Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

How do you estimate Fossil's general popularity?

Most projects are invisible to you, like in-house projects that never see the light of day. If Fossil is used by 1% of those projects, that's still a lot of projects.

How can you tell if Fossil is popular in HN? By my reading, only two people in this thread use it. More than that mention having never heard of it before, or make statements where it's clear they don't really understand what it is.



Fossil is posted to and discussed on HN all the time. It's a novelty, it differs in design from the popular tools that people actually use in their day-to-day jobs.

There's really no reason to think that Fossil is used internally at a rate higher than its public-facing usage. And I've literally never seen a project that uses Fossil aside from Fossil examples.


I use it for lots of projects, spanning years of work, including commercial projects with ~20 devs. Those projects aren’t public, and I’m sure lots of other people do use it that way. It’s still not going to rival the install base of git.


I use it too. I like to self-host my commercial projects, and when I decided to move away from Subversion about seven years ago I did consider Git, but the ease of self-hosting Fossil made the choice for me. Just plug it into apache httpd and off you go. Some time later GitLab became available, so now I'm also running a GitLab instance. But the Fossil server has much smaller footprint and is easier to manage.


The ease of use/management is a big deal, and to an extent I think it falls out of technical design decisions (proper database for repo storage that fully, easily, obviously allows first class multiple checkouts of a single repo instance). Others are philosophical design decisions, most of which I think I like, some I think I don’t, and others that just aren’t in place yet. Key is it’s nice to work with, and if I’m doing development/management, as much as possible I want it to be an enjoyable experience.


"All the time"? I counted about one HN posting per month.

From https://news.ycombinator.com/from?site=fossil-scm.org there are 20 links to fossil-scm.org in the last 2 years. I manually looked at the last 2 years of submissions with "fossil" in the name and found an additional 4 which were not to fossil-scm.org.

It's over 50% higher for Mercurial. I counted about 30 postings with the name 'Mercurial' over the last two years, of which about 7 were from mercurial-scm.org. I left out some of the obvious duplicates.

You write "I've literally never seen a project that uses Fossil aside from Fossil examples".

How much effort did you put into looking, and would you have recognized one if you did?

After SQLite, the most widely used project which uses Fossil is likely Tcl. http://core.tcl.tk/tcl/wiki?name=Index says it uses Fossil 2.7. For obvious reasons, there is an affinity between Tcl projects and Fossil.

A search for '"This page was generated in" "Fossil"' in DDG finds some non-trivial active projects: https://duckduckgo.com/?q=%22This+page+was+generated+in%22+%... . "Active" means "commits in the last few weeks." For examples:

"MySQL++ is a C++ wrapper for MySQL’s C API" - https://tangentsoft.com/mysqlpp/home

"Jsi is a C (+/-) embeddable JavaScript interpreter" - https://jsish.org/fossil/jsi/doc/tip/www/home.wiki

"Cxxomfort (cxx as in C++, comfort as in comfort) is a small, header-only library that backports various facilities from more recent C++ Standards" - http://ryan.gulix.cl/fossil.cgi/cxxomfort/index

"SquirrelJME is intended to be a Java ME 8 compatible environment for strange and many other devices" - http://multiphasicapps.net/doc/ckout/readme.mkd





Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: