Board index » delphi » Re: About Entity-Relationship Diagram in BDS 2007

2006-05-19 11:06:45 PM
delphi276
streamed and placed in a blob/clob, ala bitmap, except it is not using a
standard format - and why should it be, the OPF application is perfectly
happy with it, it is super-convenient and super-fast for it to store
things that way...
Listed other variations in another post in this thread.
Eric
 

Eric Grange writes:
way. I don't know if it is changed since I looked at it.
--
Craig Stuntz [TeamB] ?Vertex Systems Corp. ?Columbus, OH
Delphi/InterBase Weblog : blogs.teamb.com/craigstuntz
Useful articles about InterBase development:
blogs.teamb.com/craigstuntz/category/21.aspx
 

Cheers,
Jim Cooper
_____________________________________________
Jim Cooper XXXX@XXXXX.COM
Skype : jim.cooper
Tabdee Ltd www.tabdee.ltd.uk
TurboSync - Connecting Delphi to your Palm
_____________________________________________
 

going to mention them, anyway. I am afraid I don't quite understand the point of
your example either, or what you meant by the "too cheap" comment.
irritating and one reason why people often don't take you seriously.
If I had wanted to make such a ridiculous claim I'd have done so. I clearly
did nothing of the sort.
Cheers,
Jim Cooper
_____________________________________________
Jim Cooper XXXX@XXXXX.COM
Skype : jim.cooper
Tabdee Ltd www.tabdee.ltd.uk
TurboSync - Connecting Delphi to your Palm
_____________________________________________
 

Cheers,
Jim Cooper
_____________________________________________
Jim Cooper XXXX@XXXXX.COM
Skype : jim.cooper
Tabdee Ltd www.tabdee.ltd.uk
TurboSync - Connecting Delphi to your Palm
_____________________________________________
 

techniques in procedural languages.
This is not something over which there can be disagreement - it is fundamental
computer science.
Cheers,
Jim Cooper
_____________________________________________
Jim Cooper XXXX@XXXXX.COM
Skype : jim.cooper
Tabdee Ltd www.tabdee.ltd.uk
TurboSync - Connecting Delphi to your Palm
_____________________________________________
 

RDBMSs. It is a problem all over the place.
Set theory is an issue in the sense that it is why there is the "impedance
mismatch", since that is why there are the structures (data and logical) there
are in RDBMSs.
slow external storage for our persistent data. I am not convinced about OODBMSs
either. If memory was non-volatile and sufficiently large, I think we would use
other techniques again.
not, isn't it? And an RDBMS doesn't precisely solve all issues to do with that
either.
an RDBMS as the backend, otherwise there would be little value in the OPF.
Cheers,
Jim Cooper
_____________________________________________
Jim Cooper XXXX@XXXXX.COM
Skype : jim.cooper
Tabdee Ltd www.tabdee.ltd.uk
TurboSync - Connecting Delphi to your Palm
_____________________________________________
 

whilst cutting a piece of wood, it does not mean that a saw is no good for
cutting wood.
certainly always been able to get at data, and I have certainly used badly flawed
OPFs. Try explaining what you mean, instead of repeating the same statement.
can make out.
discussion go by without leaping hugely to incorrect conclusions.
sufficiently. I'd argue that you should only do that when you are accessing
an existing database. Otherwise it is better to design your objects first, since
they are the reflection of your problem domain.
I don't understand what you mean about guaranteeing the accessibility of the
data. How could that fail to be the case?
Cheers,
Jim Cooper
_____________________________________________
Jim Cooper XXXX@XXXXX.COM
Skype : jim.cooper
Tabdee Ltd www.tabdee.ltd.uk
TurboSync - Connecting Delphi to your Palm
_____________________________________________
 

obviously *would* speed up things considerably.
Do you not understand why there are such things as databases in the first place?
Cheers,
Jim Cooper
_____________________________________________
Jim Cooper XXXX@XXXXX.COM
Skype : jim.cooper
Tabdee Ltd www.tabdee.ltd.uk
TurboSync - Connecting Delphi to your Palm
_____________________________________________
 

does it with lists of objects). I also know of one RDBMS vendor that has done
that in the past.
Cheers,
Jim Cooper
_____________________________________________
Jim Cooper XXXX@XXXXX.COM
Skype : jim.cooper
Tabdee Ltd www.tabdee.ltd.uk
TurboSync - Connecting Delphi to your Palm
_____________________________________________
 

problem the evaluation went no further.
Cheers,
Jim Cooper
_____________________________________________
Jim Cooper XXXX@XXXXX.COM
Skype : jim.cooper
Tabdee Ltd www.tabdee.ltd.uk
TurboSync - Connecting Delphi to your Palm
_____________________________________________
 

Jim Cooper writes:
the call stack, to judge from the quizzical replies I get when I ask
people to post the call stacks from problems they're having. The CPU
window is perhaps the most terse of all of the power tools, but I'd
hate to have a version of Delphi which included only what the "average"
user needed.
breakpoints midline, reset the EIP, etc.
--
Craig Stuntz [TeamB] ?Vertex Systems Corp. ?Columbus, OH
Delphi/InterBase Weblog : blogs.teamb.com/craigstuntz
How to ask questions the smart way:
www.catb.org/~esr/faqs/smart-questions.html
 

Jim Cooper writes:
from the ground up, on a theory-grounded approach to provable
concurrency. They are designed to make it difficult to implement
systems which can not scale. With OO frameworks, it seems to be that at
best you get some features which are useful when you want to think
about concurrency, but you are never required to use them. When you do
choose to use them you are typically required to connect a lot of
plumbing (think XA) and approach your application in a completely
different way than you would do hi world.
* Yes, I am excluding MySQL and the like. You get the point!
relational database can be expected to be aware of normal forms and
Codd's principles, there is no widely-agreed upon, theoretical
foundation for OO design. Yes, there are a lot of bad database designs
in the world, and any tool can be misused. But you can chalk that up to
simple ignorance. If you don't understand relational principles, you
don't understand the tool.
With OO systems, OTOH, you can not call someone ignorant if they've read
the GoF but not Meyer. it is not fair to do so because not everyone
agrees with Meyer (or the GoF). And even if you do, their output can
best be described as "well-reasoned," not grounded in theory and
provable.
Keep in mind, I am a fan of OO. I am not trying to give it any
disrespect or suggest that we all revert to procedural designs. But it
has substantial, longstanding shortcomings which RDBMSs address very
well. Why hasn't OO learned more from RDBMSs? To me it is in no small
part because of the silly notion that OO is "better" than relational
architectures, and hence couldn't possibly learn anything from them.
speed. The day you're waiting for came years ago. It hasn't made
OODBMSs any better.
tools you can use to solve the problem, and OO frameworks, for the most
part, don't.
ECO. But I don't pretend that the DB isn't there, which is the point of
my initional reply in this thread.
--
Craig Stuntz [TeamB] ?Vertex Systems Corp. ?Columbus, OH
Delphi/InterBase Weblog : blogs.teamb.com/craigstuntz
Everything You Need to Know About InterBase Character Sets:
blogs.teamb.com/craigstuntz/articles/403.aspx
 

implement some form of concurrency control, as far as the server ap itself is
concerned.
here? There is more to OO than OPFs, which are only needed if you want to store
objects.
foundation for procedural languages in general.
can be used. Comparing learning relational principles with reading 2 books is
not comparing the same type of thing at all.
And if you want to read something by someone who doesn't understand basic
principles, read anything Date on OO.
comparing things that are not particularly similar, and are intended to address
different problems.
same speed on my network. Although I'd like one at home.
all. You always have to do extra work yourself.
And since OPFs generally use RDBMSs for storage, they can take advantage of what
the RDBMS offers in this regard. I think this is a red herring.
of the time. Nobody is suggesting you should pretend the database is not there
all the time, but the more you can do that the better.
Cheers,
Jim Cooper
_____________________________________________
Jim Cooper XXXX@XXXXX.COM
Skype : jim.cooper
Tabdee Ltd www.tabdee.ltd.uk
TurboSync - Connecting Delphi to your Palm
_____________________________________________
 

Jim Cooper writes:
old papers simply packed with information useful to anyone doing this
sort of work. They're theoretically solid and haven't really aged.
Here's a great place to start:
research.microsoft.com/pubs/ccontrol/
how they approach concurrency. Papers by D.P. Reed, Jim Grey, Philip
Bernstein. What I write above is no hyperbole; whether or not the
systems are in fact successful at their aim they are quite literally
designed around these principles. This is *not* true of OO frameworks
in general.
Lisp is based on the lambda calculus, for example.
The reason it is a problem to have structures not based on a solid
theoretical foundation is that it becomes difficult to impossible to do
things which are provable.
J. Armstrong in the February 2006 issue of the JACM, a comprehensive
(and peer-reviewed) attempt to establish such a taxonomy -- from a wide
ranging survey of computing literature, not just people she happens to
know or what people say on certain newsgroups. There is, frankly,
precious little which is universally accepted, and *no* standard
notation for it. *None* of it deals with concurrency, and only
passingly does it deal with data structures.
claim that one is just a relic which will be obsolesced by the other.
the other, which started this discussion, is so absurd.
you like -- in-memory databases or whatever. The notion that HDDs are
why we need RDBMSs is based around the fallacy that they're merely a
storage system.
problem we're trying to solve is persistence, which I don't agree with.
closer to the mark. Read the first link in this post. Yes, it is long,
but worth it.
makes it *possible.*
database. It has features you need.
--
Craig Stuntz [TeamB] ?Vertex Systems Corp. ?Columbus, OH
Delphi/InterBase Weblog : blogs.teamb.com/craigstuntz
Useful articles about InterBase development:
blogs.teamb.com/craigstuntz/category/21.aspx