Re: [HACKERS] Re: Notation for nextval() (was Re: Several small patches)

Поиск
Список
Период
Сортировка
От Tom Lane
Тема Re: [HACKERS] Re: Notation for nextval() (was Re: Several small patches)
Дата
Msg-id 7410.945411883@sss.pgh.pa.us
обсуждение исходный текст
Ответ на Re: Notation for nextval() (was Re: Several small patches)  (Jeroen van Vianen <jeroen@design.nl>)
Список pgsql-hackers
Jeroen van Vianen <jeroen@design.nl> writes:
>> I'm dubious about the other two patches also.  Evidently there is some
>> variation across platforms in the desirable switches for ctags --- but
>> diking out the ones not wanted on a particular platform is no answer.
>> Perhaps the proper fix is to make the ctags flags a configurable
>> macro...

> About ctags: is no one using Linux and ctags on the Postgres sources? Am I 
> the first one to find this bug?

Apparently you're a little new to the world of portable software.
I don't use ctags myself, being an Emacs man rather than a vi'er,
but a few minutes' research yielded the following results:

GNU ctags (from Emacs 19.34 distribution): -a, -d, -t, -f accepted.

HPUX ctags (which claims to be based on the original UCB code and
compliant to XPG4 standard): -a, -t, but no -d nor -f.

SunOS 4.1: same as HPUX.

RedHat 4.2 Linux: comes with something called "Exuberant Ctags, Version
1.5" which accepts all four (apparently this is NOT the same code as the
GNU distribution).

Whatever Linux you're running: evidently only -a and -f.

I don't know which variant of ctags you're running, but it's definitely
odd man out as far as not accepting -t goes.  I'd certainly want to use
-d (index #defines) anywhere it was accepted, too.  Other side of the
coin is that -a is the only one of these switches that works on all the
ctags versions I was able to lay my hands on in five minutes plus yours.
That should give you some pause about asserting that if -a -f is the
right incantation for the version you have, then it must be the right
thing for everybody.

Bottom line here is that what we probably really need is a configurable
makefile macro for the ctags switches.  (In fact, what I'd personally
like is another macro to determine whether we're using ctags or etags in
the first place ;-).)  But short of that, I'd definitely lean towards
the GNU definition as being the most widespread code.  I'm pretty
surprised that your Linux distribution (which one is it?) seems to
contain a non-GNU-compatible ctags.
        regards, tom lane


В списке pgsql-hackers по дате отправления:

Предыдущее
От: Ed Loehr
Дата:
Сообщение: Re: [HACKERS] ordering RH6.1
Следующее
От: Tom Lane
Дата:
Сообщение: Re: [HACKERS] Re: Notation for nextval() (was Re: Several small patches)