May 15, 2010

Delphi Roadmap, May 2010 Update

Mike Rozlog, Delphi product manager, has released an updated roadmap for the product.

Mike Rozlog, Delphi product manager, has released an updated roadmap for the product. You can find it at http://edn.embarcadero.com/article/39934.

The update is far from a revolution, with the focus for the coming versions on multi-platform (first) and 64bit support (next). However, we now have more details, some dates, and a set of 5 ongoing projects. These are not in a strict temporary order, as has been clarifies on the newsgroups). Here are some info on the 5 projects:

  • Fulcrum: Native cross-platform support for the Mac. With a cross-platform VCL-like library, a remote debugger, source code management (that is, full version control integration), sequence diagrams support, Microsoft Azure integration. This means, only the first two elements are for Mac support, while many other features address needs of Delphi Win32 developers.
     
  • Wheelhouse: Adding Linux support. This was optionally supposed to be part of Fulcrum in the past, now it is clearly another step of the product. Notice the emphasis on server-side development support and support for Apache.
     
  • 64-bit Compiler Preview: This is another new project, clearly a limited command line compiler, but has a date attached: 1st half of 2011.
     
  • Commodore: Complete 64bit support for Delphi, with RTL, VCL, and cross-platform library support for 64 bit, enhanced multi-threading and multi-core support (parallelization support in the RTL), social network integration (I wonder what that is?).
     
  • Chromium: This is far from clear. It mentions "access to computer infrastructures" (that is, cloud computing), and talks a lot about user interface. It seems they'll work on natural input (location, voice, motion) but also think of ARM support and Rich Internet Applications. My guess, this is just an assorted group of possible directions, but you should not take them too seriously.

A few comments. First, it is good to have updates about future projects, even better to have some date attached to them. For sure, this is a very partial list, and it seems projects more far away in the future are just sketched. Overall, the projects are not much different from what was announced a few months back. I agree with a forum comment that hopefully we don't "have to wait all the way until Commodore to get better documentation".

An interesting comment (for a long thread look here) I saw was about the fact Embarcadero is "more interested in hunting for new markets than keeping existing customers". Yes, it might as well look like that, but on the other hand I certainly want Delphi to let my applications (and knowledge) move beyond the current OS and embrace new technologies (and some new over-hyped ones, as well). So, while I partially agree with the comment, I really hope Delphi can attract new users as well, to remain healthy. As another person put it they might be "chasing easy consensus instead of tackling real issues", but this consensus only result in better perception at large.

By the way, as an active Kylix user, as much as I would like to have a 64-bit compiler Linux server support is top priority for me. But I know this is a personal point of view... more or less like any other personal point of view asking for the opposite.

Another complain is that " Embarcadero is missing very important emerging platform - rich mobile. Now iPhone OS and Android are already true rich platforms for business applications." That's partially true, but remember that Embarcadero has also a Java development tool, that iPhone has a nice rule disallowing applications written in Pascal, and that Android has no real native SDK! Also, moving existing code to these platform is far from a trial stance. But notice, again, ARM support. This is brand new in the roadmap.

Finally from a comment by Nick we can also learn that J# dependencies in the IDE (used by the modeling) are going to be remove.

 





 

24 Comments

"Some Dates" would imply a plurality of dates - there is precisely ONE... 

...and that is for the preview release of a deliverable 
originally scheduled on a previous roadmap for LAST YEAR 
(Commodore was to be the next release after Tiburon... 
Tiburon was 2009, Commodore should have been 2010).

Embarcadero Roadmaps aren't worth the PowerPoint slides 
they are printed on, and any dates are worth even less. 
Comment by Jolyon Smith [http://www.deltics.co.nz/blog] on May 15, 23:07

Delphi Roadmap, May 2010 Update 

My reading of Fulcrum is that for OSX it's mainly
looking at client apps wirth dbExpress and DataSnap.
The cross platform VCL-like library is only "under
consideration".
Comment by Rob McDonell [http://arkangles.com] on May 16, 02:16

Delphi Roadmap, May 2010 Update 

 <as much as I would like to have a 64-bit compiler
Linux server support is top priority for me>

And me.

Cheers Marco - and thanks
Comment by Robert Horbury-Smith [] on May 16, 02:57

Delphi Roadmap, May 2010 Update 

Marco!

This is not only your personal view ...

By the way, as an active Kylix user, as much as I 
would like to have a 64-bit compiler Linux server 
support is top priority for me. --> "Something like " 
the Datasnap running under 64bit Linux would really 
have been an argument for the Enterprise Edition.

For fast-cgi and related things there are still lots 
of specialiced servers that run on 32-bit only...

Mike
Comment by Michael Thuma on May 16, 08:22

Delphi Roadmap, May 2010 Update 

"but remember that Embarcadero has also a Java
development tool"

JBuilder? Is that still alive?
Comment by Moritz Beutel on May 16, 16:10

Delphi Roadmap, May 2010 Update 

Go to Delphi User Voice and vote for the Delphi
features that matters to you.

http://delphi.uservoice.com/

Also Tim DelChiaro from Embarcadero has started a
survey Delphi developer to participate in, about "what
challenges developers face this year and day to day,
the technologies and tools they’re using, and where
they see opportunity".

http://www.surveymonkey.com/s/embt-developer

Doei Rif
Comment by Rif on May 16, 17:36

Delphi Roadmap, May 2010 Update 

I wonder if they are on the right road to gain new 
users. They took that road twice (Kylix and 
Delphi.net) and both times failed, because they 
delivered subpar products unable to be recognized as 
leaders in their respective markets. I am afraid they 
took the same approach with Fulcrum. First, a cross-
compiler (requiring a Windows license) and Qt-based 
framework IMHO will be very little appealing to Mac 
developers. And a Delphi-based, Qt-based solution 
will be little appealing as well to Windows 
developers willing to develop for the Mac natively. 
It looks that once again real customers could only be 
that part of actual Delphi customers wanting to 
expand somehow their market towards Apple, although I 
am almost sure it was internal needs to have pushed 
this choice. Development also requires an investment 
in Apple hardware, because you can't run MacOSX on 
your PC or a virtual machine without hacking it, and 
that raises the price as well. At least returning to 
Linux would have taken advantage of server side 
development (which does not exist on Macs), and the 
availability of free distributions that could be run 
inside a VM.
IMHO that will be another big failure, and the "real" 
product will pay once more the price of decision took 
dreaming instead of assessing Delphi market needs.
Comment by Luigi D. Sandon on May 16, 19:50

Delphi Roadmap, May 2010 Update - Why it sucks 

So, looking at this ‘roadmap’ and the Embarcadero’s
release pattern we can expect to get Fulcrum in the
second half of 2010, Wheelhouse in 2011, Preview
(command line) 64 in 2012 and (finally) Commodore in
the second half of 2013. I’m a ‘little’ disappointed…
Comment by J. Verbeeck on May 17, 08:12

Delphi Roadmap, May 2010 Update 

Luigi,

I agree with you at 100%!!

Claudio
Comment by Caudio Piffer on May 17, 15:32

Delphi Roadmap, May 2010 Update 

J. Verbeeck,

  no I don't think this is correct, since the only 
official date is the Preview of 64 bit compiler, in the 
first half of 2011 as indicated in the docs and the 
blog, and not in 2012 as you suggest. For any other 
date, we have no specific information, so my guess is 
worth your guess.
Comment by Marco Cantu [http://www.marcocantu.com] on May 18, 02:37

Delphi Roadmap, May 2010 Update 

J. Verbeeck,

  no I don't think this is correct, since the only 
official date is the Preview of 64 bit compiler, in the 
first half of 2011 as indicated in the docs and the 
blog, and not in 2012 as you suggest. For any other 
date, we have no specific information, so my guess is 
worth your guess. 
Comment by Marco Cantu [http://www.marcocantu.com] on May 18, 02:41

Delphi Roadmap, May 2010 Update 

J. Verbeeck,

  no I don't think this is correct, since the only 
official date is the Preview of 64 bit compiler, in the 
first half of 2011 as indicated in the docs and the 
blog, and not in 2012 as you suggest. For any other 
date, we have no specific information, so my guess is 
worth your guess.
Comment by Marco Cantu [http://www.marcocantu.com] on May 18, 02:47

Delphi 'Roadmap', May 2010 Update  

Totally agree Luigi !
At this rate I can't see a Windows x64 compiler/VCL 
been delivered before H2 2012 . Will probably need
first update applied before its really usable so H1
2013 I suspect is real delivery date for production use.

Its laughable to even call it a 'roadmap' IMO. 
Has ONE date in it for a 'preview 64bit compiler' an
item in previous 'roadmap' that was talked about for
about H2 2009 NOT H1 2011 ! Will believe that when I
see it to. They now have ZERO Credibility here.

Comment by FromEurope on May 18, 05:30

Linux 

 I think adding Linux support should be the 1st priority

I'm not a linux user , but It would be best if I deploy 
my software on a stable OS  which normal users won't be 
able to mess with it or infect it with viruses

also linux is free, this would reduce the cost of the 
software , great for small businesses 
Comment by Ahmed Rih [http://vega.ozone.com.ly] on May 18, 11:37

Delphi Roadmap, May 2010 Update 

I am disappointed in EMBC progress. They keep chasing
a dream that might affect its bread and butter. 
My SA is up for renewal in November and I am not sure
if the new roadmap encourages me to renew. 
Comment by Chris on May 18, 17:32

Delphi Roadmap, May 2010 Update, Delphi and C# 

I have to comment on the statement that Embarcadero 
is "more interested in hunting for new markets than 
keeping existing customers" as this is exactly what 
it seems to me and it is useful to have it confirmed 
so that we don't waste timeing hoping and waiting for 
our needs to be met. I am a Delphi 'classic' (Delphi 
5) user of many years although I also have Delphi 
20xx (which I seldom use). I recently got a new 
client who wanted me to do the project but would not 
allow it to be done in Delphi due to their perception 
that it was dated and not widely used/supported etc 
etc (sadly a common view). I am doing the project in 
Visual C# which I am learning at the same time. I 
have to say that while Delphi is still my main and 
first choice language, I really enjoy C# and to be 
quite frank I have found it easier going from 
Delphi 'Classic' to Visual C# than to Delphi 20xx. In 
fact I find myself frequently chuckling to myself at 
the similarities, both obvious and subtle, between C# 
and Delphi Classic (no doubt this has alot to do with 
Anders Heljlsberg). I keep getting special offers 
from Embarcadero but all of them offer me things I 
don't need, and none what I really need, which is 
basically a fully upward compatible (including 
Unicode conversion assistance), Classic Delphi 64-bit 
with Web capabiities (Intraweb will do but must have 
improved documentation and support) and no .NET (.NET 
is fine with C# but one of the advantages of Delphi 
is not having .NET). I am probably alone in this view.
Comment by John on May 22, 10:44

Delphi Roadmap, May 2010 Update 

@John: 'I have found it easier going from 
Delphi 'Classic' to Visual C# than to Delphi 20xx. In 
fact I find myself frequently chuckling to myself at 
the similarities, both obvious and subtle, between C# 
and Delphi Classic'

That makes no sense whatsoever: in terms of both
language and libraries, D2010 is a superset of D7,
with what it adds being very much influenced by C#
(records having class-like abilities, the whole
Unicode support stuff - TEncoding etc., generics,
anonymous methods, attributes, rich RTTI).

With respect to the IDE though, I'd agree the defaults
are now a bit annoying, and so in various ways.
However, the D2007+ IDE is IMO definitely superior to
D7's once you do a bit of (supported) customisation.
(The less said about 'even' D2006 - let alone D8 or
D2005 - the better though. Buggy rubbish!)
Comment by Chris on May 22, 20:24

Delphi Roadmap, May 2010 Update 

 "Its laughable to even call it a 'roadmap' IMO. "

I disagree. It is indeed a roadmap - the road simply 
isn't paved. ;-) 

What is laughable about the roadmap is its wishful 
thinking. Given the inability of the Borland/Codebase/
Embarcadero team to deliver 64bit in 2010 on one 
operating system, it's unbelievable that they'd be 
able to make a leap to offering ahead-of-the-curve 
cloud computing, multithreading or embedded 
development features, concurrent multiple platform 
support, etc. The development team ceased being 
leaders when they failed to begin work on 64-bit 
compilation BEFORE the widespread adoption of 64bit 
processors or even being able to unveil 64bit Delphi 
with the release of AMD's first 64-bit CPU. Certain 
folks who are now Embarcadero employees ridculed on 
usenet folks who suggested they needed to be working 
on 64bit before its arrival, and even afterwards made 
posts saying Delphi 64 bit existed right now (in 
other words, Delphi.net). It would be nice if Delphi 
had advanced support and optimization for newer 
instruction sets like SSE3/4/etc. 

Someone posted a link on Embarcadero's forums talking 
about Microsoft working on 128bit versions of Windows
 http://arstechnica.com/microsoft/news/2009/10/
microsoft-mulling-128-bit-versions-of-windows-8-
windows-9.ars

The same names came out of the woodwork saying that 
no one uses 128 bit now, 128 processors aren't here 
yet, 64bit would run on a 128bit processor, 128bit 
won't be an advantage, etc. That's the same stuff we 
heard about 64bit. Some people don't realize that 
Embarcadero needs to start working on these things 
NOW to be ready for when they appear in a few years' 
time. Look at the enormous time 64bit is taking 
because they didn't start until 64bit was mainstream. 
AMD is working on new instructions for its 
forthcoming Bulldozer processor that we'll see in a 
year or two. All the specs have been released to 
developers already. I know for a fact some open 
source compilers have already either begun or 
completed work on adding these instructions to their 
compilers. I'd wager $1000 it hasn't even crossed 
Embaradero's mind yet, and won't until there's a 
Bulldozer in at least 25% of PCs. They left CUDA and 
OpenCL (using the GPU for added number-crunching 
performance) off their list - sadly, the ability is 
here today and they're behind the curve on this again 
anyway. AMD is also talking about having a GPU 
directly integrated with the CPU by 2015, which'll 
probably be about the same time Embarcadero gets 
around to looking at OpenCL or CUDA. 

Given this history and mindset of only working on 
things like Unicode or 64bit when every other product 
on the market has it and people are walking away from 
the Delphi platform in droves because they need these 
features, I think it'll be a cold day in Hell before 
we see Delphi add features that are ahead of the 
curve. Sadly, I think those days have been over for 
several years now. 

I think you're seeing Embarcadero chase new users 
because I'm sadly also willing to bet that Delphi is 
either no longer profitable or barely profitable. 
Most people don't even know Delphi exists (when I was 
working at one of the biggest retailers in the US, no 
one in the corporate IT department even knew what 
Delphi was), can you imagine there are any new Delphi 
developers appearing? Maybe it's the same as before - 
the existing Delphi base will be milked for funds to 
produce other tools like that PHP/ruby stuff they 
came up with.

For those who are interested in 64bit Linux support 
before 128-bit cpus arrive - has anyone tried 
Freepascal & Lazarus? 
Comment by Joseph G. Mitzen on May 23, 03:51

Delphi Roadmap, May 2010 Update 

and for me Linux and create daemon app for linux is
main feature need by me ...

and can not see native supprt of ARM cpu 
(for me cortex-m3)
Comment by maziar Navahan [http://farapox-shiraz.com] on June 1, 15:25

Delphi Roadmap, May 2010 Update 

This is very disappointing. We have a number of large 
enterprise applications for enterprise deployment. Not 
having 64bit (on any OS!) is a big deal, technically and 
from an application marketing perspective.

Linux would be nice, but an enterprise customer wants an 
enterprise grade OS and so will pay for RedHat so they 
get support etc etc and that now costs as much as 
Server2008...... 

In the context of VS2010, Delphi2010 is again playing 
catchup. They should be focussing on the current 
developer base and making sure they have minimal reasons 
to abandon the ship to C# (as we have started to 
do....).
Comment by Andrew on June 4, 05:49

Delphi Roadmap, May 2010 Update 

We have also started to abandon the Delphi ship for 
Visual C#. We will continue to watch and see what 
happens with Delphi and using it for existing Delphi 
projects, and hope that Embarcadero deliver what the 
current Delphi user base require soon, at which stage, 
if we are not too far down the C# road, we will look at 
returning. 

Delphi - beginend + {} = C#
Comment by John on June 7, 15:36

Delphi Roadmap, May 2010 Update 

Embarcadero makes decision to kill Delphi, it is a
clear for now. There is no other explanation for such
one behaviour.
Comment by Best dev tool in the history goes to die on June 15, 12:49

Delphi Roadmap, May 2010 Update 

There is a lot of negativity about Embarcadero.  We 
have to remember they are all we have got if we want 
to keep our existing delphi apps in the market.  
Strategically Apple is gaining on MS.  Being able to 
develop for MS & Apple using our existing code will 
give us a major strategic advantage.  So what if they 
are a little slow on the curve.  As long as we don't 
end up in a dead end we should be happy.  Delphi will 
not attract new users.  It is no longer "cutting 
edge".  There is no resources to be "cutting edge".  
Our advantage is our existing code and the ability to 
port it.  As long as our aps run on the latest and 
most common operating systems and appear up to date 
to the users we should be happy.
Comment by commercial reality on June 17, 09:35

Delphi Roadmap, May 2010 Update 

I'm not sayig here that embarcadero is not doing 
enough. I don’t say they do not work hard, i’m sure 
they do, but they miss the key advantage they could 
have in the cross platform internet/intranet battle. 
delphi .net was a mistake, but prims is not the 
solution because it requires nearly as much job to 
switch to prims then learn c#, so développers will 
leave the boat and go to c# or java;
if the problem with delphi.net was the big changes Ms 
made to .net, then the evident solution should be to 
write their own C# semicompiled code, their own java, 
and call it HERACLES !
This way nobody would make unwanted “big changes”, 
they would masterise the virtual machine, be cross 
platform, allow us to webify existing apps in a 
nutshell, be 100% compatible with delphi and all 
browser .. etc .. only advantages.

I’m sure Ingeneers would love to work on a new 
virtual machine.
And you know what, oracle just buyed sun, there are 
probably high level ingeneers who worked on java 
leaving sun now … but ok it would cost money to hire 
them.

money is the key, and that’s why whe have the 
incompatible delphi prims, because it’s cheap.

just my 5 cents
Comment by Wchris on July 15, 10:14


Post Your Comment

Click here for posting your feedback to this blog.

There are currently 0 pending (unapproved) messages.