Edit: Wow! I really appreciate everyone’s perspectives here!
Edit: Wow! I really appreciate everyone’s perspectives here!
For that, the ultimate tier is now competing with a Jira suite, or something that PMs use made by Microsoft or a similar type of product.
The licensing for the project management tool can easily run $50+/month/user
Add in the "business/customer access to the issue tracking" and if you've got Jira or similar, that adds on a gain a bit more $$.
Next, adding the artifact scanning tools. Yes, you can get them 3rd party... but its another thing to set up.
In the end, picture the 50-100 employee shop where the sysadmin is weighing in on "get GitLab Premium vs GitLab ultimate" and the reduction of additional installs and licenses that need to be managed.
Yea, it it adds $80/user/month, but if it saves an hour or two per user per month in productivity gains - that can make a difference.
In the organization I currently work for, we were considering the options of (GitLab Ultimate) vs (GitLab Premium + Jira + other tools) and ultimately went with the premium and other for a number of reasons... but GitLab Ultimate made a strong showing.
I feel that it wasn't made to host enterprise products but rather FOSS in competition with GitHub and then made a pivot to enterprise products but not quite moving away from the "this is for FOSS on the web".
Things like "can modify own comments" not being something that can be modified (people updating their own comments and then not being able to go back over it for an audit or information request of some sort) or having a manager role that can administer the members of a project but not the code.
And yea... having to pay $100/month for a product/project manager in a larger org is a rather firm "nope." Have a dozen of those and that's a lot to pay for not much functionality.
A previous company that I worked at... there was a project manager. Just one. It was a small company - I sat across from the other Java developer. In there, were you've got a small organization where $100/month for a PM is a lot to swallow, but its only $100/month for one PM (and then you get all the other features that saves you from having to hire another operations person) - that's not as hard a price to swallow.
There the thing would be "ok, we've got 10 people at $20/month and all this other stuff that we need to set up on our own, so $200/month vs 11 people at $100/month and a bunch that we don't need to set up on our own $1100" -- the "is it worth $900 for the rest of the infrastructure done for us - along with dumping trello for gitlab issues" and the answer would probably be "yes."
a) how much money they have
b) how little understanding of tech they have
For many non-traditional tech companies, they feel spending more money, will help offset tech deficiencies. In the same way that, novice golfers equate spending more money on golf equipment will make them better golfers.
No other industry has made themselves so available at such affordable pricing than tech has?
Where are all the blog posts/free software/free advise from those finance companies?
Where is all the free legal/free court counsel info from all lawyers?
It's about time tech started charging for commercial use.
Software engineers worldwide deserve more of the pie and letting traditional industries eat Michelin quality food at fast food prices has to end.
Retailers don't sell at cost - they never have. They sell at as much profit as they can get (and yet free ERPs exist/cheap e-commerce exists).
All so the non-tech business people can have greater profit? Fuck that.
Based on my observations so far, Ultimate is quite expensive when you have an "all in" approach. Every user, despite whether they would use the feature set must be licensed as Ultimate under the current model.
I really wish there was some flexibility in how the licensing model works — right now, the all in approach is prohibitive to bringing on new users, as you need to justify the $1,200 upfront cost per user added — not all users / projects need to be licensed under the same tier — some user's needs are basic and cannot justify the cost of a license.
--
Some thoughts on how this could be addressed...
A) Assign a licence tier to a user (ie: Some users get Ultimate features, others are on premium with limited access to features)
B) Assign licenses based on projects — if a project has "Ultimate" features enabled, all users in that project count towards an "Ultimate" seat
If you want everything integrated together this is probably fine, but GitHub Enterprise + Jira Premium is $35 per user per month (and every other project management tool is cheaper and better than Jira, just about). I don't know what the value add is here, other than self-hosting. (Or at least, the $65 value add).
Preferably with an emphasis on speed?
We're building Linear, focusing on speed and providing sane defaults that sense for software companies. The app offline first so all interactions are <100ms. Sprints, projects, roadmaps, multiple teams, GitHub/Gitlab/Sentry supported etc supported. Lot of YC startups and growth companies (100-300 engineers) use us.
If you are building a docker image for example, you likely are doing a pull from docker hub, the security scanning software can help you catch security issues there.
It can also help with things like dependencies in your node projects, python projects and more (at least, that is what the security scanning software we use at $work does, I assume gitlab is similar).
It's not about employees pushing malicious code, it's about catching issues with dependencies further up the stack, to make sure that the end result you are pushing to your servers/users is not vulnerable.
Today, we’re making GitHub Enterprise Server 3.0 available as a release candidate. Announced in the GitHub Universe Keynote, it’s the biggest ever change to Enterprise Server, bringing customers:
* Actions – developer-first workflow automation and CI/CD * Packages – publish and consume packages together with code * Mobile Apps (beta)- iOS and Android apps for collaborating from any device
For companies interested in automating code security with GitHub Advanced Security, we’re also bringing:
* Code scanning – scan every pull request for known vulnerabilities with CodeQL * Secret scanning (beta) – detect credentials in code before they hit production