|
Boost : |
Subject: Re: [boost] Answers re BENUM library for automated enum streaming
From: Tony D (tony_in_da_uk_at_[hidden])
Date: 2008-09-26 21:19:43
Hi Marcus,
Thank you for your guidance, and my turn to apologise for slow turn-around and mistaking the name of EnumIO. Noted the heads-up re boost code formatting standards - will have to do something I guess :-(. As you say, my library relies on "exotic features", and while I was happy to discover Sun Studio has supported __VA_ARGS__ for a long time, VC++ support only came is 2005, and I'm not so happy to be unable to find explicit mention of it in the discussions for C++0x... so avoidance is definitely a strength for EnumIO.
I did receive one suggestion elsewhere - think it's worth sharing as you might consider it for EnumIO. Idea is: "inheriting" from an existing enum and adding new values. An "extendable" enum would be packaged in a class that actually stores the value in an int and has a protected set_from_int(int) method. A derived class defines an enum with additional symbols/values, constructors from the base's enum and their own enum - both of which call the base set_from_int(). (A protected constructor from int() could be used - and might be more efficient - but in the case of BENUM would complicate end-user control over whether construction from int is implicit or explicit.) Thought that was a pretty neat idea.
Lots to do on the docs, examples etc..
Regards,
Tony
----- Original Message ----
From: Marcus Lindblom <macke_at_[hidden]>
To: boost_at_[hidden]
Sent: Friday, 19 September, 2008 17:53:24
Subject: Re: [boost] Answers re BENUM library for automated enum streaming
(Sorry for not replying sooner)
Tony D wrote:
> Hi all,
>
> Should I expect any further feedback re BENUM? David, Marcus: we've discussed specific issues (portability of variadic macros, smart_enum features) - do you have further questions, comments re my replies, or recommended actions before submission for formal review? Should I be doing more to pursue suggestions or solicit feedback? Is anyone else currently reviewing the code in the vault, experiment with usage in a real system, or planning to do so? If so, can we keep in touch re progress? If I don't hear anything to the contrary this week, I'll start doing the preparations for requesting a formal review, specifically my TODO list currently looks like:
Small detail: My little embryo is called EnumIO, not smart_enum.
Smart_enum is probably something different altogether.
I'm quite happy with the answers, and I think it suits as a good
replacement for my lib. I took a brief look at the code, and besides it
not being formatted according to boost standards, it looked pretty ok.
I still think my little lib is useful, mainly because it doesn't rely on
exotic features and it can also be used in a non-intrusive fashion.
But I see no bigger conflict as your will probably win the future.
(Besides, I'll have to figure out a few things before I can publish mine
for review, as noted in the link I posted a few mails back. If I ever
get there there will probably be some discussion about integration, etc.)
Cheers,
/Marcus
_______________________________________________
Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk