On 11/07/2010 05:57 AM, Jacob Carlborg wrote:
> On 2010-11-06 15:15, Lutger wrote:
>> Jacob Carlborg wrote:
>>
>>> On 2010-11-06 14:43, Lutger wrote:
>>>> Stephane C wrote:
>>>>
>>>>> Hello,
>>>>> Looking at the company blog, It seems that the people from BITPROX are
>>>>> about to abandon development of their IDE.
>>>>>
>>>>> blog address : http://bitprox.com/blog/blog5.php
>>>>>
>>>>> Sad news... I am personally convinced that there is little chance
>>>>> for a
>>>>> programming language to stand against Java or C# without a
>>>>> professional-grade IDE.
>>>>
>>>> Perhaps, but bitprox ide was not professional grade by far. Plus, you
>>>> need an outstanding ide in the commercial space, just professional is
>>>> not good enough.
>>>>
>>>> Descent was the one ide that had the potential to compete with the very
>>>> best C#/Java IDE's and beat every C++ IDE out here. (and it did so in
>>>> some respects). But there are more projects, like Visual D and DDT.
>>>
>>> Descent was/is far better than any C++ IDE I've seen. Now I'm waiting
>>> for Xcode 4 which contains a complete frontend hoping it will be as good
>>> as JDT.
>>>
>>
>> Descent did have some polish problems and of course lacked in
>> contributors
>> too much. I hope that ddmd may eventually provide the semantic
>> plumbing for
>> D tools that is required to achieve such excellence as descent. I'm also
>> looking forward to try DDT.
>
> It had/has a lot of bugs, crashing several times every day but it
> had/has more features than, for example, CDT and in the end I thought it
> was (still is) worth it because of all the features it has (most of the
> times).
>

Yeah... Most of those bugs were my innability to deal with the ported parser/compiler, and I can't imagine another IDE with those semantic capabilities without using DMD's source code. Or... you could try to understand the compiler's logic and implement it in another way, more useful to an IDE, except that the logic is incredibly complex, even though the website says another thing. Or making a simpler IDE with not many semantic features, but that was a boring task for me.

Are there any other implementation of a D compiler that don't use DMD as the frontend?

> Are there any other implementation of a D compiler that don't use DMD as the frontend?

dil and another still non-public project.