Software disenchantment
-
@Gąska said in Software disenchantment:
@Jaloopa the problem domain for software development software is software development. So the knowledge of problem domain consists of knowledge of software development. You have to have knowledge of software development if you want to develop software. There's no way around it.
The biggest leap from "normal" domain knowledge to making software to support the domain is the typical lack of ability to explain to someone how to do something. In this case, it's (more or less...the analogy isn't perfect) the computer to whom you're explaining.
But most people have a very difficult time thinking explicitly about the steps required to do some nontrivial task. I'm sure we all do that even though it's our jobs to do it all the time. I know I do. I imagine (and design) something and then when I go to implement it I discover all sorts of details that I didn't consider or maybe even know anything about.
-
@Gąska said in Software disenchantment:
the problem domain for software development software is software development
Who's talking about software development software? An accountant shouldn't need to know software development to make a tool to help them with their accountancy. An architect shouldn't have to know about software development to be able to make an architecture tool. Sure, the people developing the tools to develop these tools need to know about software development, but as much as is possible they should be trying to abstract away the specific details of software development and just expose what's needed to make useful tools
-
@Jaloopa I should have a pony.
-
@blakeyrat said in Software disenchantment:
This is where some idiot who doesn't understand the concept of an ideal comes into the thread waving their arms like Kermit the Frog greeting Kenny Rogers and says, "BUT THAT'S UNPOSSIBLE! THAT WILL NEVER HAPPEN!" and then we have to get into the long discussion about the definition of words like "should" and "ideally".
-
@Jaloopa What? Do you want a pony, too?
-
@boomzilla I had a pony as a child. The one time I sat on him he bucked me off. To be fair, I was the first person ever to sit on him.
Anyway, the point is that this utopia of development tools with no programming knowledge required and a pony for every lawn will never be realised, but that doesn't mean it's not a goal that should be worked towards
-
@Jaloopa said in Software disenchantment:
Anyway, the point is that this utopia of development tools with no programming knowledge required and a pony for every lawn will never be realised, but that doesn't mean it's not a goal that should be worked towards
I think that's a horrible goal. The goal of making better tools and so forth is good, but most people are retards and should not be making stuff like that. Sure, sure, quote me blakey's idiocy about ideals but then come back to the real world.
You're talking about people who can't even begin to describe their requirements and you think they should be building the end product? Ridiculous.
-
Just noticed the second link
This is why I stopped pleading to the developers and started pleading to the users. I want users to demand more and be angrier with the promise of computing.
No, fuck you, I'm already angry with enough things. I don't want to be angry with things that aren't bothering me.
-
@Jaloopa said in Software disenchantment:
@Gąska said in Software disenchantment:
the problem domain for software development software is software development
Who's talking about software development software?
@blakeyrat. And because of that, me. And since you've joined our discussion, I assumed that you too.
-
@Gąska it was dkf and boomzilla that started with that.
Now blakeyrat is probably angry with you for the useless notification. He isn't a member of @by-joining-this-group-you-agree-to-be-mentioned-randomly-for-no-reason-is-that-okay-yes-no
-
@Jaloopa said in Software disenchantment:
An accountant shouldn't need to know software development to make a tool to help them with their accountancy. An architect shouldn't have to know about software development to be able to make an architecture tool. Sure, the people developing the tools to develop these tools need to know about software development, but as much as is possible they should be trying to abstract away the specific details of software development and just expose what's needed to make useful tools
And just what exactly do you mean by "architecture tool"? Should we make tools that let an average architect with no knowledge nor experience of software development be able to reproduce the entirety of AutoCAD suite from scratch, with just a handful of libraries and one or two frameworks? Or do you mean writing their own macros for repetitive parts of dealing with the already existing AutoCAD, which is orders of magnitude smaller in scope than "making tools"?
-
@Gąska said in Software disenchantment:
Oh, great, semantic nitpicking. So let me ask you differently: do you believe that the defining characteristic of "perfect development software" is enabling people without any knowledge about anything whatsoever to create working software?
No, but a elementary school level of knowledge should be sufficient. They have to at least be literate, I would say. Since society's decided everybody should be literate already, that shouldn't be much of a bar to jump.
@Gąska said in Software disenchantment:
If so, do you believe it should be goal of every programming tools developer to strive for "perfect development software" as defined above?
I'd just be happy at the moment if they even spent 1% of their time on UX and usability testing.
I don't know if it should be the goal of "every" tools developer. We just need a couple competing tools that meet the requirement. There's no reason the more complicated tools couldn't also exist, except they'd be at a competitive disadvantage.
@Jaloopa said in Software disenchantment:
You need knowledge of the problem domain you're writing software to solve. Any required knowledge of "programming" on top of that is extra cognitive load that stops people from being able to just solve the problem at hand
Well said. That's pretty much exactly what I believe.
@Gąska said in Software disenchantment:
the problem domain for software development software is software development.
Bullshit. The software's written to solve some kind of problem.
I mean some people do it for "fun" and more power to them, but no software that is given to the general public to actually use should be from developers who wrote it for "fun". That's how you get Git.
@Gąska said in Software disenchantment:
You have to have knowledge of software development if you want to develop software.
Right; but you shouldn't. Or rather: any knowledge of software development you need should be provided by the tool itself. HyperCard came DAMNED close to this ideal.
@Gąska said in Software disenchantment:
I'm all for easy to use software.
Doesn't sound like you are.
@Gąska said in Software disenchantment:
But making software development easy can only get us so far.
Far towards which direction?
@Gąska said in Software disenchantment:
Because most problem domains are orders of magnitude simpler than software development.
That's only true (if it is true, and I'm not sure it is-- I've done insurance billing before!) because nobody's put much effort into making it no longer true. Some people did in the past, but now nobody does.
@boomzilla said in Software disenchantment:
The biggest leap from "normal" domain knowledge to making software to support the domain is the typical lack of ability to explain to someone how to do something. In this case, it's (more or less...the analogy isn't perfect) the computer to whom you're explaining.
Quite possible, but surely a software tool could be created to assist in this?
I mean, Apple back in the OS 7 days had a macro recorder that could record every action the user tool in every application they took it in across the entire OS. It can be done. It has been done. It's not much of a leap from recording that macro to saying "ok now fill in X from this spreadsheet and run it until you run out of rows".
@boomzilla said in Software disenchantment:
@Jaloopa I should have a pony.
See now we get into "people come into the thread who don't understand the concept of an 'ideal' and start shitting all over everything".
-
@Gąska said in Software disenchantment:
Should we make tools that let an average architect with no knowledge nor experience of software development be able to reproduce the entirety of AutoCAD suite from scratch,
Sure why not?
@Gąska said in Software disenchantment:
Or do you mean writing their own macros for repetitive parts of dealing with the already existing AutoCAD, which is orders of magnitude smaller in scope than "making tools"?
- Macros are still tools, your definition of "software" is extremely narrow if you don't agree with that, and
- Sure why not?
-
@sockpuppet7 said in Software disenchantment:
Now blakeyrat is probably angry with you for the useless notification. He isn't a member of @by-joining-this-group-you-agree-to-be-mentioned-randomly-for-no-reason-is-that-okay-yes-no
I wanted to respond but then I've noticed you only said that so you can have excuse to mention @by-joining-this-group-you-agree-to-be-mentioned-randomly-for-no-reason-is-that-okay-yes-no.
-
@blakeyrat said in Software disenchantment:
@Gąska said in Software disenchantment:
Because most problem domains are orders of magnitude simpler than software development.
That's only true (if it is true, and I'm not sure it is-- I've done insurance billing before!) because nobody's put much effort into making it no longer true. Some people did in the past, but now nobody does.
It is to the extent that developing software is really open ended problem solving. If you can cut the domain of problems down enough to something relatively simple then you end up with, say, Ruby On Rails. As soon as you start to wander from its closely prescribed boundaries you're in big trouble.
@boomzilla said in Software disenchantment:
The biggest leap from "normal" domain knowledge to making software to support the domain is the typical lack of ability to explain to someone how to do something. In this case, it's (more or less...the analogy isn't perfect) the computer to whom you're explaining.
Quite possible, but surely a software tool could be created to assist in this?
I honestly think that's going to require strong AI outside of (as above) very small and well defined domains.
I mean, Apple back in the OS 7 days had a macro recorder that could record every action the user tool in every application they took it in across the entire OS. It can be done. It has been done. It's not much of a leap from recording that macro to saying "ok now fill in X from this spreadsheet and run it until you run out of rows".
Yes. That definitely fits my idea of what I was talking about above. And I'm sure a lot of us around here have used similar things in Excel and then graduated to actually editing the generated VBA and eventually writing VBA from scratch. But there's a whole lot you can do once you get down to "raw" VBA vs a macro recorder and it's a lot more difficult to get it to do what seem like relatively simple tasks when a person is driving the action. Compare "go to the end of the data" with how you might have to detect that in software, considering things like blank columns or even entire rows. All those stupid edge cases that lead to software bugs down the road.
@boomzilla said in Software disenchantment:
@Jaloopa I should have a pony.
See now we get into "people come into the thread who don't understand the concept of an 'ideal' and start shitting all over everything".
No, we're just making fun of people with dumb ideals.
-
@Jaloopa said in Software disenchantment:
ri out
-
@Jaloopa said in Software disenchantment:
@boomzilla I had a pony as a child. The one time I sat on him he bucked me off. To be fair, I was the first person ever to sit on him.
Anyway, the point is that this utopia of development tools with no programming knowledge required and a pony for every lawn will never be realised, but that doesn't mean it's not a goal that should be worked towards
That’s how we get stuff like:
-
@boomzilla said in Software disenchantment:
It is to the extent that developing software is really open ended problem solving.
Ok.
But any problem can be broken down into steps, yes? Who's to say there's a monopoly, that only software developers are capable of that.
@boomzilla said in Software disenchantment:
I honestly think that's going to require strong AI outside of (as above) very small and well defined domains.
I highly doubt it.
Every business has that one person who knows how everything goes together. Usually it's a middle-aged woman who's worked there for decades. I'm stereotyping but you get the point.
If you sit down with her, she can explain it all, and she can tell you the times when something happened that nobody accounted for and she can tell you exactly how to write the logic of your program.
That person always exists.
@boomzilla said in Software disenchantment:
Yes. That definitely fits my idea of what I was talking about above. And I'm sure a lot of us around here have used similar things in Excel and then graduated to actually editing the generated VBA and eventually writing VBA from scratch. But there's a whole lot you can do once you get down to "raw" VBA vs a macro recorder and it's a lot more difficult to get it to do what seem like relatively simple tasks when a person is driving the action. Compare "go to the end of the data" with how you might have to detect that in software, considering things like blank columns or even entire rows. All those stupid edge cases that lead to software bugs down the road.
But Excel isn't designed with any of this in mind, and it's also ossified so it barely improves year over year. (Access shares the same problems. Access is closer to what I'm talking about though than Excel is.)
@boomzilla said in Software disenchantment:
No, we're just making fun of people with dumb ideals.
Is it less or more dumb than "the US should nuke Cuba"? Help me calibrate here.
-
I don't work as a programmer, so I can choose whatever technologies and language I want to use, and work whenever I like, as slow as I like.
While I would say my skills are pretty good when writing code, I feel that recently the whole process has become ridiculously complicated. For example, I've been extending Blockly for a while now and wanted to try to do the same with Scratch-blocks. Just reading the quick start page hurts my head
Home
Scratch Blocks is a library for building creative computing interfaces. - scratchfoundation/scratch-blocks
and I still haven't got it working, but I have a Google Closure zip file that can't be deleted. Even when I manage to get it to work I have no confidence I will be able to keep it working after an upgrade.
Maybe I was spoilt by Delphi and Visual Studio, where after a couple of clicks I would be actually creating something.
Microsoft's PowerApps seems to be going in the right direction but I haven't used it yet
-
@coldandtired Back in the day we had HyperCard, Access, Filemaker, RealBasic, Runtime Revolution... those are the ones off the top of my head. Even the more "complicated" languages like C++ had Borland Builder and C# had WinForms.
Now almost all those things are dead or dying.
PowerApps looks interesting, I might take a gander at that.
-
@blakeyrat said in Software disenchantment:
@Gąska said in Software disenchantment:
Oh, great, semantic nitpicking. So let me ask you differently: do you believe that the defining characteristic of "perfect development software" is enabling people without any knowledge about anything whatsoever to create working software?
No, but a elementary school level of knowledge should be sufficient. They have to at least be literate, I would say. Since society's decided everybody should be literate already, that shouldn't be much of a bar to jump.
Let me ask even more precise question because you do everything in your power to not actually address the point of my questions: do you believe that the defining characteristic of "perfect development software" is enabling people with knowledge equivalent to American 8th grade graduate and no additional knowledge about anything whatsoever to create working software?
@Gąska said in Software disenchantment:
If so, do you believe it should be goal of every programming tools developer to strive for "perfect development software" as defined above?
I'd just be happy at the moment if they even spent 1% of their time on UX and usability testing.
Does it mean we SHOULDN'T strive for "perfect development software" as defined above? God, getting you to actually answer the questions I'm asking is so hard.
I don't know if it should be the goal of "every" tools developer. We just need a couple competing tools that meet the requirement.
There's no reason the more complicated tools couldn't also exist, except they'd be at a competitive disadvantage.
Only if the only thing that matters was ease of use by people who have no fucking clue what they're doing. But for people who actually know what they're doing, it's far more important for tools to be easy to use by people who know what they're doing. And these two goals are often mutually exclusive.
@Gąska said in Software disenchantment:
the problem domain for software development software is software development.
Bullshit. The software's written to solve some kind of problem.
And software development software is written to solve the problem of developing software.
I mean some people do it for "fun" and more power to them, but no software that is given to the general public to actually use should be from developers who wrote it for "fun".
There are also people whose actual job is to develop software that helps with software developers. Visual Studio division at Microsoft, for example.
@Gąska said in Software disenchantment:
You have to have knowledge of software development if you want to develop software.
Right; but you shouldn't. Or rather: any knowledge of software development you need should be provided by the tool itself.
You've just said something completely different from the thing you said in your earlier post.
HyperCard came DAMNED close to this ideal.
HyperCard was only about writing simple scripts that wired up to buttons of existing programs. It's a completely different scope than general software development. It's a tiny, tiny fraction of general software development. So tiny that it's actually possible to be made easy, and I'm in full support of developing software that makes it that easy (or even easier). But that's completely different thing than general software development.
@Gąska said in Software disenchantment:
I'm all for easy to use software.
Doesn't sound like you are.
How so?
@Gąska said in Software disenchantment:
But making software development easy can only get us so far.
Far towards which direction?
Towards making software easy to use for people whose job is to use computers to solve actual real life problems, not just problems with computers. For problems with computers, we have software developers, hardware designers, helpdesk, and all the other branches of IT. I agree that average people shouldn't have to know anything at all about IT to solve their actual real life problems using computers. But making them write software isn't going to achieve that - quite the opposite.
@Gąska said in Software disenchantment:
Because most problem domains are orders of magnitude simpler than software development.
That's only true (if it is true, and I'm not sure it is-- I've done insurance billing before!) because nobody's put much effort into making it no longer true.
Do you seriously believe software development in general can ever become as simple as, say, retail cashier's work? Because that's what I meant when I said most problem domains are simpler than software development.
@boomzilla said in Software disenchantment:
The biggest leap from "normal" domain knowledge to making software to support the domain is the typical lack of ability to explain to someone how to do something. In this case, it's (more or less...the analogy isn't perfect) the computer to whom you're explaining.
Quite possible, but surely a software tool could be created to assist in this?
Purpose-built software for one specific discipline? Sure, we do this all the time. General software to solve all such problems once and for all? You must be kidding me.
-
@blakeyrat said in Software disenchantment:
@Gąska said in Software disenchantment:
Should we make tools that let an average architect with no knowledge nor experience of software development be able to reproduce the entirety of AutoCAD suite from scratch,
Sure why not?
Do you know what AutoCAD is? Do you realize how much work it requires to create?
@Gąska said in Software disenchantment:
Or do you mean writing their own macros for repetitive parts of dealing with the already existing AutoCAD, which is orders of magnitude smaller in scope than "making tools"?
- Macros are still tools, your definition of "software" is extremely narrow if you don't agree with that
They are. But when you say "tools", it's only natural you mean all kinds of tools, or at least the majority of tools. And majority of software is huge application suites, not macros.
-
@blakeyrat said in Software disenchantment:
@Gąska Yes, with the proviso that you're defining "should" as "should" and not something like "will be".
I also believe there SHOULD be no wars in the world, that nobody SHOULD go without healthcare, etc.
This is where some idiot who doesn't understand the concept of an ideal comes into the thread waving their arms like Kermit the Frog greeting Kenny Rogers and says, "BUT THAT'S UNPOSSIBLE! THAT WILL NEVER HAPPEN!" and then we have to get into the long discussion about the definition of words like "should" and "ideally".
But you never have the ability to recognize that other people are talking about current issues and not ideals.
-
@Gąska said in Software disenchantment:
Let me ask even more precise question because you do everything in your power to not actually address the point of my questions
I'm not being difficult on purpose, I just don't agree with a lot of the premises of your questions. For example:
@Gąska said in Software disenchantment:
do you believe that the defining characteristic of "perfect development software" is enabling people with knowledge equivalent to American 8th grade graduate and no additional knowledge about anything whatsoever to create working software?
In a perfect world there wouldn't be any such thing as "development software" because you'd just ask the computer for what you want and it'd do it. Like in Star Trek TNG shows.
But generally yes, any additional knowledge you need to work your development tool should be taught by the tool itself. And of course it should do as much as possible automatically so there's no boilerplate or anything.
@Gąska said in Software disenchantment:
Does it mean we SHOULDN'T strive for "perfect development software" as defined above? God, getting you to actually answer the questions I'm asking is so hard.
No? I'm not sure where you got that from.
- Doing UX testing = good
- Writing development software anybody can use = better
It's not hard.
@Gąska said in Software disenchantment:
Only if the only thing that matters was ease of use by people who have no fucking clue what they're doing. But for people who actually know what they're doing, it's far more important for tools to be easy to use by people who know what they're doing. And these two goals are often mutually exclusive.
They aren't mutually exclusive at all.
@Gąska said in Software disenchantment:
And software development software is written to solve the problem of developing software.
"developing software" Isn't a thing though. It's a means, not an end.
@Gąska said in Software disenchantment:
There are also people whose actual job is to develop software that helps with software developers. Visual Studio division at Microsoft, for example.
Yeah; they're doing a shitty job.
@Gąska said in Software disenchantment:
HyperCard was only about writing simple scripts that wired up to buttons of existing programs.
You never used it obviously.
@Gąska said in Software disenchantment:
It's a completely different scope than general software development.
Untrue.
@Gąska said in Software disenchantment:
But making them write software isn't going to achieve that - quite the opposite.
Maybe; but what we're doing now certainly isn't helping, is it?
@Gąska said in Software disenchantment:
Do you seriously believe software development in general can ever become as simple as, say, retail cashier's work? Because that's what I meant when I said most problem domains are simpler than software development.
"most problem domains" == "making change". Ok, good to know? Glad you clarified?
@Gąska said in Software disenchantment:
Purpose-built software for one specific discipline? Sure, we do this all the time. General software to solve all such problems once and for all? You must be kidding me.
Saying it's the ideal state doesn't imply that it's possible to achieve. That's the thing people get all in a huff about whenever we have one of these threads.
Guess what? Ideally-- in an ideal society-- there's no murder ever! That's impossible to achieve though. But it's still an ideal. See how that works?
That's why these discussions are so stupid.
Me: "Ideally X"
You: "But X will never happen!!!!"
Me: "Well ok. But still ideally X"It's retarded. I'd prefer that the people who don't understand the concept of an "ideal" simply wouldn't reply.
-
@blakeyrat said in Software disenchantment:
@boomzilla said in Software disenchantment:
It is to the extent that developing software is really open ended problem solving.
Ok.
But any problem can be broken down into steps, yes? Who's to say there's a monopoly, that only software developers are capable of that.
No one. But who thinks that everyone can do that?
@boomzilla said in Software disenchantment:
I honestly think that's going to require strong AI outside of (as above) very small and well defined domains.
I highly doubt it.
Every business has that one person who knows how everything goes together. Usually it's a middle-aged woman who's worked there for decades. I'm stereotyping but you get the point.
If you sit down with her, she can explain it all, and she can tell you the times when something happened that nobody accounted for and she can tell you exactly how to write the logic of your program.
That person always exists.
Yes and I've worked with those people but they're still terrible at getting all that out of their head and into the open. Now get them to describe what sort of data[base] structure they need to support their stuff.
@boomzilla said in Software disenchantment:
No, we're just making fun of people with dumb ideals.
Is it less or more dumb than "the US should nuke Cuba"? Help me calibrate here.
More dumb. Saying that we should make tools the make it easier to develop software is very different than trying to make a world where anyone can build a software system. Either you have strong AI (which may not even be possible) or you end up with something super restricted that makes everyone mad when you have to color outside the lines even a little bit.
In any case, I don't support nuking Cuba. Maybe targeted assassinations and a popular revolution.
-
@Gąska said in Software disenchantment:
Do you know what AutoCAD is?
I know what it does generally.
@Gąska said in Software disenchantment:
Do you realize how much work it requires to create?
Nobody here brought up "amount of work" until just now.
Even if it took him his entire lifetime, the answer is still "sure why not?"
@Gąska said in Software disenchantment:
They are. But when you say "tools", it's only natural you mean all kinds of tools, or at least the majority of tools. And majority of software is huge application suites, not macros.
... ok? Not sure what point you're trying to get at here.
-
Is there even any other domain where end users being able to produce their own solutions is seen as desirable? If I want a chair, I don't expect to be able to create one starting with no knowledge of woodworking to reasonable final product in a matter of months. I mean, IKEA sells stuff that only requires assembly (with no options for customizing it), and people still mess those up. And I don't expect the carpenter to fabricate his own hammer and nails, he goes and buys some from a hardware store. He's an expert at using the tools, despite not being able to create the tools.
Software should be the same, it's a tool. It should make tasks easier for an end user, without the end user being involved in creating the tool itself (beyond knowing how to explain his requirements to the tool-making expert). In the example of the carpenter, if some carpenter determined that he needed a special kind of hammer not available at the hardware store, he'd go to an ironsmith and explain the kind of hammer he wanted. The ironsmith might not know a lot about carpentry, but given a description of the shape and use of the new hammer, would know what techniques to use to make the hammer. He would then provide the hammer the carpenter needed.
We wouldn't see the ideal of ironsmithing being to have an automatic machine that given instructions by a layman can create any kind of tool. You could create some kind of 3d printer that could approach that for very limited domains, but the same machine probably wouldn't be able to create both a hammer and a battleship, despite both being made of metal.
-
@Kian said in Software disenchantment:
If I want a chair, I don't expect to be able to create one starting with no knowledge of woodworking to reasonable final product in a matter of months.
First of all, it doesn't take months to learn enough woodworking to make a quality chair, most people did it during one semester of middle school.
Secondly, the point isn't whether you can do that now, the point is whether people in the woodworking industry believe that ideally everybody interested ought to be able to do it and I wager if you interviewed them, they'd all say "yeah! That'd be great!"
-
@blakeyrat said in Software disenchantment:
In a perfect world there wouldn't be any such thing as "development software" because you'd just ask the computer for what you want and it'd do it. Like in Star Trek TNG shows.
Oh, I get it now. Perfect software is the kind of software that won't only never exist, but it's also literally impossible for anything even remotely close to it to ever exist?
In other words - we can safely ignore all your posts containing the word "perfect" since they don't say anything about the universe we live in?
-
@blakeyrat said in Software disenchantment:
@Gąska said in Software disenchantment:
Do you realize how much work it requires to create?
Nobody here brought up "amount of work" until just now.
Work and software development expertise. You need a lot of software development expertise too.
-
@blakeyrat said in Software disenchantment:
@Kian said in Software disenchantment:
If I want a chair, I don't expect to be able to create one starting with no knowledge of woodworking to reasonable final product in a matter of months.
First of all, it doesn't take months to learn enough woodworking to make a quality chair, most people did it during one semester of middle school.
That's literally months of learning.
-
@blakeyrat said in Software disenchantment:
@Kian said in Software disenchantment:
If I want a chair, I don't expect to be able to create one starting with no knowledge of woodworking to reasonable final product in a matter of months.
First of all, it doesn't take months to learn enough woodworking to make a quality chair, most people did it during one semester of middle school.
Really? I remember people struggling to make birdhouses that didn't have gaps all over the place.
Secondly, the point isn't whether you can do that now, the point is whether people in the woodworking industry believe that ideally everybody interested ought to be able to do it and I wager if you interviewed them, they'd all say "yeah! That'd be great!"
This paragraph is fucking outrageous. Those people in the woodworking industry would expect those people to work and learn the specifics of woodworking. And no one here is saying that we should keep people out of software development. Just that it's something that requires hard work and has a body of knowledge all its own.
-
@Gąska said in Software disenchantment:
And just what exactly do you mean by "architecture tool"?
I don't know, I'm not an architect.
This gets towards the assertion that most people are stupid. They're not, they just have expertise in a different field to you. I've had to modify accounting modules in software before, it involved talking to the accountant, having him explain what he wanted compared to what it already did, and a quick primer in things like double entry bookkeeping. I probably looked like an idiot to him because I didn't understand the business domain, but I had to use my incomplete knowledge to try to fix the issue. Imagine if the tool was created in such a way that knowledge of accounting was the main requirement, the actual expert could have made the changes, probably with fewer dead ends and rounds of testing because he understood the requirements and logic underpinning it. All I knew was the language the software was written in and the brief introduction I'd had in the subject matter
-
@Gąska Look you guys are just engaging me on this so you can use it as an excuse to call me an idiot over and over again. Why not just call me an idiot and skip the pretense.
You obviously don't give a shit what I think and aren't interested in discussing it.
-
@blakeyrat said in Software disenchantment:
Look you guys are just engaging me on this so you can use it as an excuse to call me an idiot over and over again.
No we aren't but you're certainly providing opportunities for that and trying to be a martyr yet again just makes it worse.
-
@blakeyrat said in Software disenchantment:
@Gąska Look you guys are just engaging me on this so you can use it as an excuse to call me an idiot over and over again.
No, I'm doing this so you will finally stop saying "everyone should be able to make software without first learning about making software" when you actually mean "modern programming tools are horrible and I wish we as an industry spent less time wanking over technology and more time thinking of the end user".
-
@Gąska said in Software disenchantment:
No, I'm doing this so you will finally stop saying "everyone should be able to make software without first learning about making software" when you actually mean "modern programming tools are horrible and I wish we as an industry spent less time wanking over technology and more time thinking of the end user".
Those two things aren't mutually exclusive and I simultaneously believe both of them.
"What? Simultaneously believing two not-mutually-exclusive things?! Blakeyrat must be an idiot! Quick post and let him know!"
-
@blakeyrat said in Software disenchantment:
most people did it during one semester of middle school.
If that were true, most people would use chairs made by their children (or themselves) in their homes. You can learn enough to maybe become a hobbist, but not enough to prefer it over paying a professional to make something good for you.
@blakeyrat said in Software disenchantment:
believe that ideally everybody interested ought to be able to do it
Ok, that is a constraint I missed in your earlier comments. But "interested" is hard to define. One could argue that anyone interested enough can already learn software development and produce the software they want for themselves. Granted, not in a couple months with most tools. A big problem with teaching people is figuring out what degree of motivation they have to learn. I won't deny there's a lot that could be done to improve the current situation, I see a lot of ways in which I would like my tools to be better, where there's no reason that they couldn't be better except for the developers not having cared about it when they made them.
The problem is, most people don't care about most things. If they cared about development, they would probably be developers already. Unlike most other fields, the cost of entry is very low, the information is available for free if you have an internet connection, the cost of "materials" is basically the electricity already flowing into your house, and there are good job opportunities in most places for entry level positions even if you don't have a degree. What's left are the people that don't care about creating programs, and there's no way to make a tool that will feed them the knowledge if they don't care to learn it. It doesn't matter how easy you make the path if they don't want to get on it.
@blakeyrat said in Software disenchantment:
I wager if you interviewed them, they'd all say "yeah! That'd be great!"
I don't know, wouldn't that put them out of a job?
-
@blakeyrat said in Software disenchantment:
Is it less or more dumb than "the US should nuke Cuba"? Help me calibrate here.
Cuba should nuke North Korea.
-
@blakeyrat said in Software disenchantment:
@Gąska Look you guys are just engaging me on this so you can use it as an excuse to call me an idiot over and over again. Why not just call me an idiot and skip the pretense.
You obviously don't give a shit what I think and aren't interested in discussing it.
'the sky is green'
'no it's blue'
'look we all realize you think I'm an idiot'
welcome to blakeyland
-
@blakeyrat said in Software disenchantment:
@Gąska said in Software disenchantment:
No, I'm doing this so you will finally stop saying "everyone should be able to make software without first learning about making software" when you actually mean "modern programming tools are horrible and I wish we as an industry spent less time wanking over technology and more time thinking of the end user".
Those two things aren't mutually exclusive and I simultaneously believe both of them.
And I believe only one of them. And when I argue against the thing I don't believe in, you think I'm arguing about the one I believe in, which is the reason why you think I'm against easy to use software, and for all the other confusion in this topic.
Even if A and B are both true. Stop saying A when you actually mean B.
-
@pie_flavor He isn't saying anything remotely that strange. He is, as he always has been, advocating that we try to make things less bad, and make things easier for end-users, of which we are a part.
He wants everything to just be easy, but doesn't expect that to just happen. He wants us to try. That's what all the stuff he's been saying about ideals is about.
He's not @Gribnit . Pretty much anyone can understand what he's saying.
-
@Magus said in Software disenchantment:
He's not @Gribnit . Pretty much anyone can understand what he's saying.
Well except the people who have to pretend they don't so they can justify calling me an idiot.
-
@Jaloopa said in Software disenchantment:
I had a pony as a child
Finally coming out as a brony huh?
-
@Luhmann his name was scallywag and he never liked being ridden by men. Eventually he was OK with women. I think the farmers who owned the land near us ended up with him
-
@Kian said in Software disenchantment:
The problem is, most people don't care about most things. If they cared about development, they would probably be developers already.
I'm not convinced that all or even most people are even really capable of it.
-
@Magus said in Software disenchantment:
@pie_flavor He isn't saying anything remotely that strange. He is, as he always has been, advocating that we try to make things less bad, and make things easier for end-users, of which we are a part.
He wants everything to just be easy, but doesn't expect that to just happen. He wants us to try. That's what all the stuff he's been saying about ideals is about.
He's not @Gribnit . Pretty much anyone can understand what he's saying.
@pie_flavor isn't saying that no one understands what he's saying. Yes, he says he wants difficult things to be easy. Duh. Who doesn't? That doesn't make him insightful.
-
@Magus said in Software disenchantment:
@pie_flavor He isn't saying anything remotely that strange. He is, as he always has been, advocating that we try to make things less bad, and make things easier for end-users, of which we are a part.
He wants everything to just be easy, but doesn't expect that to just happen. He wants us to try. That's what all the stuff he's been saying about ideals is about.
This is only part of what he's saying. And we all agree with that part.But there's also the other part where he says software development is only hard because the tools we made are hard, and that if we made tools easy to use by average users, then programming would be as easy as A-B-C and virtually everybody would be capable of making software that solves arbitrary problems, and that the only reason programming tools are hard is because no one ever tried to make them easier, except for Apple in the 90s but they abandoned that idea completely and joined the "programming must be hard" elitist circlejerk. This is what I have a problem with.
He's not @Gribnit . Pretty much anyone can understand what he's saying.
Which is why it's so much more annoying.
-
@blakeyrat said in Software disenchantment:
PowerApps looks interesting, I might take a gander at that.
What happened to microsoft lightswitch? I didn't even get to look how it worked.
-
@dkf said in Software disenchantment:
embedded C [OK, there's a profile of C++ that works, but the hardware is too limited even to use the C++ standard library so the benefits of C++ are very slight]
I disagree with that. Without touching the STL at all, you can still write powerful abstractions that compile down to the same machine code as more low level C stuff. You'll not get anywhere near something as nice as C#, but still better than the bug prone shit that is C.
See this for a toy example. At this level it might look overcomplicated compared to directly writing the low level code, but the abstractions will help in the long term.
LabVIEW - Wikipedia
