Nick Hodges

Random Thoughts on the Passing Scene #131

09 Oct
  • We are looking for folks to submit their Unicode migration stories to the Unicode Migration Best Practices Cookbook.  If you have a tip or two, or a good story to tell, please let us know and help out the Delphi community.  Looks like you might win an iPod Touch if you submit something.  You’ll also get fame and glory and credit on our web site.  Nothing to blow your nose at.
  • If you are on RAD Studio 2005 or earlier and haven’t yet upgraded, now is the time to do it.  Your time for getting upgrade pricing is running out.  After January 1, only users of BDS 2006 and up will be eligible for upgrade pricing.   And with the release of RAD Studio 2010, there’s never been a better time to upgrade.  I may be biased, but I not so humble that I’m not willing to say it out loud:  Delphi 2010 is the bestest, most powerfulest and coolest Delphi ever.  :-)
  • I love IDE Insight.  You can get get to it two ways, F6 or CTRL-<period>.  (I tend to use F6, since that was the “original” keystroke during the beta.)  Getting in the habit
  • StackOverflow Post of the Week:  What are good arguments to convince management to upgrade to Delphi 2009 / 2010?

32 Responses to “Random Thoughts on the Passing Scene #131”

  1. 1
    Mason Wheeler Says:

    I think you mean that there’s never been a better time to upgrade?

  2. 2
    Rich Says:

    I’d love to see a D2009/D2010 IDE plugin that helps with unicode migrations. Is such a thing available? Ideally, it should work like Refactoring features. It doesn’t have to be 100% automated, as long as it identifies 100% of the code that potentially needs to be modified from D2007 to D2009/D2010.

    Camtasia videos about Unicode migration might help convince skeptics to finally upgrade to D2009/D2010.

    ps

    PLEASE do Camtasia videos about the UML & Design Patterns features in the Enterprise edition. It will boost sales of Enterprise edition. And it will help users get a jump start, preventing us from downgrading to Professional edition. Like many, I use (and love) RemObjects, so the UML & Design Patterns feature is the most prominent advantage of Rad Studio Ent. over Pro. edition.

  3. 3
    Kyle Miller Says:

    Rich, this might be helpful in assessing Unicode migration effort.
    http://blogs.embarcadero.com/ao/2009/10/09/39042

  4. 4
    Nick Hodges Says:

    Mason –

    Fixed, thanks.

    Nick

  5. 5
    Pratt Says:

    Guess not enough D7 users are upgrading to D2010 so Embarcadero is trying to "blackmail" them to upgrade.

  6. 6
    Jolyon Smith Says:

    Sadly I suspect the most common migration story will be how begin/end were replaced with curly braces.

  7. 7
    Alexandre Machado Says:

    Can you please explain us WHO is the genius behing this decision? Please, don’t blame Borland management!!!

    Close the door for a possible future upgrade is not a wise decision. "Give us your money now, or give more in the future" can turn to "no money in the future".

    The company I work for has no means to do a massive upgrade now. If the upgrade prices are considered exorbitant nowadays, imagine how the upper management will just LOVE Embarcadero when you close the door for an upgrade next year! Very very wise, Nick. I guess that after all those ups and downs during Borland times, Embarcadero will finally kill Delphi. Congrats!

  8. 8
    Mark A Says:

    Jolyon Smith said:

    "Sadly I suspect the most common migration story will be how begin/end were replaced with curly braces."

    Yeah, because re-writing the entire frigging application is much easier than just adding Unicode support.

    Or did you mean people will be commenting out the offending blocks?

  9. 9
    Carlos Says:

    After so many years of Borland mistakes, I was thinking that finally Delphi was in good hands with Embarcadero, who I was considering smart people that actually "think" before taking wide decisions. But now, seeing this new upgrade rule, I’m afraid I was mistaken.

    There is always a lot of complaints about the product prices being too expensive, and now you want to cut away thousands of developers who still can’t upgrade?

    Maybe Embarcadero has some MS people infiltrated, trying to make more people to leave Delphi and go with VS…

    I hope there is still time to go back with this decision and let people upgrade when they are ready to do it.

    If you offer me an automated tool that with just one click will convert all my code (including 3rd party components) to a 100% reliable unicode version, I’ll be glad to upgrade now. As you don’t have such tool, and some of my main 3rd party components are still stuck with Ansi, and I don’t have enough resources to update the code by myself, I just still can’t upgrade to D2010 (believe me, I would love to do that!). For sure there is a lot of people in the same position as me.

    Please, stop the "I wanna your money now!" attitude, and support the developers who still trust in your product.

  10. 10
    Bruce McGee Says:

    muttering to himself "don’t feed the trolls…".

    The single biggest thing to remember when upgrading to Unicode (or writing new software) is to simply treat strings as strings (just like you would in .Net), and treat the cases where you need ANSI strings as the special cases they are.

  11. 11
    John Says:

    Thanks Nick & Co. for trying to blackmail me into upgrading! As a satisfied user of Delphi 7 I’m still waiting for the release of a 64 bits version, probably in 2011 or 2012. For me Unicode, cross platform and some fancy new IDE / language features are not important. I’m going to wait several years before upgrading. Saves me nice bundle.

    Thanks you Embarcadero for reshuffling the roadmap again and again!

  12. 12
    Jim McKeeth Says:

    The longer people using older versions of Delphi wait to upgrade the more difficult it will be for them to upgrade. Most likely, as long as they can "keep their option open" then if they haven’t upgraded yet, they probably won’t. Embarcadero is doing them a favor by making them decide now. They either need to quit lying to themselves and realize they are not going to upgrade, or upgrade now and port their code.

    Yeah, we don’t like having our options taken away, but in all honesty most people get decision constipation if they have too many options and then never make up their mind. Of course this means that some people may decide to switch away, but like I said, they may have never upgraded anyway.

    Instinctively I don’t like the cut off either, but it is probably the right thing to do. Although it would have been nice if they extended the BOGO until that expired.

  13. 13
    Jim McKeeth Says:

    BTW, if someone has a strong opinion on this (either-side), and wants to debate it on the podcast, then let me know and I will have you on the next episode.

  14. 14
    John Says:

    @Jim McKeeth

    The problem is not users having upgrade constipation, it’s Embarcadero reshuffling their roadmap again and again. The beta 64 bits compiler was scheduled for this year. Deliver the goods and I’m upgrading. I’m absolutely no fan of Nick & Co. and their excuses and empty promises.

  15. 15
    David Heffernan Says:

    Yeah, instant upgrade for me the day the 64-bit version is ready. It looks like MS will have released 3 different versions of Windows by the time there is a Delphi compiler ready (XP, Vista, 7)!

  16. 16
    Jim McKeeth Says:

    @John & @David: So you are still on Delphi pre 2006 and the *only* thing you are waiting for is support for more then 4 GB of addressable memory? Yeah, 64-bit is a big deal, but there is a whole lot more to software development.

  17. 17
    David Heffernan Says:

    @Jim,

    D6 does me right now. An upgrade to D2010 is an investment of time and effort. If there is not 64 bit version soon then I’ll have to port to a different platform. Why would I port to Unicode and then port away from Delphi? That’s one more port than I need. I admit that the Unicode port is orders of magnitude easier than the away from Delphi one.

    You say that there’s more to SW dev. than 64-bit. But you don’t know anything about my app. It turns out that for my app 64-bit is crucial. It may not be for the work that you do, but it is for me. Perhaps I’m the odd one out!

    Anyway, since it seems that Embarcadero are now moving on 64-bit (even though they frustratingly refuse to give us any progress updates) then I’m planning on getting D2010 (before Xmas and hence cheaper) and porting to Unicode. Then the subsequent 64-bit port will be simple (I hope).

  18. 18
    Jim McKeeth Says:

    @David: I realized that after I posted that there were probably a large number of applications they desperately need 64-bit, so I totally understand.

    I for one am glad for the added transparency that Nick and Michale have brought to publishing a road map and letting us know when things change. The whole reason companies don’t like publishing a road map with their goals is when things change, and they change for everyone, then people get grumpy.

    I would rather be told what direction they are headed in, and when that direction changes, then be left in the dark until the last minute and not know what they are ultimately working towards.

  19. 19
    John Says:

    @David

    You are absolutely right, there is a whole lot more to software development. But that doesn’t mean that we have a need to upgrade to Unicode, cross platform, fancy features, etc.

    The most important question for us as an ISV to ask is, is an upgrade beneficial to our customers? For us the answer is no, our customers don’t need (for example) Unicode support. Delphi 7 Enterprise gives us all the features we need at the moment.

    Upgrading to Delphi 2010 is not only paying Embarcadero the upgrade price. It also means we have to upgrade components, patch and test exist code, etc. Two or three years from now we have do this upgrade cycle once more for the 64 bits version. For us it’s a lot cheaper to skip a few releases and do it all at once.

    It’s great that, after years of neglect by Borland, Embarcadero is finally investing in improving Delphi. But it’s really frustrating that, after the !@$* dot Net detour, it now takes the cross platform detour instead of using it’s limited resources to release a great 64 bits windows version. As far as we can see Delphi is foremost a tool for developing Microsoft Windows application. Developing for other OS platforms is a nice to have option. It’s better to do one thing really great instead of doing lots of things so so.

  20. 20
    John Says:

    Oops, previous post was for Jim McKeeth

  21. 21
    Alexandre Machado Says:

    @Jim McKeeth:
    So, let’s assume that you have better knowledge of my reasons to stay with an older version of Delphi than me… and let’s assume that I’m lying to myself, and I will NEVER upgrade, ok? So Embarcadero will lose nothing right? If I will never upgrade anyway, what are they losing?

    Best regards

  22. 22
    Andrew Says:

    Embarcadero are running a business too: they need to employ people, pay salaries and infrastructure, and to do this they need to *sell* their product. The attitude of people who last paid for something 10 years ago and try to threaten Embarcadero because they "are the real Delphi users" annoys me.

    That said, I believe the price of Delphi is very high because of the generous upgrade policy. It should be that you qualify for a good discount price if you move one major version up, but upgrading from anything older pays the full price. You can then sit on your version as long as you like and then if and when you feel justified in getting the latest version, you don’t get preferential prices for your almost non-existent support that brought about the newest version.

  23. 23
    Jolyon Smith Says:

    @Mark A: Migrating to Unicode is NOT going to be straightforward for a lot of people.

    Demonstrably it ISN’T straightforward.

    At the very least you have to upgrade all your components, even if you have source. Yes, we’ve been through such pain before, but making Unicode a compulsory aspect of keeping current adds just one more complication.

    PLUS, it’s a complication which for many people is "running to stand still". If they neither want nor need Unicode, Delphi doesn’t care. Htey have to deal with it whether they like it or not.

    At the same time, Embarcadero are getting heavy handed with people left behind as a result of mis-steps with the product in the past, and who in many cases are finding themselves still unsatisifed with current directions the product is taking.

    2 years ago, 64-bit was promised in Delphi 2010.

    Today we were told by Malcolm Groves that 64-bit is 2 years away. That is, ANOTHER 2 years away.

    C# has been beckoning to a lot of Delphi shops already. This could be the nail in the coffin.

  24. 24
    Bruce McGee Says:

    Jolyon,

    Promised? I’d be curious to see if anyone officially committed to a delivery date for native 64 bit support. I could be wrong, but I suspect you’re thinking of informal potential release plans that started out with a healthy disclaimer that plans may change.

    I know it wasn’t what you intended, but your post almost reads as a complaint about forced Unicode, and suggesting that C# is the solution. Unintentional, but still funny.

  25. 25
    David Heffernan Says:

    Is it true that an embarcadero employee says that 64 bit is still 2 years away?

  26. 26
    David Says:

    Sort of. We were told it was an ESTIMATED 2 years away. Malcolm said it could be less but realistically with the current focus on cross platform to Mac and Linux I don’t personally think it is very likely.

    We’re not all that fussed about 64 bit compilation, we don’t really need it at present (I know others do). HOWEVER I see 64 bit as a progression of core Delphi. Therefore I put it head and shoulders (and torso, legs and feet) over Mac and Linux support which is definitely a step sidewards with a new VCL etc. It is reminiscent of Kylux and Delphi.NET and I really believe Embarcadero are deluding themselves if they think cross platform is an instant moneyspinner.

    I’m also puzzled at the desire to produce a 32 bit cross platform compiler… aren’t Mac and Linux installations a larger proportion 64 bit than Windows? I would have thought doing 64 bit Windows compilation THEN cross compilation to 64 bit Max/Linux would make more sense for everyone?!

  27. 27
    David Heffernan Says:

    @David

    Cross compilation is probably easier since it means that they can leave Delphi and the compiler as 32-bit applications. It’s also a significant benefit for the end-user since they can compile 32/64 bit targets from a single machine and run/test them both from that same machine.

    I do agree with you regarding Linux/Mac. We’ve been this way before with Kylix. We also had the Delphi .net debacle. We now appear to be doing Mac. During this time MS basically abandoned native development (they did very little with MSVC and dropped native VB). What an ooprtunity for Borland/Codegear/Embarcadero to capitalise on.

    Well, those are my views and I imagine everyone here is getting bored of hearing me repeat them!

  28. 28
    David Says:

    Back @David.

    As far as I am aware even the 64 bit compiler version is planned to be written in 32 bit so that a developer can develop 64 bit apps on a 32 bit OS.

    My comments were more about the resulting exes being 32 bit vs 64 bit rather than Delphi and the compiler.

  29. 29
    Jolyon Smith Says:

    @Bruce: Yeah it’s easy to mock complaints about Unicode in Delphi and attendant suggestions that people will switch to the All-Unicode-All-the-Time C#. "Easy", but entirely missing the point.

    Unicode itself isn’t the problem - as you well know.

    The problems with the Delphi Unicode implementation are that in trying to appease the non-Unicode crowd whilst playing to the Unicode audience they have created an almighty mess.

    Delphi, and owners of Delphi application source code, have to deal with a TRANSITION from working, non-Unicode code to untested potentially NOT working Unicode code without having taken any deliberate steps (in many cases) to effect that transition.

    C# is Unicode through and through.

    Faced with a not insignificant job to migrate code and 3rd party components (some of which may not be supported any more, leaving you with the job of fixing any Unicode issues in that source yourself), and then realising that what you have after that effort is not as reliable or predictable as what you had before, you might consider your options.

    Now certainly, it’s a big job to port a Delphi app to C#, but it’s a job that is done explicitly and with attention to detail.

    If you are teetering on the edge of a loss of confidence in Delphi and the direction it’s taking (as I know many people are, especially in the Enterprise license holder space) then this could simply be the final straw, and the C# bullet finally bitten down on hard.

    Because yes it will be a lot of effort, but the result will be a better solution for that effort - the benefit for the cost of that effort is significant.

    Migrating legacy Delphi code to Unicode is not *as* big an effort but it comes with costs and risks and very little relative benefit.

    You don’t have a better application as a result. You don’t have an application written in a tool that you don’t have doubts about the future of. You don’t even have an application that you can be sure is going to work as expected or required.

    So yes, it’s entirely realistic to expect that the "Unicode issue" will push people over the C# edge.

    I’ve seen it nearly happen - it was largely the expectation of 64-bit support that saved the day for Delphi in that case. Even so, in that case a parallel project has started to eventually replace the Delphi application with a C# one.

    Previously the C# project was seen largely as a "hedge bet" effort. With the further delay of the 64-bit delivery I am fairly sure that priorities will now have switched. The app in question goes back a long way and has only recently been migrated to Delphi 7.

    I doubt very much whether the company involved will be migrating to any later version of Delphi than that now.

    (I am no longer involved in that company so can’t say for sure what their thinking is, but I was significantly involved in convincing them to stick with Delphi over the last two years - now that I’m no longer there there are others in that organisation I know who will be pushing the C# agenda very aggressively).

  30. 30
    Bruce McGee Says:

    Jolyon,

    I wouldn’t take it too seriously. It’s easy to mock silly (or unintentionally funny) complaints about Unicode. Legitimate concerns are another story completely.

    On a similar note, I remember an old rant on the news groups where someone was complaining about .Net being required by the Delphi IDE and saying how he was going to dump it for Visual Studio. Also funny.

    Personally, I’m glad Delphi jumped into Unicode with both feet instead of compromising to make the porting of code that shouldn’t be making assumptions about the size of a character in the first place. It’s not like we’ve never been told this.

    However, my main point was that I’m pretty sure no delivery date was ever promised. The more serious point is that if people (not necessarily you) insist that any mention of an unofficial or estimated time frame constitutes a promised delivery date, then it’s little wonder Embarcadero is reluctant to give us more detailed information.

  31. 31
    Bruce McGee Says:

    Jolyon,

    I wouldn’t take it too seriously. It’s easy to mock silly (or unintentionally funny) complaints about Unicode. Legitimate concerns are another story completely.

    On a similar note, I remember an old rant on the news groups where someone was complaining about .Net being required by the Delphi IDE and saying how he was going to dump it for Visual Studio. Also funny.

    Personally, I’m glad Delphi jumped into Unicode with both feet instead of compromising to make the porting of code that shouldn’t be making assumptions about the size of a character in the first place. It’s not like we’ve never been told this.

    However, my main point was that I’m pretty sure no delivery date was ever promised for 64 bit. The more serious point is that if people (not necessarily you) insist that any mention of an unofficial or estimated time frame constitutes a promised delivery date, then it’s little wonder Embarcadero is reluctant to give us more detailed information.

  32. 32
    Bruce McGee Says:

    Hmmm… Duplicate post (almost). Not sure how that happened.

© 2014 Nick Hodges | Entries (RSS) and Comments (RSS)

Your Index Web Directorywordpress logo
Close