VCL Feedback Abound
I expected there to be alot of feedback, but man you guys really surprised me. I didn’t think anyone really read these things :). The good news is alot of you are putting the same two or three big ticket items as your number one requests. The bad news of course being that theres so much other totally valid requests in there that it will take me years to get to it all. Reguardless, I have your feedback now and fully intend to push as hard as I can on product management to make your requests part of the VCL in the future. In case any of you are wondering, heres what came up the most:
Unicode for Win32 - Yeah this is pretty much a no brainer. I myself just had a fun bout with upgrading our internal translation system (for speed, accuracy, and sym ship), and I can tell you not having unicode in Win32 makes things a bit of a pain. I imagine it will be easy convincing the crew that this one needs to be done asap.
ThreadSafe VCL - With the influx of multicore technology into the market I can see the need for VCL to be threadsafe. We will definatly want to take advantage of the shifting architecture of processing power.
Remove Application Main Form - Wow would I love to do that. Unfortunatly I think I’m the only one who does. Being the junior VCL guy means I can make suggestions, but if I went tearing through the codebase to make a major change like that I might be in for it. I will continue to press the team for the removal of this, well, hack.
Thats the major stuff anyway. There were some requests for better databinding, as well as new controls like the ribbon controls. These things are important too, but theres a limit to the amount of work Seppy and myself can get up to inbetween releases. Anyway, thanks for the feedback. We will do as much as we can to keep VCL meaningful and useful in the future.
Share This | Email this page to a friend
Posted by David Lock on June 6th, 2007 under Uncategorized |
RSS Feed
June 6th, 2007 at 9:48 am
UNICODE VCL? Omg! this is a very good news! so when is the exact date that we can see and taste this???
ThreadSafe VCL is not really a big request though since i can live with the current style using the synchronize function.
Removing application’s main form? is there any known problems with this property? cannot figure out why you must remove it.
btw, thanks for heading codegear (especialy Delphi) to the correct path again.
June 6th, 2007 at 10:02 am
can delphi copy one good feature from VS?
creating a new custom component inside different project can be immediately seen inside the component list without installing it first into the delphi IDE?
note: this is not that important issue compare to the UNICODE VCL!!!!
June 6th, 2007 at 10:13 am
Win32 Unicode: "I imagine it will be easy convincing the crew that this one needs to be done asap."
David, you must have been hiding under a rock these last few years
Seriously, I’m glad the message has finally got through to TPTB about this one, although I expect it’s not the "no-brainer" to implement that your post implies…
June 6th, 2007 at 10:32 am
Replace hiding under a rock, with finishing my college degree and you’d be accurate :). I’m fairly new to CodeGear (at least compared to my coworkers). A little over 3 years in the company, with one of those as an intern as I was finishing up college. All things considered I have some impact on the decision making process around here. Well, a little.
And yeah, the no brainer is that Unicode must be done. The implementation is going to be quite intense, especially since we would like to break you guys as little as possible. I doubt we can make it seamless, but I don’t want a Unicode switchover to devastate existing delphi codebases.
June 6th, 2007 at 12:25 pm
Sooo, are you going to ignore ActionsBars for this release as well? What the hey, its only been broken, what, 5-6 years?
June 6th, 2007 at 1:08 pm
It looks like you have very small team for VCL development. This is bad message for Delphi user.
June 6th, 2007 at 4:06 pm
I would agree that it is a small team. However the choice is a small team or no team. We simply don’t have the resources to employ a large team to work on any area of the product. We do what we can with what we have.
Also please note that I am not a product manager, as such this is not a list of features going into future releases. So don’t burn me for the presence of items which don’t make the next release, or the absense of things you think should really be fixed (eh Lasse?).
June 6th, 2007 at 4:28 pm
While there will always be those of us that have our pet peeves about the VCL, I think you have hit on the a couple of good starting points (Unicode and Threads).
Ignore the clatter from the sidelines, any work done to improve the VCL and the tools we use is welcome.
Thanks
Michael
June 6th, 2007 at 6:10 pm
Unicode VCL… that’s good and how about porting ECO to Win32 after the improved data binding? I know this is not a VCL issue but ECO is the only reason pushing me to the .NET camp…
June 6th, 2007 at 8:16 pm
David said:
>>> I would agree that it is a small team. However the choice is a small team or no team. We simply don’t have the resources to employ a large team to work on any area of the product. We do what we can with what we have.
(hint) Let the showmen stop touring around the world for product launches and you’ll have plenty on money to put on good spendings…(hint)
People sees and cares about what you (I mean CG) do with the money they pay on licenses.
June 7th, 2007 at 5:32 am
For me it is a clear signal to leave Delphi. I’m not happy, it is a lot of work (Delphi -> VS), but I don’t want to wait for time, when CodeGear begin to produce stable and usefull product.
If you compare Delphi x VS, I’m afraid that Delphi cannot reach VS in a few years.
June 7th, 2007 at 7:01 am
Peter Schutz: Delphi 2007 is the best Delphi IDE ever. Even better than Delphi 7. And VS is no competition whatsoever. It doesn’t even natively support Vista development, and won’t until 2008 according to Microsoft. But if you have users that don’t need the latest features, then bye bye now.
Sternas Stefanos: Lazarus+FreePascal is so not ready for real use. Just doing a File|New Application and th en F9 to run takes a ridiculously long time (Pentium D 3.6 Mhz, 2GB RAM); there’s no way you can do serious development using these tools. If you can get by with them, you’re not doing any real work.
And for all of you people complaining about not having Delphi 64 bit yet… What would you use it for? The only applications that benefit from64 bit are servers and extremely large memory hogging applications. How many people are writing those?
June 7th, 2007 at 7:32 am
Ken White: Delphi 2007 is the best Delphi IDE ever.
Yes, you are absolutely right. But try thinking about product (Delphi) development. Try to remember VCL. What new features for 5 years? How many new visual components? A lot of distributed 3rd party components (Nevrona, Indy) instead of own solution. Is it a solid and dynamic platform? What I miss is >progress< in VCL.
I’d like to have robust and supported platform, but VCL is not in fact. And if the development is 2 or 3 men job, I can wait for a long time.
June 7th, 2007 at 9:24 am
Sure, the latest VS is a very fast & stable IDE. but what type of languages that VS.Net can produce for a a super RAD language like delphi? both c# & vb.net can only be run on net platform and not as fast as native (btw, how many commercial applications are out there written for net? even M$ produces their main products using native! and .net is not cross platform like java!). If you are talking about vb (not sure whether you can compile it in VS.net though), this is the language that MS doesnt want to continue anymore, they already abandoned vb totally! how about c/c++? c/c++ language can produce a super fast native code but there is no RAD!
oh! how about 64bit? i totally aggreed with ken white! only server applications currently looking at 64-bit! I bet most of ‘developers’ that saying out loudly to run away from delphi for 64bit are just to become the first developer to release a small utility or small game in 64bit?
June 7th, 2007 at 9:34 am
Ken White:
"there’s no way you can do serious development using these tools."
"64 bit yet… What would you use it for?"
Who is serious? &
What is serious?
Only time can say…
June 7th, 2007 at 11:37 am
When Delphi appeared on a market for a first time it was a need to create basic components to "define" the development. Now many different vendor offer better treeview, better grid etc. I believe that CodeGear VCL team should focus on a basic work rather then compete against 3rd party commercial or freeware vendors. Basic work is Unicode, 64bits, improved VCL.NET, more collections in Contnr, Canvas based on GDI+ (or preferable optional GDI vs GDI+)
June 7th, 2007 at 3:40 pm
Yes, but imagine the following situation. You have big project with about 400 forms and 500 grids from different vendor. One day the different vendor close its business. Did you ever try to change all grids in such application? There is a problem with dependency on severe vendors. One Delphi version is supported but next one - who knows? Do you have receipt for choosing right vendor?
June 7th, 2007 at 3:51 pm
Dear David Lock
I have MANY Delphi VCL code patches and bug fixes + new VERY HANDY features implemented in Delphi that I think you could use.
How may I help you there?
Please leave me an Email if you’re interested…
June 8th, 2007 at 3:48 am
Please focus on the base! VCL is a system which can easily be extended by others with components. So let the others do this! Why build a ribbon when there are stable versions out there?. But what cannot be done by others is changing the base and core features. That’s where CodeGear has the strength. So do things like UniCode, Threadsafe VCL and the Canvas-Implementation.
June 8th, 2007 at 6:10 am
Second CORE and INFRASTRUCTURE.
Leave components to 3d-parties.
June 8th, 2007 at 1:12 pm
After a difficult conclusion we decided to difinetely leave Delphi and move our project to VS. When I see current "progress" of Delphi, I don’t regret. Another activities of CG (PHP, Ruby) is good, but it causes small resources to Delphi, what is highly visible.
June 9th, 2007 at 6:15 am
Borland bought a lot of companies that developed third-party tools that integrate with their main products (Together, Caliber, StarTeam, etc.). Why not do the same thing about components? There are several companies that do a great jobs on developing great components for a long time, components that, in most of cases, were developed in the lack of functionality of Borland’s ones.
I think Borland could give them a try.
September 10th, 2007 at 9:20 pm
Third CORE and INFRASTRUCTURE !
No need to compete with vendors.
Concentrate on Win32 as much as .NET
( I do not find .NET more useful )