2019 FreeBSD Community Survey

ucomp said:
17. At home, I have approximately the following number of FreeBSD installations:
Option >10000 😂
Click to expand...
There are probably some people who have that many FreeBSD installations at work. However, they would be smart enough to not speak in public for their employer, unless explicitly authorized. I would also think that the foundation would know about those large users, and would be in close contact with them.
 
ralphbsz said:
Explain? Why? Just out of curiosity?
Click to expand...

You're welcome. It's a bit difficult to coin in proper words, I'll give it a try.
The survey seems to imply a kind of, lets say, "business use"; it implies that one has an "employer" or a "business". (What about retirees, what about people suffering unemployment?)

But then, concerning business - when I had a job (at a computer company) I found that these people simply hate computers. Nobody would be interested in what kind of operating system one would use (that was just a label tag to be passed on unreflected: it's this OS, so we need to quarrel with that company); the only thing that was of interest were so-called "business cases", as they contain dollar figures. A dozen times a day I got to hear "that's too technical" - concerning everything that did not end in a dollar figure.
That's why ITIL was invented - ITIL enables companies to use computers while guaranteeing 100% skillfree operation, by strictly dismissing all technical matters and focussing on pure buerocracy.
No, employers don't give a damn about which OS to use - if anything at all they want to use Windows, for whatever reasons I don't want to understand.

So, when I came home from work and could sit at my FreeBSD, that was recreation: finally to handle the real thing: a proper computer, built by those great and admirable people like Kirk&Eric!
And on occasion FreeBSD actually helped me to solve work problems. For instance there once was a storage (filesystem) problem with an AIX system. One cannot look into the source there, because they don't give you the source. But I could look into the FreeBSD source, and that explained the problem.

So, its all just copied. Putting aside the minis and hosts, how much of the currently running installations are in one way or another copied from what was initially created in that garret at Bell's and then refined at Berkeley? I'd suppose: almost all of them. And who is conscious about that? Almost nobody.
And this is not a matter of using one OS or another, this is something that started a process that is currently transforming our whole society, the entire world - to the better or worse, nobody knows.

Then another thing.
I don't know how FreeBSD preferres to separate so-called "developers" from NPCs. But it seems to me that such separation gets more and more expressed. I don't know what to make of that, and even less do I know where to place myself. Three weeks ago I helped identify a problem with postgres, last week I made a bugfix for apache, recently I sent a bugfix for bacula (nobody cares about, as usual there), and so on - but I am definitely not a developer, because a developer produces something (he is a production slave in the sense of the communist manifest).
Whereas I just do what was always normal for people to do: when the egyptian fellach farmer in 2000 b.c. would experience a torn harness on his plough, he would not call for a "specialist", he would just take the proper tools and fix it. For 100'000 years, when humans did use tools, they were in control of these tools, they could in most cases make and fix them on their own. Why should we give up this control, this power-over our circumstances, just because now there are computers?
So, if something does not do what I want, I just go to /usr/src, understand it and eventually change it - the same way as it was always done at Berkeley. (If you would run a Berkeley, you would have the source. Yes, there were copyright issues, but there was never a stance that this source is only to be handled by so-called "developers".)
Well, maybe, I could do the job of some developer as well, as they also seem to just understand and work the code.

So, in my eyes that whole separation between developer and NPC is highly obscure and simply doesn't exist: if something doesnt do what you want, you change it, and if nothing does what you want, you write something. (Actually even Zuckerberg did just write something he wanted: keep track of his different f***buddies he couldn't distinguish otherwise).
 
PMc said:
The survey seems to imply a kind of, lets say, "business use"; it implies that one has an "employer" or a "business".
Click to expand...
You are at least half correct, perhaps three quarters. Some of the survey takes about home (or hobby) use, and a lot about business use. I answered the survey, but only the parts about home/hobby use. While I work with lots of computers at work (and have for the last 25+ years), I don't feel comfortable speaking on behalf of my current employer, so I did not answer questions regarding by use of computers at work.

What about retirees, what about people suffering unemployment?
Click to expand...
BTDT; about 2 years ago, I retired (quite early) from computer company A, took about a year off (not being employed), then went back to work at computer company B.

But then, concerning business - when I had a job (at a computer company) I found that these people simply hate computers. Nobody would be interested in what kind of operating system one would use (that was just a label tag to be passed on unreflected: it's this OS, so we need to quarrel with that company); ...
Click to expand...
I have worked at some of the largest computer companies in the last 25 years, and my experience is very different. Sure, everyone hates computers for 30 seconds at a time. Matter-of-fact, that's one of the things I tend to mutter into the room when something is broken again (and it always is, doing R&D and software development is mostly the task of making good ideas be non-broken in their implementation). But I find that my colleagues are passionate about what they do. While they all program in the languages that are needed at the moment, they all have strong opinions on what a better programming language would be. They all like and dislike OSes, for good reasons (and no two ever agree on their preferences or their weighting of reasons). For example, in the last ~15 years I have professionally mostly used Linux, with some AIX and HP-UX thrown in, and I personally don't like any of these OSes. If I had my choice, I would do server development and deployment on *BSD (OpenBSD or FreeBSD, in that order), and desktop/UI development and use on MacOS. But I don't have my choice; I'm a professional, so I do what the majority of my colleagues need to do.

No, employers don't give a damn about which OS to use - if anything at all they want to use Windows, for whatever reasons I don't want to understand.
Click to expand...
I have used Windows as a desktop OS for about 10 years (at work), and it is perfectly functional. Personally, I find it very annoying, but I respect its advantages for mass-production corporate desktop use. I try to minimize my use of it; at home I'm about 99% successful (occasionally, I still need to boot Windows, sadly), at work about 70%.

Where you are right: Corporations don't care about which OS to use. At the very basic level, they simply care about making money. Which not only makes perfect sense, it is even logical and the law: investors give money to corporations, with the expectation that the officers of the corporations will use the investment for the highest possible return. If I were a stockholder in a corporation (which I am), and I would find out that the bosses use an OS that they happen to personally like but that is less good at being productive and profitable, I would see to it that they (a) get fired, and (b) go to jail for mishandling my investment. This is exactly the same thing as when they hire someone not because the new hire is expected to be a good employee, but because they think the person looks sexy and they expect to sleep with him/her.

So, when I came home from work and could sit at my FreeBSD, that was recreation: finally to handle the real thing: a proper computer, built by those great and admirable people like Kirk&Eric!
Click to expand...
I still don't know who you mean by Kirk and Eric ...

I even agree with that. For the server I use at home, I use the OS that gives me the most joy. I get joy out of it being easy to administer, logical, productive, and not annoying. The reason I abandoned using Linux for my home server about 10 years ago was that it was getting too disorganized, too fluctuating, and too unreliable with regular major changes. That's when I switched to OpenBSD and later FreeBSD. On the other hand, I have discovered that for my Raspberry Pi, FreeBSD doesn't work well enough (it really goes on my nerves when it only works 90% right), so there I have switched back to Linux (Raspbian to be exact), and it is mostly delightful to use. I even discovered that I like configuring systemd (even if I continue to think that systemd is the worst software engineering train wreck of the last 10 years).


... how much of the currently running installations are in one way or another copied from what was initially created in that garret at Bell's and then refined at Berkeley? I'd suppose: almost all of them. And who is conscious about that? Almost nobody.
Click to expand...
Yes and no. Many good ideas have been taking from Unix and BSD. And many bad ideas that started in Unix and BSD have also been rejected. And Unix in particular took many good ideas from its predecessors, in particular Multics (even the name Unix is a parody of Multics). So "almost all of them" is a vast exaggeration. But where I agree: Both the Bell Labs and the Berkeley systems research groups greatly pushed the state of the art forward.

Where I disagree: Everyone who knows about the history of computing is conscious of that. You may not find it in the marketing literature; for example Microsoft ran around in the mid-90s claiming that NT was going to be a great new operating system, when in reality it was mostly the ideas (and management and implementation teams) from Digital's VMS merged with the 16-bit Windows tradition; and VMS itself was a hybrid of the RSX-11 tradition within Digital, with stealing all the better ideas from Unix and Multics.

But that doesn't mean that I want to use a research prototype operating system. I also don't drive a car that is a prototype out of the research labs of the big car companies, but instead a boring and inexpensive Honda, which is reliable, practical, and gets me from A to B with no hassle and little gasoline. (Funny story: A good friend of mine, who has a PhD in computer science, used to work in Daimler-Benz' R&D lab, and on weekends he used to borrow research prototypes, and use them to visit his family, or do mini-vacations on the French Riviera. Crazy times.)

... sorry, ran out of time to respond more, have to do something else for the evening ...
 
ucomp said:

Online Survey Software: Closed Survey

This survey is currently closed. Please contact the author of this survey for further assistance.
www.research.net www.research.net
17. At home, I have approximately the following number of FreeBSD installations:
Option >10000 😂
Click to expand...

I've had hit someone running with 60K jails for research purposes (at home). Also, just for the matter of curiosity, some associates of a IRC fellow have mainframes at home for fun...

You never know what people do at home! ;)
 
ralphbsz said:
There are probably some people who have that many FreeBSD installations at work. However, they would be smart enough to not speak in public for their employer, unless explicitly authorized. I would also think that the foundation would know about those large users, and would be in close contact with them.
Click to expand...

Not really possible, from time to time some corporation has a problem look for help, and just then the Foundation/Project become aware they are using FreeBSD for years. ;)
 
ucomp said:
I have approximately the following number of FreeBSD installations:
Click to expand...
I felt guilty at 6-19 (Seemed like a strange range of numbers in survey)
It was a good question. It made me think. I have around 8 in a rack and 4 in various other spots counting laptops.
Not counting VM's.
 
rigoletto@ said:
I've had hit someone running with 60K jails for research purposes (at home). Also, just for the matter of curiosity, some associates of a IRC fellow have mainframes at home for fun...

You never know what people do at home! ;)
Click to expand...
As long as you do not end up in jail yourself, that's a good value for an attack, so the option should be extended to > 60000 😂
 
Phishfry said:
...It was a good question.... It made me think..
Click to expand...
yes, I realize that too, e.g. :

rigoletto@ said:
Also, just for the matter of curiosity, some associates of a IRC fellow have mainframes at home for fun...
Click to expand...
e.g those associates could spend some Tier2(sparc64 etc.)- VMs for FreeBSD-developers...
since my power was once turned off because of too high an electricity-invoice, I do not operate servers in private rooms anymore and there are no hosters for sparc64 or Power 8/9 e.g and Qemu is a load of crap for that purposes
 
I did notice RISC-V as a question choice for other platform support. That space might go somewhere soon.
 
Phishfry said:
I did notice RISC-V as a question choice for other platform support. That space might go somewhere soon.
Click to expand...
AWS recently supported Risc-V for FreeRTOS, so your RiSC-V thing should be addressed for FreeBSD.
Also, I have an aarch64 instance running on AWS (FreeBSD13), but Power8 / sparc64 , eventualy Mips is simply nowhere to be found in the cloud.
 
ralphbsz said:
You are at least half correct, perhaps three quarters. Some of the survey takes about home (or hobby) use, and a lot about business use. I answered the survey, but only the parts about home/hobby use.
Click to expand...

Yes, I then figured that it does not seem to harm to leave certain questions unanswered.

I have worked at some of the largest computer companies in the last 25 years, and my experience is very different. Sure, everyone hates computers for 30 seconds at a time. Matter-of-fact, that's one of the things I tend to mutter into the room when something is broken again (and it always is, doing R&D and software development is mostly the task of making good ideas be non-broken in their implementation). But I find that my colleagues are passionate about what they do. While they all program in the languages that are needed at the moment, they all have strong opinions on what a better programming language would be.
Click to expand...

Oh no, that is something normal! What i experienced was entirely different. First of all, one was not supposed to do any coding: all development has to be done in India. Certainly there were processes in existance how one could request the approvals for such, could request the funding, could then request some developer from India to be assigned to the task, and so on. But consider some custumer who wants some non-out-of-the-box reports or some modified mailer config or such. Normally you would just grab perl and write it, or go into the mailer config and tune it. But you were not supposed to do such, nor would anybody in charge understand it or even know such skills might exist or might be necessary.
Interesting thing was if something failed or didn't work. Then one would have endless phone conferences with some ten or more people, all of them designed "managers" of some kind, all producing massive amounts of stress and telling how important the customer was, but none of them being able to contribute substantial information on what actually was the problem (on a technical level) - it could take hours until one would get something useful from that (for instance a hostname), to be able to go and look what actually was the problem.

But then, if You ever did look into project-management - I am using the PRINCE2 scheme, but PMI might not be very different in that regard -, it is clearly stated that
  • everything is a product (including all the paperwork),
  • the actual technical things to be created are just a certain kind of product, called a "specialist's product",
  • those "specialist's products" are out of the scope of the business; they are created by some foreign or subordinate people termed "specialists".
You can use that scheme to build a house, or a car, or a software - it does not make a difference; the method is de-coupled and abstracted from the actual task and is purely administrative (aka buerocratic): to correctly run a suftware-development project, the project manager would not need to know anything about computers!

So, what these folks at the company were doing was correct and in line with the specs. And when I asked around among colleagues for other experiences, it was confirmed that other companies would work in quite the same style.

Where you are right: Corporations don't care about which OS to use. At the very basic level, they simply care about making money. Which not only makes perfect sense, it is even logical and the law: investors give money to corporations, with the expectation that the officers of the corporations will use the investment for the highest possible return.
Click to expand...

That's correct, but my impression is, this is a very American thing which has spread all over the world. Originally a company would have a product (considered a great product) and primarily want to make that product available to the masses - and if done right and if the product is indeed a great product, that should produce a good return on investment. But in the modern way of understanding, as it is currently taught in business sciences, the company is reduced to only produce maximized return on investment, while the product is secondary and can be choosen more or less freely.
i think such a philosophy brings along a couple of problems - but that would lead into a discouse about economic policy, so lets leave that out, and agree that this is indeed the current understanding.
 
hotaronohanako said:
This looks exactly like the Mao way to let the people speak in china during the cultural revolution .
Click to expand...

Really? What gives You that impression, if I may ask?

I rather think these people have a problem, and I am now beginning to understand the nature of their problem.
 
Looking thru the survey questions more deliberately, I think I get a clue on what is the intention behind: the originators seem to look for a task description fo themselves, i.e. how FreeBSD (as a product!) should be targeted in the market and which services it should provide, which use-cases it should support.

This is obviousely necessary, and it explains the idea behind the questions, for instance, how long releases should be supported, how FreeBSD is used to build other products, how important predesigned ("opinionated") configurations are, how security mitigations should be handled, etc.

So I'm indeed a bit at the wrong place in that survey, as I don't need or expect any such "services". I remember when I looked at my first release CD, I was thinking "wow, how many man-hours of work have gone into this little shiny thing and are now made available?!?". And I was immensely thankful, and I still am.

But today you need to provide a marketable product, otherwise you will just vanish in the flood of all the "cool&new" things. So, my personal interest (just having this wonderful OS and codebase to do with it whatever I like) is kind of someplace-else to the current mechanisms of the market.

And, as ucomp rightfully stated: developers should not starve.
 
A word of caution about filling out this survey...

The FreeBSD Foundation (FF) put out a survey once before in 2016 (I think). I diligently filled it out taking some time to put in as much detail as possible. Start to finish probably took about 30-40 minutes. And then never heard any more about it. Fast forward 18 months and I bumped into Deb Goodkin of the FF at a BSD conference, manning the FreeBSD stall. Remembering the survey I asked her what had happened to it. Her response astounded me. She told me that they had not had enough time to go through the survey responses as there were too many of them. She said that they had printed the responses off and put them in a drawer and that she had completely forgotten about that survey until I reminded her.

So, if anyone on here is from the FreeBSD Foundation, can you please (at a minimum):

1. Not put out a survey unless you have the manpower to follow through (too late now I guess)
2. If you do put out a survey and suddenly find that you don't have the manpower, please be fully transparent and let people know
3. Take the time to publish the responses and results somewhere, as cross-pollination of ideas can often have a ripple effect

I wonder how many good ideas were lost in that original survey? Or what other direction the project could have taken if the original responses had been taken into account? I guess we'll never know.
And just for the record, I will be filling out the survey again, but only because my affection for this OS is greater than my desire to not have my time wasted.
 
xavi said:
I think OP is referring to Kirk McKusick and his partner Eric Allman.
Click to expand...

Yepp! I thought using the ampersand symbol (which is often used for people in a personal relationship) should make things quite clear. I for my part had recognized the work of both for long, and then much later noticed that personal detail, and just thought, wow, lucky both of you! :)
xavi said:
Remembering the survey I asked her what had happened to it. Her response astounded me. She told me that they had not had enough time to go through the survey responses as there were too many of them.
Click to expand...

That's sad indeed, but can always happen. If somebody asks us question, we can never force them to do something sensible with our answers. (One gets used to that.)
And, commonly, as Dylan Hunt used to say: "if you want something done, you gotta do it yourself."

So, lets tackle this from a broader viewpoint: I am not involved, I usually don't meet people at conferences, and I don't know anything about the ongoings in this core team. But since Your description, like a spotlight, gives a little glimpse of it, I think it is really grotesque that for instance some Mr. Zuckerberg makes a fortune by creating a bad web-interface for people's self-expression, while we, having the probably best OS, and all of them copying (or, as I didn't dare to say, but ralphbsz did: stealing) from us, don't have the funds to staff the usual crap, like interpreting surveys (which is indeed crap work and no fun, so you have to pay somebody for).

I am not sure on how a remedy could look like, but as other people with a lot more unsubstantial agendae seem to achieve it, it should be possible. Maybe we should immanentize a bit of a thinktank to brainstorm approaches of marketing and fundraising. And I don't mean a thinktank of hackers, but people knowledgeable, business analysts, lawyer, that kinda folks - proably we should steal a round from them...
 
xavi said:
put out a survey once before in 2016 (I think)
Click to expand...
Yes this was it here:
forums.freebsd.org

FreeBSD Community Survey

The foundation is doing a survey and I am sure they would appreciate everyones input. https://www.surveymonkey.com/r/freebsd2016 https://lists.freebsd.org/pipermail/freebsd-announce/2016-June/001727.html Sorry for the short notice. This was from June.
forums.freebsd.org forums.freebsd.org

I am sorry it was handled that way. I still am glad I filled out one this time.
Hopefully it gets used. Somebody put some time into the questions.
 
You must log in or register to reply here.
Back
Top