Board index » delphi » Re: Delphi Bugs

Re: Delphi Bugs


2004-07-31 03:11:05 AM
delphi76
Martin Waldenburg writes:
Quote
where ?
By Iman's post.
--
Nick Hodges -- TeamB
Lemanix Corporation -- www.lemanix.com
Read my Blog -- www.lemanix.com/nick
 
 

Re: Delphi Bugs

On 30 Jul 2004 12:21:05 -0700, Nick Hodges [TeamB] writes:
Quote
It's my view that the whole response arrogant and unhelpful.
Did you actually read the response? I think you need to take another look.
Several implementations were offered, and it is stated that it is by
design.
<quote>
Is it a bug or a feature (request)?
Bruce Perens points out to Thorpe, "That you might figure out if this
particular form of error checking is required by the ELF standard, or
any other standard that applies to the linker. If it is not part of a
standard, this is a feature-request rather than a bug."
In a later message, Perens said the loader used by Linux behaves
exactly the same as the loader he used when developing on an SGI 3D
machine, meaning that it is a Unix thing, not a Linux thing.
Miguel de Icaza seconds that when he notes "Loading a shared object in
Linux is very flexible, but I am afraid the people from Borland are
not aware of how this is done on Unix. In Unix, you have to establish
your own convention on 'plugins', this convention is something that
each application establishes."
Eric Raymond said, "Part of it may be that Thorpe still thinks in
(Borland) Pascal. In C, libraries don't self-initialize -- that is, we
call library initialization routines explicitly at runtime rather than
implicitly at library load-time."
</quote>
 

Re: Delphi Bugs

Nick Hodges [TeamB] schrieb:
Quote
Martin Waldenburg writes:


>where ?


By Iman's post.
as "designed" is not a bug,
Borland has told us that quite often.
Martin
 

Re: Delphi Bugs

On 30 Jul 2004 12:25:40 -0700, Nick Hodges [TeamB] writes:
Quote
Thomas Edison writes:

>It seems to me that there is a small group of people here who will do
>just about anything to bash Linux and OSS in general because they are
>afraid of it.

Who's in that group? Not me -- I don't "bash" Linux. I just comment
on it. It seems to me that you view any criticism at all of Linux as
"bashing".
Well, I take Jake's seemingly continuous pot-shot taking at Linux/OSS as
bashing it. I guess with his comments and maybe a few others' here, I was
presuming you were doing the same. Sorry.
 

Re: Delphi Bugs

Martin Waldenburg writes:
Quote
as "designed" is not a bug,
Borland has told us that quite often.
Here's Danny's summary:
"The Linux module loader doesn't provide a way for shared object
libraries to indicate that they were unable to successfully initialize
themselves. Danny Thorpe of the Delphi/Kylix R&D team explains how this
affects your Kylix applications."
Now, we can argue if this is a bug or a design flaw. I think it is a
bug. Saying that it is "as designed" is a total copout. Clearly it is a
problem that needs to be addressed. Linus seems utterly uninterested in
addressing it, preferring instead to insult the messenger. I note that
the response URL basically argues "There's nothing wrong here, this is
as designed, and people that point out that something is wrong here are
stupid whiners".
It's my view that the whole response arrogant and unhelpful. And of
course, the response, and indeed the response from folks here, both
strongly support Jake's original point: Not all Linux bugs get fixed
either, and if you try to say that Linux has bugs, you will be shouted
down.
And the response here confirms in my mind even more that Linux/Open
Source advocates will quarter /no/ criticism of their philosophy or
their precious OS.
--
Nick Hodges -- TeamB
Lemanix Corporation -- www.lemanix.com
Read my Blog -- www.lemanix.com/nick
 

Re: Delphi Bugs

Martin Waldenburg <XXXX@XXXXX.COM>wrote in news:410a9e09$2
@newsgroups.borland.com:
Quote

as "designed" is not a bug,
Borland has told us that quite often.

Martin

I'm sorry, I am not sure that I understand. Do you mean that explicitly
ignoring and returning the exit codes is "as designed"? If so, then I
assert that linux has a "design" bug, wouldn't you?
 

Re: Delphi Bugs

Quote
Not all Linux bugs get fixed either
may be true, however:
as "designed" is not a bug,
Borland has told us that quite often.
Martin
 

Re: Delphi Bugs

Thomas Edison writes:
Quote
It seems to me that there is a small group of people here who will do
just about anything to bash Linux and OSS in general because they are
afraid of it.
Who's in that group? Not me -- I don't "bash" Linux. I just comment
on it. It seems to me that you view any criticism at all of Linux as
"bashing".
Quote
It's annoying to read Jake spinning Linux/OSS in a bad light almost
every time the subject comes up. I am sure you guys wouldn't like it
if I did the same with Windows/CSS.
I'm sure it is annoying to you. Jake has some valid points, though, in
my view.
I don't appreciate you painting me with the broad brush of "I'm sure
you guys...". If you have valid criticisms of Windows/CSS(?) I might
agree with you, night not. I certainly have plenty of complaints about
Windows.
--
Nick Hodges -- TeamB
Lemanix Corporation -- www.lemanix.com
Read my Blog -- www.lemanix.com/nick
 

Re: Delphi Bugs

XXXX@XXXXX.COM (Larry Drews) wrote in
Quote
Martin Waldenburg <XXXX@XXXXX.COM>wrote in
news:410a9e09$2 @newsgroups.borland.com:

>
>as "designed" is not a bug,
>Borland has told us that quite often.
>
>Martin
>

I'm sorry, I am not sure that I understand. Do you mean that explicitly
ignoring and returning the exit codes is "as designed"? If so, then I
assert that linux has a "design" bug, wouldn't you?

Mea culpa, I meant to say "ignoring and not returning the exit codes".
 

Re: Delphi Bugs

Larry Drews schrieb:
Quote
ignoring and returning the exit codes is "as designed"?
It's simply the way Unix does it.
Linux is an Unix clone.
Martin
 

Re: Delphi Bugs

Martin Waldenburg <XXXX@XXXXX.COM>wrote in
Quote
Larry Drews schrieb:
>ignoring and returning the exit codes is "as designed"?

It's simply the way Unix does it.
Linux is an Unix clone.

Martin

OK, but to me that is like saying "My uncle is a thief, but that just the
way he is" to justify why he isn't locked up.
 

Re: Delphi Bugs

Thomas Edison writes:
Quote
Did you actually read the response?
Please don't insult me like that, okay?
--
Nick Hodges -- TeamB
Lemanix Corporation -- www.lemanix.com
Read my Blog -- www.lemanix.com/nick
 

Re: Delphi Bugs

Martin Waldenburg writes:
Quote
as "designed" is not a bug,
Borland has told us that quite often.
Yes, this is true, and it won't get any truer by you repeating it.
I'll say again -- calling this "as designed" is a copout. It may very
well be "as designed", but if so, it is a really bad design.
--
Nick Hodges -- TeamB
Lemanix Corporation -- www.lemanix.com
Read my Blog -- www.lemanix.com/nick
 

Re: Delphi Bugs

Thomas Edison writes:
Quote
I guess with his comments and maybe a few others' here, I was
presuming you were doing the same. Sorry.
No problem.
--
Nick Hodges -- TeamB
Lemanix Corporation -- www.lemanix.com
Read my Blog -- www.lemanix.com/nick
 

Re: Delphi Bugs

"Nick Hodges [TeamB]" <XXXX@XXXXX.COM>wrote in news:410a9c6a$1
@newsgroups.borland.com:
Quote
Yes, I do -- Isn't that what the open source community constantly
touts?
No, OS touts if there are bugs or features you don't like, here's the
code fix it yourself. If you can not do that, pay someone to fix it for
you.
Quote
I don't agree -- it is a bug. Danny quite clearly showed how the
loader fails to honor valid exitcodes.
Nick it does not work the way windows works, it works the way unix
works. That does not make it a bug. Several work arounds were
provided. Apparently it wasn't too big a deal, Borland didn't want to
implement any of the work arounds.
You'll see "as designed" in QC too. If you read all the comments
you'll see changing the loader would break existing code. you will see
that as reason not to change something in QC also.
Calling someone stupid is IMO childish. I do have a rule about online
communication: "Take everything you read online with a pound of salt."
It could have easily been a tease.
--
Iman