More than one commenter to recent posts has trotted out the same tired old myths that, in their view, prevent [insert current owner of Delphi here] from being able to compete fairly on the pricing front. Some of these myths are as old as the Enterprise Customer thinking that is the real problem, and just as stale, not having been updated to reflect current circumstances.
But First: Where’s Sugar ?
The promised post on Sugar and Oxygene is coming, but getting up and running with it to the point of being able to blog about it has proven a little less straightforward than I had anticipated.
Now, for those myths.
Myth #1: Microsoft Give Away Their Dev Tools
This is simply not true.
Visual Studio is a little cheaper than Delphi and quite a bit cheaper than RAD Studio. The Microsoft NZ $ price for Visual Studio Pro is NZ$990 vs US$999 for Delphi, for example.
Worth noting is that if there is upgrade/renewal pricing for Visual Studio as a stand-alone product, I can’t find it in the current Visual Studio store.
Adding an MSDN Subscription makes Visual Studio significantly more expensive, as are the renewals as compared to Delphi SA, but you get quite a bit in return and being a content subscription rather than a maintenance agreement, what you get for your subscription/renewal is immediately tangible and predictable.
Yes, there is Visual Studio Express which is free. But if Visual Studio Express was all that any and all developers ever needed from Visual Studio then the very real, very expensive Visual Studio Editions wouldn’t even exist.
I wonder, is the MS developer community rife with Visual Studio Pro subscribers bitterly complaining that those Visual Studio Express users are some sort of free-loading ingrates that just wanted a cheaper edition of Visual Studio for themselves rather than paying their dues ?
Or do they perhaps see those Express Edition users as the future Pro subscribers ?
Myth #2: Microsoft Force You to Develop For SQL Server, Exchange etc etc
I had honestly not seen this flatly nonsensical assertion before, but since it seems to exist in some people’s minds then it needs to be examined.
The only thing I can think was in the mind of the commenter making this claim is that when you take out an MSDN subscription you get a lot of Microsoft technology included for the specific purpose of developing applications. SQL Server, Windows Server etc.
Would MS like you to develop applications that demand these technologies ? Probably. But how many people buy Windows Server/SQL Server specifically to run an application compared to the number of people that buy specific applications because they run on the Windows Server and/or SQL Servers they already use ?
In any event, are you forced to develop only applications that demand these other technologies ?
Of course not.
Myth #3: Microsoft Give Away Tools To Undercut The Competition
This I think is the crux of the problem, betraying the tendency to jealous resentment that seems to characterise the thinking behind these myths.
Microsoft really don’t care whether they “beat” Delphi or any other Windows development tool. What they do care about is that people are developing Windows applications, and to facilitate that they make developing for Windows attractive.
Microsoft want Windows developers to sustain Windows. A free/cheap edition of Visual Studio is one way to recruit people to that cause, and is something that has only really been available since other, free/cheap alternatives have appeared and become a significant presence. Not alternatives to Windows development, but alternatives that – from Microsoft’s perspective – risk recruiting developers to platforms other than Windows.
The same is true of Google with respect to Android of course, and Apple and their iOS and OS X platforms.
Myth #4: [competitor] Has Other Revenues, [current Delphi owner] Does Not
This one isn’t so much a myth (although in one respect it is *), so much a complete irrelevance.
Yes Microsoft has other revenues. Yes so do Apple. And Google, obviously.
What about RemObjects ? SmartMobile Studio ? And numerous others that could be mentioned.
Where are the significant other revenue streams for these companies ?
Certainly RemObjects has other products, but the idea that these are comparable or on a scale with those enjoyed by the likes of Microsoft, Apple and Google etc is of course just as ridiculous as the idea that the RemObjects developer tools “division” is a loss leader driving sales of their other products.
Yet they manage to produce a development tool for not one, not two, but three platforms (.NET, Java and Cocoa – more if you break out mono and Android from .NET/Java and consider iOS and OS X as separate from the Cocoa umbrella) and at a price that is lower than the “Mobile Add-on” which adds just two (and only two) platforms to a product which costs twice as much again at the outset.
This excuse for the pricing of Delphi started in the Inprise era and, like the pricing strategy that really started at that time, hasn’t really been updated since.
* The myth of course is that Embarcadero does not have other revenues. Where did they get the $30m from to buy the CodeGear products in the first place ?
Myth #5: Pay A High Price Once or Give Up A Cut Forever
This is a relatively new one, with the advent of the app store model.
There seems to be a perception that the choice is as simple as this: That we either have to accept a high, initial price for our developer tools or give up a significant percentage of the sales of our products forever, in exchange for a lower price.
This is nonsense of course.
First of all, the percentage taken from sales of apps through the Google/Apple stores is not in exchange for free/cheap developer tools. It is for services rendered.
Xcode and the developer program subscriptions are the same price (free and $100 respectively) whether you are developing for iOS or OS X. OS X has an app store but you aren’t forced to use it. As far as I know, you aren’t even forced to use the store for iOS if your app is an Enterprise app as opposed to a commercial one.
And even if it is a commercial iOS app or you choose to distribute your OS X app through the store, it’s not as if you don’t get something in exchange for the 30% of revenue that you “give up”.
You aren’t “giving up” that revenue at all. You are paying it, in exchange for:
- Distribution and update infrastructure
- Marketplace with product presence and ratings
- Payment processing
Is it really unreasonable to pay for these services ? Is the app store model uniquely draconian ?
If you sell products through any form of distribution channel, the eventual customer will be paying more to the reseller than the reseller pays you, representing a cost of sale per unit to you.
Marketing, or even simple hosting for promotional content (or direct distribution if you are cutting out the resellers) is rarely, if ever, free and will often have a minimum base cost to be recouped through 100% of some initial number of “sales”, as well as scaling costs associated with increased volumes (traffic) if you do particularly well. Costs which you can average out per transaction only after the fact and may never recover entirely if this channel doesn’t prove effective.
Payment processing is never free. If you take credit card payments or even PayPal payments, there will be fees to be paid.
But further than that, even if you pay a high initial cost for your tools you will still have these costs to deduct from your revenues if your product is to reach any sort of market.
And again, there are other companies that manage to produce tools at a reasonable price that do not “clip the ticket” on any sales arising from the efforts of their customers. The benefit they gain is by those customers continuing to be customers.
Myth #6: There Is No Easy Solution
Yes there is. Possibly many different solutions in fact. But they require a change in thinking. A different mindset. And that is where Embarcadero seem singularly unable or unwilling to apply themselves.
At the highest level perhaps, the various options fall into two distinct groups:
- Free/Cheap entry level product properly distinguished from the higher priced editions. Let’s call this the “Visual Studio Model”
- An upgrade subscription model where you pay a single annual fee entitling you to upgrades to a perpetual license. Let’s call this the “Xcode Model” (since with Apple your ability to productively develop is contingent more on your annual subscription to the relevant Developer Programs than it is on Xcode itself, which is free).
Yes, Visual Studio with MSDN is essentially a subscription model as well. Would you look at that ? Microsoft actually have feet in both camps. It’s almost as if they actively think about pricing and how to make it work for them, instead of just relying on some uniquely comfortable position and sticking dogmatically to an outdated mindset!
Next Time
This post has been what my detractors would no doubt characterise as “negative”, despite the fact that the intention is not to be negative toward Delphi/Embarcadero but to illustrate why some of the fatalistic thinking in this area is flawed.
For that is the real negativity I think: to believe that nothing can be done and that people just have to suck it up, get onside and stop complaining.
So next time (or possibly the time after that, depending on how I get on with Sugar) I shall revisit my proposal for a Community Edition of Delphi, and look at exactly how and why I think Starter fails to meet this need and what I think could be done to fix that.
I’m in pretty much the same boat as you with an employer that uses Delphi extensively (but doesn’t really have a need for the newer upgrades) and my side-line interests that can’t justify hundreds of dollars for upgrades. I’ve been reading (and agreeing) with most everything you’ve written lately and also decided to get Oxygene for personal exploration–although I haven’t had time to get into it much yet.
This blog was interesting to me because I had not heard of most of these myths before. It sounds like Delphi proponents are reaching further and further to the absurd for reasons to continue their feelings of superiority–or perhaps hiding their anguish at being unable/unwilling to shift their loyalty to a different environment.
You dispelled these arguments very well. The only one I might question is the first one. With VS Express 2012 (and possibly others), you can actually write web services and database applications with the full feature set of .NET available. I’ve actually wondered how Microsoft can do this, but like you said, there must be enough of a reason to get the license that it still sells well. I guess the support for TFS, 3rd-party add-ons, unit-testing and refactoring, among other benefits which I’ve read about are enough.
But it really does make one wonder why on earth Embarcadero doesn’t make an entry-level Delphi product to entice people to at least try it out. As it is, they’ll try Lazarus instead and then once you’ve tried that, well why buy Delphi?
Yes, you can. But when it comes to commercial web services where you will be having invest in providing your web service infrastructure, and if you reach that point you almost certainly will also invest in upgrading to Pro editions for the additional benefits you mention. Similarly for any database application, if it is at all successful and/or complex or large, then you will need a bigger team to support and maintain it going forward, and that’s where – again – those “Pro” features come in.
Beyond the simple issue of “scaling”, I would say that a key feature that makes Pro appealing for, well, a pro, in the modern world of increasing diverse but inter-operating platforms, is the integration of the different capabilities, removing the need for separate Web, Mobile and Desktop development environments (in addition to things like plug-in support).
True. I haven’t gotten very far into VS and would likely hit those walls soon if I did. The point that EMB should take note of is that I can write many types of small applications with VS and can really familiarize myself with the environment for free and WITHOUT A TIME LIMIT. I can’t remember how many times I’ve downloaded a piece of software with the intention of really getting into it during the trial period and something comes up and I don’t get back to it until there’s only a few days left. I keep VS Express installed all the time and occasionally open it up to try something.
#1: VS free editions exist because of #3. They didn’t exist until Linux got real attention, and several free dev tools arose, especially the LAMP stack. First MS only released the compilers, then the whole IDE – the aim was to keep people on Windows, not sustain the Pro sales. Probably with that move it lost some Pro sales, but for MS was more important to sustain its true revenue generators, which are not dev tools. Just look at the numbers.
#2: Microsoft doesn’t force you, of course, but its tools *are designed* to work best with its complete offering. Do you use asp.net with Linux and Apache? Do you use C# with Firebird? Obviousvly, not.
#3: Exactly, MS sustain Windows sales and all its ecosystem: and why? Because it is a huge revenue stream Emb has not.
#4: Are you comparing Oxygene to Delphi? What IDE Oxygene use? What framework Oxygen use? It’s just a .NET compiler. Emb has a fairly more complex tool to develop, and thereby far higher costs. Those expenses should be paid somehow, don’t them?
Emb has Thomas Bravo behind, but I guess it’s not an equity who wants to lose money delivering free dev tools. It also has its database product business, but for what I see they are inferior for example to Quest’s ones (and Quest was bought by Dell). I think TB spent $30M to make more money from Delphi, not to lose more cutting Pro sales.
#5: Apple get a lot of money from its “service”, which is compulsory and thereby not a service, and paid dearly – because there is no competition. Sure, internal apps can be delivered outside the store, but it also cost $299 a year – and it’s only for internal app. You need another license for external ones.
Anyway developers were happily selling their apps on the web before Apple forced them to pay it for selling. And no credit card circuit asks you 30% of your sales. Rating system? It can actually kill your app.
Would you like it for your Delphi application? And still remember Apple has huge revenues from its hardware business. Yes, I wish Emb adopts the Apple model. High-end, well designed, well implemented but expensive products…
#6: the problem with the “free/cheap” edition is that developers like you don’t accept any limit. You want VCL source, you want components, etc. etc. basically you want a Pro edition for free. And if fact you forgot to say what the “cheap/free” edition should be, and what it could have not.
You and other still live in the myth that a powerful free edition of Delphi is feasible, but I’ve still to see a business model that could work for Embarcadero, not Apple, Google or Microsoft. Sure. if they just make just an ObjectPascal compiler for VS and.NET maybe they can sell it at $99, but is this what you want? Sorry, but your business model looks the Underpants Gomes one:
1- Give away software for free
2- ?
3- Profit.
Maybe the reason is that you still think – or pretend to – that a free version of Delphi would bring back a lot of developers, and many new younger ones. What you fail to understand – or pretend to – it’s that it’ll never happen. There are plenty of free/cheap tools which remain small niche markets, because they’re not fashionable. Pascal and Delphi are no longer fashionable, and no matter what the price is most developer will keep on ignoring them. Only gray bearded actual Delphi developers would be happy of a free Pro version. It’s time they admit it.
Just as MS considered it important to sustain their money making business, so Embarcadero need to sustain theirs. And in their case, their dev tools are the business that their dev tools need to sustain. It frankly defies any sort of rational thought to suggest with a straight face that being if not the then one of the most expensive options whilst – by your own admission – providing technology which is far from perfect, is the way to go about doing that.
Lowering the barriers to entry, growing the user base… these are what Embarcadero need to do and people like you frankly need to grow up and accept that the people coming after you might get a better deal than you did. Because the important thing is that there are any such people coming after you at all, not whinging that they are cheap-skates.
Alternatively you can support Embarcadero as they turn away potential new users, safe in the knowledge that nobody is getting a better deal than you – how dare they?! But nobody much will care either because Delphi will have become a footnote in the history of development tools.
As for some of your other comments. With regard to Oxygene being:
Um, no. it isn’t. It is also a Java compiler which can produce native Android applications as a result, and an LLVM compiler that can produce native Cocoa (and Cocoa Touch) applications. Note that there is no need to qualify or dissemble the definition of “native” in either case.
I think perhaps you need to do some research. 😉
As for the notion that I just want a free Pro edition and haven’t said what I think a free edition shouldn’t have, well frankly this is laughable given that in this very post on which you are commenting I linked to a previous post that went into exactly those ideas in a great deal of detail and furthermore promised to revisit and revise that soon.
You are not doing yourself any favours with this comment I think. You are simply confirming that your position is one seemingly based on prejudice and not on a consideration of facts, given that you are seemingly unequipped with a number of key ones and have an entirely closed mind when it comes to drawing any conclusions from those that you do have.
I just upgraded Delphi Pro and FireDAC … not saying I do have a specific need at the moment.
Don’t tell me that MS would be driven by altruism. But you still get the VS Prof. without MSDN.
http://www.microsoftstore.com/store/msde/de_DE/pdp/productID.283471200
From what I understood is that you have to buy VSXXXX every time again. Considering a yearly release cycle that has already been announced it’s similar to Delphi Enterprise.
I agree the offering is hidden, because in the Visual Studio product pages MS describes that the primary way to receive the VS Prof would be via MSDN subscription. MS always try … it’s not the first time. They try to push you into MSDN
a) You use the things you paid for
b) Afterwards they use the trick – sunk cost which is a psychological phenomenon (considering former investment costs). 1300 EUR vs. 856 EUR (list ex VAT*)) renewal = little more than the full price for VS Prof (inkl VAT) . Azure is no longer free. Maybe in a next step they don’t allow you to install VS as often as you like. The license said, iirc, one developer several machines. I think VS Pro standalone is limited by times of installation too. I am not sure about this.
*)I found offering of VS Pro MSDN for 2 years … at very much the same price.(1300 EUR). But a fair price for an upgrade is something different. If you compare editions only the hire editions offer a 1/3 (upgarde/first time purchase)
http://www.microsoft.com/visualstudio/eng/buy.
They cannot force you to develop for one of their products but if you are a partner you have pass tests. Otherwise you don’t get their products in an affordable way. Maybe licences installed at a customers site for a certain period of experiencing your solution or comparing the solution to competitive products.
The whole pricing is aiming strong towards pushing candidates attracted by free versions towards the partnership or a at least MSDN + renewals per seat and max 2 installations.
Honestly for a developer who at least indirectly drives Microsofts license business that’s a poor perspective.
You will never get anything that is a business opportunity still for free from MS.
Why does VS Pro compare to Delphi Enterprise
a) Yearly release cycle with quarterly updates announced – so most of the fixes will be shipped ‘in the next quarter’ but always the next.
b) A Windows Server is about 500
c) Client Operating systems – you have to buy anyway you cannot register them as of as you like – or add MAPS. (then the whole bunch gets expensive in summary)
At this moment the whole open-source infrastructure offerings combined with Delphi are already a lot cheaper. Assuming the ARM stuff works you are already a lot better of. Include Oxygene and you have every possibility or use SharpDevelop and C#. Assuming you are looking for a better alternative to Winforms.
.net Desktop is/will be about WPF … there is no doubt about that. Any kind of XAML but never something like VCL in fashion of a more advanced or improved MFC.
Another (perhaps) viable option exists. It might be possible for EMBT to keep the current pricing model, and still win back users who have departed or (as in my case) have not upgraded recently. (On the issue of “recently”, I upgraded to XE a bit more than 2 years ago, and have now skipped XE2..XE4, and soon XE5. Pretty short delivery cycle, to be missing anything I would want to buy!)
OK, here is my notion:
– maintain pricing
– seriously decimate the long-standing defects
– deliver high-quality *printed* documentation
– maintain a commitment to near-zero defects
When Delphi 1 shipped, it was notable for a few key items:
– a really good visual design model
– a strong set of core components (despite lacking serial I/O)
– the excellent exception handling
– the feel of being almost entirely free of defects (built with Delphi!)
– terrific documentation
Barring application of significant counter-force, entropy always increases.
We now have (as of XE, which is my latest):
– a long list of defects
– no printed docs, and embarrassingly thin wiki
– no apparent commitment to quality
And users are departing. Hmmm…..
You still fail to undertand a free/cheap edition of Delphi *won’t bring new develoepers*.
Why should it? As long as Delphi is perceived an old legacy subpar tool struggling to survive, managed by a mostly unknown company, and delivering so-so features and quality, why people should use it instead of VS, GCC, XCode or whatever? That’s the question mark in the business plan above – and noone of you explain what would brough Pascal back to its glory days. Just the price? C’mon!
You have prejudices, that what Delphi only needs to get back to be a widely used tools is a free version, and that’s not true at all.
You still believe there are lot of developers just waiting to use a cheap or free Delphi, I bet there are none, or almost. Maybe you should start to talk to young developers, they don’t care about Pascal and Delphi at all. Because it looks old and legacy like COBOL not because it looks expensive. They don’t care about xplatform when they can do already with C++, Java or Python, they don’t care about VCL or FM controls when they’re using Javascript – you’re dreaming about a world that doesn’t exist.
Do you know what happened to the once mighty Watcom C++? It’s now an open source (free…) project no one no longer cares about. Becoming free didn’t save it.
Please, explain me how could you sustain your dev tool business giving away the same very tools for free. MS & C. sustain their *profitable* business giving away *other* tools – not the same ones. Do you give away your main software for free to sustain it?
The free edition of JBuilder started to kill the product even before Eclipse was released – it was enough for many Java developers they never minded to buy the non-free versions. Why do you believe Borland bought Visibroker and later Starteam and Together? Trying to expand the portfolio and find some high-revenue products to offset the lack of income from the dev tools.
And the free edition didn’t save JBuilder from being forgotten as a Java dev tool.
PS: I know about the other Oxygene version, but we are talking about a Pascal tools – not a Java or other languages one. Let’s see how much it will be used to deliver Android and Cocoa apps. And still, it’s just a compiler, VS IDE, no framework on its own. Don’t compare oranges to apples, please.
Free/cheap editions will bring users and – perhaps more importantly – help retain users. This is an evidence based opinion since I know of people who have walked away from Delphi precisely because it is too expensive.
Where is your evidence that users will not be attracted ? I am not asking seriously for such evidence because it is asking for “proof of a negative”, but that alone should be enough to make you question the strength of your conviction in this matter. How can you possibly know what won’t happen ?
I was a Watcom C++ customer, but as much as I loved it it always felt like an also ran, even if it was “mighty” in some quarters. Perhaps that was because I came to it when it was already in decline (it would have been around 1994/5 I think). In any event, I think going free was the effect of it’s eventual failure rather than the cause.
Aha, and now I think I see what you are up to…. a Community/Standard edition would serve to postpone the day that Embarcadero have to give up and you then get an open source and free version of Enterprise/Architect! Very clever! Well played sir. 😉
On the off-chance that this isn’t your game, please stop latching on to the “free” part of free/cheap. It’s a straw man.
If you actually bother to read what I have said on the matter, as opposed to simply assuming my position based on your own prejudiced opinion of me, you will see that a free edition is only one part of what I believe is needed, the other part is a small reduction in the Pro price and the creation of a Standard edition below Pro.
The “sustenance” this would provide is in ensuring that there will continue to be Delphi users in the future, instead of an ever dwindling group asked to pay ever more to keep that reducing and aging population going. The fact that you don’t see this beggars belief.
Clearly you don’t know that much otherwise you would know that Oxygene for Java and Oxygene for Cocoa are both also Pascal even to the extent of being able to share code between all the supported platforms (.NET, Java incl Android and Cocoa [OS X and iOS]).
Still some research needed.
Your other points:
Yes it’s just a compiler. So is Delphi (it comes bundled with a bunch of other things, but take those things away and all you are left with is… a compiler. [GASP!]). So how about those other things… ?
The IDE – yes Oxygene is hosted in Visual Studio rather than a home-brew effort. But guess what? That’s actually not a bad thing at all.
The frameworks – nope, Oxygene doesn’t have it’s own framework. But then it doesn’t need one in the same way that Delphi did, because the target platforms all have very rich frameworks of their own.
Yes Delphi has it’s own IDE and yes it has a framework (two actually). But these things are largely already bought and paid for by past license and maintenance revenues. It doesn’t cost to already have an IDE, it costs money to create one. There is some on-going maintenance cost of course, if you are using your own, but since one of the biggest complaints about Delphi in recent years has been the lack of attention in this area it is hardly realistic to suggest that the extra cost of Delphi is justified by the quality in these areas.
Or more simply put: Grocery based complaints aren’t really valid. 😉
I think I have to side with LDS on the core issue of new users.
>Free/cheap editions will bring users and – perhaps more importantly
>– help retain users. This is an evidence based opinion since I know
>of people who have walked away from Delphi precisely because it is
>too expensive.
Someone who walked away from Delphi because of cost isn’t the same as a new user (nor did you say it was). Of course, once they walk away they’re going to discover other tools and it’s questionable whether price alone will bring them back. Market research shows that most people don’t look to switch products unless they’re unhappy. If they’re happy with product A they’re not going to look at product B even if it’s better. This is why Bill Gates famously said he’d rather people pirate Windows and figure out how to monetize them later than have them use another OS. He knew that users rarely come back.
But for Delphi to remain viable it needs to attract new blood, not merely slow the defection of existing developers. Lower price won’t do that.
>Where is your evidence that users will not be attracted ? I am not
>asking seriously for such evidence because it is asking for “proof of
>a negative”, but that alone should be enough to make you question
>the strength of your conviction in this matter. How can you possibly >know what won’t happen ?
I could write an essay on this(and probably have on the EMBT forum and Reddit), but simply look around: no more commercially published Delphi books, no magazines, few physical conferences (none in America), few if any user groups left, almost complete disappearance from academia… people simply aren’t interested in Delphi anymore. Look at EMBT’s own marketing material on their site – absolutely no comparison to other products, unlike most other development tools (even RealBasic and Eiffel have material targeted at users of other products). In an interview with The Register, EMBT’s CEO declined to answer a question about what Delphi had to offer that C# didn’t because C# is “a .NET IDE” and “we don’t consider it a competitor”(!!!). When the interviewer rephrased the question in terms of VS C++ instead, CEO Williams hand-waved the question away again, saying that product “wasn’t their focus” (I’m sure C++ Builder users loved that). Delphi is way behind the technology curve today and younger folks are naturally going to want to learn/use something up-and-coming or established today (ideally both), not invest in a technology that’s had a significant decline from its hayday.
COBOL has a free version now – open source OpenCOBOL – and COBOL even added object-oriented extensions (!!!) in 2002. I haven’t seen COBOL re-capture the marketplace though. As someone who spent a little time programming in COBOL in the early 90s – a free version would not be enough to get me to use it, OO extensions or not. A paid version (where I was paid to use it) wouldn’t be enough. 🙂
>I was a Watcom C++ customer, but as much as I loved it it always felt
>like an also ran, even if it was “mighty” in some quarters. Perhaps
>that was because I came to it when it was already in decline (it
>would have been around 1994/5 I think). In any event, I think going
>free was the effect of it’s eventual failure rather than the cause.
True enough, but do you see the relevance here? Going free didn’t revitalize Watcom C++. If better products exist and the community has “moved on”, going open source won’t save a dying product.
>you will see that a free edition is only one part of what I believe is
>needed, the other part is a small reduction in the Pro price and the
>creation of a Standard edition below Pro.
Jolyon, you’re going to have a heck of a job convincing me that that’s all that’s required. What about the jobs? The young users? The books? The courses? The libraries? The *ecosystem* by which a language lives or dies? Without a plan to jumpstart that, it’s a doomed effort. It would be like trying to revive a person with no heartbeat in a room without air. Getting the heart started again won’t do you much good without dealing with the reason they ended up that way in the first place. Without fixing what’s caused the decline of Delphi, you could attach $100 bills to Delphi DVDs and not appreciably grow the user base.
Let’s put it this way – do you think if I dragged a group of young C++, C#, Java, Ruby and Python developers to this blog you could get them to use Delphi merely by lowering the price?
>The “sustenance” this would provide is in ensuring that there will
>continue to be Delphi users in the future, instead of an ever
>dwindling group asked to pay ever more to keep that reducing and
>aging population going.
You’re going to sell a language with “features” like no automatic memory management, no type inference, no design by contract, no functional programming features, lots of code redundancy (aka “separation of interface and implementation”), closed source, marginal web support, and locked in to one IDE which can’t run natively on OS X or Linux with an environment with no conferences, books, academic use, significant libraries or JOBS to a group of young people by making it cheap?
This is 2013. Kids have not had it like we have. They’ve grown up with the ability to slap Ubuntu on a computer and add gcc (C and C++), Java, Eclipse, Mono, MonoDevelop and C#, valgrind, Ruby, Rails, Python, Qt, Apache, MySQL, SQLite and PostrgreSQL and develop whatever they want (after having learned with free books and online courses) for free, with complete control over their system and software. Perhaps more importantly, they’ve been able to LOOK INSIDE their compilers and interpreters, learn how they worked, and even change them and contribute back to the project. Heck, Python made someone a release manager when he was FIFTEEN YEARS OLD. This is first generation growing up with open source. They’ve probably never even used a compiler they had to pay for and everything they’ve used outside of VS will have been open source. How do you sell the idea of paying for a compiler (cheap or not) with a handful of bean counters and suits in CA telling them what they can do with it, closing off access to the source, not fixing bugs, not using their ideas/letting them submit patches, etc? That’s as antiquated an idea to them as the horse and buggy are to us.
Oh, and just try showing possible new users of Delphi interfaces with GUIDs and see if they can keep a straight face. They’re liable to ask if the classes require bar codes. 😉 So much of what we DO have in Delphi is like 21st century car parts bolted onto a 1970s automobile. That’s why languages go into and out of favor. Sometimes you just can’t bolt stuff on anymore and you have to start from the ground up with a language based on the latest ideas in computer science.
Yep. You seem to be on the same page as LDS in so far as accepting that nothing can be done to revitalise Delphi. I was hoping that there might be a last chance and I still think that price is an issue. Not the only issue, obviously, but it is an issue.
It seems to me that when a problem has no simple, single solution then there are two things you can do: Either address the things you can address and hope to be able to address the other things eventually, or simply throw your hands up and cry that if you can’t solve everything then there is no point trying to solve anything.
I find it ironic that I – repeatedly accused of being negative – seem to be one of the last, desperate optimists. 🙂
But after yesterday’s presentation I am now resigned to accepting that Embarcadero themselves are more on your and LDS’s page than mine and there is no point any more in trying to convince them of the error of their ways. It’s not that they don’t see it, they simply don’t seem to care. And, that being the case, why should we, and why should I ?
It’s time to move on.
Now I feel all sad and guilty. 🙁 I appreciate your ideas and I think there’s definite merit in lowering the barrier to entry. I do believe that if a solution is possible it’s going to involve thinking long term and being willing to trade immediate short term profit in exchange for growing the user base. And I did feel the way you did until relatively recently when some issues with the language itself and some actions on the part of EMBT sort of hit me all at once.
Let’s put it this way: revitalizing Delphi, if possible, would probably require a coordinated multi-pronged attack on several fronts, of which your pricing/SKU changes would be an important part.
In regard to interpreting Embarcadero, I think it’s a combination of things: it’s quite possible the business management doesn’t care (focusing solely on meeting their VC owner’s immediate quarterly number requirements by any means necessary and not really being attached to a product that was forced on them by their owners that they never asked for, and the old hands on the development end (including a certain literal greybeard) being in complete denial. And the TeamB stalwarts – they’re in the biggest denial of all (believing that there literally are 3 million Delphi users, it’s bigger on Windows than Java, believing the spin that World Tour attendance is up – technically true given they added more dates but quite a fib otherwise – etc.).
An impression I have is that the development side repeatedly tell themselves that the blogs, the forum, etc. are not representative of the “real” user base (something that’s actually been said to me, come to think of it). I think this lets them hand-wave away all the complaints, without stopping to wonder why no other language’s users seem this unhappy. Bauer basically chastises people who show up on his blog and start complaining or taking issue with him, saying this “isn’t the way to start a conversation”. When one user described for me the pitiful turn-out at a California world tour event in what’s supposed to be the tech capital of America and EMBT’s own backyard (including EMBT selling raffle tickets to pay for the pizza they provided at the event!!!!) a forum moderator I’d never heard of before appeared and essentially demanded that the poster now list all the good things at the event! There’s this weird sense of them being entitled to some sort of praise that I don’t get. The Delphi team’s opinion of how they’re doing seems to differ wildly from the users’, so there does appear to be a disconnect that just can’t be breached. But the short of it is this: they’ve expressed absolutely zero intention to listen to us or any of our suggestions. They often seem to do things and then count the pitchforks to see what they can get away with. It’s the same old faces and they seem determined, almost out of spite, to continue doing the same thing and expect different results. We never got an apology over the state of FireMonkey in XE2, the closest being Marco using the phrase “rough edges” at one point. They seem blissfully immune to criticism or to the fact that to some degree they’re openly laughed at.
I think you, Eric, Arnaud, and a host of others could run the business and technical ends of the Delphi product a lot better than it is now and quite possibly save the product. But sadly you don’t and I guess I am guilty of kicking it when it’s down but nothing seems to change. I’d love it if they’d at least learned one lesson and realized they need a rock-solid, bug-free Android release in XE5 to remove the reputation for poor quality, but Embarcadero Insider suggested the implementation is incomplete, so we’ll see when it gets released. If EI is correct, I’m sure they’ll remain completely deaf to complaints as usual.
Maybe it’s not kicking Delphi while it’s down so much as… being mad that one can see the potential it could have had vs. where it’s been taken today.
Very well put! And apologies for making you feel sad and guilty. That was not my intention. 🙂
Unfulfilled potential – singing to the choir, but I’ll not elaborate on the other potential goldmines mismanaged and exterminated to satisfy ego.
There appears to be a group of ex- and current-EMBT personnel on Stack Overflow who follow each other around, voting-up each others’ every utterance and co-ordinating attacks on anyone outside of their little clique who may for instance post a “code-only” solution on the basis that code-only without a comment is “weak.” Fact that it contains //comments within the code is irrelevant – and they take joy in voting-down and voting-to-delete any contribution of which they don’t approve (‘though they’ve been known to object strenuously to similar behaviour directed against their pontifications.)
The main perpetrator has been known to respond to a question from someone who was obviously struggling with English, never mind Delphi with a canned response that I, as a native English-speaker with 40+ years’ IT experience and a humble Bachelors’ IT degree had problems deciphering, never mind relating to the problem raised.
What’s really sad is this : If those pundits spent their time publishing their solutions under EMBT’s banner, then it would actually be useful to the user community. Revitalising perhaps. Solutions to problems raised. You know, what the community has been asking for for decades. The downside as far as EMBT is concerned would be that there would then be an expectation that the solutions formulated would be rolled into the product and would be an admission of problems with prior versions.
Can’t see it happening though. There seems to be a stubborn refusal from EMBT to incorporate any “outsider’s” ideas like the fixpacks and extensions designed and kindly published by well-known (and appreciated) third parties. Unfulfilled potential again. Maybe the same attitude applies to any clever ideas that EMBT insiders have. Not a management idea, so must be excluded. Do as you’re told – don’t try to have any ideas yourself, that’s management’s prerogative. And hence the basic VCL/RTL toolkit remains unmaintained – simply beneath the managers’ dignity to admit that there are problems and allot the personnel to fix them. Met that elsewhere …
Perhaps this pursue-the-figures attitude is endemic at EMBT. To return to the pricing issue – EMBT does not make any distinction between upgraders who jump four or more versions and those who faithfully collect each version. Inevitably, there’s a “special offer” as the next release – or perhaps end-of-quarter approaches. Lowest price can be obtained by waiting until the periodic panic to achieve the figures hits – and not an attitude confined to EMBT either.
.net does not exist because of VS. It’s the other way around. That’s the comparison. For the developer it does not matter because Windows and .net already exist. You don’t pay for the OS and .net because ‘billions’ of others do. Things EMB cannot do cannot be compared. If it’s not enough no reason to compare. So every comparison especially the price can only make sense if applicable alternatives can be compared.
The EggLayingWoolMilkPig – jack of all trades – does not exist.
Why is Oxygene not Pascal? JVM, .net, Cocoa … is not about the language.
I personally think that the ‘procedural’/. C/C++ interpretation of OO way could be an attractive alternative still. Giving Delphi away for free or a lot cheaper will not expand the customers in the west beyond those who are satisfied with what exists already. Never the young not at a large scale.
G.I.G.S. – grass is greener on the other side. But every solution that serves a larger scale does come with trade-offs that maybe hurt you maybe not.
On the young. They simply look for something different and not only the young.
Mostly agree, except that Microsoft _DOES_ give away their tools for some time if you are a new company (BizSpark). Not only that, they provide free unrestricted versions of Windows, SQL Server, Office… just name it.
Of course, after a couple of years, you “graduate” from BizSpark and have to pay a small fee and then start paying for your software, but the MSDN subscriptions are killer deals (you can get basically ALL of Microsoft software for less money than ONE of the Emb enterprise tools!)
I’m working a lot with MS tools lately and I must say they are great. Every minute I miss the coherence of developing a desktop Delphi application inside the IDE, but if Embarcadero doesn’t get their game straight they may go the way of the dodo. I really hope this doesn’t happens because I’m invested in, but I’m resharpening (no pun intended) my skills just in case.
This is true, but why do Microsoft do it if it is such a dumb idea for everyone else ? The fact that they can afford to is only half of the equation. The fact that they choose to do it when the “received wisdom” is that it is giving product away is the point that needs to be considered.
Microsoft can afford to do it. Embarcadero perhaps can’t afford not to.
Certainly Embarcadero can’t afford to give their products away, literally. But what they do have to do is invest in building a user base for the future, otherwise their gravy train is simply headed for a cliff.
Microsoft have their BizSpark (which sounds a lot like what used to be called ActionPack) for a very good reason. If they did not have this option for new/small companies, then those companies would find free/cheap alternatives. Microsoft are looking ahead to when even just some of those new/small companies grow and need more than BizSpark offers.
If they are already comfortable using Linux, OpenOffice and MySQL etc, then they will likely stick with those as they grow. And equally if they are comfortable with MS Office and SQL Server, then this is where Microsoft get the return on their investment.
The companies that do not grow beyond the needs of BizSpark were never going to be big earners for Microsoft in the first place so there isn’t much lost in those cases. And having even those companies still on-board and likely recommending BizSpark to other new/small companies who still might provide a return is better than dismissing them as not worth the bother/cost at the outset.
Rather than any similar forward looking approach, Embarcadero – and those advocating their approach – seem more inclined to keep their head buried in the sand.
Retain users? What’s the advantage of retaining users who are not willingly to pay? Those are the customers you have to lose!
At least you confessed that those who need a free/cheap Delphi are those gray beard developers who can’t – or don’t want to – pay for Delphi anymore.
Should Emb target this group? No, there are no revenues to be made there. It would accelerate the day Emb has to give up because the business is no longer profitable.
What? A cheaper Pro, am even cheaper Standard (of course with little or no limits), and a free one. C’mon, what is Emb? A charity for old poor Delphi developers?
Smart business plan – Emb would go bankrupt before they could say “XE6”.
My evidence that new users would not be attracted is that every young developer I interview or hire doesn’t know Pascal, doesn’t know Delphi and doesn’t bother to know them. If you ask you to learn it, they feel somewhat demoted.
Just say “Pascal” and their eyes go to the ceiling. You could say “COBOL” or “VB6” and get the same reaction. Give them a free license of the Architect SKU and they will leave it dusting on a shelf while playing with everything else is fashionable today.
The actual fashion is elsewhere, and people don’t like unfashionable things even at sales price. Only nostalgics do.
I have a ten years old application using PascalScript as its scripting engine. Next version we will need to add Python because more and more users don’t know Pascal and don’t want to learn it. They want a language they already know.
What’s your evidence a free version would bring many new developers? It won’t bring back developers who left Delphi years ago, all the ones I know are busy writing applications with other technologies and would not look back.
Emb could deliver a cheap/free edition while investing little in the product, and make happy some old Delphi addicts for a while, or can change its attitude, invest heavily in the product, and target professional who can pay for it with an high-end, high-quality product although expensive.
It will get rid of some low-end developers, but probably could find a niche where it could survive happily. Another useless free product wouldn’t go far.
Although I’m sure Emb would do nor the former nor the latter – it will keep on selling a limited, low quality product at high prices – a lose-lose situation anyway.
PS: maintaining and improving a framework and IDE has a cost, and not marginal. It means R&D, and investments. Sure, if you rely on someone else is far easier.
The Delphi IDE would need a huge refactoring and some rewriting, to get rid of .NET and keep pace with the other IDEs. Unless you’re suggesting that a free Delphi version should be only the compiler inside VS IDE… after all Windows has a rich set of APIs you can call yourself…
You seem to be saying that the game is already over, so the question isn’t “What good would it do?” but rather “How could it possibly hurt to try?”
I acknowledged the cost of maintaining an IDE, but asked you to explain how accounting for that cost in the price can stand realistically alongside the complaints precisely about the lack of effort that goes into that maintenance ?
Your comments are not worth responding to any more since you clearly don’t even bother reading those responses.
Enjoy your gilded, ivory tower. It will be increasing lonely up there, but think of all the space you’ll have… 😀
>You seem to be saying that the game is already over, so the question
>isn’t “What good would it do?” but rather “How could it possibly hurt to
>try?”
It seemed he was explaining how it would hurt to try, whether we agree with it or not. I think the argument just shows that EMBT really doesn’t have any good options right now with a reasonable chance of success. They don’t seem to have any interest in gaining new customers (at least they aren’t trying anything to gain them).
I may be wrong on this, but I seem to recall Embarcadero Insider or another insider stating that after Codegear began offering their free Turbo Delphi version, revenue plummeted, suggesting that most remaining users simply didn’t need the bigger/more expensive versions. Something to consider in support of LDS’ arguments.
BizSpark is made for startups, and have limitations. You can’t stay in BizSpark for ever. Although much larger, is something alike the Starter. It ends, and when you outgrew it you have to pay for standard licenses.
Yep, and when you reach those limitations you have to move on up to the more expensive product offerings.
Or, if BizSpark wasn’t a viable option (or not available at all) those startups would instead use alternatives from someone other than Microsoft and Microsoft would have lost a potential, bigger customer.
Gee, see how that works ?
LSD is right.. Pascal and Delphi is something from 20th century. Current young minds do not want to know anything about it and would not care to learn anything about it. Why ? Because it is not user friendly by current industry standarts: the code is too verbose; no, it is not easiest language to program in; defining variable on top of function/procedure simply sucks; begin/end sucks; “double” definition of function/procedure in interface and implementation sections suck; this one is rather subjective – case insensitive; Delphi IDE is slow and is a bug fest; bellow crappy support for everything WEB – websites, web services.. And many more. Like LSD already told: new and young programmers will not use Pascal or Delphi. It is like asking a teenager to wear his grandfather’s shorts.. It looks and smells funny, in a bad way..
With this opening, this comment needs no reply. 🙂
Let’s not forget lack of functional programming techniques, no memory management (sounds like not for XE5 either), no operator overloading of classes, no first class types, no generics for first class functions, possibly generics vs. templates, no slice notation, still a lack of iteration in many places (such as database query results), no power symbol, poor serialization support, no built-in logging, can’t replace a record helper without losing all the other ones, no tuples, no generators, poor multiprocessing support, no lambda, no step for the for loop, relies on third parties for unit testing and profiling, no bigint, decimal or fraction support; sets limited to 255 values, sets confined to a few discrete data types, lack of RTTI support for enumerations with custom values, GUID needed for some interface functions…. one could go on and on. 🙁
It’s the elephant in the room no one ever wants to address. We’re REALLY behind, and whenever Delphi gets a “new” language feature, it’s something that other languages have had for 4-5 years or more. And in cases like iteration, we often get a little bit but no one ever goes back through the standard library and VCL and refactors to take advantage of the new feature (which hinders adoption by the community). And when something new IS added, the community screams at the thought of not being able to compile Delphi 1/Turbo Pascal code, so nothing is ever deprecated either. Then we end up with situations like there being at least SIX different ways of reading or writing a file in Delphi! (Much more if you consider that there are in some cases separate classes just for reading or writing). You’ve got TDirectory but FindFirst/FindNext is still there. By the way, EMBT says of TDirectory: “Most of the methods exposed by TDirectory are signature-compatible with the ones exposed by the Directory class in the .NET Framework.” They’re essentially lifted verbatim from .NET, only seven years after .NET introduced them. I believe it was Harry Truman who said, “When a Republican runs against a Republican, the Republican will win every time.” How do we convince .NET users to use Delphi if they can get the real deal, earlier and cheaper?
I’ve seen quotes from David I and Allen Bauer talking about adding language-specific features to ease parallel programming as early as eight years ago, but obviously nothing’s been done. Other than Barry Kelly’s work, it seems the language has been left to languish for quite some time. The interest that’s been shown in modernizing it a little with the advent of a new compiler has been met with torches and pitchforks by the “old guard”. The remaining Delphi users are almost by definition highly resistant to change, so there doesn’t seem to be a good option when going after new users could lose old ones (although I have no idea what they would go to instead).
It would hurt me because I don’t neeed a subpar cheap tool. I need an high-end one, even if that’s expensive. If ever Emb would listen to you, it would start to deliver cheap software with little or no improvements, and that’s not what I need – I’m no longer an hobbyist programmer.
That’s why more and more of my code is migrating to C/C++ (and some even to FPGAs), while Delphi is more and more used for desktop GUIs only.
Just today we inked a deal to buy specialized NICs, they cost several K each – plus SSD PCIe disks, many K each – running on 32 core servers. Of course the software using them won’t be written in Delphi, because it can’t deliver the expected performace (without much trouble). Do you believe that to support that hardware we are afraid of a development tool costing a couple of K?
But this way I need to buy both Delphi and other tools, and if I have to choose, well Delphi will be the one going away, even if it was cheaper or free.
About the IDE, you wrote that it costs to create one, while it doesn’t cost to have one. Of course that’s not true. Sure actual features don’t justify the actual price, but I want better features, not a cheaper product without them. It looks you would be happy just with a cheaper product to keep on playing with Delphi. I wouldn’t.
But I guess my comments are not worth responding just because you don’t know how to answer them.
I asked you what’s your evidence a free version would bring new many developers, and it looks still you have not an answer.
LDS, one last attempt to make you understand how your thinking is fatally flawed. Fatal for Delphi that is.
I used to advocate Delphi. My advocacy was directly responsible for a number of license sales. That advocacy started only because I was able to afford to get interested in the first place. If new people aren’t able to afford to get interested then eventually all that are left are those current users unable or unwilling to switch until they absolutely have to. Those people that are left either eventually reach the point where they too can no longer afford not to switch. Or they die (well, they’ll probably retire before they die, but you get my point).
But either way, if you don’t have new blood coming in then the inevitable conclusion is that Delphi will become just another SAS, Watcom C++, PowerBuilder, SQLWindows, Omnis, SuperBase etc etc (comparisons to COBOL aren’t really valid since it is to exaggerate the significance of Delphi).
It really isn’t about how much you can charge, i.e. what the market will bear, it’s about how many new users are coming on board to be charged anything at all.
You dismiss suggestions that pricing should be set to attract new users as unrealistic without any supporting evidence (Pot? Kettle? Black?), as trying to recruit people who don’t want to pay, but that’s just your blinkered hyperbole, not what is actually being suggested.
As for prices being justified by the cost of developing your own products rather than taking advantage of someone elses. What about when Embarcadero chose to put Prism in their box? Or FireMonkey? Or taking advantage of Xcode and the Android SDK instead of developing their own products to meet the needs they fulfil? Of course, when they did that their prices came tumbling down because they don’t need to cover the cost of developing these things and investing in original, home-grown, innovative, and high quality R&D, right ? Hmmmm.
But it’s about quality as well right ?
Well you yourself bemoaned the quality of FireMonkey. And now here you are using supposed high quality to justify high prices (and when that argument is cut out from under you you turn to something else, and then something else ad nauseum).
But then inconsistency and rank hypocrisy is rife in your commentary so what else would we expect ?
Enjoy having your cake and enjoy eating it.
>If new people aren’t able to afford to get interested
This puts the cart before the horse though. While I agree with your argument, the bigger problem is that there has to be an interest in the first place. What if people get this affordable version of Delphi and then encounter glacial start-up times, a buggy IDE, lackluster documentation, or another rushed release on the order of XE2?
Raymond Hettinger has given several versions of a talk usually titled “What Makes Python Awsome”; one version spent the opening part discussing what any language needs to succeed based on his study of successful and unsuccessful products. I’ve written of this much more in-depth on the forum, but Delphi is lacking in several of those areas. We don’t have unique selling points anymore like we did in 1995, we don’t have “zen” (the language/RTL is a mess and looks like Delphi has passed through even more hands than it has), we don’t have a large quantity of easily available learning materials (as mentioned earlier, we’ve been “off the shelves” since at least 2005), we don’t have a friendly license, the community is a shadow of its former self and of course we’re not open source or even standardized. How do you sell that? (Hint: “native” ain’t it).
>then eventually all
>that are left are those current users unable or unwilling to switch >until they absolutely have to. Those people that are left either
>eventually reach the point where they too can no longer afford not
>to switch. Or they die (well, they’ll probably retire before they die,
>but you get my point).
I have a theory that Rudy will offer to euthanize them with laughing gas. 😉 But I think you’re describing where we are right now. There is no new blood, as you’ve made a point of. I’d completely agree with your points about an accessible version, but only if I felt we had a competitive product that had a list of features we could point to as unique and advantageous (and we have to be pretty advantageous for even a cheaper Delphi to compete with “free”).
>But either way, if you don’t have new blood coming in then the
>inevitable conclusion is that Delphi will become just another SAS,
>Watcom C++, PowerBuilder, SQLWindows, Omnis, SuperBase etc
I think we’re already PowerBuilder. PowerBuilder is still released every 18 months by Sybase, can now target the web and .NET, etc. But outside of the few people remaining who support legacy PowerBuilder apps, no one cares (or probably even remembers what PowerBuilder is). Outside of Delphi Island, it’s basically the same way. I’ve mentioned before that at a billion-dollar company in 2005 only the director of the IT department even knew what Delphi WAS. And today the few who know what it is express surprise that it still exists (often thinking it disappeared with Borland).
>It really isn’t about how much you can charge, i.e. what the market
>will bear, it’s about how many new users are coming on board to be
>charged anything at all.
For the record I agree with everything. But recently I’ve come to the conclusion that with no remaining viable ecosystem and no one up to the CEO able to make a case for the product, making Delphi more accessible won’t change anything.
>But then inconsistency and rank hypocrisy is rife in your
>commentary so what else would we expect ?
Isn’t this getting a little bit personal and mean between you two? That’s what the EMBT forum is for. 😉 😉 😉
Actually if LDS = Luigi D. Sandon I’m kind of baffled as I believe a few months ago we had a big argument in which I was the one arguing that Delphi was outdated and he was dismissing the features I was talking about as fads. 🙂 Now he’s embracing Python, which I was championing in that discussion along with D and R. 😉 😉
I think LDS is in the same camp (to some degree) that I am, believing that first you need a product that young programmers want to have; pricing/segmentation strategies come second. Remember Barry Kelly’s sign-off:
>…But more than anything else, I’d fallen out of love with Delphi, and
>could no longer motivate myself to try and make it better – the gap
>I’d try to bridge would be a gap too far for the market to bear.
>…by gap, I mean the distance between the established product of
>Delphi, and a language I’d be happy and proud to continue working
>on.
>
>It’s a fact that Delphi has been around for long enough that
>changing it significantly isn’t really an option.
>…Delphi is very procedural. It grew out of Pascal, a language
>designed in an era when memory was very expensive. So most of its
>core runtime is based around mutation and destructive updates.
>
>But the longer I’ve been coding, the greater and greater benefit I
>see to more functional approaches – which pretty much require
>garbage collection – and persistent data structures like you see in
>Clojure.
>
>I’ve also lost some of the object orientation religion…. these days, I
>see the bureaucracy and busywork involved in creating class
>hierarchies, how it can fool you into thinking you’re doing productive
>work when you’re filling out various idioms and “patterns”. Some
>problems – like GUI widgets – work really really well with OO. But
>others work far better with the functional approach, where the set of
>data structures is closed but the set of methods is open.
>Shoehorning these into OO results in ugly architectures with extra
>indirections.
>
>Suffice it to say, if I was creating a language I was truly in love with,
>it would look quite different to Delphi.
If Barry Kelly encounters the new paradigms and falls out of love with Delphi, how do you get people who’ve never even written a line of Pascal in their lives (virtually all young programmers today) to fall in love with it?
You don’t understand me at all. I wrote Pascal/Delphi are no longer fashionable and perceived legacy. For the matter C and C++ are from the 20th century also, and they’re not going away.
I don’t believe what you write about Pascal are defects, they are good things often. I don’t care if Delphi is not good for the web – after all who writes web applications in C/C++ or Objective-C? Nobody. .NET and LAMP are good for the web? Perfect, but I don’t write web applications.
The problem is Delphi lost ground on its own territory, Windows native applications. It was born to develop desktop applications and it never became really a tool to write *any* Windows applications – including those who have no GUI. It multiplatform attempts are crippled by hurried up implementation and bad design choices. And quality – overall quality, including optimazation and performaces.
You don’t need a product to be fashionable to be a good – or great – product. I don’t care if young developers don’t bother about Delphi (and they may be right as long as it doesn’t deliver nothing outstanding), but I do bother about a product that is becoming a true Jack-of-all-trades-and-master-of-none. Maybe good for small ISV with basic needs, but useless for more advanced products.
I’ll challenge “The myth of course is that Embarcadero does not have other revenues” assertion from publicly-available data.
According to Wiki, http://en.wikipedia.org/wiki/Embarcadero_Technologies states that EMBT has revenues of $100M (both in the summary box and under the Company heading, dated 2008 and 2012 respectively.) EMBT’s “about us” http://www.embarcadero.com/company/about-us claims “Over 3 million users”. Wiki claims “400” and “450” employees; EMBT, 500.
So – $100M might seem a lot – but the last company I was working for had revenues of $100M and 100 staff – $1M revenue per staff member against the implicit $200K if (and I emphasise IF) those figures are correct.
Taking that same $100M and dividing it by the “average” price of a Delphi upgrade – well, we have sales outside of the US apparently greater in number than US sales, despite the well-publicised loading. My last few PRO upgrades have cost ~$700 or so each, so if we take $500 as a very conservative average, given the lower US price and many Enterprise or Architect upgrades pushing the average in opposite directions. That’s 100M/500 = 200K renewals per annum. Again, from EMBT’s page, a claim of 3M users.
So, if every cent of EMBT’s revenue was derived from Delphi upgrades, then upgrade-uptake is less than 7%. If EMBT has other revenue sources, then this rate would be decreased…
The goal is to ensure the future viability of EMBT – which claims to have made a profit for 58 successive quarters in 17 years of existence. I’d assign a small team to completely re-writing the entire VCL and RTL to ensure that the basic toolkit is even usable in the multithread multiprocessor world of 2013, not the environment in which it started in the mid-1990s. Release a monthly update to patch every post-Borland edition. Trawl the internet for mentions of problems and fixes. This Herculean task is only going to get harder to solve the longer it is postponed. Until EMBT starts, customers will assume the same old deny-and-ignore philosophy continues.
Documentation and tools need to be brought up-to-date. Nick Hodges’ test-bed for a start. And documented official recompile-your-rtl/vcl routines for those who want to IMPROVE the product at their own expense of time and effort. Documentation on using the IDE interface, and how to reduce the compiled size of the object modules. “Hello World” using Oxygene compiles to 4K – yes, plus .Net, but the Delphi version is huge (and didn’t use to be.)
None of this is being negative. Ideally, we should be able to restore Delphi to the pre-eminence that TP enjoyed.
And for the future – well, we have some criticism in this thread about interface and implementation – this could be solved very easily. Allow an optional “exported [name]” keyword to allow a method/variable to be seen from the outside as if it was in the interface area, with the option of changing its name (for case-sensitive DLL entries, for instance.)
EMB are not poor. They are growing. No one who sells software and IT services is working poor. This will not change so soon.
There are some mourners who always complain. The only thing that needs fixing is the FMX imo. Secret sources revealed that the ARM support – the ‘compiler portion’ – works very satisfactory.
I will hop toward the goal and not toward what is not the goal.
(@bunnybuddhism)
>I’ll challenge “The myth of course is that Embarcadero does not have
>other revenues” assertion from publicly-available data.
Cool! I’ll challenge your anti-assertion then. 🙂
>According to Wiki, http://en.wikipedia.org/wiki/Embarcadero_Technologies
>states that EMBT has revenues of $100M (both in the summary box and
>under the Company heading, dated 2008 and 2012 respectively.)
The problem is that the figure from 2008 may be accurate as it was taken private in 2007 and may have put out a final financial statement in 2008. The claim that the revenue is also $100 million in 2012 in the article has no footnote associated with it, so we don’t know what the source is for that conclusion.
>EMBT’s “about us” http://www.embarcadero.com/company/about-us
>claims “Over 3 million users”.
Oh, this is a favorite gem of mine. A few years ago an EMBT employee put out a calculation (I don’t want to elevate it to “a study”) that there were two million licenses of Delphi ever sold, but that figure also included educational use and an estimation for “piracy”. No work was shown, and somehow this became distorted such at CodeRage TeamB presenters were claiming that there were two million users of Delphi worldwide. No one bothered to correct this assertion and this claim still gets batted around today.
It got worse when they then ran with the two million figure (which of course even if done accurately doesn’t tell us how many people use Delphi today) and began claiming that Delphi was the largest Windows development tool outside of Microsoft’s. At this point Java developers did a collective spit-take. 🙂 Team-B have attempted to spin this by highlighting the Windows qualification and claiming that Java developers don’t target Windows (!!!).
Finally, someone at EMBT must have figured that as long as they were pulling numbers out of their you-know-what, why not go long? 😉 In 2011 they were claiming two million; now they’re claiming three million (while the about us page doesn’t specify products so could include all the EMBT database products, I have seen one marketing press release from this year claiming three million Delphi users, which shows that marketing are still enumerate or they’re being encouraged to make up numbers at this point). What this would mean is that the worldwide Delphi base grew 50% in two years! With absolutely no explanation as to why that would suddenly happen or any sign in the real world that it did happen, of course.
Let’s imagine what this would mean if it were true. If a Pro edition of Delphi is $1000USD, that would mean additional revenues of ONE BILLION DOLLARS over two years, with minimal additional cost, for almost pure profit. All from a company they purchased for $30 million. If EMBT has seen revenue increases of half a billion dollars a year, their venture capital owners would have done an IPO so fast your head would
spin and it would have been as big as Facebook. Since that DIDN’T happen, I think the idea of three million Delphi users (or two million, especially since that was the estimate of total licenses ever sold from Delphi 1 onwards) is ridiculous. Embarcadero Insider estimated 150,000 Delphi users worldwide.
>So, if every cent of EMBT’s revenue was derived from Delphi
>upgrades, then upgrade-uptake is less than 7%.
I think the figure changes somewhat if we use more realistic user numbers, but at this point it’s really conjecture.
> I’d assign a small team to completely re-writing the entire VCL and
>RTL to ensure that the basic toolkit is even usable in the multithread
>multiprocessor world of 2013, not the environment in which it started in
>the mid-1990s.
Promise me you’ll refactor that RTL while you’re at it to take advantage of all the features that have been added over time (e.g. no reason in the world to have Pos and PosEx when Pos with a named/optional parameter would do the job of both, iteration needing to be everywhere including database query results, enumerations, etc.) and getting rid of a lot of features that should be deprecated like all the ways we have to open a file and you have my vote. 🙂
> Release a monthly update to patch every post-Borland edition.
Amen! My linux distro is free and is supported for 2.25 release cycles; PostgreSQL is free and supported for 5.
>Trawl the internet for mentions of problems and fixes. This Herculean
>task is only going to get harder to solve the longer it is postponed. Until
>EMBT starts, customers will assume the same old deny-and-ignore
>philosophy continues.
You join my pantheon of “Delphi developers I’ve never met but I know could run the company better than those doing it now.”
>Documentation and tools need to be brought up-to-date.
Another big mess that needs cleaning up; agreed.
I’ve floated a few ideas before, including EMBT putting together a free online course on how to write a tech book. Those who take the book would be provided with the services of a technical editor by EMBT for free to edit any manuscripts they produce. This acknowledges we’re not going to see any more commercially published books and keeps self-published books at the same quality as what we had before. EMBT could also develop a style guide and such and possibly even have this books branded with their own publishing moniker – I’m partial to “Embookadero”. 🙂 EMBT could also commission community members to write books on topics their research shows the community desires, help promote the books, etc.
We also need a central repository for Delphi libraries. Note this is NOT the “software store” their survey was asking about, where they were salivating to know how much we’d pay to have our software in it. I mean something like Torry but modern with up-to-date code and that indexes open source Delphi libraries hosted anywhere on the net. It will also need an API which should be embedded into the IDE so that users could search for, install, uninstall and update libraries right through this. Python, Perl, R and many other languages already have such a service; we really need one too.
>None of this is being negative. Ideally, we should be able to restore
>Delphi to the pre-eminence that TP enjoyed.
It’s a huge hill to climb, but it’s certainly possible. There’s been little sign of the will to polish the product (maybe some from Marco) or rebuild the ecosystem though (no signs of that at all).
>And for the future – well, we have some criticism in this thread about
>interface and implementation – this could be solved very easily. Allow
>an optional “exported [name]” keyword to allow a method/variable to be
>seen from the outside as if it was in the interface area, with the option
>of changing its name (for case-sensitive DLL entries, for instance.)
Do you mean something like what you can do in Python with, e.g.,
>from SomeLibrary import HTMLparser as parser
At the RAD Studio tour event on Friday they kicked off with a “Who are Embarcadero” slide right on which they state 3.2 million users across all products.
Thank you for this information! I await the next TeamB’er to tell me that there are three million Delphi users. 😉
“The goal is to ensure the future viability of EMBT – which claims to have made a profit for 58 successive quarters in 17 years of existence.”
NO. That is a goal for EMBT management, not for us as customers. As a developer, my goal is to have the most capable and bug-free environment I can, to better support my delivering that same level of quality to my clients.
My profitability is my problem, not a responsibility of EMBT; equally, their profitability is not my problem. However, their management needs to realize that their failure to resolve long in the tooth defects does affect my productivity, and therefore my profitability. And things which detract from my desire for the most productive development environment will, or necessity, influence my decision about future expenditures.
XE2..XE5 have failed to address my interests and needs, but have galloped off in several directions in pursuit of the mobile market. I do not assert that this market should be ignored, but rather that it is not an excuse for ignoring the needs of the customer base. Moreover, leaving Android to last was, at best, a questionable choice.
There are many applications which will never move to a mobile device. Most of those I have written in the last 25 years are among them. The popular delusion that mobile devices will replace the desktop is just that: a delusion. Even leaving out the many applications which require specialized hardware inside the device, I know of no one who writes seriously who would consider for an instant writing all day on a mobile device.
Too much emotional response seems to cloud the decisions made in SFO. Hard logic and rational analyses are needed.
“Hard logic and rational analyses are needed.”
You’ll never be taken seriously making assertions like that. 😉
>The popular delusion that mobile devices will replace the desktop is just
that: a delusion.
IT will evolve beyond the day you retire;) The change process is not a linear one.
I think we would do no one a favor if we continue to tie people to their desk sitting in front of a computer.
Displays are getting better and better …
http://www.isiqiri-multi-touch.com/products/
These displays are not ‘mobile’ devices that’s true…
‘Appstore’ already existed – C64 – there have been various cards that speed up to ports for the floppy or the printer and provided various programs that way..
Now I am asking you – What’s this
http://windows.microsoft.com/en-us/windows-8/preview?ocid=GA8_O_WOL_Hero_Home_BlueRP_Pos2_01
Is this a car
http://en.wikipedia.org/wiki/Go-kart
http://en.wikipedia.org/wiki/Dune_buggy
http://en.wikipedia.org/wiki/Formula_One_car
http://en.wikipedia.org/wiki/Rover_(space_exploration)
Are these cars everyone drive … ?
I have tested wonderful ARM Thin Client solution, ARM based industrial devices …
CAD/CAM is a good candidate for a hosted solution and does already work in the clouds as well as games. A German trusts send engineering data to China … for example. The most simple standardization of data structures in IT is about using the same application. Even the very special case CAD/CAM and similar things. Those applications had been ported from UNIX workstations 15 years ago … almost no one believed. Architectural change because of advanced logic is machines (production – autonomous learning – autonomous optimization … lowering energy consumption by optimizing the logistics steps – works already – Frauenhofer Institut, Germany – humans no longer required).
EMB is a lot less dependent on profitability than you are dependent on theirs. This weakens your position. EMB cannot simply say – we forget about the future.The world is not built around the individual – impression created for you by the matrix that owns you.
I totally agree that Delphi for about 3 years provides science friction – that’s not a typo condisdering Delphi 2010 and XE as an improvement.
EMB is a lot less dependent on your profitability than you are dependent on theirs.
Sry…
> I do not assert that this market should be ignored, but rather that it is not
>an excuse for ignoring the needs of the customer base.
The reality is they’re too small and simply have too few developers to do all they want to do. I think it’s estimated you’ve got 20-something Delphi developers while Digia has 10 times that working on just the Qt cross-platform framework. EMBT is trying to handle a language, two compilers, an IDE, and two frameworks, one cross-platform, with 1/10 the manpower. Then management insists on shipping whether the product is finished or not, then this affects future sales (Embarcadero Insider suggested XE3 sales plunged after the debacle of XE2). Then more layoffs come. This has the makings of a death spiral a la Blackberry.
>Moreover, leaving Android to last was, at best, a questionable choice.
I think they still had painful memories of their last foray into Linux.
> The popular delusion that mobile devices will replace the desktop is just
>that: a delusion.
Let’s not go down that road I’ve seen other Delphi users go down. Over the last few quarters we’ve seen both the biggest single-quarter drop in PC sales since records have been kept and also the longest string of consecutive declining quarters for PC sales in history. The juggernaut, the monopoly, Microsoft certainly sees the end of the PC era and is attempting to transition to a devices and services company (their own words). The biggest PC maker, Lenovo, reported selling more smart phones and tablets than PC the last quarter! “Lenovo’s mobile devices business had a revenue increase of 105 percent year-over-year for the quarter” as well.
As Adrian Kingsley-Hughes of ZDNet put it last month:
>I don’t say this lightly either. …my career has, for several decades, been
>firmly rooted in the PC industry, and at the core of that has been
>Windows. …I … possess the superpowers/voodoo/magic/power needed to
>make PCs do what I want them to do. Give me an IT-related problem, and
>with a little time and adequate resources, I can put together the PC
>equivalent of the cotton gin.
>
>I can safely say that I don’t want to see the end of the PC. Ideally, I want >things to stay just as they are for a long time.
>
>But they’re not, and people who think that the downturn in PC sales is
>temporary, or those who think that Windows is just as relevant and
>popular as it ever was, are either kidding themselves, deluded, or just
>desperately trying to plaster over the cracks in a crumbling ecosystem.
>
>But as with most things in life, wishing for something doesn’t make it so.
Mobile devices do what most average users use(d) a PC for. In addition, much is moving to the web to the point where the browser may become more important than the underlying OS. It’s the beginning of a whole new world. PCs won’t go away completely… mainframes never did. But their relevance will continue to decline.
> Even leaving out the many applications which require specialized
>hardware inside the device, I know of no one who writes seriously who
>would consider for an instant writing all day on a mobile device.
Think bigger. In the near future your mobile device will be your mobile device and your PC. You’ll get to work, plug your phone into a dock that will be connected (wired or wirelessly) to a a keyboard, mouse and monitor and the OS will switch to a desktop OS and you’ll do your work that way. Asus already has a phone that can plug into the back of an otherwise brainless tablet and I’ve seen Canonical demonstrate a prototype of this docking arrangement with Android and Ubuntu (with Ubuntu even having the same browser tabs open when it switches over OSes!).
>Too much emotional response seems to cloud the decisions made in
>SFO. Hard logic and rational analyses are needed.
Hard numbers show PC sales are in decline. Emerging markets are at best delaying purchase of a PC by purchasing a cheaper mobile device (at worst, never going to move beyond the mobile device). Rational analysis has made Apple (iPad), Microsoft (transition to devices/services) and Google (ChromeOS, Chromebook) believe we’re at the dawn of a post-PC era.