GitLab raises $20M Series C round led by GV

Could you clarify what you mean by the difference between colo vs on-prem?

We run their on-prem in our colo, and it works great for us.

Yes, this is true to an extent. But price is one of many factors that go into GitHub's design decisions for infrastructure planning. (I used to work on the Git Infrastructure team at GitHub, which is responsible for exactly this.) Nor is it the _first_ consideration. Price is important (obviously) but only after performance and resiliency.

Putting repositories on bare metal on-premises makes sense for GitHub's architecture. They run git itself, which is happiest packbuilding with as many IOPS as it can manage. But they've also built out a number of custom performance monitoring and load balancing applications, and any single Git repository is replicated across a number of servers using spokes, their custom Git load balancing system. They run on-premises because they've done a detailed analysis of their needs and hired a staff of developers and infrastructure experts to deliver it. And as a result, they're the world's largest Git hosting provider.

But you _can_ be perfectly happy in the cloud.

Visual Studio Team Services hosts its Git repositories in Azure. Since VSTS didn't start out hosting Git repositories, but Git repository management was added _after_ it already had a cloud-first architecture, it didn't make sense to host repositories on-disk. (Nor would relying on Git for Windows have been performant five years ago.) Instead, VSTS built out custom Git server implementation to take advantage of storing repositories in SQL Azure and Azure Blob Storage. And this implementation scales up to the needs of Microsoft's customers - both external and internal - who are hosting the world's largest Git repositories.

So I think that you can be happy on-premises _or_ in the cloud. I think it comes back 100% to your point that you need to hire for that expertise.

I have been thinking about whether GitLab should opt for the M version Xeons when they finally get their servers so they can upgrade to 128GB TSV LR-DIMMs in the future.

We are very afraid of spreading ourselves too thin too. Recently we've been encouraged by how GitLab CI is doing, it is winning developers gitlab-ci-is-the-most-popular-next-generation-ci-system and industry analysts over https://about.gitlab.com/2017/09/27/gitlab-leader-continuous...

I think what makes GitLab different is that more than 1800 people contributed to it to increase the depth. Our team members are focussed on increasing the breath. From https://about.gitlab.com/strategy/

"We realize our competitors have started earlier and have more capital. Because we started later we need a more compelling product that covers the complete scope, that is integrated out of the box, and that is cloud native. Because we have less capital, we need to build that as a community. Therefore it is important to share and ship our vision for the product. The people that have the most knowledge have to prioritize breadth over depth since only they can add new functionality. Making the functionality more comprehensive requires less coordination than making the initial minimal feature. Shipping functionality that is incomplete to expand the scope sometimes goes against our instincts. However leading the way is needed to allow others to see our path and contribute. With others contributing, we'll iterate faster to improve and polish functionality over time. So when in doubt, the rule of thumb is breadth over depth, so everyone can contribute."

I don't see how GitHub will ever be dethroned due to the mind share and market share of Open Source projects. It is a simple process to move to GitLab BUT all the developer tools for many projects relies on GitHub.

When I get a package in Racket or R they normally come from github automatically.

You don't remember any downtime with GitHub? Lol.

I stopped using Github for private repos for three years. I get them for free from Bitbucket and Gitlab. So, I think the actual contender to Gitlab is Bitbucket and not Github.

However, Bitbucket feels a bit more organized and mature (the UI). What I also like: Bitbucket has the real Trello integrated which is great. I still like Gitlab, it's always good to have competition.

I haven't experienced any downtime with their public service - which I think is great for hosting private repos (as others have mentioned). I actually prefer it to bitbucket.

In other professional experiences, hosting our own instances of Gitlab was a great solution. Setup/updates were easy and we had the piece of mind knowing that we were in control of the maintaining uptime. I was very impressed by how simple it was to their implement CI service as well.

Really happy for GitLab and wish you guys all the best (and luck).

We've been using Gitlab for all our company's development, however, one major issue that is pushing us to switch over to GitHub is the fact that GitLab goes down almost every other day (or sometimes every day) due to deployments. Although during this often the site remains available, when CI is not executed or triggers are not performed, it is still extremely disruptive. I really hope that you prioritise having disruption-free deployments.

Quick question: for my side project I currently pay money monthly for both Cloud9 and GitHub, but I am very open to switching tools. One of the things I love about storing code in the cloud and manipulating code in the cloud is that absolutely nothing needs to touch my local machine, which means I'm free to use (almost) whatever machine I want with (almost) any browser or (almost) any platform I want. It also means switching costs are next to nil, since I don't have to worry about if it runs on a Mac or Windows or what native tooling I have.

However, a big problem I've discovered with Cloud9 is that it doesn't run on iOS and there are seemingly no plans to make that happen. I'm beginning to transfer over a lot of my work to an iPad, and with the incredible power offered by modern iOS systems, not supporting the platform seems a huge miss.

I see in your slides you're working on a web IDE for GitLab, and in some of the comments on the open issue, iOS is mentioned. Can you answer if mobile platforms like iOS are really in the roadmap?

As a paid customer, I'd switch in a heartbeat if I could use the same IDE on my Mac, PC, and iPad.

Canadians are criminally underpaid. I made significantly more than many full-time employees (I'd wager the majority) on more than half of my internships. Adjusted for rent and cost of living.

I do love London, Ontario though. Beautiful place even if it's too homogenous and quiet outside of around Western university.

GitLab is amazing but I wouldn't host with them until they fix their frequent downtime issues.

> What I'm most surprised by in this thread is the toxicness. As of right now there are only two threads that I can see, one that is slamming Gitlab's salaries and one that is slamming their reliability.

After many year on the internet, it's not that surprising. It's best summed up as, "Haters gonna hate.".

The jabs at the cloud offerring's reliability are valid. That includes both uptime and the horrendous data loss event they had earlier this year. I don't recall anything major after though not knowing about internals, can't tell if they learned from it or have just gotten lucky.

As a CE user I love GitLab. It does what it says on the tin and for the most part I don't have to think about it. The UI changes every other release are bit haphazard but not enough for me to truly care.

I've always thought that slamming their open comp policy is hilarious. They're one of the most open companies about how they handle comp and the result is they get insulted for it. I'm not saying I agree with the formulas (in particular the non-NYC COLA adjustments are out of whack) or the general approach, but to say they're worse than the voodoo guesswork that goes into figuring out how much a company is going to pay you is disingenuous. Especially if it's after a number of interviews only to find out they're going to low ball you.

At least in GitLab's case I know, at the current comp levels, I'd never work there or even interview there. And that's beyond fair for the worker. Plus it seems to be working fine because they've got a bunch of people that are happy to work there.

These "_______ raises $__MM Series _ round" posts always end up like this. They are essentially rah-rah-hooray posts with little or no technical content. HN has always been down on blatant marketing and that's what these posts feel like.

I think you would have been seeing a completely different discussion if the post was "GitLab launches _______".

What "toxicness"? I don't see any in this comments section? Did HN delete most of it or did the community down vote those entries into oblivion? Due to a few comments in here I did some digging on the compensation they calculate publicly. Honestly, I can actually see why they are getting negative feedback from it. Paying people 221k in San Fran for the same job they would pay max 102k for in Minnesota, regardless of the city, seems quite unfair. Minneapolis is actually quite expensive. Why not pay for the value a developer provides independent of location? Plus, if someone moves to a less expensive location, they lower the pay. Looking at all their compensation rules, I would never feel secure about knowing how much I might be paid month to month. It seems they can make new rules as they go along. Might they want to reduce my pay if my average monthly grocery bill goes down after learning to cook?

Used to be a fan, but the change in licensing is worrying me. We started with the enterprise edition, which was the top tier at the time. Since then another tier has been added above and another one is in the works. The road map is not encouraging, unless you are on premium or ultimate editions. And the license costs at least five times as much. Also the road map shows a very Oracle-esque thinking, that only the big guys need advanced features and they are willing to pay what it takes. I hope they find a better way, like scaling of the features rather than the features themselves. Reluctant to place more eggs in our GitLab basket.

Anecdotally, I've seen this same sort of phenomenon play out in plenty of other discussions. The worst ones are typically conversations about JS frameworks, iOS vs Android, messaging apps with "private" chat features (Telegram/Signal/etc), and battles about Gitlab/Github. For some reason these subjects seem to create an open invitation for haters of all stripes to list their grievances about the product under discussion.

I used to think it was mostly astroturfing -- and there may be some of that on occasion -- but lately I've come to understand this phenomenon as a "feature" of humanity's tribal nature, much like sports team preferences.

(Edit: that's not to say that fair criticism of a product is a bad thing. It can be healthy! But these conversations often go beyond balanced criticism, at least from my perspective.)

More likely Matt knows a lot about running a free software business, which is what GitLab is for the most part.