Andreas Hausladen has released his latest DevExtensions and IDEFixPack for Delphi 10.2, just the latest release of fixes for things that Embarcadero should have fixed before pushing the “It’s Ready” button.
Andreas’ work always presented something of a quandary to me. Whilst it was great that someone like him was able and willing to improve the product in the way that he did, it was also a constant source of disappointment to be reminded that there was even the need for it in the first place.
Now I should be clear… the title of this post is a paraphrased reference to the title of the 4th book in the Hitchhiker trilogy by late, great Douglas Adams. But it does not mean that Andreas is leaving Delphi behind for good.
Sadly however, for those still keeping up with the release cycle, future releases of Delphi are unlikely to be supported by anything he releases in the future as he has chosen not to renew his update subscription. As a result he will not be able to compile or verify/validate his code for any releases after 10.2.
For me personally this is no great loss, as XE4 is the last release I stumped up for and even then is used increasingly infrequently. Never-the-less hopefully Andreas will continue to support that for a while longer at least.
As noted in his reply to comments on his post/announcement, it’s not simply a matter of cost (in case anyone was thinking of rousing a “Hit the Donate button” campaign with thoughts of changing his mind).
Having said that, if – I’m somewhat ashamed to say: like me – you have used, and/or still use, even only occasionally, the results of Andreas work but for some reason have never hit that button before now, why not take this opportunity to say “Thank you” and do that right now.
I for one am going to do exactly that, right now.
Thanks Andreas.
Man if he put all that work into Lazarus instead of the Delphi money pit. Don’t miss Delphi at all, I can do all the same stuff I was doing with Lazarus.
Then you must be writing some very vanilla stuff 🙂
Not necessarily. Lazarus is roughly the equivalent of Delphi 7 in terms of the IDE. And with third-party components, you can do just about anything you need with an old version of Delphi.
I recently saw several very capable and highly used business-class applications connecting remotely to SQL Server, automatically updating themselves from an FTP site, and display beautiful bar and pie graphs embedded in reports. It was written in Delphi 5 and using QuickReports!
Unless you actually want to use Delphi for cross-platform stuff, there’s not much you can’t do with early versions of the product. I only have a little experience with Lazarus, but it certainly was not limited to vanilla stuff.
That’s a very odd observation to make about an implementation of Pascal which has a far greater and more diverse reach than Delphi.
Unless you meant vanilla in it’s original sense of “exotic”, rather than the relatively modern equivalence with “plain”, “standard” or “boring”. 😉
I think Andreas is focused on the C++ Builder especially. See Andreas just needs a CLI and a C compiler.
What’s a community. It’s a mixture between a market(place) and a premise. Value the masters work with coins.
The download in the concept the community tells something different. It tells Andreas that he is still regarded as a master.
Our income system, especially in the North of Austria, cannot so easily be expand income even on a temporary basis. Without a sustain increase of the domestic demand you don’t get an increase in salary. Our incomes are socialist incomes (money base). Industrial income with no growth which means the income is always a basic income because in our world you cannot pay back consumer debt you could not save. A second income is required.
In Austria we have 2 incomes handed over in order to save or temporarily expand our personal money supply (income).
In the model of everyone is employed no can save anymore. If someone starts to save another one drops out (and received a basic one). In return the one goes to his lab and tries to successfully finish activity without a material constraint from the outside world. So how to determine if your activity is successful. On the premise level finishing is enough. In the market the download is the opinion of the receivers (count 1) but not the valuation.
EMB are definitely not the masters other wise his extension wouldn’t be required.
In general either the free starter is a valid option or an open source alternative. Open Source is not the perfect model but it’s the best we have.
He tried. It’s been years ago now, but he’s mentioned trying to submit a patch that would have made Lazarus compile using DCC instead of FPC, and the attempt was not well received. Certainly understandable from the FPC and Lazarus developers’ perspective, but it did close the door to using a faster compiler and better debugging experience. Given how obviously Andreas prioritizes speed and productivity, it’s no surprise he stopped trying after that.
It may have closed the door to using DCC, but was that door ever open in the first place ? Given the raison d’etre of FPC and Lazarus I don’t think it was.
But I’m not quite sure how that relates to a better debugging experience as that surely would have required a significant amount of work on the debugging subsystems of FPC/Lazarus, not just a change in the compiler. Andreas could have instead worked on improving the speed of the FPC compiler itself, what with it being entirely open source.
That’s not a criticism, just an observation that perhaps the reasons for Andreas not getting involved with FPC were a bit more complex than having baulked at an initial meeting with understandable resistance.
In 2003 I had a patch for Lazarus that allowed me to compile the LCL with DCC32. Lazarus’s source was never part of that patch. The main reason was that I could use the Delphi debugger to debug the LCL code. Lazarus uses GDB and that debugger is horrible compared to Delphi’s debugger, which itself isn’t that good but worlds better than GDB, especially under Windows.
By compiling the LCL with DCC32 I also found bugs without looking at the code because the compiler had better hints/warnings/errors at that time.
The “problem” with the Lazarus maintainers, that many speculate about, was that they left me with the decision to either maintain an extra set of patches or to stop this journey. Maintaining an extra set of patches would have cost me too much time to keep up with the development of the LCL, so I stopped that journey. I reported all bugs that I already had found and haven’t returned since. I also stopped using Linux around the same time because Kylix was dead and so there was no need for me to use FPC/Lazarus.
And then, in 2004, DelphiSpeedUp and later IDE Fix Pack ate all my spare time.
Some time ago I compiled Lazarus with FPC and thought about finding performance bottle necks in the compiler but instead I started working on the DCC64 SSE4.1 performance optimizations, that are still in BETA.
Fun fact: The folder that contains IDE Fix Pack and all my other tools/projects is named “DelphiFPC”.
Wow, thanks for the background and info Andreas. 🙂
I for one, would love to use the Delphi debugger with Lazarus. The GDB debugger experience 14 years later is still horrid, and one of the primary reasons I have not spent more time with FPC/Lazarus. If any of your work could still be used to that end, I would implore you to try to get that code into Lazarus repo. Perhaps things have changed since your last attempt.
BTW, I appreciate your work on Delphi SpeedUp/IDE Fix pack.
You might want to post this on Andreas’ own site, to make sure he sees it. 🙂
why not open source the code so other people can continue his work ?
I don’t think it’s an issue any more. Since this post it looks like Andreas has resumed his great work to at least some extent. \o/