Hey, I got a pretty cool link.  Jason Stamper of Computer Business Review linked to my blog in his blog.  Pretty cool. He links to the post right before this one which discusses Allen Bauer's comments about DevCo's relationship to Borland.   Jason also mentions my little row with ComputerWorld editor Paul Brislen.   Hopefully this is a step in the right direction towards the press understanding DevCo a bit better -- i.e. that Borland and DevCo are now pretty much distinct entities and that there is a lot of interest in purchasing DevCo -- as it seems Jason understands the situation a whole lot better than Breslin did.

Thanks for the link love, Jason.

There was a bit of confusion yesterday about the Borland layoffs and what they mean for DevCo.  Allen Bauer posted the following in non-tech:

None.  DevCo folks are isolated and running separately at this point.  This is a from a product/marketing/sales perspective.  Due to SOX rules, the financial aspects must remain intertwined until the divestiture is complete.

It is also my understanding that the Seque portion was "right-sized" prior to the close of the deal.

For the record as well, DevCo is hiring and has many openings available today, and will have many more when the separation completes.

So it would seem that there really is a pretty clean distinction between Borland and DevCo.  Nice to see that they are hiring, too. 

More, <ahem> "interesting" news from Borland today:  "Borland Unveils New Organizational Structure"

I've long since given up knowing if this sort of thing is good or bad for business.  I do realize it looks bad for business, but cutting costs isn't always a bad thing.  It's certainly non fun for those who lose their jobs.  (I should note that I think Borland would do well to line up all the vice presidents in the company, have them number of by 1 and 2, and then fire the 1's and tell the 2's to get back to work, but that's just me.  You can't swing a dead cat at Borland without hitting three VP's.)

What I do know is that we Delphi people need to start realizing that this isn't something for us to be concerned about anymore.  Borland and DevCo are de facto separate entities -- soon to be de jure separate.  It's been a long time, and it will likely take some getting used to, but Borland's fate really isn't ours anymore.  While I'm sure we all wish Borland well, DevCo is on a separate path, doing separate things, with a different market and a separate business model to follow.  Borland laying off peole doesn't really seem to affect DevCo. It appears as though DevCo is pressing forward.

This seems to me to be one of those cases where we all wish that things could speed up and DevCo could stand totally on it's own two feet as soon as possible.  According to the press release, it's coming.  I know we all are looking forward to it. Well, at least I am. 

If you haven't already, you should head over to Hallvard Vossbotn's blog and catch up on his series of blog posts about the inner working of the VCL and the compiler.  He's been running a series of very, very interesting, in depth articles about how Delphi's RTTI works, how the Virtual Method Table is working, how interfaces are referenced and implemented -- i.e. really cool, geeky stuff.   Best stuff on this topic since Danny Thorpe's Delphi Component Design. 

This is the sort of stuff that Hallvard ought to wrap up into a book and publish on http://www.lulu.com. 

I'm nearing the end of a pretty big project at a client very near our office.  I've been going in ever day full-time for the past few months.  I've really enjoyed working there, as the folks are nice, the development process is very professional and quitegood, and the comeraderie quite enjoyable. 

Now that the project is winding down, I've had a chance to actually do a bit more visiting and kibitzing with the folks there about general topics concerning development (plus various forays into the Civil War, American Idol, etc.)  The topic of Delphi books came up, and one of the guys, Randy, made a good point.  He noted that the reason that there aren't that many Delphi books out there is that there doesn't need to be that many good Delphi books out there. 

The books that are out there are complete, comprehensive, and well done.  For instance, there are really two books out there right now for Delphi:  Delphi for .NET Developer's Guide by Xavier Pacheco, and Mastering Borland Delphi 2005 by Marco Cantu. Once you've bought one (or both) of these books, you've got pretty much everything you need.  Xavier's book covers pretty much everything you need to develop in Delphi for .Net, and Marco's book covers both .Net and Win32 development.  If you have Marco's book, for instance, you really don't need any other books.

Another factor in limiting the market for Delphi books is that Delphi is a mature product.  It's users are to a large degree familiar with the language and the tool, and thus the demand for books that cover the basics is limited.  There just aren't that many folks out there that need to buy a book about Delphi anymore.  (Now, we can argue that we all need to figure out a way to increase the number of folks that want to buy books, but that is another blog post.) And since there are so many facets to Delphi, any book on an advanced topic would be of limited interest.  And for topics like ASP.NET, there are so many C#/VB.NET based books out there, the content of which is 99.9% useful to a Delphi developer, that there's little reason for a publisher to go to the effort of putting out, say, an "ASP.NET for Delphi Developers" book -- as much as I'd like to see such a thing.

In addition, the presence of resources like http://groups.google.com makes finding answers to questions and problems easy.  It's a rare day indeed when I can find an answer to a question using the newsgroup archives.  This is an invaluable resource, when you stop to think about it.  Shoot, just Peter Below's responses alone are worth their weight in gold.    It might very well be that the excellent support received in the newsgroups has limited the market for Delphi books as well.

The bottom line, it seems to me, is that there aren't that many Delphi books out there because we just don't need that many Delphi books.

So what is the future of Delphi books?  I think it's with something like lulu.com.  Lulu.com couldn't be easier -- you send them a PDF, and then they'll print out a real book whenever someone pays for one.  No muss, no fuss.  Basically you are 100% in charge of the PDF, and you get 80% of the selling price. Someone like Ray Konopka, who sells his class Delphi Component Development book online as a PDF, would be perfect for something like this.  You want to publish your own Delphi book?  There you go.  The size of the audience matters not.  You want to buy a book?  Encourage authors to use a service like Lulu.com to publish their materials.  Heck, publish one yourself.  The overhead costs seem quite small, and overall, it seems pathetically smart and simple.

I suspect that services like lulu.com will florish and only get better and cheaper.  The stack of Delphi books at your local Borders (weird -- that link basicall goes to amazon.com...) might not be that tall, but the amount of information available to Delphi developers is still staggering.  I've never been a big believer in the "books on the shelf equals market strength" theory -- at least in the sense that I think Delphi's market strength is much, much stronger than one might conclude from a trip to Barnes & Noble.  Delphi book publishing need not be dead, it just needs to shift to a new way of doing business.

This week's Post of the Week goes to Dennis Landi for his excellent words on being a Delphi Advocate:

Isn't this moot?  Is Delphi really in that good a shape in the market place  that it can continue to withstand blanket condemnations/dismissals/insults?  I not so sure it can.

When the Borland/DevCo divestiture occurs, DevCo will then be directly responsible and directly accountable for the Marketing of Delphi.

Let's look forward and not backward.  We need to put the mistakes of the past behind us; and deliberaretly take an up-beat attitude for the future.

But I would add that DevCo, alone, isn't solely responsible for the Advocacy of Delphi in the world.  Each of you who depend on Delphi for your livelihood should also feel a responsiblity to keep the Delphi eco-system alive.  This means promoting Delphi at your location.  If you are a hobbyist then promote it in your hobby community.  If you are a professional then promote in your profession.   Talk it up; and don't gratuitiously put it down.

Ask yourself what you've done lately to promote Delphi; before you publicly flog DevCo for poor performance in this area.

Get creative.  If you can't be creative then at least be active.  Start a blog.  Put something in it.  Start with some little thing that you love best about Delphi.   Write a about it.

And What if....

What if each of us actually recruited ONE new Delphi developer a year? Doesn't sound impossible does it?  And yet, that would effectively double our numbers every year....

Especially after the divestiture, I think this will be my automatic question to self-styled CRITICS.  And woe unto you if it turns out that you've done nothing at all to help Delphi except wallow in the oh-so-easy role of CRITIC.

If you care so much about Delphi to spend hours bashing the employees who work on the Delphi product, then perhaps you might have ONE hour a week to give to the promotion of Delphi in your own personal sphere of influence.

Just remember, that you also represent Delphi in the world.  From now on, every time you bash Delphi publicly you are undermining it in the world.  Do you have another choice?  Which choice would serve your interests better?

Boy, that's almost enough to bring a tear to your eye, eh?  Dennis and I have had our disagreements over the years, but man, I couldn't agree with him more here.  Sitting around whining about Delphi's (alleged) loss of market share isn't going to help anyone.  Whining that there are no Delphi jobs isn't creating any Delphi jobs.  (And I'll repeat again, I know at least four places here in the Twin Cities alone that are looking for Delphi developers.  "There are no Delphi jobs" is just plain false.)  What it takes is passionate users, of which Delphi has plenty, getting off their tushies and doing something about it. 

Well said, Dennis.  I applaud your post and your sentiment.  Keep it up.

A while back, Larry O'Brien wrote an SD Times column about DevCo and the future of Delphi, JBuilder, etc.   The article was pretty fair, though it did have a few unfair digs, most notably the statement that "The Delphi language itself is well past its peak, and with its Pascal roots is on the wrong side of trends in syntax."  I beg to differ on that point, as you would guess. Generally, though, the column was pretty upbeat, and recognized that there might be some spring in the ol' gal that is Delphi and the rest of the DevCo toolset.

Well, DavidI, being DavidI, gave Larry a call, and Larry wrote about the results in his blog.  (Hat tip to Mike Gunderloy)

Larry gives a fair reading of the future of Delphi, though I don't necessarily agree that book sales are a measure of potential popularity.  (I've discussed the notion of dead tree book sales a bit on this blog, and my general view is that the "buy it at Barnes & Noble" model is a dinosaur industry).  It is true that C-like languages are gaining in popularity, but languages like Visual Basic (and Delphi) are still viable alternatives.  Not everyone requires C-like syntax, and some of us like the explicit syntax. In addition, some of us like being able to develop Win32 applications.  (I note that if you want to use a C-like syntax to develop Win32 applications, you are pretty much stuck with C/C++.  Java and C# don't do Win32)

In addition, the popularity of C-like languages doesn't necessarily mean that people prefer them.  It may be a case of not knowing any better.  I agree that C-like syntax is very widely used, but it might very well be that many C-like programmers would prefer Delphi if given a decent demo of it or made a serious comparison.  Maybe I'm wrong, but I'm guessing that many of the C-like developers (not all, but many) just simply haven't considered the difference.

The most interesting nugget out of the rest of Larry's notes on the conversation was David I mentioning the possibility of expanding out into other languages.  That would be sweet. Though DavidI didn't "bite on the idea", I like Larry's notion of "TurboRuby".  I think there is a broad  market for some new languages that could be added into BDS.  A Python or Ruby personality in BDS would be a great differentiator for DevCo tools.

Anyway, it was nice to see a columnist respond in his blog to a phone call about a column.  That's the way it is supposed to work.

(One additional note:  If DevCo becomes a private company, it wouldn't be subject to Sarbannes-Oxley, would it?  I was under the impression that SOX applied only to publicly traded companies.)

(Another additional note:  I really enjoyed reading Larry's rather frank bio.  Quite a run.  I'd love to be able to get to the point where I could say "I won't work on a project of more than 10,000 lines of code if it does not have..."  That's cool.)

Marco pointed out a cool new website for Delphi 2006.  Looks like David I will be traveling the world to bring Delphi 2006 to a town near you.  Looks like all European sites so far, but I bet the list grows in time.

Nigel Brown is the guy in charge of what will become DevCo.  He's the "General Manager of the IDE Business" inside Borland, and one could presume he'll be the CEO of the new company that ends up getting spun off.  He recently gave an interview to The Register -- not one of the more reputable periodicals, but the article seemed pretty well done.

The part I found interesting is that he is looking into pursuing the RedHat model -- sell product and services together.  I was also glad to hear that he's looking at selling subscriptions instead of merely licenses.  I'm a  big fan of the subscription model of selling, as it's a win/win for developers and DevCo. DevCo can smooth out the development cycle and get away from the "monolithic release" model of development and selling, and customers can count on a steady stream of improvments, updates, new features, etc. 

I'm personally pretty clueless about the Java tools market, but I'm intrigued by the idea of JBuilder on Eclipse.  Eclipse is obstensibly 'free', but I presume that 'out of the box' it isn't very feature-filled.  I'm guessing that people and companies are willing to pay for value-added plugings to Eclipse.  Presuming that all the development that formerly went into the infrastructure of JBuilder can now be pointed at features for Java development, it seems like JBuilder might end up being a very strong, very feature rich tool that is merely based on Eclipse.  I'll be following this more closely in the future.

The time has come for me to become a stone-cold Javascript expert. 

Do you have any good recommendations for On-Dead-Tree resources (i.e. a book) for learning Javascript well? (I've found a few excellent online resources, but I'm wanting a book)  I'm particularly interested in it with an eye towards AJAX development. 

 

And while we are on the topic of job shifts, I note that Danny Thorpe is now at Microsoft.  Looks like he's working on Windows Live.  Good for Danny, if that's what he wants.  (Well, clearly it is, I suppose.  )   I wish Danny well in his new endeavor.  I also wish he's come back to NewCo, but that doesn't appear to be in the cards.  The part I find funny is that this was some sort of crisis in the newsgroups.  (I note that Danny did take a swing by *.delphi.non-tech to clear up a few questions and conspiracy theories.) 

Danny, as always, I wish you well in all the new endeavors in your life. Don't be a stranger.

I note with interest that Julian Bucknall, long-time Delphi-ite from back before Delphi even existed has taken the position of Chief Technology Officer at DevExpress. This is a great match I think. I've known Julian for a while, and if I may presume to say so, I don't think he's cut out for the "crank out the billing hours at a customer site" kind of job.  No insult intended to the reader, but he's way too capable and smart for that.  Julian is an idea man and super-duper smart, and clearly he has a terrific position now to let those great ideas flow.  I look forward to some of the cool things that will come from DevExpress now that Julian is at the technological helm.  (Not that DevExpress doesn't already have a lot of cool things for sale.)  Congrats to Julian and DevExpress for making this happen. 

There has been an interesting conversation going on in *.delphi.non-technical about Delphi user groups, why they may or may not be struggling, the role of such groups, and if they are even necessary in the age of the internet.  Here's my $0.02.  We have a small but dedicated Delphi User's group here in the Twin Cities.  On an average night, we get 8-10 folks out.  It's almost always the same crew, but we get an occasional new person every once and a while.  However, when we do a Delphi launch event, we can easily attract up to seventy or eighty folks from the area.  Our DUG attendance represents maybe about 10% of the people I personally know could or might show up.  In other words, there are a lot of Delphi folks out there, but they just don't come to the DUG.  And that is fine.  I'm a believer in the "The internet killed the user group" theory.  User groups were historically places to share knowledge, and since the Internet doesn't that far more efficiently than a User Group can, the utility of a user group has decreased.  I'm still a fan of user groups, though, and attend regularly.  We host our user group in our local Dunn Solutions office, and I'm a pretty regular presenter.    Anyway, the conversation in non-tech is really good.  Dennis Landi and Jim Cooper have had some interesting things to say.

Update 2 for Delphi 2006 is out.  It closes out quite few QC reports, as seen here.  Naturally, the release of any update with numerous fixes and improvements brings out the boo-birds in the newsgroups who didn't get their particular pet bug fixed. Alas, not all bugs can be fixed in an update.  First of all, any bug requiring an interface change will not be fixed in an update.  Breaking the interface side of things is far, far too costly for a number of reasons, not the least of which third-party controls would all break and have to be recompiled.  Secondly, I'm guessing that other bugs aren't done in updates because they represent too much of a risk -- that is, fixing them will have broader implications for the product than Borland wants to deal with or take a chance on with a released product.  Updates simply can't fix everything. If you have particular reports that you think should have been dealt with, mention them in the newsgroups and rattle few cages, especially when John Kaster puts out the "We might be doing an update so put in your QC reports" message.  An update will never please everyone, but it certainly is better than not doing one at all. 

The other part I always find amusing is then someone says that any given bug is "an easy fix".  I've said it before and I'll say it again:  There is no such thing as an "easy fix" in a product like Delphi. The effort that goes into fixing, verifying, testing and releasing a single bug is far more complicated than people realize. And to be quite frank, if you don't think so, then you aren't as sophisticated a developer as you think you are. 

Another thing people need to realize is that QC is filled with a lot of chaff.  First, when dealing with bug reports, you can't compare the number of reports fixed to the total number of QC reports there are.  Many of the reports in QC are feature requests.  Many of them are actually requests for technical support.  Many of them are simple expressions of frustration.  Many are just, well, really crappy bug reports.  For a better feel on what is really going on inside QC, then take a look at the stats application.  There is quite a bit of interesting information to be found there.  For instance, just on the opening page you can see that only about 16% of QC users have actually cast a vote.  I wonder what we as a community can do to raise that number.  Thoughts?

Well, it's been a while since I last posted. As many of you know, I had abdominal surgery about two weeks ago to take out my appendix and to work a bit on some irregularities in my intestines.  I'm glad to say all the biopsies were benign. I guess I'm getting to the age where you have to start worrying about that stuff.  I'm resting comfortably at home, and should be back at work tomorrow morning if all goes well. 

I'm very grateful to the large number of you who expressed your concern while I was away. It really means a lot. It's strange how you build up friendships based more on online communications than face to face time.  Many of you I know in person, but many I don't.  But I find that many of the people that I know well and consider close friends are those whom I've spent very little physical time with, but with whom I have spent lots of time online.  I suppose that is just a sign of the times.  In any event, thank you for your thought, prayers, calls, cards, emails, and even a few blog entries out there.  I really appreciate it.

I had no Internet connectivity while I was in the hospital, and so I've been spending the last few days getting caught up in the doings of the Delph world.  Of course, me being me, I'm compelled by forces beyond my control to comment on them. So, I'll be spending my morning blasting out a bunch of posts.  Enjoy, and always feel free to comment.

Chris Hesik, the Delphi Debugger Dude, found a pretty bad memory leak in the .NET framework which was effecting Delphi to the tune of about 400K every time you compiled your app.  I wonder if that isn't the main problem with the memory consumption that we see in D2006.

The fix will be in D2006 Update 2, which I would guess will be out pretty soon.

We all know those language geeks.  You know, those dudes that actually use and like vi or Emacs for programming.  I personally think they are nuts, but hey, there must be something to it or nobody would do it.  I'm a "to each his own, knock-yourself-out" kind of guy, so if that is what you are into, hey, that's cool with me. 

Oliver Steele has a great article called "The IDE Divide" that discusses the difference between what I assume is most of you and the "Language Maven" types.  Me?  I'm a Tool Maven -- the tool is as valuable to me as the language. 

Steele argues that it is hard to be a Language Maven and a Tool Maven.  I think it is somewhat true, but becoming less true as IDE's become "studios", i.e. as they host more and more languages.  Right now, you can use the same IDE, BDS, to do Delphi, C#, C++, and even VB.NET if you want.  IDE's are beginning to let you learn more than just one language, and thus it's easier to at least lean towards the Language Maven way of life.  Language Mavens, on the other hand, unless they start using a multi-language IDE, can't really lean towards the advantages of being a Tool Maven. 

In any event, it's an interesting read about something I've seen but never really recognized before.

1. Message boxes that ask a Yes/No question, but give you Ok/Cancel buttons.   I mean, come on. If you are asking a "Yes or No" question, how tough is it to tell the dialog to have Yes and No buttons? Not tough at all, that's how tough it is.
2. Ok buttons that are enabled when a dialog is not properly filled in.   This is basic User Interface design. If pushing a button will result in an error message, don't let the user push the button.
3. Non-sizable dialogs.  Argh. This one drives me nuts. It's especially galling when there's a list box or something that is so small you feel like you are looking at it through a straw.
4. Dialogs that don't remember their size and position. Related to the previous item. Sometimes a dialog is too small, and when I size it, I want it to stay sized. Sometimes it blocks stuff I want to see. It should stay where I put it, not where the developer thinks it should go.
5. Windows that insist on putting themselves in front when I am doing something else. This is absolutely, unequivocally the most irritating thing about Windows. It bugs me more and more each day.  I decide what I am looking at, not some shareware programmer from Wisconsin. If I am typing or otherwise working in a Window, no other application should ever be able to steal the focus, unless it's warning me that my house is on fire or something equally serious. 
6. File directory trees the size of postage stamps. Related to the issue above. Ever get one of those slightly older applications that won't let you size the directory lookup tree? With ever expanding hard drives and increasingly complex file directory structures, looking at your harddrive through a fixed size treeview that's only 150 pixels square feels like being shoved in the trunk of a Yugo.
7. Crappy error messages, especially when they are sentences and don't end in a period. "List index out of bounds". Great -- which list? The name or even type of the list is known. Tell us! "Error 332322". This isn't a problem because I have, of course, memorized all the error codes for your application. What's wrong with "In the DoSomething method, the length of the input from the name edit box is too short.  It needs to be at least five characters."  Feel free to write a novel in your error messages. Believe me, no one will complain.
8. CAPSLOCK keys. The person who thought putting the CAPSLOCK key above the SHIFT key and right below the TAB key should be rubbed vigorously with rough sandpaper and then placed in a bathtub full of lemon juice.
9. Unnecessary modal dialog boxes that I have to click when it doesn't make any difference. I love these. "You've done something really stupid. Press Ok to continue". Great. Thanks. I couldn't have made it through the day without that totally, utterly meaningless and pointless message.
10. Dialog boxes that have the negative answer on the left and the positive answer on the right. OK buttons go on the left. Cancel buttons go on the right. Don't put the Delete button on the left and the Approve button on the right. It's a gross violation of the laws of nature.

As I mentioned below, I had a really bizarre email with that editor from ComputerWorld.  It was so strange, I felt the need to write it up.  It ended up being really long -- I reproduce the entire email conversation -- so I put it in an article.  I still can't believe the guy can't tell the difference between "No Interest" and "No strong interest yet", nor can I believe the level of cynicism you apparently need to be a tech magazine publisher.  It seems that if a software executive says it, it must be false.

Oh well, read if you can stand it.

I think I'm just going to name this the Bruce McGee Memorial Post of the Week Award. 

Bruce wins yet again with this excellent post:

<rambam@bigpond.net.au> wrote:
  not much

Are you kidding me?  Are you seriously trying to dismiss David I as
some kind of feel-good hippy mascot that's acting as a mouth piece for
Borland management to hype up the sale?  Have you ever seen him speak?
Have you ever spoken to him?  Do you have any idea what role he plays
at Borland and with partners?  Everyone is entitled to their opinion,
but dude, you are way out in the bushes on this one.

I'm glad he's actively involved in the details of the sale, that he's
sharing some of these with us and that he'll be moving to the new
company.  His involvement in and enthusiasm for the venture adds
credibility.

Allen Bauer is involved with and commenting on details of the sale as
well.  Care to take a shot at him?  Or anyone else in DevRel?  How
about any of the R&D people who have spoken up on the subject?

Some person who apparently is too cowardly to use his (her?) real name in the groups, "rambam," took a totally ignorant, uncalled for, and totally out of line shot at the indefatigable, inestimable, and incomparable DavidI, the face of Borland for a good twenty years.  Truly, Delphi developers have no better friend than DavidI, and taking pot shots at him like the pusillanimous "rambam" did is beyond the pale.  Bruce, in his typically excellent fashion, knocked ol' "rambam" down about seventeen pegs.  Well deserved and well done.

So, all in one little thread on non-tech, we learn how to be a total jerk within the Delphi community, and how to take one down. 

SourceForge is running a little awards contest for different areas of their project database.  There are, as far as I can tell, three Delphi projects among the nominees.

JVCL in the Development Category

InstantObjects in the Database Category

TurboCash Accounting

Feel free to support the Delphi open source community by casting your votes for these projects.

Wow, this blog entry:

http://www.grimes.demon.co.uk/dotnet/vistaAndDotnet.htm

is sure getting a lot of play in the newsgroups and elsewhere.  This is adding fuel to the fire of the "Microsoft doesn't even use .Net, so why should I crowd?".  Folks seem to think that unless Microsoft immediately dumps their whole code base and rewrites everything in .Net that this proves that .Net is a fraud.

Overall, I find it entertaining.  Arguing about whether or not you should use .Net is a bit like arguing whether you should go to Harvard or Yale.  Both are great schools, and you should go where you want.  .Net is a cool, powerful framework with strengths and weaknesses.  Win32 is a cool, powerful framework with strengths and weaknesses.  So is Win64.  So is the Mac.  So is QT.  So are a number of other platforms and frameworks.  So, use what you want to use.  Don't use what you don't want to use.  It's really not that big a deal. 

Me?  I like ASP.NET alot.  It's the platform I've chosen to use for web development.  Now, if you want to use PHP or Ruby on Rails or notepad or AJAX or whatever, that's cool.  I don't much like WinForms, so I don't use that much.  I still use the good ol' VCL for Win32 for most of my client-side stuff. 

Microsoft of course promotes the heck out of .Net.  They want all developers in the world to use it.  Folks seem really bent out of shape that Microsoft isn't using .Net in every single new thing that they are building.  They are in fact using it for some new development (see SQL Server 2005), but as the blog post above shows, they aren't using it for their new OS, Vista.  Frankly, I couldn't care less what MS themselves use it for.  I only care what it can do for me, and so far, I like using .Net.  But if you don't, then don't use it.  I really don't see why it is such a big deal.  Long, tedious threads go on in *.delphi.non-technical about why .Net is (slow/fat/bloated/great/cool/whatever).  I find it all tiresome, really.  Like I said, if you want to use .Net, use it.  If  you don't want to use it, don't.

Then there are folks who rail at Delphi for spending effort on developing for .Net.  They argue that more time should be spent on (Win32/Win64/Linux/whatever).  Well, the notion that Delphi should ignore .Net is almost as silly as the notion of open-sourcing Delphi.   If you want Delphi to really fall off the face of the planet, argue that it should fall behind on .Net.  Just look at the reaction to the fact that Delphi is behind on the .Net 2.0 framework to see that.  The marketplace clearly demands that you keep up.  Fall behind too far and you'll die.  It's that simple.  Accept it.

And it's not like Delphi is ignoring Win32 development.  They aren't, not by a long shot.  In fact, Delphi is the only RAD framework and tool that is being updated on the Win32 side, for both Pascal and C++. Just ask Visual Basic developers about that.  Some took the "fact" that Microsoft isnt' using much .Net in Vista to argue that they don't even need to upgrade their Delphi version, with one newsgroup member proclaiming "....then why upgrade for GUI programming? I love my delphi 5 enterprise!."  Now if you think that you can be as productive in Delphi 5 as you can in Delphi 2006, then you haven't spent one second using Delphi 2006.  Delphi is a great Win32 platform,and the Delphi roadmap shows that this trend will continue for the foreseeable future.

So, please, let up on the "MS isn't using .Net to build MSWord, so why should I use it?" discussions, ok.  If the fact that Excel is still a native app leads you not to want to use .Net, then don't.  Just quit telling me why you aren't  using it.   

 

Yet another yellow technical press article.  Check out the sub-head on this article.  Great stuff.  It says:

"Borland plans to sell off its faltering JBuilder Java IDE business, as well as its Windows tools platform, including Delphi, **but so far nobody's interested**" (Emphasis mine)

Which is the exact opposite of what Erik Frieberg says later in the article:

"We have had, I think, over ten, I think the number is up to 12, interested parties right now looking at this. So, there's a lot of interest there."

Twelve interested parties hardly warrants the comment "nobody's interested".

I've had a running discussion in email with the editor in question  (I wrote him by using the "Feedback" link at the top of the page) and his rather bizarre position is that ten or twelve groups of investors showing interest means that "so far nobody's interested".  Go figure.

Just getting started with ASP.NET?  http://aspxdelphi.net/ is a great place to do so.  Lot's of good tutorials and other stuff to get you going.

I'd like to extend a big hearty thanks to everyone who came to our Delphi 2006 event today.   John Kaster gave a whirlwind tour of Delphi 2006 and gave a great explanation of what is going on with DevCo.  Thanks again, John.  We had about fifty folks at the event, and given the weather and the road conditions, I'd say that is pretty good.  Delphi is definitely alive and well in the Twin Cities!

NixBlox has a new blog and a new control, the ShadowBox.

Hey, we aren't going to let a little snow get in our way, are we?

John Kaster is on his way, and we are all set to go.  Borland swag, great discount coupons, good info, and good refreshments! 

See you tomorrow.

 

I think we all have a moral obligation to read and take note of all rants against the supposedly perfect Visual Studio.NET 2005.  (Link via Larkware's Daily Grind).

Eric Sink has a good article on the death of magazines for software developers. Given the death of Delphi Informant, and The Delphi Magazine going to online content only, this is interesting reading for Delphi folks.

The weather here in the Twin Cities is pretty tough, but I hope you all can still make it to our Delphi 2006 event tomorrow.

I'm on a roll here.

When you create Live Templates, you end up putting a lot of <point>s in them.  Well, of course that requires a live template!  (Deborah Pate posted this one first, but I took the liberty of making a few changes and improvements)

<?xml version="1.0" encoding="utf-8" ?>
<codetemplate xmlns="http://schemas.borland.com/Delphi/2005/codetemplates" version="1.0.0">
<template name="point" invoke="manual">
   <description>
     Creates a point in code templates
   </description>
   <author>
     Deborah Pate, enhanced by Nick Hodges
   </author>
   <point name="pointname">
    <text>
      Point Text
    </text>
    <hint>
      Text that will invoke the action for the point
    </hint>
   </point>
   <point name="pointtext">
    <text>
      PointText
    </text>
    <hint>
      Text that names the point
    </hint>
   </point>
   <point name="pointhint">
    <text>
     Hint Text
    </text>
    <hint>
    Text for the Hint
    </hint>
   </point>
   <code language="XML" delimiter="|"><![CDATA[<point name="|pointname|">
|*|<text>
|*||*||pointtext|
|*|</text>
|*|<hint>
|*||*||*||pointhint|
|*|</hint>
</point>|end|]]>
   </code>
 </template>
</codetemplate>

.Net properties are a little more restrictive than regular Delphi properties.  Properties must have explicit getters and setters (you can use the old field shortcuts in Delphi for .Net, but the compiler will just create the stub functions for you.)  They also have to be declared starting with get_ and set_ (and that's case sensitive by the way).

So, here's a code template that does all that grunt work of properly declaring a .Net property in a class.  It even automatically invokes class completion once the template is done.  This saves me a lot of time when building NixBlox controls.   Note that this declares the propname point, and then actually fills it in in three different places in the template all at the same time.  Live Templates just rock, don't they?

<?xml version="1.0" encoding="utf-8" ?>
<codetemplate xmlns="http://schemas.borland.com/Delphi/2005/codetemplates" version="1.0.0">
 <template name="netprop" invoke="auto">
  <description>
   Creates the proper declaration for a .Net property
  </description>
  <author>
   Nick Hodges
  </author>
  <point name="propname">
   <text>
    Property Name
   </text>
   <hint>
     The name for the property
   </hint>
  </point>
  <point name="proptype">
   <text>
    Property Type
   </text>
   <hint>
     The type of the property
   </hint>
  </point>
  <script language="Delphi" onenter="false" onleave="true">
   InvokeClassCompletion;
  </script>
  <code language="Delphi" delimiter="|"><![CDATA[property |propname|: |proptype| read get_|propname| write set_|propname|;]]>
  </code>
 </template>
</codetemplate>

 

... is easily the silliest idea that I've heard since the invention of Microsoft Bob.  Martijn Tonies agrees.

For the record -- I think Open Source is great.  If you want to open source your stuff, good for you. If you want to use Open Source, more power to you. 

What bugs me is when someone lobbies for the open sourcing of someone else's stuff.  I think it is a bit disingenuous to call for the open sourcing of Delphi.  To me, that translates into "I want a free copy of Delphi".  It also translates into "I want Delphi to die".

The notion that Delphi would continue to be developed as an open source project is ridiculous.  Open-sourcing Delphi would be the surest way to kill it.   The amount of lost revenue would be staggering and wouldn't allow for the continued employment of the R&D team.  No single organization would have a vested interest in Delphi anymore.  The project would simply languish with no one working on it. 

It's a silly idea all around.  For example, efforts like this are a complete waste of time.

Yes, this event is still on!  We are amazed and pleased with the response so far.  It isn't to late to signup and come.  The event is being held on 14 March 2006 at the Bloomingtion Marriott.  It will start at 0800  -- I know that is early, but John Kaster has a lot to cover.

The Bloominton Marriott is the one near the Mall of America by the IKEA store.  Here's the Google Map.  Remember, we'll have Borland swag, copies of Delphi 2006 to give away, and food and more.  Plus, it'll be a good chance to network. If you are even thinking of coming, please do.  We'll be happy to a big strong contingent of Delphi folks next Tuesday.

 

Joe Mele tossed his name out there as the potential CEO of DevCo.  Well, I'll throw my hat in the ring as well.  And not only that, I'm here and now putting out my official plan of action for my reign as the leader of the premier development tools company.

My plan is divided into three different courses of action:  What I'd do immediately, what I'd look into doing, and wild-ass ideas.

Here's what I'd do for sure right away, no questions asked:

• Put a full page ad in every issue of SDTimes -- I've asked about this before.  I can't understand why Borland hasn't been doing this all along.
• Put the help in a Moderated Wiki -- As I understand it, this is in the works,  but it really needs to happen.  We all together can build a better help file than the Delphi Docs Team.  No offense to them, but it's a big job.
• Release a "Delphi for .Net SDK". -- This totally needs to happen.  The lack of a public compiler is killing the third-party Delphi market.  Just killing it. 
• Double, no Triple, no QUADRUPLE the budget for what is now BDN -- BDN will be even more important to DevCo.  It's totally underfunded now. As the CEO of DevCo, I'd fix this right away.
• Double, no Triple, no QUADRUPLE the size of the documentation team. -- Somebody is going to have to manage all this inputs that we community members put into the Help Wiki.
• Provide a competitive upgrade price from any other development tool -- How are you going to woo VB6 developers without a competitive upgrade?
• Give every person in the company, all the way down to the cleaning crew, a copy of The ClueTrain Manifesto  -- Oh man, this really, really, really needs to happen.  Could there be a company that was less ClueTrain-ish that Borland?
• Plan, and announce as soon as possible, the new, improved, exciting DevCon -- Get the faithful together as soon as possible. 

These are the things I'd look in to doing.  I wouldn't necessary do them, but I'd throw them out there and look seriously at them:

• Find a way to aggressively pursue VB6 developers -- The field is still rich for the sowing. How can we get to these guys?
• Look into the right way to release a Standard/Personal/Free version of Delphi - This is hard.  Sell a low end version, and people buy it instead of the more expensive versions.  Don't have a low end, and you don't get students, hobbyists, and other new folks on board.  There has to be a way to solve this.
• Look into instituting a subscription selling model -- Smooth out the development cycle and the revenues.
• Sell licenses (not boxes, just licenses) of any version of Delphi -- There are folks out there using older versions of Delphi, and DevCo should make it easy to expand the size of their development team.  Or, put another way, if people want to buy something you have, sell it to them.
• Revitalize and release NDatastore -- This product would simply rock, and the entire .Net development world would buy it, not just Delphi users.
• Open up "DevCo Press", using on-demand printing technology -- On-Demand printing will keep costs way down, make keeping multiple titles available, and allow for more esoteric topics to be available. 
• Institute an Affiliate program, allowing people to sell Delphi from their websites -- the more folks selling and promoting Delphi the better. 
• Look at harnessing the power of the community
  • Club DevCo -- figure out ways for folks to earn points by being good community members.  Points can earn free stuff.
  • Free marketing materials to anyone who asks -- put together a package the ordinary folks can use to promote Delphi.  Give it to anyone who asks.
  • Free bumper stickers and other swag -- get the word out ASAP.
• Look into buying/merging with/closely partnering with FinalBuilder, AutomatedQA, Altova, Modelmaker and Castalia -- These folks make products that Delphi developers want.
• Add syntax highlighters for Ruby and Python for starters -- These to languages are just getting too popular to ignore
• Release a VS.NET version of ECO - ECO is awesome technology.  Consider letting the VS.NET guys in on it.  They'll pay. 

Here are a few ideas that are really out there, but worth at least thinking about:

• Integrating a VB6 personality into BDS -- why not?  That would be one way to get them involved.
• Create a personality for Javascript/AJAX development -- This is getting hotter and hotter everyday.  Someone will do this first.  It ought to be DevCo.
• Add a personality for Python and/or Ruby -- Do this after the AJAX personality.

Just a few thoughts for the new CEO, whoever it ends up being. 

I predict that DevCo will not be bought by any of the following companies:

• Apple
• IBM
• Microsoft
• Oracle
• Adobe
• BEAS
• Any other publicly traded company
  

In fact, I predict that DevCo will be bought by an organization that most of us, if not all of us, have never heard of. 

I have to snicker when I see folks in the newsgroups say that any particular release date for any particular product is tied to a desire to ensure that folks on Software Assurance get a specific release in a specific time range. 

Software Assurance is sold every day of the year, and people's contracts thus expire every day of the year.  Contracts are good for a year from the sale, not the end of any given calendar year.  So contracts are expiring all the time.  Thus, any particular release date for Delphi will be good for some people and neutral for others.  No particular date matters with respect to Software Assurance.  Some people will get their update a week before their SA expires, and some people's SA will expire a week before the release.

In addition, if you are viewing Software Assurance as just a "free copy", then you are thinking wrong.  Software Assurance makes sense for some people and doesn't make sense for others.  It makes sense for people committed to Delphi over the long haul.  If you pick and choose your upgrades, it might not make sense.  Over the long haul, renewing your Software Assurance every year, and not paying any attention to release dates will save you money.  Buying software assurance and then being upset about a particular release date is foolish.

Finally, I also snicker when I read about people saying "They should have waited a few more months and released a version for .Net 2.0".  Wow.  First, the .Net 2.0 update is a major undertaking, not something that they will be able to knock out over a weekend.  Secondly, it's ironic to hear people complaining about releasing a product earlier than they wanted.  If they release to late, they get complaints, and now, apparently, if they release too early, they get complaints. 

And finally, remember, you don't have to upgrade, so if they do release something on a date that you don't like, then you don't have to buy it.