|
Boost : |
Subject: [boost] Fwd: Re: [log] review part 1
From: Andrey Semashev (andrey.semashev_at_[hidden])
Date: 2010-03-13 16:16:12
Due to technical reasons this reply went off-list. For sake of keeping
the conversation public, I'm reposting it.
-------- Original Message --------
Subject: Re: [boost] [log] review part 1
Date: Sat, 13 Mar 2010 16:08:14 +0300
From: Andrey Semashev <andrey.semashev_at_[hidden]>
To: Steven Watanabe <watanabesj_at_[hidden]>
On 03/12/2010 07:22 AM, Steven Watanabe wrote:
> I mostly have documentation notes at this point. Andrey, since I have so
> many, and they're mostly fairly minor, would it be better if I made
> a patch for the documentation source?
Yes, that would be great. I quickly reviewed your suggestions, and they
all look reasonable (I always have a hard time with articles). Thanks!
> I think newbies should probably be replace globally with new users,
> since I kind of feel that the term newbie is somewhat derogatory.
Ok. I didn't mean to offend anyone, of course. :)
> http://boost-log.sourceforge.net/libs/log/doc/html/log/defs.html
> ================================================================================
>
> For Log record:
> I don't like the expression "A single pack of information," although
> I can't think of anything better off hand.
A collection of data,...?
> ================================================================================
>
> Issues:
>
> Instead of compiling a narrow or wide library, why don't you use
> libboost_log
> and libboost_wlog?
There are parts that would be the same in both these libraries. Most
notably, singletons of file_collector repository and some constants.
This might lead to linking problems if someone tries to use both
versions of the library.
> I agree with other comments that I saw about Attribute being
> a poor choice for what it does. I would prefer that the term "Attribute"
> should be used for what you call an "Attribute Value," and that
> there should be a different term, like "Attribute Extractor" for
> what you call an "Attribute."
My understanding of the term "attribute" is that it's some kind of
meta-information that defines the concrete values that will be attached
to log records. I'm ready to agree that the word "attribute" doesn't
bring my intention precisely, but the alternatives don't quite get
closer. "Extractors" are reserved for another tool in the library.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk