Ping?..
Igor
On Sat, 1 Feb 2003, Igor Pechtchanski wrote:
> On 2 Feb 2003, Robert Collins wrote:
>
> > On Sun, 2003-02-02 at 07:37, Igor Pechtchanski wrote:
> >
> > > Yes, it does recompile everything when resource.h is modified, but that
> > > only happens once. A necessary evil...
> >
> > > No. I like Igors approach better.
> > > For yours, we end up with duplicate resources : you should split the
> > > 'installation incomplete' and the log to check into two separate
> > > messages.
> >
> > Then you have to tangle with temporary buffers to combine them.
>
> That should be triv
I've sent this a couple of hours ago, but it didn't appear in the
web archive, so I'm resending it. If the original does get through,
apologies for the duplicate.
On 2 Feb 2003, Robert Collins wrote:
> On Sun, 2003-02-02 at 07:37, Igor Pechtchanski wrote:
>
> [snip]
>
> > Incidentally, there are
On 2 Feb 2003, Robert Collins wrote:
> On Sun, 2003-02-02 at 07:37, Igor Pechtchanski wrote:
>
> > Yes, it does recompile everything when resource.h is modified, but that
> > only happens once. A necessary evil...
>
> Even so... resource.h urrghh.
>
> > Sorry about the (null) bit. Should be fixe
On Sun, 2003-02-02 at 08:44, Max Bowsher wrote:
> > No it's not, because the current setup code *always* uses a log file
> > of the appropriate level. It's a fallback to catch future changes.
>
> Well, at least make it "" (with the brackets) to make the
> message easier to read.
Fine by me.
>
Robert Collins wrote:
> On Sun, 2003-02-02 at 08:22, Max Bowsher wrote:
>> Robert Collins wrote:
>>> IDS_MISSING_LOG should be "No log was in use", not "setup.log.full".
>>> It'll be ugly in the output, but at least accurate.
>>
>> "Check No log was in use for details."
>> That's worse than what we
On Sun, 2003-02-02 at 08:22, Max Bowsher wrote:
> Robert Collins wrote:
> > getFileName doesn't take a minlevel, it takes an exactLevel - your
> > parameter name is misleading, or you've got the logic in the loop
> > wrong :}.
>
> Or, you could say that it gets the file whose minlevel is what you
Max Bowsher wrote:
> Robert Collins wrote:
>> getFileName doesn't take a minlevel, it takes an exactLevel - your
>> parameter name is misleading, or you've got the logic in the loop
>> wrong :}.
>
> Or, you could say that it gets the file whose minlevel is what you
> request. In which case its corr
Robert Collins wrote:
> getFileName doesn't take a minlevel, it takes an exactLevel - your
> parameter name is misleading, or you've got the logic in the loop
> wrong :}.
Or, you could say that it gets the file whose minlevel is what you request.
In which case its correct :-)
> IDS_MISSING_LOG sh
On Sun, 2003-02-02 at 07:37, Igor Pechtchanski wrote:
> Yes, it does recompile everything when resource.h is modified, but that
> only happens once. A necessary evil...
Even so... resource.h urrghh.
> Sorry about the (null) bit. Should be fixed now. I also took your
> suggestion regarding "l
Igor Pechtchanski wrote:
> Incidentally, there are two backslash() functions defined: one in
> filemanip.h, and another in concat.h. This is the only function
> defined
> in concat.h, and it doesn't seem to be used anywhere anymore. Is
> concat.h dead? Should it be removed? Ditto for concat.cc
On Sat, 1 Feb 2003, Max Bowsher wrote:
> Igor Pechtchanski wrote:
> > On Sat, 1 Feb 2003, Max Bowsher wrote:
> >> What advantage do resources give us over string literals, for
> >> messages that are only used once in setup?
> >
> > Resources have the advantage that whole resource tables can be
> >
Igor Pechtchanski wrote:
> On Sat, 1 Feb 2003, Max Bowsher wrote:
>> What advantage do resources give us over string literals, for
>> messages that are only used once in setup?
>
> Resources have the advantage that whole resource tables can be
> substituted at once. If we ever want to internationa
On Sat, 1 Feb 2003, Max Bowsher wrote:
> Igor Pechtchanski wrote:
> >> just do: < return "the log"; >, and remove the bad_file variable?
> >
> > Hmm. Personally, I prefer to use named constants whenever possible.
>
> It seems a little silly to use a constant when it is required only once.
I know
> From: Max Bowsher
>
> General question:
>
> What advantage do resources give us over string literals, for
> messages that
> are only used once in setup?
I18N
J.
Igor Pechtchanski wrote:
>> just do: < return "the log"; >, and remove the bad_file variable?
>
> Hmm. Personally, I prefer to use named constants whenever possible.
It seems a little silly to use a constant when it is required only once.
> However, static variables aren't the way to go -- I mus
Max,
Thanks for the feedback. Replies inline below.
On Sat, 1 Feb 2003, Max Bowsher wrote:
> Igor Pechtchanski wrote:
> > Ping? (Just making sure this was seen).
> >
> > On Fri, 24 Jan 2003, Igor Pechtchanski wrote:
> >> 2002-10-17 Igor Pechtchanski <[EMAIL PROTECTED]>
> >>
> >> * res.
Igor Pechtchanski wrote:
> Ping? (Just making sure this was seen).
>
> On Fri, 24 Jan 2003, Igor Pechtchanski wrote:
>> 2002-10-17 Igor Pechtchanski <[EMAIL PROTECTED]>
>>
>> * res.rc (IDS_INSTALL_INCOMPLETE): Change hard-coded
>> log filename to %s.
>> * LogFile.cc (LogF
18 matches
Mail list logo
