Message to programmers

After over a week of fighting with a barely functional computer, I have a few suggestions (from someone who first learned programming in the early 70s):

1: Make a sign and put it on your wall: “If it’s not broken, DON’T FIX IT!”
Making changes and obsoleting “old” software because it’s not the latest (and greatest) in your mind must be approached with extreme caution. I’ve had VERY USEFUL software be made useless by the “I gotta change something!” crowd.

I remember that sign being posted where the mainframe area was at my school. It was a problem even then. Someone would ‘fix’ something - and it would take programmers weeks or longer to figure out why suddenly certain packages no longer would work.

2: Another sign: “Just because you don’t use it, doesn’t mean it’s worthy of being removed!”

Many times (over the years) I’ve heard people claiming to be programmers and insisting that workspaces were useless and needed to go. They usually turned out to be gamers. I don’t do much gaming at all. I use the different workspaces, especially when I’m dealing with some heavy programming (like analyzing archaeological data). It helps to keep things organized.

3: Pay more attention to users and less to corporations. I’ve encountered issues caused by the profit-driven motive of corporations. For instance, back when W95 was coming out, I was using OS/2 Warp 3. (About equal in power to WXP or W7) When I tried to install W95 on dual boot, W95 TRASHED my OS/2! I finally learned of a workaround that prevented that and allowed OS/2 - W95 dual boot. I have other examples I could relate - videos I took which were deleted because “You don’t have the right to have this video!” (actually happened!) Programmers, please think long and hard about any changes requested by a corporation - their only good, after all, is increasing profits.

Just my two cents worth - and thanks for reading!
Bob

6 Likes

@Archaeometrist I started programming back in the 70’s… Sinclar ZX81’s… then onto a VIC-20, built a modem (to hit those local bulletin boards) and also a tape recorder interface to save all that typing… Back in the day when things like magazines and veroboard (I still use it for RPi3 projects) was the rage…

Unfortunately these comments will fall into somewhat oblivion here, it’s the wrong place for feedback…

See < https://lists.opensuse.org/archives/list/factory@lists.opensuse.org/message/QSFECIZRORDBBRM7F5XBYF5BOPLYOOZL/>

Likewise the openSUSE base is now maintained by our Primary Partner SUSE, so Community input is structured somewhat different now for packages and integration as it was in previous Leap releases.

2 Likes

back in the 70’s… Sinclar ZX81’s… then onto a VIC-20

Wasn’t this early 80’s?

On topic: don’t split an OS platform into several hundreds of variants.

1 Like

I agree. Make the OS more customizable, and less conformity demanded of the user.

Which then makes supporting an operating even more complicated, with every customization that the user does.

Its your machine, and you can do whatever you want with it, it’s one of the four freedoms, but nobody is required to support it when it breaks, or help you do it.

Throwing your 98 cents of change back over the counter… :slightly_smiling_face:

You clearly haven’t worked in any software development capacity for any length of time, have you?

You haven’t met any of the whiners, and of the “good idea” people, the brand-new product managers about to change the world (and their annual bonus, of course), the “but we want to build the best Linux of the world” crowd, have you?

I am now old enough to have met enough samples of any of them to make some sweeping general judgements.

Whenever you introduce anything new to a piece of software because some user begged you for it, and you saw that it was a very reasonable request, and you told that user who wanted it “but it comes with some caveats: it’s limited to this and that use case; it will never work in the general case for every possible scenario”, and that user agreed, and you implemented it with that caveat and announced the limitations, it will take about 20 nanoseconds for the wiseasses and the “but we want the best Linux of the world!” crowd to pop up and start whining and nagging.

It starts with a seemingly harmless question or an enhancement request in your issue tracker; something like “hey, I noticed that the finaglify feature doesn’t work in this case” (and describing said case), and you answer with "yes, that was the limitation that I announced and documented when the feature was introduced! - some more “…but couldn’t you just…” back and forth arguments later you start giving up on mankind, and at some point, that user falls silent. In the bug tracker, that is.

He is now moving on to Reddit, the Great Cesspool of the Internet, where all the wiseasses and whiners meet; and starts publicly spreading FUD and badmouthing that piece of software that he got from you for free; because Open Source, and damn all those “limited warranty” paragraphs in the license legalese.

Somebody in that forum will finally shut him up (yes, there are some of those people even in Great Cesspool Land), and he will disappear for the while. Some weeks later, a new user name will pop up in the same place and spread the same bad-mouthing and FUD again with the exact same wording.

You have made an indirect penpal friend for life; or at least for the next ten or so years.

Been there. Seen that. Got the T-shirt; the whole edition in different colors.

About removing features: You may or may not have read my long post about it, but this is exactly what ultimately killed YaST, and why nobody wants to touch YaST anymore with a 3 meters (the civilized part of the world uses metric, not some medieval king’s foot size) pole. Because a gazillion such people brought in such “good ideas” or worse, contributed patches to implement stuff that keeps adding technical debt.

It took 25 years, but now YaST is on the way out. It’s not exactly dead yet, but the smell has become unmistakable. Because we could never remove any of the counterproductive BS “features” that accumulated over the years.

So keep your old piece of junk software and have a lot of fun with it - as long as it lasts. When it breaks, don’t expect anybody to come to the rescue; because those who have the know-how also know the whiners.

As for “pay” and “corporations”, who do you think pays for software development in the first place? Who is funding Linux kernel development, Linux distributions, the tools that make a distribution special?

Is it grateful users who keep donating to Open Source with the PayPal account that many have on their GitHub page? - Okay, just kidding; in 10+ years of QDirStat development I received something like $15 of donations on average by year, amazingly often enough from people who I am sure couldn’t really afford it.

No, of course it was one of those “evil” corporations who paid me a monthly salary over many years. Without it, there wouldn’t be a SUSE distro, neither SLES nor openSUSE.

7 Likes

Huh? Tell that to MS and Apple. In Linux land, literally every aspect of the system is customizable; so much so that even seasoned experts have to double-check, read man pages and experiment before giving a user any answers.

And that aspect makes maintenance hell, and it’s why the number of officially supported packages that business customers pay as SLES etc. is so tiny in comparison with Factory.

5 Likes

IMNSHO you should be sentenced to rewriting an entire distro in ANS COBOL75. :rofl:

2 Likes

Variety is not a value by itself. It becomes a positive value if it is advantageous. I can see little advantage in making several hundreds of variants of one OS platform. It increases maintenance, testing and documentation effort, decreases compatibility, divides user groups into smaller communities etc. I can imagine it would be better if all community effort could focus on, say, 3 variants (HPC/server, desktop, mobile, for example) and one package system.

Customizability is a quite different question to me. Customizability (programmability, scripting, theming) is a good general concept, IMHO.

1 Like

It must be nice to be rich and able to afford the “latest and greatest”. Not all of this are that lucky - and in spite of the good ole American stereotypes, poverty is usually NOT the fault of the poor.

1 Like

Are you talking about openSUSE or is this some general statement?

You seem not to know where linux is nowaday used. It is everywhere: smartphones, cars, cloud computing, servers, home appliances, tv, servers, desktops, super computers, real time computing (medical, automotive, finance, public), …
Limiting the almost infinite usecases of linux to “3 variants” would be suicidal…

3 Likes
  1. Absolutely agree.
  1. Again, agree.

Retired software engineer here, with about 40 years experience, and published author of five computer books.

The OP (to me) has a long way to go to claim any “dev” experience. The first post (to me) makes no sense.

“Rich”?? “poverty”?
Wow - a final non-nonsensical reply that has ZERO to do with programming. Probably time to lock this thread :+1:

Are you talking about openSUSE or is this some general statement?

general

It is everywhere: smartphones, cars, cloud computing, servers, home appliances, tv, servers, desktops, super computers, real time computing (medical, automotive, finance, public), …

Aren’t there hundreds of variants for desktop already? Correct me, if I am wrong.

Furthermore, if the OS components/packages were “orthogonal” enough, a minimum system could be installed, just enough in order to boot and install packages just as needed in order to build a system for the respective application area (embedded, mobile, desktop, server/HPC).

1 Like

Technical discussions (even the off topic open chat ones) do not benefit from wildly varying claims about possible experience.

4 Likes

The minimal system consist of at least the kernel. But already the kernel build needs to differ between a desktop, server, super computer, embedded system. In other words, it is nearly impossible to provide a minimum system which suits all use cases without heavy disadvantages for 95% percent of the systems.

Linux systems can be and are heavily tailored and adapted for the special use case. The same applies for MS Windows systems, even if the average desktop user never heard of the special versions. But as Linux does not necessarely need to be sellt like a MS product, the amount of niche variants is much , much higher, due to the fact that many devs can and want to do it.

1 Like

Agreed. Discussions work best when egos and self-endorsements are left at the door.

2 Likes

@M43K8M yes, end of 1980… think I go the VIC-20 in 1982… Then with work was TRS-80’s and PET 64’s for testing telecommunications equipment…

1 Like

it is nearly impossible to provide a minimum system which suits all use cases without heavy disadvantages for 95% percent of the systems

What disadvantages would it have if e.g. server users would install a base system, configure the kernel for optimized network I/O performance, say, and install server related packages; workstation users would install a base system, configure the kernel for optimized graphics and tool chain performance, say, and install desktop packages. I thought this is what package systems are actually meant for.

G O A H E A D !!. openSUSE offers you all the tools, infra etc. to build that distro. Your users would finally be able to use all those redundant USB keys, since they will have 64GB isos to install from. Since it will be perfect, most likely openQA testing will not be needed, and to speed up downloading your team will of course invent that wonderful new compression algoritm. I’m also looking forward to finally have the choice between 24h-clock and the new decimal one, between vertical and horizontal time zones, between traditional and RAM- and kernel-free booting, the feature that allows usage of rpm, deb, pkg, dmg, apk all on one system, run AMD drivers on NVIDIA GPUs and v.v. , what would be the disadvantage of that aside from the > Pbyte repos.

2 Likes

What would be the disadvantage of a 1m³ box that can contain 6m³ of water?

1 Like