| |||||||||||||
|
| |||||||||||||
SnarkSnarkPosted Jun 18, 2008 18:24 UTC (Wed) by ncm (subscriber, #165)Parent article: Converting GCC to C++ Jon, I didn't find any justification in the slides for the addition, "while, with luck, avoiding the C++ language's worst problems." If you don't personally like C++, that's your right and your handicap. Please don't project it onto Ian. In my 22 years' experience, the language's worst problems are in its C compatibility subset. Thus, whatever "worst language problems" the conversion might encounter are already manifest in the Gcc codebase. Converting to C++ offers a route to avoid them. That point was central to his presentation; 8 out of the 17 the slides illustrate it. A scrupulous summary would note that instead.
(Log in to post comments)
Snark Posted Jun 18, 2008 18:40 UTC (Wed) by corbet (editor, #1) [Link] Jon, I didn't find any justification in the slides for the addition, "while, with luck, avoiding the C++ language's worst problems." I was thinking of slide 13, which notes "we would only use features which are worthwhile," and 14, adding "Maintainers will ensure that gcc continues to be maintainable." These are answers to charges that C++ is too slow and too complex. If I've overstated what Ian meant, I apologize.
Snark Posted Jun 18, 2008 20:51 UTC (Wed) by ncm (subscriber, #165) [Link] Thanks, Jon, for the apology, but I'd rather see the summary corrected. The charges Ian answered in those slides were assertions, by others, that he was contradicting. We should not take them as his own statements. As an aside, he also didn't suggest that luck would be needed. He seems completely confident that use of "features which are [not] worthwhile" and any unmaintainably complex constructs can be deliberately eschewed. Since a chief goal is to replace unmaintainably complex C with maintainable C++, I agree with his confidence.
|
|
||||||||||||