Two quick questions

Two quick questions

[Andrey Smirnov]

Hi,

I'll be quick:
     1. Since neither
     http://www.sourceware.org/gdb/current/onlinedocs/gdbint/Coding-Standards.html
     nor http://sourceware.org/gdb/wiki/JoelsCodingStyleCheatSheet
     mentions this, or at least yours myopic truly haven't noticed it, I
     wonder what is the GDB's allegiance in the Tabs, Spaces, Tabs +
     Spaces holy war? Who are the infidels we must crush? :) The reason
     I'm asking is that I don't want to generate false diffs just
     because I re-indented the line in different style.

     2. The second question is about zroot@sourceware.org.
     I've been told by it, and I quote:

     > You have been assigned edit rights to the gdb GNATS bug reporting
     > database. Your password is "fluttershy". Please change it as soon as
     > possible.

     Where is that "gdb GNATS bug reporting database"? I can't help but
     feel guilty for being a security threat, but I just can't find it.
     There is a Bugzilla, I've seen it, but those two don't seem to be
     related.

Andrey Smirnov

Re: Two quick questions

[Joel Brobecker]

>      1. Since neither
>      http://www.sourceware.org/gdb/current/onlinedocs/gdbint/Coding-Standards.html
>      nor http://sourceware.org/gdb/wiki/JoelsCodingStyleCheatSheet
>      mentions this, or at least yours myopic truly haven't noticed it,
>      I wonder what is the GDB's allegiance in the Tabs, Spaces, Tabs +
>      Spaces holy war? Who are the infidels we must crush? :) The
>      reason I'm asking is that I don't want to generate false diffs
>      just because I re-indented the line in different style.

We use tabs mostly, optionally followed by spaces if a line of code
doesn't exactly start at a tabulation. Although it's perfectly fine
to tabify a line of code that you're modifying, I would personally
prefer that you don't touch lines just for the sake of tabifying.
Others might disagree (hence the "personally"), but these little
changes don't bring anything in terms of readability, mess up the
indentation when reading diffs, and also potentially create merge
conflicts for changes that are essentially no-ops.

>      Where is that "gdb GNATS bug reporting database"? I can't help but
>      feel guilty for being a security threat, but I just can't find it.
>      There is a Bugzilla, I've seen it, but those two don't seem to be
>      related.

I don't know much about that one, and who might have created an
account for you. For GDB, we've transitioned to buzilla, as you
have found. We used to use gnats, but no more.

-- 
Joel

Re: Two quick questions

[Christopher Faylor]

On Mon, Nov 28, 2011 at 07:38:09AM -0800, Joel Brobecker wrote:
>>      1. Since neither
>>      http://www.sourceware.org/gdb/current/onlinedocs/gdbint/Coding-Standards.html
>>      nor http://sourceware.org/gdb/wiki/JoelsCodingStyleCheatSheet
>>      mentions this, or at least yours myopic truly haven't noticed it,
>>      I wonder what is the GDB's allegiance in the Tabs, Spaces, Tabs +
>>      Spaces holy war? Who are the infidels we must crush? :) The
>>      reason I'm asking is that I don't want to generate false diffs
>>      just because I re-indented the line in different style.
>
>We use tabs mostly, optionally followed by spaces if a line of code
>doesn't exactly start at a tabulation. Although it's perfectly fine
>to tabify a line of code that you're modifying, I would personally
>prefer that you don't touch lines just for the sake of tabifying.
>Others might disagree (hence the "personally"), but these little
>changes don't bring anything in terms of readability, mess up the
>indentation when reading diffs, and also potentially create merge
>conflicts for changes that are essentially no-ops.

I think most people would agree that gratuitously reindenting lines that
haven't been touched is really a bad idea.

>>      Where is that "gdb GNATS bug reporting database"? I can't help but
>>      feel guilty for being a security threat, but I just can't find it.
>>      There is a Bugzilla, I've seen it, but those two don't seem to be
>>      related.
>
>I don't know much about that one, and who might have created an
>account for you. For GDB, we've transitioned to buzilla, as you
>have found. We used to use gnats, but no more.

Wow, that's ancient.  I need to move that message into the 21st century.
Also, I suspect that the script shouldn't be creating bugzilla accounts
at all.

cgf

Re: Two quick questions

[Andrey Smirnov]

>>I don't know much about that one, and who might have created an
>>account for you. For GDB, we've transitioned to buzilla, as you
>>have found. We used to use gnats, but no more.
>
> Wow, that's ancient.  I need to move that message into the 21st century.
> Also, I suspect that the script shouldn't be creating bugzilla accounts
> at all.
>

I tried to use the credentials in the message to log into Bugzilla
and it didn't work. So I think it doesn't.

Andrey Smirnov

Re: Two quick questions

[Christopher Faylor]

On Tue, Nov 29, 2011 at 09:33:24AM +0600, Andrey Smirnov wrote:
>>>I don't know much about that one, and who might have created an
>>>account for you. For GDB, we've transitioned to buzilla, as you
>>>have found. We used to use gnats, but no more.
>>
>> Wow, that's ancient. ?I need to move that message into the 21st century.
>> Also, I suspect that the script shouldn't be creating bugzilla accounts
>> at all.
>>
>
>I tried to use the credentials in the message to log into Bugzilla
>and it didn't work. So I think it doesn't.

I wasn't implying that it was silently creating bugzilla accounts.  I meant
that I don't think that the script which generates the welcome message should
try to generate any account for bugzilla in any circumstances.  It does this
for gcc but, generally, people already have a bugzilla account when they
gain access.

cgf

Re: Two quick questions

[Andrey Smirnov]

On Mon, Nov 28, 2011 at 10:38 PM, Joel Brobecker <brobecker@adacore.com> wrote:
>>      1. Since neither
>>      http://www.sourceware.org/gdb/current/onlinedocs/gdbint/Coding-Standards.html
>>      nor http://sourceware.org/gdb/wiki/JoelsCodingStyleCheatSheet
>>      mentions this, or at least yours myopic truly haven't noticed it,
>>      I wonder what is the GDB's allegiance in the Tabs, Spaces, Tabs +
>>      Spaces holy war? Who are the infidels we must crush? :) The
>>      reason I'm asking is that I don't want to generate false diffs
>>      just because I re-indented the line in different style.
>
> We use tabs mostly, optionally followed by spaces if a line of code
> doesn't exactly start at a tabulation.

So Tabs + Spaces, OK.

> Although it's perfectly fine
> to tabify a line of code that you're modifying, I would personally
> prefer that you don't touch lines just for the sake of tabifying.
> Others might disagree (hence the "personally"), but these little
> changes don't bring anything in terms of readability, mess up the
> indentation when reading diffs, and also potentially create merge
> conflicts for changes that are essentially no-ops.
>

Oh no, that not the meaning I was trying to convey in my message, I'm
sorry I wasn't very clear and was misconstrued.

I was asking exactly because I wanted avoid generated unnecessary
"junk" in the patches just because I idly hit the tab on some line in the
code an my editor re-indented it differently(for example with all
spaces, which is default for me now) while writing a patch. It is just
that I've already seen one such occurrence in one of my patches and
I would like to avoid having them in further ones.

So, just to be clear, I'm neither planning nor thinking about planning
to re-indent the code.

Andrey Smirnov

Re: Two quick questions

[Stan Shebs]

On 11/28/11 7:25 PM, Andrey Smirnov wrote:
> I was asking exactly because I wanted avoid generated unnecessary 
> "junk" in the patches just because I idly hit the tab on some line in 
> the code an my editor re-indented it differently(for example with all 
> spaces, which is default for me now) while writing a patch. It is just 
> that I've already seen one such occurrence in one of my patches and I 
> would like to avoid having them in further ones. So, just to be clear, 
> I'm neither planning nor thinking about planning to re-indent the 
> code. Andrey Smirnov 

99+% of the code has been indented with Emacs using its old standard of 
8-space tabs.  So if you use that setting, the surprises will be fewest 
in number. :-)

Stan

Re: Two quick questions

[Andrey Smirnov]

On Tue, Nov 29, 2011 at 11:09 AM, Stan Shebs <stanshebs@earthlink.net> wrote:
> On 11/28/11 7:25 PM, Andrey Smirnov wrote:
>>
>> I was asking exactly because I wanted avoid generated unnecessary
>> "junk" in the patches just because I idly hit the tab on some line
>> in the code an my editor re-indented it differently(for example
>> with all spaces, which is default for me now) while writing a
>> patch. It is just that I've already seen one such occurrence in one
>> of my patches and I would like to avoid having them in further
>> ones. So, just to be clear, I'm neither planning nor thinking about
>> planning to re-indent the code. Andrey Smirnov
>
>
> 99+% of the code has been indented with Emacs using its old standard
> of 8-space tabs.  So if you use that setting, the surprises will be
> fewest in number. :-)

So, what you're saying is

(setq tab-width 8)
(setq indent-tabs-mode t)

is the way of fewer surprises. Got it. :-)

Andrey Smirnov

Re: Two quick questions

[Eli Zaretskii]

> Date: Tue, 29 Nov 2011 10:34:17 +0600
> From: Andrey Smirnov <andrew.smirnov@gmail.com>
> Cc: gdb@sourceware.org
> 
> > 99+% of the code has been indented with Emacs using its old standard
> > of 8-space tabs.  So if you use that setting, the surprises will be
> > fewest in number. :-)
> 
> So, what you're saying is
> 
> (setq tab-width 8)
> (setq indent-tabs-mode t)
> 
> is the way of fewer surprises. Got it. :-)

Yes.  But these are the defaults, so you could simply not change them.