3 ms·
Why not just test your stuff on CentOS Stream or even just Fedora? RHEL is for big business having bespoke applications; unless you need the stability nobody e
by tpush 3y ago
Why not just test your stuff on CentOS Stream or even just Fedora?
RHEL is for big business having bespoke applications; unless you need the stability nobody expects normal open source devs to make RHEL their testbed.
- tannhaeuser 3y ago> Why not just test your stuff on CentOS Stream or even just Fedora? Because that doesn't say anything about stuff running on RHEL, and you can test on Debian just as well so why bother?
- tpush 3y agoWhat is the use case for needing to test stuff in RHEL that isn't an Enterprise App where you want/need to buy a license anyway? Like, why do you care about RHEL at all in a way that isn't covered by CentOS Stream?
- tannhaeuser 3y agoAs a developer, I never considered buying an enterprise license anyway. CentOS/RHEL with 10-years stability was kindof the canonical base for development of backend apps when you have better things to do than keeping dependencies up to date. Especially when RHEL themselves change things all the time in a zero-sum fashion without new capabilities (cough systemd cough). Now I don't care about RHEL at all indeed.
- mistrial9 3y agothere are conflicting or at least mutually differing goals for an OS and the products. It seems from this desk that part of the extended systemd unrest is actually about spending resources for one set of aims, not another, when both are legitimate. It seems obvious the companies acquire tech with finance, and finance often cares about security and adding new locks, fences and cameras in the product fabric. systemd and the EFI boot seem to becoming more vehicles for new locks, fences and cameras than adding features to the product.
- geerlingguy 3y agoBecause end users run on RHEL, so testing on RHEL as a service to those users is a nice thing to do.
- JonChesterfield 3y agoI used to test on centos because red hat was _the_ enterprise Linux platform and I wanted my code to work there unmodified. It also helped avoid hardcoding assumptions from the Debian world. I don't test on centos stream because I'm skeptical anyone deploys it to production and code working there wouldn't tell me much about working on RHEL. If my code crashes on their bleeding edge, do I try to fix it or wait to see if something under it is broken? Could waste a lot of time there for zero benefit. I personally think this is the beginning of the end times for RHEL. It will not survive IBM.