MichaelT

Think of errors retrieved from an external source mapped to Delphi exceptions in case of Oracle or SAP via Connect For SAP (formerly SAPx) for example. The abuse of exceptions to hint at errors in the business logic is no good idea at all. I can remember applications in the 1990s (Delphi flower power area) that were designed to hint at business logic errors via NIL pointer exception or better said EAccessViolations which show up calling virtual methods in Smalltalk 'Message Not Understood' style. I can live with retrieving error numbers raised and additional information gathered on the source system via Delphi exceptions and if people like they can include call stacks from the application. In a 24*7 server side environment an error raised at night and the applications has to continue working unattended (service, servers-side application, ...) errors mentioned are exceptional. Agreed. Even such exceptions have to be fixed as soon as possible and logged on both sides.

Shift + F2 makes the IDE focus on the Object Inspector and puts a property selected in a certain kind of edit mode. No idea if that feature is a useful one in practice or required, welcome, widely used or whatsoever, since I never came across this keyboard shortcut before.

Did you have IDE Fixpack 2007 or any other helpful extension/add-on installed before that made problems go away?

The whole MS Microsoft ECO-System gained it's competitive advantage by off-shoring to Russia in the 1990s at the time of Y2K (in comparison to IBM). It has always been about (big) Enterprise(s) and nothing else. Afterwards there have been several economic development programs in the U.S. aiming at moving companies to the West in a legal sense at least (subsidies for founding companies). It's more or less a copy of the British off-shoring (deindustrialization of England in the 1970s) idea. As far as business is concerned no change with Russia so far. If you think of the prime minister of Estonia, she's touting extreme Anti-Russian propaganda on one hand and her husband still does extensive business with Russian companies on the other still. A second reason was a gain in reputation by setting up the company in the U.S. and within the U.S. in the state with the most restrictive terms and laws concerning business (SQL Detective & Co for example). The one who setup the business portion is a former director of Oracle Marketing manger from Austria and 'friend', a big word, of mine. The economy of Russia in contrary to the E.U. is far more liberal and younger, so it's no surprise that one-man shows and smaller companies have a fair chance to succeed. When we grew up in Austria during the so called steel-crisis in the mid 1980s the situation was not that bad as in Russia in the 1990s but from the perspective of the material flow the Germans put their goods into their shelves setup up here (Aldi/Hofer). At the peak of this evolution we had three kind of shops for food. Discounters, luxury stores and market stands at the backyard of the parts in the city populated by the Turkish community offering the best fruits and vegetables one could grab. Pretty much the same happend to the vast majority third-party vendors after 10 years as far as Delphi was concerned. Things don't work out this way, such transition things happen in a well organized manner in order not to run into shortages. Local products disappear for about 10 to 15 years from the Business to Customer part of an classic industrial consumption society an move into B2B, especially when evolving from non-industrial lines to industrial ones (carpenter vs. . Another trick in Austria was to invest free business and the matching licenses and everyone with money on the savings account could get one and access the product from B2B. Follow the model and introduce exceptions (to the rules) whenever applicable. In the meanwhile some carpenters turned their small companies into furniture factories (more automation, machines instead of heavy tools) and tourism as well as the business with Russia was the supporting driver of this evolution. What was needed, was a higher price level. A third tendency was to cut off the growing dominance of shops from Russia (in a sense of protection) on the Huge Java stacks for example in middle of the first decade after Millennium shift.

That language 'feature' is a useful relic. A file in the context of the 'C' or 'Pascal' world is a compilation unit and a module is defined at a different level of abstraction. The keyword unit is about compilation units. Modula introduced modules. That's more or less the 'Host world' & Friends, no compilers or what we call a compiler today. These systems loaded a module for example into an execution context/process and the code run and/or were interpreted under the control of the executing system. A PC application or an program on home computers is controlled by a minimalistic runtime and utilizes the OS way more actively. Modules can have parameters. Have a look at PL/SQL. If you think of MTS or COM and apartments, said sloppy, it's about putting those old machines to the Windows PC. There is good reason for having interfaces and functions/procedures there and not classes in general. This COM and COM+ world is a mixture of IBM-host and DEC computers I think (appartments and friends). In remote scenarios a parameterized module does it's job pretty well (Oberon - Modula successor). On way to mimic modules are those kind of procedures. Object orientation also means modularization and a class is usually responsible fot this. In the context of above an object is a loaded module ready to execute and the application is already a 'host' computer and with a GUI involved in a certain kind of 'single user' mode (kinda MS-DOS). With the help of threads or co-routines this single user application turns into a multi-user application but without a GUI on the server side. Another example would be package vs. package library. The library is simply a concatenation of complied compilation units and the package library goes beyond. In Delphi you still have compilation units in fashion of a module, for example initialization and finalization, module variables and so on, classes, and procedures acting as a module pretty similar to Pascal on 'host' computers. On the other hand people tried out different kind of ways to experiment with the visibility and accessibility of variables and recursion on the other hand. Some problems in theoretical computer science can be solved easier or more readable when stack a function call is read a different way from left to right, right to left or from the middle. In practice those cases are rare, but at the days of upcoming multi purpose languages there have been approaches to mimic the capabilities of domain languages focused on math problems and expressed in the U.S. style of function definitions (Haskel and other functional languages from the syntax perspective). In Europe we use another standard way of describing mathematical functions, different notations. Agreed, without ever having seen these old, in the 1990s most of this was already pretty old stuff just mentioned in our scripts at the university by the guys who invented PL/I at IBM, Mr. Rechenberg (Left Linear programming languages LL(1 aka Pascal for example), just on e symbol needed to decide what comes next) and his buddy at the university in Berlin who focused on LR aka. C, a guy from the Netherlands. The difference was that the LL compiler were hand crafted and the LR languages were those built with Lex and Yacc. The difference is not the grammar solely but the way the compilers are built. Later also C and especially C++ syntax and compilers were harmonized into a more left linear appearances (2 pass no longer required or as substantial as before - from necessity to feature). I hope I crumbled this very past knowledge together correctly. I never built a compiler, we built Pascal interpreters or better said Modula Interprets aka. Mini-Modula later replaced by Mini-Java interpreters with the focus on recursive decent parsers.

On 1) LOL. I was not aware of this fact, since I don't use third party controls that add actions. I see the advantage. Maybe they have/had no other alternative left, because caching means listening and updating. No doubt that such a delay sucks and EMB has to fix the issue urgently (without delay) and make the whole thing fast. Showing hour glasses is nice, but in that case I'd guess, if possible to implement a speed up is required. The situation is not just cumbersome for you and your daily work, nowadays such thing are just annoying. Putting the blame on EMB solely doesn't make sense, since those who add such actions have to take care that the Delphi IDE stays responsive. If the dialog broke afterwards, we'll see. Agreed. It's been a hot discussion these days, a storm in the tea cup, because of the no coding touring of marketing and speed up in development. Karma is returning. In fact I just wanted to let you know to lower expectations and the outcome of the discussion these days. Question? Does the problem show up in actions manager too? Looks like the same dialog is used.

Cat lover? Feel free to cuddle a bunny of your choice, but in the end bunnies are bunnies with all their very specific lumps and 'd'umps 🐇 As far as I have read you already recognized that the problem with the default actions in the RAD Studio IDE cannot be solved in Visual Studio for example. So it's wise to file in a QC report or as Uwe calls it a QP report and please consider that a new portal has been set up by EMB formerly known as EMBT, but for a while now the technology went ashtray to a certain but in the end minor degree than expected. I didn't come across this issue in the past and thank you for posting your experience. Since take it or leave it is the new normal almost everywhere or better said take things the way they are and leave'em that way, simply don't use the default actions as long as the fix is not considered and/or implemented. I have no idea if it's possible to separate actions into separate BPLs, If I remember correctly that topic or a similar related to actions is not totally new. At least there as something in the past with 'too' many actions too and the outcome as not to use the default action list. I don't recall if that topic came up at TMS. The result of the discussion simply was that people using component set XYZ should use the Action Manager provided and in case of the Action Manager (dialog) provided don't use the feature you are talking about.

So you are looking for a certain kind of poor man's PrintDAT! in order to provide list-reporting, list-printing or formatted list-printing. The situation is similar to poor man's partitioning on Oracle. It works well, but it's work and to a certain degree more in both cases. Take the data from the dataset, use the TPrinter's canvas in order to format the document, print the result to a PDF printer installed into the OS for example. I'd not bother with 'arguments' /thoughts like 'What if there is no such printer installed'. A computer cannot boot to no OS installed, so a program can fairly print to no appropriate printer installed. - In case of designing real reports and passing data from a Delphi data source the story is a different one. Fast Report does integrate into Delphi and not the other way around. - Indeed I'm more infrastructure focused rather than a fan of application centric doing things, that can/should be done another way. An application should not handle the transfer of messages for example, it should just send it. It's just in the early beginning or in case no reliable (messaging) infrastructure can be found or set up. Users tend to get used to this situation. I mentioned messaging just because it's a prominent example or the most prominent one.

Yes. FireDAC definitely does and did from the very beginning.

Great!

No idea. Wondering about the title="Printer-friendly preview' with a " at the beginning and ending with '. Maybe that's a result from providing the manual description. No idea what 'Remoting SDK' refers too, but in general 'Software caused connection abort' doesn't tell a lot and the socket error is a general one for any kind of failure or protocol error. I think you are on the right track. It's very likely about a not well formatted template or a problem with handling the template in most cases. But that doesn't come as a surprise in your case.

Why do you want to switch? As far as relational databases are concerned I do fairly see an advantage in switching at all. Just in case you build your own data access framework as Arnaud mentioned Zeos is one alternative and indeed FireDAC was designed for this and the component layer is really 'just' put on top in order to make the layered library more easy to use, to lower the learning curve. FireDAC as far as the command is concerned is/was more ADO.net stylish compared to other component sets. If I remember correctly UniDAC was introduced pretty much around the time AnyDAC (predecessor of FireDAC) showed up in public. The difference is in the nifty detail. From the bird's eye perspective both are more or less pretty much the same. Under the hood both work completely different. Another question is about having being in the position to utilize access to different database systems by one application since the database is already provided by the customer or 'just' being in the position to use a different database for 'the next application'. In the first case have a focus on the unification layers, for example the unified SQL offered by the 'components' and the statements issued to the database. FireDAC in some cases emulates functionality missing in/not offered by the database system. No idea in how far UniDAC goes beyond what's offered. The authors of FireDAC e.g. in case of Oracle had a focus on doing things the way they are done on the Oracle database. Which for example allowed to have 3000 users accessing a database about 8 to 10 years ago. Working with UniDAC is easier and those components perform very well too. Their advantage is being in the position to access many databases without the need to have a client library installed. As long that is accepted by the customer this direct mode works (very well already) today. There is no thing such as a dedicated support forum. In general what is written in places like this is brought to the attention of the authors or file a request in the official Embarcadero Quality portal. In general: If you are a hobbyist for example feel free to use ZEOS. These component set is easy to use and pretty straight also in case of for example someone who doesn't care a lot about specifics of the underlying database and ZEOS components are pretty straight and well layered too. If you have the choice rely on one database ('forever) and use specific components. I like DOA for Oracle access simply because that's really Oracle just in Delphi and not TDataSet style put on top of the Oracle database. It's ORACLE like when it comes to treating things the ORACLE way in case of CODE Blocks vs. procedures vs. queries in a way things are presented at the database level - kinda 'SQLPlus script' just coded in Delphi. Today few people care about such things.

I usually don't use both or better said never bothered with one of those. So please consider my 'suggestion'/guess carefully. If I recall correctly midas.dll has to be registerd using the regserver utility (syswow64 directory on 64-bit Windows) or use the midaslib unit. Not totally sure, but the first suggestion found in a thread on Stack Overflow. The point with midaslib seem(s|d) to be a registry entry allowing the library to be located. Frankly, I'm just guessing.

Benchmarks are benchmarks and usually compare the capability of a certain piece of hardware or a system as a whole to solve a task. That's what they do and speed up is the point, because slower is easy to achieve in general. The granularity of data is growing and considering time series gets more important than ever before, since hardware became fast enough. I'm also doing hard to find an application for just sorting key figures/values regarding an array as an independent dimension not connected to descriptive values. Maybe my brain is too strongly tied to non-math database thinking and I'm no technician or mathematician. These guys make use of such data-structures. Sorting values should result in faster lookup times acting as an index and extracting ranges in series of values for example. No idea why someone would need that, but as mentioned short before ... In general we are used to, me at least, to know what I'm looking for. If values are just retrieved searching and creating ranges can help to find out what should be looked up. Integer (just normal) as well as floating point values can act as describing dimensions too. In technical applications or the analysis of production line data regarding or planning worked that way. In a business application almost everything that can be related to a surrogate id is more or less discrete and searching starts with specifying describing values retrieved from describing dimensions assumed to have an impact. In case of finding the impacting/influencing dimensions values measured have to be regarded/analyzed first, values grouped and so on. One of my former business partners simplified optimization by removing criteria that had no impact to the optimization of the product mix in production first instead of considering a fixed set. Richard Werner (German economist) used pretty much the same trick to remove the impact of the interest rate to the demand for credit in business for example unless the plug is pulled by the central bank. It's no surprise because 'a line'/company providing goods in the classical Mid-European industry model used the interest rate after app. 35(40) to 50 years to decide if the line should keep producing and just repair the broken machines or build a new plant (reproduction of the line). Since the mid. 1980s even in Austria everything has been pushed into direction the 'Anglo-Saxon' industrial/business model. Those guys simply rebuilt the plant after 45 years and never really bothered with this situation. The myth of the impact of the interest rate and the believe concerning investment decisions is still alive and kept alive. The same myths do exist in many production planning systems considering a wide variety of influential factors/criteria retrieved from the past and with this for so called experience. That's why so many optimization model rely on let's say about 40 describing dimensions/criteria/influential factors taken from a pool of maybe 2000 potential influential factors while in current production, in a steel plant liquid phase, only 3 or 4 really drive the optimization software's decision and the result of the planning and optimization run. This requires the analysis of the values/key figures and describing technical dimensions first an the analysis should finish quickly, because today a run of production plan happens in de facto real-time (about 0,5 sec or less) and finding the influential factors should also not take long. In the past 20 years ago such a run was allowed to take 20 minutes.

One of the major reasons to use components from Chilkat was SFTP and/or FTPS in contrary to Indy, a few years ago. Without a real good reasons, performance concerning floating point operations for example, I'm no big fan of deploying dynamic link libraries instead of static linking. What's missing Embarcadero has to provide and improvements too. Where does deploying this and that a different way lead? It does cause headaches even on a mid term in practice. Headaches go away with a Delphi only solution after while and then problems 'Wolpertinger' style deployments start. The whole story looks different if you want one homogenous component set serving different development technologies and/or programming languages. Even in this case homebrew, (bindings) provided by a community and/or out of the box come into play. Thinking of Java with the possibility to 'just plug-in' a different socket would be an alternative (early days of SSL). After a while what makes the difference today will be caught up by a certain kind of 'out of the box' way of doing things.