Hi Tomas, Tomas Volf <~@wolfsden.cz> writes:
> Hi Maxim, > > Maxim Cournoyer <maxim.courno...@gmail.com> writes: > >> After cleaning my mailbox recently, I've noticed that guix-commits was >> very heavy on disk usage, with translation commits weighing upward of 10 >> MiB often, and with large topic branch rebases being notified as well, >> for the complete series of rebased commits. >> >> The signal to noise ratio appears very low, and the burden of archiving >> all this perpetually high for the GNU infrastructure. I'd like to >> suggest we sunset its operation. We can get a better and more focused >> history by using 'git log', e.g. 'git fetch && git log FETCH_HEAD' to >> learn what was freshly committed that we haven't pulled yet. >> >> What do you think? > > I personally find the mailing list useful for getting quick overview of > what is happening and whether there is something interesting I should > pull. Compared to git log it has the advantage I do not have to go out > of my way to check, since it is right there in my mail agent together > will all the other mail. It is also nice to be able to just "comment" > on commits via replying to the emails (I admit that I have done that > only like 3 times in past year). I used it that way in the past. It's true that it can be an effective means to keep notice on things, but the volume/noise is just too high to pay attention for a sustained amount of time, at least I tired from trying. > I am subscribed for almost a year, and I have collected 1.2 GB of emails > in that time. That does not feel like a particularly significant amount > given the size of modern hard drives. But if it is an issue, would it > be possible to not just archive the mailing list? I agree there is no > value in persistent archive here, git log is better for archaeology. We could tell the Savannah folks that we do not care about archiving this mailing list, if we want to keep using it. Since you find it useful, I guess we can should keep it for now. Let us know when if/when you change your mind ;-). -- Thanks, Maxim
