Hello,
The dll in the redist for intel fortran 2022 update 1 has the same file version number and product version number than the dll in the redist for intel fortran 2023.2.0
both are 2.0.0.1.0
On the picture attached, on the left, it is fortran 2022 update 1 dll and on the right, it is dll from intel fortran 2023.2.0
dlls are different of course but file version and product version are the same.
Is that normal?
I asked this question because I have some installer that use those numbers.
thanks in advance!
連結已複製
@Barbara_P_Intel , Isn't it a good practice to change the version numbers when something in the dll has changed ? I believe this may cause upgrade issues for certain applications. Could you please clarify ?
As a former compiler library developer, I am horrified that Intel would ship different DLLs with the same version numbers. This flies in the face of any kind of good, responsible practice. How is an installer (or a user) supposed to know whether it needs to update an already installed DLL? Comparing file sizes tells you nothing other than that they are different, and you don't know what the difference is. Who was responsible for this travesty? A halfway decent build system would automatically bump the last version field on every build.
@Steve_Lionel , having just finished two minutes ago the code of ethics for the ISO organization that I found through your Mastodon site in relation to the last meeting of the Fortran committee.
I somewhat chuckled at your choice of words.
Perhaps the use of travesty and horrified could be perhaps toned down a touch to say, a more moderate
Old bean it is a bit of a bother that the error occurred, perhaps we could try better in the future, what?
John, I think I earned the right to criticize a seriously flawed decision, given that I was once responsible for building the Fortran run-time library (back in the DEC days) and worked with many of the players here. I know this DLL version issue didn't exist when I was working at Intel, and either someone decided that version numbers no longer mattered, or a mistake was made and not caught. Regardless, it is a very serious error and should be addressed. That it is an error may not be understood by the development team.
Steve, no problem, I agree with your comments, but I was amused that I had just read the minutes of the last Fortran standards meeting. The minutes, with you as chairman, had a note saying all work should be to the following ethical standards. On reading the standard I was amazed at how strict the controls were and then I immediately read your note. It was just so funny, I had to comment.
I think the ISO standard is akin to a preschool level of conversation at voice level 1, but that is the modern world. We cannot have a decent argument.
My apologies! I misstated about the version number of the redistributable DLLs. I did some further research since @Steve_Lionel was so adamant and I know my experience with the redistributable files is limited.
The version number should change. A bug has been filed to remedy this.
When can we expect the new release with the fixed version numbers?
Thank you, @Barbara_P_Intel ! I had a feeling that it wasn't a deliberate choice., but it does suggest a possible improvement to the RTL test system.
@Barbara_P_Intel , adamant is an interesting word choice, it's meaning from the olden days, before Fortran also know as bF,
one sample., with apologies to the OED.
With a penne of yron & with an Adamant clawe.
yron is an eagle feather, would seem to fit the people here quite well.
John,
Adamant is an English singer, musician, and actor. He gained popularity as the lead singer of new wave group Adam and the Ants and later as a solo artist, scoring 10 UK top ten hits from 1980 to 1983. Just saying...
