On 14/11/11 14:23, Tor Lillqvist wrote:
Considering actual multi-user site scenarios, I think it is fairly
useless to have the name of the machine in the lock file. Consider
these examples:
1) Alice wants to edit a document. Computer says "No". Er, sorry, been
watching too much Little Britain.
Considering actual multi-user site scenarios, I think it is fairly
useless to have the name of the machine in the lock file. Consider
these examples:
1) Alice wants to edit a document. Computer says "No". Er, sorry, been
watching too much Little Britain.
I mean, LO says "document is locked by Bob
Hi
On windows you could lookup the environment variable COMPUTERNAME and put that
in the lock file.
That would solve the issue with flaky remote shares.
Could use uname on linux/MacOS.
We could also use the IP address of the machine - that should help with flaky
NFS shares.
Although in the pre
On 14/11/11 12:30, Michael Meeks wrote:
On Sun, 2011-11-13 at 12:06 +0200, Tor Lillqvist wrote:
But in general we should avoid potentially pointless DNS calls. Let's
not risk having to wait for DNS timeouts in badly configured
situations. I think there has been bug reports of OOo and/or LO bein
On Sun, 2011-11-13 at 12:06 +0200, Tor Lillqvist wrote:
> But in general we should avoid potentially pointless DNS calls. Let's
> not risk having to wait for DNS timeouts in badly configured
> situations. I think there has been bug reports of OOo and/or LO being
> very slow to start in some cases,
On 11/13/2011 03:42 PM, Lionel Elie Mamane wrote:
On Sun, Nov 13, 2011 at 03:04:57PM +0100, Arnaud Versini wrote:
I've seen it is used by using gdb on
_osl_getFullQualifiedDomainName. I don't remember how but I could
provide you the bt after the build. It's not directly but
_osl_getFullQualifie
On Sun, Nov 13, 2011 at 02:16:51PM +0100, Lionel Elie Mamane wrote:
>
> AFAIK, a well-configured machine has /etc/hosts setup so that
> gethostbyname(gothostname()) does not touch DNS, but is resolved
> entirely from reading /etc/hosts. Doesn't the installer *by* *default*
> set things up like tha
On Sun, Nov 13, 2011 at 03:04:57PM +0100, Arnaud Versini wrote:
> I've seen it is used by using gdb on
> _osl_getFullQualifiedDomainName. I don't remember how but I could
> provide you the bt after the build. It's not directly but
> _osl_getFullQualifiedDomainName is used during file locking.
Tha
On Sun, Nov 13, 2011 at 02:30:40PM +0100, Arnaud Versini wrote:
> there is this portion of code in _osl_getFullQualifiedDomainName :
>
> if (nLengthOfDomainName == 0)
> {
> sal_Char pDomainNameBuffer[ DOMAINNAME_LENGTH ];
>
> pDomainNameBuffer[0] = '\0';
>
> if (_
And this is the backtrace.
2011/11/13 Arnaud Versini
> I've seen it is used by using gdb on _osl_getFullQualifiedDomainName. I
> don't remember how but I could provide you the bt after the build. It's not
> directly but _osl_getFullQualifiedDomainName is used during file locking.
>
> I'm writtin
I've seen it is used by using gdb on _osl_getFullQualifiedDomainName. I
don't remember how but I could provide you the bt after the build. It's not
directly but _osl_getFullQualifiedDomainName is used during file locking.
I'm writting a patch to remove the usage of _osl_getDomainName.
2011/11/13
Arnaud is correct, I was looking at the wrong method. But where do you
see this method being used in file locking?
Lionel, I've run across a lot of machines with badly configured DNS.
socket.c: _osl_getDomainName
is only used by
socket.c: static sal_Char* _osl_getFullQualifiedDomainName (con
there is this portion of code in _osl_getFullQualifiedDomainName :
if (nLengthOfDomainName == 0)
{
sal_Char pDomainNameBuffer[ DOMAINNAME_LENGTH ];
pDomainNameBuffer[0] = '\0';
if (_osl_getDomainName (pDomainNameBuffer, DOMAINNAME_LENGTH))
{
pD
On Sun, Nov 13, 2011 at 12:46:06PM +0200, Noel Grandin wrote:
> GetDomainName is only used by
> GetYPDomainName
> in the same file
> which in turn is only used by
> const rtl::OUString& SubstitutePathVariables_Impl::GetYPDomainName()
> in
> /core/framework/source/services/substitutepathvars.
On Sun, Nov 13, 2011 at 12:06:52PM +0200, Tor Lillqvist wrote:
>> The portable and strictly correct way seems to be to do a
>> DNS query on the result of gethostname() (gethostbyname or getaddrinfo
>> with AI_CANONNAME).
> But in general we should avoid potentially pointless DNS calls. Let's
> no
This function is also used in _osl_getFullQualifiedDomainName, I check if
the result of this function is correct, I'm checking the result of this
function.
2011/11/13 Noel Grandin
> OpenGrok says that:
>
> GetDomainName is only used by
> GetYPDomainName
> in the same file
>
> which in turn is o
OpenGrok says that:
GetDomainName is only used by
GetYPDomainName
in the same file
which in turn is only used by
const rtl::OUString& SubstitutePathVariables_Impl::GetYPDomainName()
in
/core/framework/source/services/substitutepathvars.cxx
which in turn is only used by
bool SubstitutePath
Seems that the original code was also broken of course, forking and
execing /bin/domainname is indeed a bit stupid, and the domainname
command is also specified (to the extent it is "specified" at all) to
return the NIS domain name.
But honestly, how many of the LibreOffice installations on Unix e
> For nitpickers, note that getdomainname does *not* return the DNS
> domain, but the NIS domain, which strictly speaking may or may not
> coincide.
>From my Solaris sysadmin experience (which is limited, and old), I'd
say it is a very lucky site and/or organization if they happen to
coincide.
I
On Sat, Nov 12, 2011 at 09:00:08PM +0100, Arnaud Versini wrote:
> I have noticed some forks during valgrind check, I think it will be better
> to use functions instead of the unix command. This patch replace the fork
> and exec domainnanme by a simple call to getdomainname(). It is faster and
> w
Hi,
I have noticed some forks during valgrind check, I think it will be better
to use functions instead of the unix command. This patch replace the fork
and exec domainnanme by a simple call to getdomainname(). It is faster and
we sould have less message during valgrind and strace investigation's
21 matches
Mail list logo
