On 5/3/2011 5:23 PM, Vladimir Batov wrote:
Edward Diener<eldiener<at> tropicsoft.com> writes: The review period for the Boost.Convert library has officially ended. ...
Edward,
I thought I'd express my opinion on the situation potentially making your decision easier. :-) Given the situation I feel the library should be rejected (so now you can count my vote as well :-) ). Quite a few people expressed negative opinions, criticism from function names to design, reservations or conditional acceptance with considerable re-work. Many suggestions have their merits in certain contexts and for certain usage patterns. Unfortunately, I feel that trying to address those would inevitably be done at the expense of or would have a considerable impact on other use-cases equally important to others. Given my abilities I do not see how it is possible to satisfy a considerable number of conflicting requirements/expectations to everyone's satisfaction.
In that light I do not feel comfortable trying to offer the library which copped so much criticism... and will be copping criticism unless re-done to satisfy conflicting requirements... which I personally do not see achievable.
Thank you for taking on the hard task of managing the review. Thank you everyone who expressed their views and opinions. I am throwing in the towel. I am hoping someone tougher and brighter will come up with an adequate solution.
I'm really sorry to hear this Vladimir. I think you've done a great job trying to juggle conflicting use cases and sustaining a great deal of criticism. I can certainly sympathize that at some point it's good enough for your own use and that getting criticism for its suitability in use cases you'll never need is distressing. I am thinking that this domain (lexical conversion) would benefit from a working group instead of a single developer with comments being constantly fired from outside. There are too many use cases to reasonably expect a single person to commit to supporting and defending all of them. The working group would be comprised of people with a set union of interests that included all the use cases. It could also handle merging boost.convert, lexical_cast, and boost.conversion (if Vincente's willing). -Matt