Board index » delphi » Re: [64-bit Delphi] A msg from Hairy (van Tassel)

Re: [64-bit Delphi] A msg from Hairy (van Tassel)


2005-03-18 05:11:23 AM
delphi253
Nick Hodges [TeamB] writes:
Quote
No one is arguing that.
Kind of. "Buggy compiler" was used to mean plain Delphi 7 in some places in
this subthread.
Quote
The point is that people are now claiming
that the updated compiler is buggy as a result of the bug fixes. I've
yet to see any evidence of this.
Well, you can call it what you want. The fact that it generates less
efficient code than plain Delphi 7 or D2005 still remains.
German's point is that he has to choose between a buggy compiler (plain
D7, see above), or a compile that generates innefficient code (updated
D7).
And, as I said, I kept plain Delphi 7 and I am not complaining.
--
Leonel
 
 

Re: [64-bit Delphi] A msg from Hairy (van Tassel)

Leonel writes:
Quote
Rudy Velthuis [TeamB] writes:

>The bugs were not caused by the codegen changes.

Did I say that anywhere?

>Codegen changes mean
>a recompile is necessary, and packages delivered as dcu/bpl/dcp only
>don't work.

Nope, that is not true. Delphi 7 upd1 is dcu-compatible with D7. That
means those components delivered without source still work.

Codegen changes mean that the generated code is different. DCU format
is the same.
Codegen can also mean that the dcu format is different. If Danny meant
the affaction of optimization, then people can do what was suggested:
install the update, but keep your old dcc70.dll and dcc32.exe. There
were not that many problems with D7'd compiler after all.
--
Rudy Velthuis [TeamB] rvelthuis.bei.t-online.de
"Outside of a dog, a book is man's best friend. Inside of a dog, it's
too dark to read." -- Groucho Marx
 

Re: [64-bit Delphi] A msg from Hairy (van Tassel)

Leonel writes:
Quote

Well, you can call it what you want. The fact that it generates less
efficient code than plain Delphi 7 or D2005 still remains.
That isn't in dispute, eh?
--
Nick Hodges -- TeamB
Lemanix Corporation -- www.lemanix.com
Read my Blog -- www.lemanix.com/nick
 

Re: [64-bit Delphi] A msg from Hairy (van Tassel)

German Pablo Gentile writes:
Quote
Amazing post.
Indeed, it was. Very well put.
--
Nick Hodges -- TeamB
Lemanix Corporation -- www.lemanix.com
Read my Blog -- www.lemanix.com/nick
 

Re: [64-bit Delphi] A msg from Hairy (van Tassel)

Rudy Velthuis [TeamB] writes:
Quote
Which means that all dcus, bpls and dcps generated were incompatible
with previously generated ones.
No, it doesn't. As I replied above, Delphi 7 dcus are compatible with D7
upd1.
You might be confusing it with the latest Delphi 8 update.
Quote
Not a bug, merely a nuisance, since it
means a recompile. One must do the same when changing major versions.
No, that is not point, again. The point is in QC#8695. Calling it a bug
or not, doesn't change it, it is an issue.
(And, even though I still use D7, I am not expecting it to be fixed)
--
Leonel
 

Re: [64-bit Delphi] A msg from Hairy (van Tassel)

Rudy Velthuis [TeamB] writes:
Quote
Codegen can also mean that the dcu format is different.
Yes, but in this case, it doesn't.
--
Leonel
 

Re: [64-bit Delphi] A msg from Hairy (van Tassel)

Nick Hodges [TeamB] writes:
Quote
That isn't in dispute, eh?
Exactly! Neither calling it a bug or not.
What's in dispute here is that both choices (keeping the first version,
or updating) have problems associated with them.
--
Leonel
 

Re: [64-bit Delphi] A msg from Hairy (van Tassel)

Leonel writes:
Quote
No, that is not point, again. The point is in QC#8695. Calling it a
bug or not, doesn't change it, it is an issue.
OK, OK, I got it. <g>
--
Rudy Velthuis [TeamB] rvelthuis.bei.t-online.de
"The world is a tragedy to those who feel, but a comedy to those who
think." -- Horace Walpole (1717-1797)
 

Re: [64-bit Delphi] A msg from Hairy (van Tassel)

Leonel writes:
Quote
Neither calling it a bug or not.
Actually, that is precisely what is in dispute.
--
Nick Hodges -- TeamB
Lemanix Corporation -- www.lemanix.com
Read my Blog -- www.lemanix.com/nick
 

Re: [64-bit Delphi] A msg from Hairy (van Tassel)

Nick Hodges [TeamB] writes:
Quote
Actually, that is precisely what is in dispute.
Allright, if you want to make it so, and ignore other point that I
think I made clear enough, fine by me. I am pretty that was German's
problem, not if it a bug or not.
--
Leonel
 

Re: [64-bit Delphi] A msg from Hairy (van Tassel)

Leonel writes:
Quote
I'm pretty that
pretty sure. <g>
--
Leonel
 

Re: [64-bit Delphi] A msg from Hairy (van Tassel)

Craig Stuntz [TeamB] writes:
Quote
Did it occur to either one of you that Borland just bought a company
which makes a really good VCS and might, in the spirit of eating
their own dog food, have changed over to it? Or that changing VCSs
(we just did it here) is complicated at best and back-porting changes
from one VCS and IDE system to another (remember: many of these
changes had already been implemented for D2005) is even more
interesting?
I have the feeling that you take this a little personal? Right?
For me, nothing in particular occurs at all, other than that the ideal
situation is to be able to roll back to any date, almost immediately.
I am puzzled when hearing that a company of a certain size and
importance does not have complete snapshots, let us say per major build,
of the total development environment, including tools and all settings,
all files, and so on. Making it possible to recreate the situation at
any time, like a time machine.
Versions systems based on a per file approach may end up in a mess, as
you indicate, especially when switching to another system, making the
need for snapshots even more evident.
Quote
Sorry, your messages both read as arrogant and presumptious to me.
And yours come out as if you were the dev. manager getting nagged :)
Quote
Evidence -- namely, the service pack itself, indicates that they in
fact did it.
Point is, it was not easily done, in an automated way.
Quote
But if you don't understand why it took some longer than planned to
do, then please save your laughter for things you do understand.
Please save your anger, it is a little misplaced here IMO.
--
Ingvar Nilsen
www.ingvarius.com
 

Re: [64-bit Delphi] A msg from Hairy (van Tassel)

Nick Hodges [TeamB] writes:
Quote
German Pablo Gentile writes:


>>You have to make the call stack window visible!
>
>Thats the Nick trick: every call (post) hide the original call.Keep
>the stack call window closed.


Comments like this are so unnecessary, aren't they?
FTR, I was trying to make a joke, to lighten up a little :)
--
Ingvar Nilsen
www.ingvarius.com
 

Re: [64-bit Delphi] A msg from Hairy (van Tassel)

Nick Hodges [TeamB] writes:
Quote
Leonel writes:


>Neither calling it a bug or not.


Actually, that is precisely what is in dispute.
Call it a butterfly
--
Ingvar Nilsen
www.ingvarius.com
 

Re: [64-bit Delphi] A msg from Hairy (van Tassel)

Nick Hodges [TeamB] writes:
Quote
Nope, sorry, the loss of a feature is not a bug.
I think that would depend on the intention of Borland.
Did they intend to get rid of that feature or not?
( Also, did their customers expect this feature loss? )
If they didn't, then I am very sure most people would say it is a bug.
And, honestly, they would be correct in their usage of the word.
A bug in the dictionary sense of the word is an
"an unexpected defect, fault, flaw, or imperfection"
.
(Source m-w.com, bug[2,noun])
The UNEXPECTED loss of a feature is considered a defect and therefore "a bug".