Board index » delphi » Re: The New Roadmap
Dave Nottage [TeamB]
Delphi Developer |
Dave Nottage [TeamB]
Delphi Developer |
Re: The New Roadmap2007-06-12 03:46:18 PM delphi185 Nils Boedeker writes: QuoteIs Unicode really more urgend than 64Bit ? research to determine which technologies have more demand and/or will return greater ROI. -- Dave Nottage [TeamB] |
Anders Isaksson
Delphi Developer |
2007-06-12 04:06:33 PM
Re: The New Roadmap
Michael Anonymous writes:
QuoteOne needs some kind of empirical evidence to make sound judgments. Other places have other needs, but I sincerely belive that a majority of the Delphi users are *not* in an immediate need of 64-bit compiling. Interesting? Yes. Nice to have? Definitely! Compulsory? No, not for a long time. For the moment, I don't need .NET either, but I do think it is much more important for CodeGear to catch up with .NET 2.0/3.0 than to release a 64-bit compiler right now. -- Anders Isaksson, Sweden BlockCAD: web.telia.com/~u16122508/proglego.htm Gallery: web.telia.com/~u16122508/gallery/index.htm |
m. Th.
Delphi Developer |
2007-06-12 04:11:52 PM
Re: The New Roadmap
Nick Hodges (CodeGear) writes:
QuoteCaptain Jake writes: Imho, the validity of an IT ecosystem is given by the crowd not by the scientists, even if scientists is proved that they 'knows more'. We're returning again and again to the great danger of delusion, and imho, the safest way to escape from here is asking. In this way we simplified a lot our business and found that the users usually want simple and natural things (much simpler than our in-house plans) and they are very keen and trustworthy about their *needs* but very dangerous when it comes to express their *solutions*. The key, imho, is to make them to express what is their real pain, their human problem, in human terms, not their solution. And now to judge myself: For example, (some of) our needs: We have (rather) big (imho) business applications (ERP, CRM stuff like that) on Win32 using Firebird as backend, IBO as connectivity (BO is tight integrated here), InfoPower and JVCL as PL. Because we found that our agility/response time is crucial here (and in fact, one of the few reasons which keeps us up in front of tyrannosauruses like M$ Dynamics, SAP and the like) we need on Win32 tools which can allow us: - change/refactor our ecosystem in a quick and centralized way. (DB back end, connectivity, BO and PL) - manage our code better (to know what is where - what dependencies are where, where are our weakpoints aso.) - increase the quality of our code - allow us to enhance our RAD system. (RAD = Rapid Application _Deployment_). And here with "deployment" I don't mean, for ex., a better installer but an entire enforcement of good programming practices from Pascal Language/IDE to raise the quality bar of our IT ecosystem. (code, specifications, testing aso.) (and of course we need from CG to help us to walk on water, but imho, this is obvious... :-) ) And now our solutions: You are walking now through the entire VCL, doing the Unicode. So, in order to gain time, you need coverage. So, if you are there do simultaneously (if is possible) the ECO for Win32 (QC #46720). This implies, I think, enhancing the RTTI bringing it (at least) closer to reflection or to a compile-time checked reflection (approx. QC #46581) (research.microsoft.com/os/singularity/publications/OSR2007_RethinkingSoftwareStack.pdf) also this implies enhancing the data binding layer, (but because you are on your own in VCL, perhaps the things should be easier). Also, of course, a Firebird DBX driver will be ok to have... :-) (QC #41034 + QC #40916) ...and, of course, almost forgot a new data-aware control set... (www.devexpress.com or www.tmssoftware.com or...) ...and prgraming by contract... (QC #2444 + QC #46801) ...and... ...and if you think that you don't have work to do, please tell me, I am eager to tell you more... ;-) Being serious again, I think that it is crucial to steer on users needs, not on their solutions, even if is necessary to hear also the latter ones. And you must find a way to find them. QC isn't (yet) the way, perhaps of UI (not so easy to work with), perhaps because users feel that their items from there aren't taken in consideration - here I understand CG on some items, but also I understand them. OTOH, the surveys proved (at least) sometimes a dangerous solution, so you have to talk, to observe, to judge. In fact we are just men. See for this www.codinghorror.com/blog/archives/000880.html Also read the comments and follow the links from there ...especially blogs.msdn.com/jensenh/archive/2005/11/07/489864.aspx (It's a little lesson for all of us about how much can do a team to be obedient - I hope that it will not be a complete waste of time). my 2c, m. th. |
m. Th.
Delphi Developer |
2007-06-12 04:38:29 PM
Re: The New Roadmap
Nick Hodges (CodeGear) writes:
Quotem. Th. writes: In this way you will gain the confidence which will lower the pressure upon you. And, imho, is important to have asap early builds to provide that 'something is happening'. Also, perhaps a good thing is to constantly update the roadmap with the features added/changed/removed. It will keep the community close, imho. inc(Nick, 0.02); // ;-) m. th. |
yannis
Delphi Developer |
2007-06-12 04:51:21 PM
Re: The New Roadmap
Nick Hodges (CodeGear) presented the following explanation :
QuoteDennis Cote writes: searching in the MS SDK help. In most cases we know in general that we need to find something in SDK or in Delphi help and I am used to this from an add in for Delphi 6 which searches in the SDK. For me this is acceptable and in some cases better than the current Dexplore thing. Regards Yannis. |
Daniël Mantione
Delphi Developer |
2007-06-12 04:56:24 PM
Re: The New Roadmap
Nils Boedeker <XXXX@XXXXX.COM>writes:
Quote>>It's right after the Unicode release. architecture. Daniël Mantione |
IanH
Delphi Developer |
2007-06-12 05:02:37 PM
Re: The New Roadmap
Nick Hodges (CodeGear) writes:
QuoteCaptain Jake writes: QuoteHighlander is focused more on solutions and less on QuoteHighlander is for people who want to build websites, it does: it can "consume" them..... I recognise the requirement to finally deliver a .NET2 product, but you HAVE to take a serious look at your .NET strategy after this. It has failed. Worse than that, it is compromising your ability to deliver for native developers. Ian |
IanH
Delphi Developer |
2007-06-12 05:02:40 PM
Re: The New Roadmap
Brian Moelk writes:
QuoteCaptain Jake writes: also purchase Highlander, and what proportion hold off until Tiburon? The .NET strategy is a complete wash-out. What state would native code development be in if Borland/CodeGear had been smart enough to support .NET via interop, plus nice new features like generics? They struck out on Linux. They struck out on .NET I may be English, but even I know how many strikes you are allowed before you are out - which is why I didn't mention CBuilderX ;-) Ian |
IanH
Delphi Developer |
2007-06-12 05:02:48 PM
Re: The New Roadmap
Captain Jake writes:
Quote
- committed to .NET - bit off more than they could chew - realised it was a mistake - have to push this out before they can reconsider the .NET approach Sadly, I think that they are left with no choice at this point. The lag on .NET is currently an embarrassment: they realised that native development was more important and so Delphi 2007 was released out of sequence, as it were; but they cannot do that again. They need to at least get 2.0 support out of the door before they can change direction: even though people with a real need for .NET2 support left a long time ago. In the meantime, a big bunch of Win32 programmers are wondering exactly what is in Highlander for them, and making plans to skip it in favour of Tiburon. And the Win64 crowd are realising that they are not going to see anything in the next couple of years. How many are prepared to wait that long? Ian |
Rod
Delphi Developer |
2007-06-12 05:07:57 PM
Re: The New Roadmap
Francois Malan writes:
QuoteMay I also express my wish that when the 64-bit compiler is written, |
Eric Grange
Delphi Developer |
2007-06-12 05:30:57 PM
Re: The New RoadmapQuoteI do "Business Web application development" and I told you. 64Bit is QuoteAll of our new servers are Win64 servers and we must distort us and the machines (but there is a distinct lack of enthusiasm for Vista, so they dual boot XP64). Eric |
Nils Boedeker
Delphi Developer |
2007-06-12 05:42:45 PM
Re: The New Roadmap
Hi
QuoteNot just servers are Win64 here, desktop and laptops too. Nils |
Nils Boedeker
Delphi Developer |
2007-06-12 05:55:55 PM
Re: The New Roadmap
Hi,
Quote
For our choice we need Unicode if wie want to write applications for areas with other charsets or with charsets that are not map by 8-bit, for example china. Yes... this you can not solve with the actual Delphi but there are workarounds also for this. Basicly we use Unicode symbols at the time as XML Entities. As a publishing house with a lot of printworks (Indesign based). Other situation with 64Bit... we NEED 64bit to be up to date in server bussiness and there is no really solution/workaround... Because of this 64bit is much more importend than Unicode because for Unicode exist a workaround, for 64bit exit nothing like this. Nils |
m. Th.
Delphi Developer |
2007-06-12 06:54:51 PM
Re: The New Roadmap
Brian Moelk writes:
QuoteCaptain Jake writes: Even if their way of thinking is good (*now*), and I agree with that way (see my other post), pushing .NET any further now is a very very very disputable matter from both human and technological POV. In a few years perhaps... Applying that way of thinking on native is an entire other story. my 2c, m. th. |
m. Th.
Delphi Developer |
2007-06-12 07:02:44 PM
Re: The New Roadmap
IanH writes:
Quote
Of course isn't much but it is a clear sign that they are at least looking into that. And they went deep enough to encounter some problems. hth, m. th. |