The right to fork makes open source software a breeding ground for innovation!?



  • So why the fuck ain't we seeing any actual innovation from open source developers?

    I hope his doctorate had like an asterisk next to it or something.


  • Winner of the 2016 Presidential Election

    <°)))o><

    🐟🐠🎣


  • BINNED

    I fixed your gibberish title



  • My title was indeed gibberish.



  • Open source software is everywhere. From phones, tablets, TVs, and game consoles to less self-evident examples like cars, washing machines, and the International Space Station. However, what makes open source software remarkable is not where it can be found, but rather what can be done with it. One of the most astounding rights guaranteed by all open source software licenses is the right to fork the source code. In other words, the right to copy any program, either in part or in its entirety, and use that program to create a new, modified version of it.

    The right to fork has an enormous impact on both the development and governance of open source software. Despite its significance, code forking has seen little academic study. This dissertation examines the right to fork, its impact and significance, and how it is viewed and practiced by developers.

    The study draws on data consisting of hundreds of forks, interviews with open source software programmers, and an in-depth analysis of the birth of the MariaDB fork. This dissertation is relevant to anyone seeking a greater general understanding of how open source works and why it is considered a superior software development model. It may also serve as a useful resource for firms seeking to harness the power of open source software. Furthermore, it offers important insights to those who want to better understand how code forking is practiced and viewed by developers.

    This study finds that forks are primarily started for non-competitive reasons, with unique features or goals that distinguish them from their parent projects. Competitive forks are rare but do exist, with some motivating factors being to ensure the freedom of the code and the community’s ability to contribute to it. Furthermore, though developers may not always agree with the forking of a project, they nonetheless consider the right to fork to be of vital importance, and a cornerstone of free and open source software.

    In many ways, open source can be thought of as a return to how software was developed before the emergence of proprietary licensing. The same freedoms of development and sharing that thrived back then can be found today in the open source community. Indeed, in many ways the right to fork is synonymous with freedom: the freedom to explore and experiment, the freedom to benefit from the work done by others, and the freedom to keep any project relevant and vibrant even when faced with leadership decisions that are deemed unsupportable. In short, the right to fork is open source software’s guardian of freedom and watchdog of meritocracy.

    So mariadb is their main study case. That was more like "let's save this thing from fucking Oracle".

    On the other hand, defensive forks like that can reinvigorate projects. See node.js vs io.js.



  • What "innovation" has MariaDB introduced? I mean, maybe there is something-- I honestly don't know. (But I doubt it.)



  • @blakeyrat said:

    So why the fuck ain't we seeing any actual innovation from open source developers?

    Because we you define innovation as whatever open source isn't doing.



  • So do you have an example of something open source has done recently that I might consider innovative? I mean you can at least try here.



  • Not off the top of my head. But I'm no mind reader. What sort of things do you consider to be innovative?



  • OTOH, I'd add package managers. Even MS seems to be catching on to that one.



  • @blakeyrat said:

    So do you have an example of something open source has done recently that I might consider innovative? I mean you can at least try here.

    Linux kernel version 4 is getting the functionality to be be patched while running.
    On your oh-so-loved windows, every fu***n insignificant piece that gets updated forces you to reboot.


  • BINNED

    @TimeBandit said:

    On your oh-so-loved windows, every fu***n insignificant piece that gets updated forces you to reboot.

    I'm posting this ahead of time. If it turns I was wrong, no one will be happier than me:



  • @blakeyrat said:

    So do you have an example of something open source has done recently that I might consider innovative? I mean you can at least try here.

    Docker? While they didn't invent containers, they're the runaway leader in the field of container automation. Obviously you're gonna move the goalposts on me, but a normal person would say Docker contributes new and useful technology.



  • @blakeyrat said:

    What "innovation" has MariaDB introduced? I mean, maybe there is something-- I honestly don't know. (But I doubt it.)

    Community improvements from Facebook, Twitter, Google, and Percona all roll into MariaDB sooner than they appear in MySQL.

    MariaDB also provides user statistics and better instrumentation through the data dictionary information schema, including microsecond support. If you're using TIME or DATETIME data types, you can specify precision, such as TIME(4), where the number represents the number of digits after the decimal place. MariaDB supports up to six digits -- for example, 0.000001 second or one microsecond. No more wondering how fast is fast. For those cases where precision is not specified, MariaDB defaults to 0 for easy backward compatibility.

    The query optimizer in MariaDB offers serious improvements, including better subquery optimization, as well as faster, more efficient, and more consistent joins, derived tables, and views. In addition, MariaDB gives you additional control over how the optimizer makes decisions, exposing more internal instrumentation and configuration as server variables you can set.

    MariaDB gives you two new high-performance storage engines to choose from: Aria and XtraDB.
    You also get access to a whole new clustering technology called Galera. It allows active-active multimaster updating. Here you can really start to scale writes on cloud servers. You also get access to parallel and synchronous replication features.

    Want to get NoSQL speed? Consider the HandlerSocket plug-in, which enables direct access to storage engines without going through the optimizer, boosting velocity by 10 times or more. You can also now get a row of data returned in JSON format using dynamic columns in MariaDB -- not so in MySQL.

    MariaDB includes yet another new storage engine, Cassandra SE, that allows you to read or write data into a Cassandra data store. Finally, integration between SQL and NoSQL made easy!

    Want to consolidate data from multiple master databases? Multisource replication is exactly what you're looking for. Assuming your source data is stored in multiple schemas, they can all be brought together on one instance downstream -- again, not possible in MySQL.

    Source: InfoWorld

    That is just some examples of the improvements MariaDB brings over MySQL.

    But hey, of course it's shit, since it is open source :wink:


  • BINNED

    While I'm not dissing the explanation, the way this was written makes me think of this:



  • @boomzilla said:

    OTOH, I'd add package managers.

    I grant that one, but that was ages ago and it's unclear what the "right to fork" had to do with it. It's more like, "our OS was so shitty at software distribution that we hacked this work-around which just happened to be useful in other ways as well".

    @TimeBandit said:

    Linux kernel version 4 is getting the functionality to be be patched while running.

    I highly doubt it will work. Or it'll work with the same caveats we have now, which is: "the security update is only 100% secure when you close all applications and restart all services and logout all users" which is functionally identical to rebooting anyway.

    @TimeBandit said:

    On your oh-so-loved windows, every fu***n insignificant piece that gets updated forces you to reboot.

    Until you update drivers, Windows can do a shitload more with drivers than Linux can.

    @NedFodder said:

    Docker? While they didn't invent containers,

    So far this is the best example, and you weasel-word it in the second sentence.

    @TimeBandit said:

    Community improvements from Facebook, Twitter, Google, and Percona all roll into MariaDB sooner than they appear in MySQL.

    We're not talking about improvements, we're talking about innovations. Keep it fucking straight, Linux people.

    @Onyx said:

    While I'm not dissing the explanation, the way this was written makes me think of this:

    Seriously, does he work for MariaDB or something?



  • It's copypasta. Google the phrase "MariaDB also provides user statistics and better instrumentation"


  • BINNED

    @hungrier said:

    It's copypasta.

    I believe my point still stands. I was not aiming for an ad hominem anyway.

    This time.



  • @blakeyrat said:

    @TimeBandit said:
    On your oh-so-loved windows, every fu***n insignificant piece that gets updated forces you to reboot.

    Until you update drivers, Windows can do a shitload more with drivers than Linux can.


    That just show you know nothing about Linux.
    hint : rmmod and modprobe.



  • @blakeyrat said:

    I grant that one, but that was ages ago and it's unclear what the "right to fork" had to do with it.

    I don't think it had anything to do with the "right to fork" either, but I don't give a shit about random dude's theory.

    @blakeyrat said:

    It's more like, "our OS was so shitty at software distribution that we hacked this work-around which just happened to be useful in other ways as well".

    :rolleyes:

    At least we've disproved your theory.



  • @hungrier said:

    It's copypasta. Google the phrase "MariaDB also provides user statistics and better instrumentation"

    What, was I supposed to write a novel here ? blakeyrat can't even be bothered to do some research before he spit on anything open source.



  • @blakeyrat said:

    So far this is the best example,

    Yay, Blakey said something nice!

    @blakeyrat said:

    and you weasel-word it in the second sentence.

    Damn, shouldn't have gotten my hopes up.


  • Winner of the 2016 Presidential Election

    @blakeyrat said:

    "the security update is only 100% secure when you close all applications and restart all services and logout all users"

    Nope. http://en.wikipedia.org/wiki/Ksplice



  • @boomzilla said:

    Not off the top of my head. But I'm no mind reader. What sort of things do you consider to be innovative?

    I can't come up with anything that could be considered innovative, as in, a new idea.

    The telephone?...

    @blakeyrat said:

    We're not talking about improvements, we're talking about innovations.

    ...no, the telephone is just an improvement over yelling. Hm....


  • Winner of the 2016 Presidential Election

    @Bort said:

    I can't come up with anything that could be considered innovative, as in, a new idea.

    Doesn't matter anyway, because recognizing innovations would require imagination, and @blakeyrat has none.

    Exhibit A:

    @blakeyrat said:

    I highly doubt it will work.



  • @TimeBandit said:

    That just show you know nothing about Linux.

    I know drivers are built-in to the kernal.

    @TimeBandit said:

    hint : rmmod and modprobe.

    hint: dsajdgg and weuyuytqetw



  • @blakeyrat said:

    So why the fuck ain't we seeing any actual innovation from open source developers?

    We are, but apparently you aren't.

    I would invite you to prove your case, but given that you would have to prove the non-existence of innovation in ALL open-source projects, I invite you to give up.

    @Voices in my head said:

    That's just pedantic.

    Yeah? Well that's science.



  • @asdf said:

    Nope. http://en.wikipedia.org/wiki/Ksplice

    That also kills his comment that "I highly doubt it will work.". Even if it's been possible since 2008.

    Of course, if MS ever manage to fix the "reboot every time we patch something", he will call it Innovation !



  • @boomzilla said:

    At least we've disproved your theory.

    Just as an FYI, the burden of proof was never ours.



  • @TimeBandit said:

    Of course, if MS ever manage to fix the "reboot every time we patch something", he will call it Innovation !

    They won't because it can't be done without a radical redesign of the OS kernel. (For example, to a BeOS-style microkernel where everything in the OS including stuff like networking, storage, etc is a rebootable "service". But even then, updating a service was functionally identical to a reboot in many, many cases.)

    The only difference between Windows and Linux in this area is Windows is honest about the need to reboot, and Linux fans dangerously lie to their users that you don't need to.



  • @Shoreline said:

    Just as an FYI, the burden of proof was never ours.

    Meh, we all knew he was wrong anyways. It's just fun watching him rationalize his knee jerk offs.



  • @blakeyrat said:

    I know drivers are built-in to the kernal.

    @TimeBandit said:

    hint : rmmod and modprobe.

    hint: dsajdgg and weuyuytqetw


    Drivers are not built in the kernel, they are dynamically loaded.
    You can remove a kernel module with "rmmod" and load another one with "modprobe" WITHOUT EVER REBOOTING THE MACHINE.

    Your ignorance would not be so bad if you didn't use it to trash everything not MS


  • Winner of the 2016 Presidential Election

    @blakeyrat said:

    They won't because it can't be done without a radical redesign of the OS kernel. (For example, to a BeOS-style microkernel where everything in the OS including stuff like networking, storage, etc is a rebootable "service". But even then, updating a service was functionally identical to a reboot in many, many cases.)

    Wrong. If you had read the article I linked above, you would have known you're wrong, so I'm going to link it again: http://en.wikipedia.org/wiki/Ksplice

    Feel free to ignore it again and continue writing horseshit, though.


  • sockdevs

    @boomzilla said:

    knee jerk offs

    He jerks off with his knee? Weird…


  • Winner of the 2016 Presidential Election

    @RaceProUK said:

    He jerks off with his knee? Weird…

    Thanks. Now I'll have to drink a few bottles of Scotch to get that image out of my head.



  • @blakeyrat said:

    They won't because it can't be done without a radical redesign of the OS kernel. (For example, to a BeOS-style microkernel where everything in the OS including stuff like networking, storage, etc is a rebootable "service". But even then, updating a service was functionally identical to a reboot in many, many cases.)

    The only difference between Windows and Linux in this area is Windows is honest about the need to reboot, and Linux fans dangerously lie to their users that you don't need to.

    You can't be that fuckin stupid.
    Or maybe you can.

    So restarting a single service is equivalent to rebooting the whole machine. At least now I know why every new version of windows is innovative because it can boot faster !


  • sockdevs

    @asdf said:

    Thanks. Now I'll have to drink a few bottles of Scotch to get that image out of my head.

    …and my work here is done! :laughing:



  • @TimeBandit said:

    Your ignorance would not be so bad if you didn't use it to trash everything not MS.

    Fun Fact: People who weaponise ignorance will use it to trash everything they don't like.

    I guess it's the same as people who weaponise anything.



  • @TimeBandit said:

    windows

    If they could but innovate it to fucking tell me what's locking the file I can't move/delete, that would be great, thanks.


  • Winner of the 2016 Presidential Election

    @TimeBandit said:

    So restarting a single service is equivalent to rebooting the whole machine. At least now I know why every new version of windows is innovative because it can boot faster !

    Shhhh! Stop arguing against his straw man counter-argument. He'll continue ranting about how restarting services makes live kernel patching insecure otherwise.


  • Discourse touched me in a no-no place

    Ignoring the fucking stupid reboot debate and approaching the actual question of innovation.

    In my experience, innovation is when you do something genuinely new and break new ground or apply existing ideas to non-obvious existing, previously unrelated problems.

    While, in principle, forking could allow a stuck in the mud project to break apart and a new team take it in directions nobody had ever thought of, I can't think of any examples.

    While I can think of many genuinely innovative things that, to my knowledge originated in open source, I don't think any of them directly resulted from a fork.

    It's also my experience that opensource teams more often excel at expanding on ideas than actually having them. This is probably just an observational bias, as most people in general don't have original ideas but are pretty good at elaborating once they have a concept.



  • @TimeBandit said:

    Drivers are not built in the kernel, they are dynamically loaded.

    They're not built in the kernel, they're just in the exact same repo as the kernel and have to be updated along with the kernel. Suuure.

    @TimeBandit said:

    So restarting a single service is equivalent to rebooting the whole machine.

    Depends on what the service is. A lot of them are. Some aren't.

    The real point is, if you update a widely-used library, you have to restart everything using it. Linux's design actually makes this situation worse because processes will still use the "ghost" in-memory copy of the library, tricking people into thinking their system is secure when it is not-- the insecure code is still loaded and executing, and no errors were reported during the update process.

    For this specific situation, Windows' locking of executing files is actually a plus and not a minus.


  • Winner of the 2016 Presidential Election

    @blakeyrat said:

    The real point is, if you update a widely-used library

    TIL that @blakeyrat doesn't know the difference between libraries and kernels.



  • @RaceProUK said:

    Weird…

    Some would say, "innovative."


  • :belt_onion:

    @cartman82 said:

    defensive forks

    I am envisioning an army of forks in phalanx formation.
    I am chuckling.
    Have a like.


  • Winner of the 2016 Presidential Election

    Edit: BRB, haven't played that game in ages.


  • :belt_onion:

    THIS

    IS

    SPORKTA!!!!!!!!1



  • Isn't the burden of proof usually on the party claiming the existence of something? Unless there's already a well-established body of evidence demonstrating the existence, but we don't have that in this case...

    @Voices in Shoreline's head said:

    That's just pedantic.

    I think this is how science is really supposed to work. Burden of proof is properly on...

    @Voices in Bort's head said:

    Enough talk. Get the knife.

    NO! I told you, I won't do that anymore!

    @Voices in Bort's head said:

    You know it must be done.

    NO! You're wrong! We can be reasonable about this!

    @Voices in Bort's head said:

    BULLSHIT YOU CAN BE REASONABLE! DO AS I COMMAND!

    snob I can't. All those terrified faces... all those lacerated corpses... they haunt me...

    @Voices in Bort's head said:

    Aw, come on. When you're done, we can go get ice cream.

    ...promise?

    @Voices in Bort's head said:

    Promise.

    Well, if there's ice cream in it for me...

    @Voices in Shoreline's head said:

    Hey! How'd you get in here! AAAAAAAA!!!!! Aaaaaa uuuggggghhhh...

    HA HA HA HA!!! AAAA HA HA HA!!! slurp

    @Shoreline said:

    OH MY GOD! WHAT DID YOU DO?!?

    ...do you like mint chocolate chip? No one needs to know about this..



  • @asdf said:

    TIL that @blakeyrat doesn't know the difference between libraries and kernels.

    No, I'm saying it doesn't matter. The majority of updates are going to touch the libraries, and those require shit to be rebooted.



  • @blakeyrat said:

    No, I'm saying it doesn't matter.

    Yes, we get it. You don't understand the subject. Observe:

    @blakeyrat said:

    The majority of updates are going to touch the libraries, and those require shit to be rebooted.

    The topic was updating the kernel. Not "the majority of updates."


Log in to reply
 

Looks like your connection to What the Daily WTF? was lost, please wait while we try to reconnect.