[Info-vax] Taking a break - Open Source on OpenVMS Conference Calls Resume in the FALL of 2022...

[Arne Vajhøj]

On 6/17/2022 9:18 AM, chris wrote:
> On 06/16/22 21:55, Arne Vajhøj wrote:
>> On 6/15/2022 8:23 AM, Simon Clubley wrote:
>>> The Cygwin approach of providing core Unix compatibility functionality
>>> in a library and then building Unix applications against that library
>>> would indeed appear to be the best approach for VMS, given that it has
>>> been a major success story on Windows and has provided us with a rich
>>> Unix userland environment on Windows.
>>
>> *nix compatibility is definitely a good thing.
>>
>> But a few comments.
>>
>> 1) Cygwin is not a success on Windows. It is a great thing, but
>> it has not gotten mainstream for Windows development and
>> its use is pretty rare.
> 
> It's definately a success for those  who need its functionality. As
> a developer, it allows me to run X under cygwin on windows and
> to access a whole myriad of useful apps and as much of a unix like
> environment as needed. For those who must use windows, cygwin
> adds so much functionality, and unlike some offerings, it's quite
> lightweight in resource usage as well.
> 
> I know uSoft have a linux environment package for windows, but
> they really are a bit late to the game. Cygwin don't make a big
> noise in publicity terms, but it just gets the job done...

I like Cygwin. I have always had it on my Windows PC's for more
than 2 decades. I don't use that much of it and I never use
bash, but I like many of the utilities.

But if you look at what software for Windows is being build
with then MSVC++ is by far the majority and number two is
GCC mingw*. GCC cygwin is a niche.

For whatever reason the average Windows + *nix developer
seem to prefer #ifdef'ing and building with either MSVC++
or GCC mingw* instead of GCC cygwin.

As a data point with significant impact look at
Boost.

Arne