I agree with the spirit of this article, but it's completely unrealistic. I mean things like not installing telemetry or spyware are completely reasonable, but things like: Most companies don't...
I agree with the spirit of this article, but it's completely unrealistic. I mean things like not installing telemetry or spyware are completely reasonable, but things like:
should never update a user interface without allowing the previous one to be used as the default
Most companies don't have the resources to maintain 2 user interfaces, and users aren't willing to pay the cost to make it happen. So they can say they don't want developers to do that, but unless they can afford to support it, it's not realistic to tell companies not to do that. Believe me, if users would pay for it, companies would be lining up to do it!
should never cause a system to become unusable or unstable
Well yeah, that's the ideal, but again, you only have so much money to spend on testing. You can't test every possible combination of features on every conceivable configuration of hardware, so you have to choose.
Even better, it would be nice if more software allowed users to select the exact version they want to install. One thing I love about Minecraft is that all versions, including all alphas, weekly...
Even better, it would be nice if more software allowed users to select the exact version they want to install. One thing I love about Minecraft is that all versions, including all alphas, weekly snapshots, and buggy broken versions, can be installed from the launcher. You don't need to do much work to make that happen - just keep the installation executables on a server somewhere.
OK, that's fair. Most software these days seems to have an auto-update option that's opt-out, so by default it updates itself. That's also a dark pattern that needs to die.
OK, that's fair. Most software these days seems to have an auto-update option that's opt-out, so by default it updates itself. That's also a dark pattern that needs to die.
Yep, I'm with you on this one! I've seen a couple programs which prompt me to install a new version on launch and have the option to "install on quit". That's become my favorite so far. When I'm...
Yep, I'm with you on this one! I've seen a couple programs which prompt me to install a new version on launch and have the option to "install on quit". That's become my favorite so far. When I'm launching software I want to use it, not install updates and wait, but install on quit is a good balance
Unfortunately, there are few organizations devoted to preserving software largely as-is. Debian seems pretty good at this, though. A lot of Debian packages seem to be old Unix programs from bygone...
Unfortunately, there are few organizations devoted to preserving software largely as-is. Debian seems pretty good at this, though. A lot of Debian packages seem to be old Unix programs from bygone days, largely of historical interest. Game emulators are another example.
If you decide to freeze things largely as they are, it often means that few new users will be interested, so you're going to end up maintaining things for a dwindling community. This is not always bad, but look at how much people panic about this happening to Firefox.
If you stop doing major updates, people will start writing articles like "Is X dead" and there will be debates on Hacker News about whether you are dead or not. Within the community, there will be a lot of talk about what major changes are needed to gain more users.
This should sound familiar. Even on Tildes, we sometimes worry about this even though @Deimos has reassured us that it's okay not to worry about growth.
I completely agree with this, and in particular this: if you release a new version of a program with a fundamentally different UI, you have released a broken program. What software companies...
I completely agree with this, and in particular this: if you release a new version of a program with a fundamentally different UI, you have released a broken program.
What software companies typically fail to realize is that their software is not the end-all and be-all of what the user wants. The user has an entire system in which software is a part of. If you change how the software is operated, you break the system! And the more thoroughly you break the system, the more likely your users are to search for replacements that better fit into their existing systems.
I understand why you would want to make changes to the UI, but if you are going to do that the best strategy is to make sure those changes are incremental. Look at how Chrome and Firefox have both slowly changed the paradigm of what a web browser looks like; those changes didn't happen all at once; they grew organically over time. The most dramatic change - moving the menu bar into a dropdown - was done in a fairly logical manner, included a way to get the menu back, and also was the only major change for that update.
I mean isn't the article "Why Johnny's security bugs don't get patched" - I get the sentiment and if its a paid product I think as a customer you have the right to be angry. For the rest of it...
I mean isn't the article "Why Johnny's security bugs don't get patched" - I get the sentiment and if its a paid product I think as a customer you have the right to be angry. For the rest of it there is some realism needed: mainly that there is no way in hell you can backport all new features. Or keep doing security releases for every release available. Or keep two or more UI's (depending on how old Johnny's UI is) handled as is.
Changing things incrementally is relevant - but its not always possible. Sometimes during a release process you need to rip the band-aid off and go for it. You may lose users, but if you did it for good reasons you will gain new users so it doesn't matter that Johnny is there fuming over his frozen version of the old software.
As for the bloatware and spyware etc. If you have bought something that you know will contain or possibly install bloatware or spyware - then ... I mean... you CAN write a blogpost, or you can avoid it in the future? Seems like an untrustworthy set of soft-ware that you shouldn't use?
I agree with the spirit of this article, but it's completely unrealistic. I mean things like not installing telemetry or spyware are completely reasonable, but things like:
Most companies don't have the resources to maintain 2 user interfaces, and users aren't willing to pay the cost to make it happen. So they can say they don't want developers to do that, but unless they can afford to support it, it's not realistic to tell companies not to do that. Believe me, if users would pay for it, companies would be lining up to do it!
Well yeah, that's the ideal, but again, you only have so much money to spend on testing. You can't test every possible combination of features on every conceivable configuration of hardware, so you have to choose.
Kepp in mind that it's about automatic updates. You're free change the UI as long as the user can click an "update" button somewhere in the process
Even better, it would be nice if more software allowed users to select the exact version they want to install. One thing I love about Minecraft is that all versions, including all alphas, weekly snapshots, and buggy broken versions, can be installed from the launcher. You don't need to do much work to make that happen - just keep the installation executables on a server somewhere.
Yeah. For me I wouldn't even need the exact version, but maybe only the latest patch of every minor version
OK, that's fair. Most software these days seems to have an auto-update option that's opt-out, so by default it updates itself. That's also a dark pattern that needs to die.
Yep, I'm with you on this one! I've seen a couple programs which prompt me to install a new version on launch and have the option to "install on quit". That's become my favorite so far. When I'm launching software I want to use it, not install updates and wait, but install on quit is a good balance
Unfortunately, there are few organizations devoted to preserving software largely as-is. Debian seems pretty good at this, though. A lot of Debian packages seem to be old Unix programs from bygone days, largely of historical interest. Game emulators are another example.
If you decide to freeze things largely as they are, it often means that few new users will be interested, so you're going to end up maintaining things for a dwindling community. This is not always bad, but look at how much people panic about this happening to Firefox.
If you stop doing major updates, people will start writing articles like "Is X dead" and there will be debates on Hacker News about whether you are dead or not. Within the community, there will be a lot of talk about what major changes are needed to gain more users.
This should sound familiar. Even on Tildes, we sometimes worry about this even though @Deimos has reassured us that it's okay not to worry about growth.
I completely agree with this, and in particular this: if you release a new version of a program with a fundamentally different UI, you have released a broken program.
What software companies typically fail to realize is that their software is not the end-all and be-all of what the user wants. The user has an entire system in which software is a part of. If you change how the software is operated, you break the system! And the more thoroughly you break the system, the more likely your users are to search for replacements that better fit into their existing systems.
I understand why you would want to make changes to the UI, but if you are going to do that the best strategy is to make sure those changes are incremental. Look at how Chrome and Firefox have both slowly changed the paradigm of what a web browser looks like; those changes didn't happen all at once; they grew organically over time. The most dramatic change - moving the menu bar into a dropdown - was done in a fairly logical manner, included a way to get the menu back, and also was the only major change for that update.
relevant xkcd: https://xkcd.com/1172/
I mean isn't the article "Why Johnny's security bugs don't get patched" - I get the sentiment and if its a paid product I think as a customer you have the right to be angry. For the rest of it there is some realism needed: mainly that there is no way in hell you can backport all new features. Or keep doing security releases for every release available. Or keep two or more UI's (depending on how old Johnny's UI is) handled as is.
Changing things incrementally is relevant - but its not always possible. Sometimes during a release process you need to rip the band-aid off and go for it. You may lose users, but if you did it for good reasons you will gain new users so it doesn't matter that Johnny is there fuming over his frozen version of the old software.
As for the bloatware and spyware etc. If you have bought something that you know will contain or possibly install bloatware or spyware - then ... I mean... you CAN write a blogpost, or you can avoid it in the future? Seems like an untrustworthy set of soft-ware that you shouldn't use?