Complaining that an engineering company is full of mostly white males (in an industry which is full of mostly white mails) is an example of nothing. Red Hat started programs to push LGBTQ+ people in 2014, women in leadership in 2017, neurodiversity acceptance in 2018, black leadership in 2018. You may think all of those things are late, but as driven company missions which they spend real money on (and real money on community involvement on), it's not insignificant.
A "good-old-boy's club" at a company where only 25% of engineering was even in the US is a little mindblowing. The vast majority of RH's engineers when I left were in Pune, Brno, Tel Aviv, and scattered across Europe (eastern and western), with a reasonable showing from Brazil.
Open source is, for better or worse, a place where a cult of personality matters. "This is a bad idea and we shouldn't do it" isn't personal. It's how the community interacts. It's an invitation to argue back and say "this is a good idea because XXX". This is basically how any upstream mailing list works also.
While Red Hat did have some cultural problems, they were primarily around resistance to establishing an engineering presence in China (and it's hard to fault them for that -- my team was truly geographically distributed, and adding someone in APAC would have meant we literally never could have met at the same time), and regional cultural attitudes.
I left last summer, along with a lot of other people, but your comment is apropos of nothing Red Hat was from 2013-2021.
The vast majority of people I knew at Red Hat ranged from industry-average to absolutely excellent, but overall I'd say even Facebook had a better internal culture in terms of diversity etc. Painting RH as anything more than a typical tech company in those areas is a bit absurd.
For sensitive issues, it's hard to imagine this will lead to a constructive, curious conversation. It would be interesting, for example, if you and the GP tried to examine why you have such different experiences? Maybe you were in different places, working for different people, and saw different things? Maybe different job descriptions? The world is always much bigger than what we imagine.
It's essentially GP posting on the Debian/OpenBSD/LK ML and saying "wow, these guys are arrogant". Well, they have strong opinions, and they need to justify them. If you don't have thick skin, you should probably avoid upstream discussions. It doesn't mean you should cast aspersions on the developers or their communities/companies. Linus is an incredibly nice person, but you wouldn't think it if all you saw was his mails blasting Intel for Spectre/Meltdown. You'd have a very different opinion of the man than who he is in 99% of interactions, and commenting about it on HN would be likely to get a reply similar to mine.
GP only had "anecdotal" experience with Red Hat versus actually working inside the company. When I started, I was worried that it would be different "behind the curtain". It was not. The other who mentioned junior engineers speaking up if they disagreed was spot on.
In 201x (2015? 2017? I dunno), RH changed their pay structure from being bimonthly to biweekly in the US to align with the rest of the world. There were hundreds of replies to the entire company debating whether it was better to pay us once a month and manage your money better, or once a week to be more consistent, or...
That's what the company was. It was not by the time I left. Granted, when I started, RH HQ was still on a college campus. If all you ever read was upstream mailing lists, you'd get the feeling that everyone (Linux, Theo [not that he works for RH, obviously], Lennart, whomever) other than Dan Walsh was arrogant assholes. That's not representative of who they actually are.
Painting Red Hat with a broad brush from incidental interactions versus the perspective of someone who spent nearly a decade there is dramatically different. Did it have problems? Sure. But none of the ones GP mentioned. The world is much bigger than what GP imagined from his anecdotal experience.
RH was a unicorn, and it's gone. LinkedIn last summer was essentially rats fleeing a sinking ship. IBM spent $35B on a company with no actual assets (a few patents) other than smart, passionate people. They would have had to really try to screw it up. And they did. I don't honestly know if there will be another RH in the future, but I'm hoping there will be.
>If you don't have thick skin, you should probably avoid upstream discussions
I have a humble request, please discontinue this attitude on your own projects and please encourage other upstreams to also discontinue it. I'm baffled as to why so many open source maintainers seem to confuse needing to defend technical decisions with "having a thick skin," they are not the same. It's perfectly possible to be opinionated and strongly scrutinize a technical decision, while also not being harsh and rude towards the person presenting it. If you believe those mailing lists are unsalvageably hostile, then in my opinion, they should just be shut down. That's not the kind of place to have official project communication.
To me, the intentional insistence that you want to and will remain ignorant undermines the credibility of the rest of the comment. Who knows what else have you shut your mind to and thus overlooked at RH.
Sure, but then don't enter the discussion at all. It's generally no faster to fight with the other person than to try to understand them.
> no matter how much we might think _they're_ the ones in the wrong
That assumes that's the point of the discussion. Do it to find out about the other person, help them find out about you, and to explore what they know and what you know.
> there's just no convincing them
I think that assumes, again, this adversarial fight. I'm not trying to convince them (see above). And often, people are open and curious if you approach or respond to them differently.
This adds to my suspicions that the long-term IBM C-suite goal of the Red Hat acquisition is to eventually convert Red Hat to be like their captive mainframe revenue and profit stream. I would not at all be surprised to see a gradual deterioration in Red Hat quality in the coming 10 years.
Watch the general level of support engineering skills in what will eventually be a Blue-washed Red Hat L2 Support. That is your canary. What you are used to today in that tier will be gradually relegated to the tech leads for L2. Don't look for tech leads by title; look for the leads who gates decisions to escalate to L3. Highly-skilled L2 organizations grant most if not all their support engineers autonomy to escalate to L3. Profit-squeezing L2 organizations (what I suspect may be happening with Red Hat in the coming years) will gate those escalations, and staff vacancies with relatively down-skilled engineers.
Note that I also enjoy working with these relatively down-skilled engineers. No complaints about "bedside manners", and they can address common scenarios. Given enough time and exposure to more experienced staff, they would absolutely step up to highly-skilled L2 levels. The challenge is often these situations see those more experienced staff leave too soon. In my experience you need at least 7, ideally 14 years to season new staff to those skill levels, and the attrition rate is high over that time; troubleshooting well under pressure with grace is a very difficult combination of skills and characteristics to find.
Two ways I've seen to effectively deal with this on the customer end is find an alternative solution (not really feasible in this case at this time, though I'm watching several possible developments), or up-skill your staff. You'll know you've reached the right skill level when your engineers can delve into the internals of the product sufficiently that they routinely convince the down-skill L2 support engineers to escalate their lead within a day or two of engaging.
If my suspicions bear fruit (I really hope not, the vast majority of my clients hold massive people and capital investments in the Red Hat ecosystem), then this is extremely bullish for AWS. Their customer-obsession-to-a-fault culture would swiftly exploit IBM's strategy, and there would be no contest. I can see several simultaneous attacks AWS could mount to neutralize Red Hat's position if IBM's strategy at the end of the day really is what I currently suspect.
I'd really like to hear from others that have found other workarounds for this eventuality.
Disclosure: I might be biased as a Debian user/contributor.
In the 10-15 year time frame: IBM can't. Simple as that.
In the 3-10 year time frame however, which not coincidentally happens to be in the range of CEO tenures, the momentum of enterprise infrastructure relying upon a specific platform delivers a kind of lock-in that absolutely can be exploited by IBM leadership.
Look at Oracle. There are still quite a few use cases where their arguably painful pricing, licensing, and account relationship management experience are worth it. For some use cases, there is no other acceptable solution; I've seen them, and Oracle is damn impressive if you operate at that rarefied strata. But for arguably in the Pareto distribution of scenarios where Oracle is used, much of it like CRUD applications, Postgres does just fine.
Yet Oracle today is as dominant as ever in many enterprises.
When it comes to what can and cannot become captive revenue streams in the software market, I've learned to accept that the market can often be irrational for far longer than I can remain solvent.
I really like Debian. It is what I use in my personal servers. I'd like nothing better than to see it supplant Red Hat. But in the enterprise, the Debian story is under-developed. Debian is very rarely supported by application vendors. The LTS duration is half that of Red Hat. The members of the big, profitable enterprise market move that slowly.
I think those are tough problems to solve by someone like Ubuntu basing off a Debian distro to solve on their own. All the other challenges of interfacing with enterprise customers, communicating in their lingo, meeting them half-way, *etc., are addressable by a company organized around doing that using Debian, but there are inevitable questions of resolving differences of opinion in the future between the community and the company. I suspect Red Hat being both the distro and the commercial entity traded solving a lot of those types of problems with a different, more commercially-acceptable set of problems.
The discussion was announced in a blog post, but hasn't happened yet, some of the related blog posts are worth reading:
https://sunweavers.net/blog/node/129 https://raphaelhertzog.com/2021/03/30/challenging-times-for-... https://raphaelhertzog.com/2021/03/31/challenging-times-for-... https://raphaelhertzog.com/2021/04/01/challenging-times-for-... https://raphaelhertzog.com/2021/04/02/challenging-times-for-...
To participate you'll need to subscribe to the (currently very inactive) debian-enterprise mailing list:
https://lists.debian.org/debian-enterprise/
On your points about Debian & the enterprise;
The private ELTS offering by Freexian makes it slightly longer, but indeed less than RH.
https://wiki.debian.org/LTS/Extended
I don't think Ubuntu LTS is able to compete with RH support periods either.
I used inclusive in this context to mean that a junior engineer is expected to speak up and offer input in an executive meeting if they have something to offer. Cutting through hierarchy, "good ideas come from everywhere". This could be idealistic but it helped me in eliciting input.
I don't know what it's like to be on the inside or close periphery to daily core operations at RH. I agree and understand completely what you are saying. "Free speech" has (recently?) been a guise to act terribly toward each other in a number of settings. Trolls gonna troll. It's important to maintain decorum in the workplace and a culture that defies that can turn toxic quickly.
We do seem to have either interacted with difference slices of the company, or gotten different impressions.
Which is all well and good, an aggregate of anecdotes is better than just one.
From the outside they seemed pretty open. I've got no idea about how they did with black / LGBTQ+ - but it seems a big jump (in an industry where there are not that many black female LGBTQ+ engineers) to say they were so bad.
What is the engineering company with lots of black female LGBTQ+ folks? It might be helpful to be able to look at a model that has worked rather than just trashing RedHat for not being an engineering company that gets this right (and all these companies are pushing initiatives in most of these areas).