Subject: Re: [boost] [Review] GGL review starts today, November 5th
From: Phil Endecott (spam_from_boost_dev_at_[hidden])
Date: 2009-11-20 19:57:03
Thomas Klimpel wrote:
>> Please always state in your review,
>> whether you think the library should be accepted as a Boost library.
> This is a difficult question in this case, because it is not completely
> clear what it means. I found out during my review that the library as
> submitted for review is "work in progress", at least in parts. So I vote
> for acceptance, but try to explain how to interpret that vote.
> The library as a whole is not ready for release, but it is ready enough
> for a formal review, especially if we consider that it is normally
> expected that a library author will modify his library after a formal
> review to address the most important issues that came up during the review.
> The following parts of the library to give me the impression that they are finished:
> - The user level documentation
To me it seems like a bad idea to accept libraries that are submitted
with unfinished documentation, for the following reason: it's much
harder to review a poorly-documented library. It's more friendly to
reviewers if they don't have to search through the source or ask
questions here to determine how to do basic things, and the limited
time that reviewers have could be spent on more productive things. I'd
like to discourage setting a precedent of accepting libraries with
inadequate docs, even when it's believed that it would be improved
later, so that future library authors are not tempted to submit their
libraries too soon.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk