Re: Will tree-ssa be GCC 3.5?

["Zack Weinberg" <zack at codesourcery dot com>]

kenner@vlsi1.ultra.nyu.edu (Richard Kenner) writes:
> zack@codesourcery.com (Zack Weinberg) writes:
>> This strikes me as disingenuous.  The people working on the tree-ssa
>> project have clearly and repeatedly stated their intention to replace
>> several (perhaps most) of the RTL optimizers.  That this has not
>> happened yet merely indicates that the project is not done yet.
>
> Yes, I understand the intention, but I think the jury is very much
> still out as to what extent that will be possible.

I would agree with that, although I am more sanguine about the
possibility than you; but that's not my point.

Your earlier statement

>>> We don't yet know that it will be *replacing* anything.  For the
>>> moment, it'll be adding a new class of optimizations.

can be read as saying the same thing as "the jury is still out as to
what extent that will be possible" -- but that is not the natural
reading.  The natural reading is that you don't believe replacing
low-level optimizations is the intended goal.  Your previous
statements on the same topic have read the same way, or else as bald
assertions that it will not be possible.  

As you yourself say the jury is still out, which means predicting
failure of this effort is just as premature and inappropriate as
predicting its success.  And expressions of disbelief that other
people's stated goals are actually their intended goals, in the
absence of any evidence either way, have no place in a cooperative
environment.

I must ask you to stop undercutting your fellow developers in this
manner.

zw