This is a dumb comment trying to pass the buck IMO. Of course it's a national lab, but probably a specific group at a national lab (potentially even in a Office of Science arm) going through their support channels at Cray/HPE to get a fix, presumably porting away from intel compilers or something or trying to onboard to something else (conda/spack).
Cray/HPE could/should have just do the fix or pay for it, since presumably they are packaging gcc/gfortran as part of the OS. Somebody there has to be qualified to execute such a fix.
It's crazy to me that open source maintainers offer free support to corporations or governments that they could easily extract money from for the benefit of the project or themselves.
An individual wants a feature and is prepared to pitch in, probably worth the time investment for a maintainer.
The US government wants floating point arithmetic done a certain way, fuck you, pay me.
I work for a government contractor with military sponsors. I agree with this 100%; the government should pay people to work on FLOSS code. Not only because I might get paid more, but because I strongly believe that classification and bureaucracy are tools to hide incompetence, and also that as (much as possible) public money taken from everyone should produce public code that benefits everyone. It’s really a win for everyone doing the right thing.
It's disappointing that this response linked by OP was posted at all. And even more disappointing because context gets lost every time it shows up on HN.
The linked email is from an HPE/Cray employee interacting with the upstream gcc team, not from anyone in the U.S. government.
The U.S. Government, via lots of programs, national labs, etc. does pay people (and companies) to work on open source code. This has at various points included LLVM, clang, flang, gcc, and many other projects. We like it when things get upstreamed and we also contribute ourselves to these projects.
Certain companies' willingness to put in the work required to upstream has been an issue at times, but it is improving and it's something that we push on very hard.
They afford to, but it's surprisingly non-trivial for a government to fill out all the paperwork involved and get the various approvals needed to procure/pay someone.
It's common in big companies too. My employer has to use a 3rd party supplier for Google Maps, because Google sure as hell isn't going to jump through the hoops necessary to become our supplier.
Yeah but that's why governments hire consultancies... they're not just preventing the government's payroll not explode (usually, government jobs are all but unfireable whereas contracts with consultants aka body shops can be discontinued easily), but they can also act far, far more flexible than any government entity can on their own.
Consultants are hired because hiring consultants is the first advice that consultants generate. Its a recursive loop of scamming government money on undeliverable projects.
This whole boogeyman over "unfireable" government employees is nonsense created by people looking to sell private contracts. Most consultants when hired somehow themselves become unfireable and even more expensive for basic work that could be done by competent employees with a decent paycheque and union jobs.
> Consultants are hired because hiring consultants is the first advice that consultants generate. Its a recursive loop of scamming government money on undeliverable projects.
That's more a problem of governments being unable/unwilling to properly scope projects - in construction for example, politicians love to low-ball estimates for their pet projects in order to have better chances at getting them passed. When you have competent government that is willing to work with and sell reasonable estimates, you get examples like the Luise-Kiesselbach-Tunnel in Munich, which got finished on time (IIRC, even a few months earlier than planned) and in budget [1]. In military projects, the "overestimations" and giant budgets are usually because money is diverted to clandestine, top secret projects that are not supposed to be known to the general public outside of a very few high ranking military and government officials. You gotta read between the lines here.
> Most consultants when hired somehow themselves become unfireable and even more expensive for basic work that could be done by competent employees with a decent paycheque and union jobs.
I've worked in the public, private and consultancy sector in Germany, so I can at least add some context from a German perspective.
Usually, government employment is unionized in one way or another and trades lower paychecks for long-term job security, in the case of the Germanic "Beamtentum" ("public servants") it's a literal lifetime agreement - the government will take care of you until you die.
However, that also has its downsides - in Germany, the difference for a simple programmer or IT admin in the public service to a colleague in the private sector can be as large as 4x. You simply won't get staffing for these positions outside of utter idealists and people looking to start a family. And you can't go outside the unionized agreements and pay market wages because that would result in "low level" staff getting paid more than the director of the entire agency in the worst case, and in the best case it would only blow apart the entire wage hierarchy. Additionally, the staffing cost for a project realized with consultants can be directly booked on the budget of the project, whereas assigning time/wage for government-hired employees is way more difficult.
Hence, the government is better off hiring specialists from consultancies and body shops... it allows the government to keep its existing wage structure, keeps permanent obligations from the books (similar to the capex vs opex debate), and to get actually validated experts instead of having to deal with validation itself.
Cray/HPE could/should have just do the fix or pay for it, since presumably they are packaging gcc/gfortran as part of the OS. Somebody there has to be qualified to execute such a fix.