HN Explorer

Ask HN: Experiences with the free GitLab account vs. competitors?

3 points by hellofunk ↗ HN
I was surprised to see that GibLab offers free unlimited private repos, and other interesting features, that clearly compete with Bitbucket and Github. I had previously thought they were just for self-hosting, so perhaps they have evolved recently.

Have any former GH or Bitbucket users migrated their personal or corporate work to GitLab and care to share experiences of what worked, what didn't?

5 comments

[ 3.0 ms ] story [ 25.8 ms ] thread
LurkerAbove
I used to more or less alternate between Github and Bitbucket but I've been using GitLab exclusively for the past six months. So far it's been a mostly painless transition. I've only had one hiccup, I had an IP banned for an hour or so but that's more Netbeans' fault than GitLab's .
hellofunk
I've read reports that GL is quite slow. Just browsing their site or existing issues pages causes long load times. Has that been an issue for you?
jobvandervoort
GitLab itself is pretty fast if you host it yourself. Pretty much independent of scale.

GitLab.com, our hosted instance that you're asking about, is slow. Sometimes painfully so. There are several reasons for that, but as an end-user that won't relieve the pain.

We are working on it, hard. You can get an idea of that here [0]. As you can see, we've made significant improvements over the past months, but are now facing some setbacks with our hosting provider.

At GitLab Inc, we use GitLab.com exclusively for all our development. We have all the reason in the world to make sure it runs really really fast and stable. I'd encourage you to give it a try and otherwise consider running your own instance. It only takes about 2 minutes to set one up, less so if you use pre-build versions like our Docker containers or various cloud-platform images [1].

[0]: https://gitlab.com/gitlab-com/infrastructure/issues/59

[1]: https://about.gitlab.com/downloads/

LurkerAbove
Page loads on GL.com can take a while, but I wouldn't say that I've found the actual pull/push cycle to be slow per se. I'd describe it more like "measurable."