Mozilla Shifting Firefox To A Four-Week Release Cycle

[Pajn]

I hate marketing-oriented versioning so much. It undermines the reason behind versioning, which is letting people using software know what scope of changes to expect when updating. People should just stick to semver and call it a day.

Semver is great for libraries but falls apart when applied to user applications. What is a breaking user application change? Either you treat every UI change as breaking and then you end up with the versioning scheme of firefox and chrome but with a .0.0 applied after it or you in the context treat web api breakage as breaking and is then stuck on 1. forever. In any case, just dismissing the two numbers that never change and just bump the single number is the reasonable thing to do. The other reasonable thing to do is to dismiss version numbers entirely and just go with year and month.

[kaprikawn]

Oh for the days of Firefox 2.6

[elatllat]

Maybe thry will make dev tools less buggy, or add more rust.

[Guest]

Semver is great for libraries but falls apart when applied to user applications. What is a breaking user application change? Either you treat every UI change as breaking and then you end up with the versioning scheme of firefox and chrome but with a .0.0 applied after it or you in the context treat web api breakage as breaking and is then stuck on 1. forever. In any case, just dismissing the two numbers that never change and just bump the single number is the reasonable thing to do. The other reasonable thing to do is to dismiss version numbers entirely and just go with year and month.

I'd be fine with Chrome version 1241.1.1 if it had a meaning beyond marketing. Picking what is the API is pretty tough in general, even if code doesn't directly face the user. Consider a kernel. Do you version according to usage of syscalls, or driver APIs, or something else? The choice is tough, but some information over pure noise is better than nothing.

[crystall]

*inefficient lab rats (Average Joe doesn't know how to make good bug reports)

Do they even read the automatic crash reports?

Yes, we even triage them on a daily basis and we've spent significant efforts in the past two years to detect and fix common crashes faster.

[david-nk]

Since web standards are developed and adopted at a glacial rate, I find it strange that Mozilla thinks they need to release even more often. It seems to me that even if browsers had only one major release per year, it would make no real difference, as long as bugfix releases keep happening throughout the year.