments in toto. Unfortunately, I wasn't clear. ALL of these applications, save one in Delphi 2010 and a half-dozen in XE7 are in Delphi 7. In creating the umbrella app, I want ALL the source code of the umbrella app and the 74 independent apps in ONE folder. There's a reason for this, I need to bring them all together for Delphi Parser to help change the database backbone. The end result needs to be a case of all the software working with the new backbone,, the apps being menu items in the umbrella main app. It's ballpark three quarters of a million lines of code. And it has to be ONE APPLICATION, ONE FOLDER. Oh, and working in Delphi 7 with an eye towards the lot being transitioned to XE7. 

  I LOVE the idea of frames and looked at them. But I didn't use them, preferring composite components. Bad decision. But I've been in the 'Get it to work' business for a long time, once releasing more than 250 updates in a single year (and that was one a day) back in the early days. My employer suffered from the dual scourges of not telling me enough and then having an active imagination when seeing what was possible. But those days of turning in updates two days out of three are long gone. The kludged together main program and all the satellite apps now need merging. Faster the better. I'm stuck on something I managed to change into non-usability back in 2017 but left it like that since the 2015 version of the software worked. Now, I NEED to fix what I broke. But once I get past that, I THINK I have the routine of work down pat. But that leaves me thinking I'm missing something. It's worrisome. 

  While not being able to immediately take advantage of your advice, it's in the bank for moving forward.  

  Thanks!