Sujet : Re: Improving build system
De : david.brown (at) *nospam* hesbynett.no (David Brown)
Groupes : comp.arch.embeddedDate : 16. May 2025, 13:42:16
Autres entêtes
Organisation : A noiseless patient Spider
Message-ID : <1007br8$3ookp$1@dont-email.me>
References : 1 2 3 4 5 6
User-Agent : Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101 Thunderbird/102.11.0
On 16/05/2025 12:21, Nicolas Paul Colin de Glocester wrote:
On Thu, 15 May 2025, David Brown wrote:
"There /are/ no alternatives that are more trustworthy - they
just have different failure or risk points. There can be benefits in buying a
commercial IDE, and/or a commercial toolchain, but lower risk of bugs, quirks or
installation issues is most certainly not one of them."
C and C++ compilers and codes produced thereby are not trustworthy. Use
commercial Ada compilers to avoid bugs.
No one was questioning the trustworthiness of compilers or the code they generate. The issue was about how well IDE's cope with unusual installations outside the defaults expected by the supplier. The free IDE's provided with manufacturers are vastly more commonly used than commercial IDE's (especially those that have their own custom IDE's), and you can expect them to have been tested and used in a much wider range of circumstances.
As for the trustworthiness of compilers, that's another matter. I have never seen any reason to suppose that commercial compilers are more trustworthy (in terms of accepting valid code or generating correct code) than the good open source compilers (gcc and clang). I have never seen any reason to suppose that Ada compilers are more trustworthy than C or C++ compilers. On the contrary, I find that more popular tools are less likely to have serious bugs.