5 ms·

Google Forks WebKit And Launches Blink, A New Rendering Engine

72 points by jordn 14y ago

Lightning 14y ago
https://news.ycombinator.com/item?id=5489025 https://news.ycombinator.com/item?id=5489025
ChuckMcM 14y ago
" the decision to fork WebKit was entirely driven by the engineering teams and solely based on the fact that the engineers felt constrained by the technical complexity of working within the WebKit ecosystem." Is this code for "They wouldn't take our changes so we said, 'Fork it!' and made our own." ?
- mooism2 14y ago
  “Currently, the majority of WebKit reviewers are from Google (95), with Apple coming in second (59)... Google is also currently responsible for the vast majority of commits to the WebKit repository...” Doubtful.
  - mirsadm 14y ago
    Do you have any proof that the statement is false?
    - mooism2 14y ago
      No, I was using the statement to cast doubt on the parent comment's suggestion. And then I got a good night's sleep, waking up to find that I'd communicated the opposite of what I meant and that it's too late to edit my comment.
  - coldtea 14y ago
    Numbers are doubtful? I mean, sure, one could fabricate them, but why, especially if he can easily be caught doing it? Those are things anybody can verify by looking at the repository stats. Here's what I found with a cursory look: Google has overtook Apple in commits 2-3 years already: http://blog.bitergia.com/2013/02/06/report-on-the-activity-of-companies-in-the-webkit-project/ http://blog.bitergia.com/2013/02/06/report-on-the-activity-o... Older stats (already Google on top): http://neugierig.org/software/chromium/notes/2010/02/webkit-commits.html http://neugierig.org/software/chromium/notes/2010/02/webkit-... And a TechCrunch article on the issue: http://techcrunch.com/2010/02/06/google-apple-webkit/ http://techcrunch.com/2010/02/06/google-apple-webkit/
    - mooism2 14y ago
      Sorry, I meant that the suggestion in the comment I was replying to (that Google was forking because they couldn't get their changes in) was doubtful, not that the figures I quoted from the article were doubtful (I was using them as justification for my comment).
  - hp50g 14y ago
    Reviewers and number of commits aren't really relevant stats. I'm the top committer and code reviewer at my company but it doesn't mean I write the most code :)
  - ChuckMcM 14y ago
    Reading through some of the analysis I agree, the story about the amount of glue necessary getting worse rather than better was pretty compelling. As a systems guy I find it amazing how intertwined the various parts are, between the process model to the rendering model, simplification there can only be good.
- dragonwriter 14y ago
  > Is this code for "They wouldn't take our changes so we said, 'Fork it!' and made our own." ? I think its probably more exactly what it says -- a way of eliminating the overhead of dealing with the complexity of the WebKit ecosystem. As noted in the article, "Google says, for example, that it will be able to remove 7 build systems and delete more than 7,000 files comprising more than 4.5 million lines right off the bat." That's a whole lot of complexity that isn't relevant to what Google is doing with Chrome that is worth not having to deal with.
  - stock_toaster 14y ago
    This[1] other article infers that those lines/files are related to webkit2 features of multiprocessing and sandbox code. So it seems to me that ChuckMcM may not be far off. [1]: http://arstechnica.com/information-technology/2013/04/google-going-its-own-way-forking-webkit-rendering-engine/ http://arstechnica.com/information-technology/2013/04/google...
- drivebyacct2 14y ago
  Pretty sure it has more to do with their architecture changes RE: multithreading and embedding/porting.
- justinschuh 14y ago
  Nope. If Google wanted to push things through it would certainly be possible as the single largest contributor, but that's not a productive way for anyone to participate. Instead, it's just a matter of simple project divergence. Apple landed the WebKit2 layer a few years ago, which was largely incompatible with the Chrome architecture that predated it by several years. Some minor efforts were made to unify parts, but WebKit2 continued to move in a direction that was increasingly incompatible with the existing Chrome architecture. Rather than burden everyone by continuing to maintain a codebase between these increasingly divergent approaches, it made sense to split.
sergiotapia 14y ago
_IT BEGINS_ We're switching out the cluster that was IE6 for Google Chrome. Mozilla, save us; you're our only hope. Opera is going to contribute to Blink as well, so I guess there's that.
- AaronFriel 14y ago
  On the contrary, this saves us from WebKit being the new Trident from IE6.
crimsonzagar 14y ago
This is just a personal opinion, but for a browser I'd always support and use a not-for-profit one. No matter how fast, or perceptibly fast, an offering from a business entity is. Here is why: 1. Tomorrow you, the billion dollar enterprise, will force real identity even more vehemently. 2. I certainly don't want to see a constant hum of ads. 3. I don't want to turn my computer into a wide-mouth bucket for chrome apps, for example, which have higher than normal access and rights on my disk. Your browser could potentially take me back into an era of 'download-yet-another-snoopy-app' into your machine yeah! Put it all behind a curated garden and keep me happy within. Where web on tablets and smartphones is today. Nope, not for me.
- jmillikin 14y ago
  Current popular browsers are: * Firefox * Chrome / Chromium * Internet Explorer * Safari * Opera All of them are produced with the aim of benefiting their parent corporations in some way -- for example, Mozilla Corp rents the default search engine setting to Google for hundreds of millions per year. I currently use Firefox and Chromium as my primary browsers, because they are the only popular browsers that are open-source and therefore the only browsers I fully trust. Despite coming from "business entities", neither browser shows me ads, requires me to enter a government-approved name, or requires me to install third-party applications.
  - ZeroGravitas 14y ago
    Mozilla Corp was set up by Mozilla Foundation for tax reasons. Firefox is produced with the aim of meeting the foundation's mission.
  - asadotzler 14y ago
    "All of them are produced with the aim of benefiting their parent corporations in some way -- for example, Mozilla Corp rents the default search engine setting to Google for hundreds of millions per year." You've crafted a sentence that will mislead people. Mozilla did not, and does not, make Firefox in order to generate search revenue. Mozilla makes Firefox to directly advance Mozilla's non-profit, public benefit mission to improve the Web. Firefox is not produced with the aim of generating revenue. It never has been and it never will be.
OGinparadise 14y ago
Duh! Chrome, Youtube, Maps, Android and virtually everything serve Google's $$$$ making efforts. Same applies to Microsoft and other large corps, no matter how much they spin. So I use Firefox and will use for as long as they are independent.
- teej 14y ago
  You do realize that Google pays Firefox $300M a year, making them at least somewhat aligned to the business of Google?
  - OGinparadise 14y ago
    Yes, and I suspect that Firefox is, at least, not going to work against the way it makes its money (ads) but the alternative browsers have bigger conflicts of interest.
    - Apocryphon 14y ago
      It's interesting to see open-source foundations get corporate patrons. Like Ubuntu's relation with Amazon.
      - laumars 14y ago
        That's always been the case. (Linux < IBM, Oracle, Redhat. FreeBSD has corporate sponsors and so on). For any project to succeed, it's going to need some kind of funding. And often that means turning to big businesses, who have deeper pockets than all of us little people combined.
    - laumars 14y ago
      Opera? Plus with their switch to Blink on the horizon, even the old complaints about rendering compatibility are soon to be moot.
      - myko 14y ago
yalogin 14y ago
This could also be code for we are implementing features that we don't want to contribute to WebKit.
vacri 14y ago
Please, when you make a new project, please, please don't use a pre-existing English word. Perl? Great name. Searchable. You won't be contaminated and neither will you contaminate others. Tumblr? Same thing. Python? Well, has a lot of groundswell so now tech 'owns' the name, but you contaminate other people's search. Rust? (or a million other projects without python's heft) The reverse issue. Blink also has the problem of already being a word used occasionally in web design (bad or otherwise).
- thatthatis 14y ago
  At least until a quick and convenient context setting syntax enters the mainstream search syntax. Id love the ability to search "Json parsers [python]" or "music packages [django[python[programming]]]"
- JamesAcorn 14y ago
  'Chrome' used to refer to the GUI elements that surround a webpage - scrollbars, back button etc. Google changed that. Google will effectively Google-bomb the web development word 'blink' too. It is no big loss.
dchichkov 14y ago
"Don't Blink. Blink and you're dead. Don't turn your back. Don't look away. And don't Blink. Good Luck."
ricardobeat 14y ago
"A Short Translation from Bullshit to English of Selected Portions of the Google Chrome Blink Developer FAQ" http://prng.net/blink-faq.html http://prng.net/blink-faq.html
yardie 14y ago
Well, talk about taking your ball and going home. News like this is something I would have assumed from Apple rather than Google (based on the amount of code they've contributed). Maybe it's political, maybe it's technical but this is a bit of news I expected yet didn't see coming.
chesh 14y ago
By forking WebKit Google is fucking Apple

72 points by jordn 14y ago

https://news.ycombinator.com/item?id=5489025 https://news.ycombinator.com/item?id=5489025

" the decision to fork WebKit was entirely driven by the engineering teams and solely based on the fact that the engineers felt constrained by the technical complexity of working within the WebKit ecosystem." Is this code for "They wouldn't take our changes so we said, 'Fork it!' and made our own." ?

mooism2 14y ago
“Currently, the majority of WebKit reviewers are from Google (95), with Apple coming in second (59)... Google is also currently responsible for the vast majority of commits to the WebKit repository...” Doubtful.
- mirsadm 14y ago
  Do you have any proof that the statement is false?
  - mooism2 14y ago
    No, I was using the statement to cast doubt on the parent comment's suggestion. And then I got a good night's sleep, waking up to find that I'd communicated the opposite of what I meant and that it's too late to edit my comment.
- coldtea 14y ago
  Numbers are doubtful? I mean, sure, one could fabricate them, but why, especially if he can easily be caught doing it? Those are things anybody can verify by looking at the repository stats. Here's what I found with a cursory look: Google has overtook Apple in commits 2-3 years already: http://blog.bitergia.com/2013/02/06/report-on-the-activity-of-companies-in-the-webkit-project/ http://blog.bitergia.com/2013/02/06/report-on-the-activity-o... Older stats (already Google on top): http://neugierig.org/software/chromium/notes/2010/02/webkit-commits.html http://neugierig.org/software/chromium/notes/2010/02/webkit-... And a TechCrunch article on the issue: http://techcrunch.com/2010/02/06/google-apple-webkit/ http://techcrunch.com/2010/02/06/google-apple-webkit/
  - mooism2 14y ago
    Sorry, I meant that the suggestion in the comment I was replying to (that Google was forking because they couldn't get their changes in) was doubtful, not that the figures I quoted from the article were doubtful (I was using them as justification for my comment).
- hp50g 14y ago
  Reviewers and number of commits aren't really relevant stats. I'm the top committer and code reviewer at my company but it doesn't mean I write the most code :)
- ChuckMcM 14y ago
  Reading through some of the analysis I agree, the story about the amount of glue necessary getting worse rather than better was pretty compelling. As a systems guy I find it amazing how intertwined the various parts are, between the process model to the rendering model, simplification there can only be good.
dragonwriter 14y ago
> Is this code for "They wouldn't take our changes so we said, 'Fork it!' and made our own." ? I think its probably more exactly what it says -- a way of eliminating the overhead of dealing with the complexity of the WebKit ecosystem. As noted in the article, "Google says, for example, that it will be able to remove 7 build systems and delete more than 7,000 files comprising more than 4.5 million lines right off the bat." That's a whole lot of complexity that isn't relevant to what Google is doing with Chrome that is worth not having to deal with.
- stock_toaster 14y ago
  This[1] other article infers that those lines/files are related to webkit2 features of multiprocessing and sandbox code. So it seems to me that ChuckMcM may not be far off. [1]: http://arstechnica.com/information-technology/2013/04/google-going-its-own-way-forking-webkit-rendering-engine/ http://arstechnica.com/information-technology/2013/04/google...
drivebyacct2 14y ago
Pretty sure it has more to do with their architecture changes RE: multithreading and embedding/porting.
justinschuh 14y ago
Nope. If Google wanted to push things through it would certainly be possible as the single largest contributor, but that's not a productive way for anyone to participate. Instead, it's just a matter of simple project divergence. Apple landed the WebKit2 layer a few years ago, which was largely incompatible with the Chrome architecture that predated it by several years. Some minor efforts were made to unify parts, but WebKit2 continued to move in a direction that was increasingly incompatible with the existing Chrome architecture. Rather than burden everyone by continuing to maintain a codebase between these increasingly divergent approaches, it made sense to split.

sergiotapia 14y ago

_IT BEGINS_ We're switching out the cluster that was IE6 for Google Chrome. Mozilla, save us; you're our only hope. Opera is going to contribute to Blink as well, so I guess there's that.

AaronFriel 14y ago
On the contrary, this saves us from WebKit being the new Trident from IE6.

crimsonzagar 14y ago

This is just a personal opinion, but for a browser I'd always support and use a not-for-profit one. No matter how fast, or perceptibly fast, an offering from a business entity is. Here is why: 1. Tomorrow you, the billion dollar enterprise, will force real identity even more vehemently. 2. I certainly don't want to see a constant hum of ads. 3. I don't want to turn my computer into a wide-mouth bucket for chrome apps, for example, which have higher than normal access and rights on my disk. Your browser could potentially take me back into an era of 'download-yet-another-snoopy-app' into your machine yeah! Put it all behind a curated garden and keep me happy within. Where web on tablets and smartphones is today. Nope, not for me.

jmillikin 14y ago
Current popular browsers are: * Firefox * Chrome / Chromium * Internet Explorer * Safari * Opera All of them are produced with the aim of benefiting their parent corporations in some way -- for example, Mozilla Corp rents the default search engine setting to Google for hundreds of millions per year. I currently use Firefox and Chromium as my primary browsers, because they are the only popular browsers that are open-source and therefore the only browsers I fully trust. Despite coming from "business entities", neither browser shows me ads, requires me to enter a government-approved name, or requires me to install third-party applications.
- ZeroGravitas 14y ago
  Mozilla Corp was set up by Mozilla Foundation for tax reasons. Firefox is produced with the aim of meeting the foundation's mission.
- asadotzler 14y ago
  "All of them are produced with the aim of benefiting their parent corporations in some way -- for example, Mozilla Corp rents the default search engine setting to Google for hundreds of millions per year." You've crafted a sentence that will mislead people. Mozilla did not, and does not, make Firefox in order to generate search revenue. Mozilla makes Firefox to directly advance Mozilla's non-profit, public benefit mission to improve the Web. Firefox is not produced with the aim of generating revenue. It never has been and it never will be.

OGinparadise 14y ago

Duh! Chrome, Youtube, Maps, Android and virtually everything serve Google's $$$$ making efforts. Same applies to Microsoft and other large corps, no matter how much they spin. So I use Firefox and will use for as long as they are independent.

teej 14y ago
You do realize that Google pays Firefox $300M a year, making them at least somewhat aligned to the business of Google?
- OGinparadise 14y ago
  Yes, and I suspect that Firefox is, at least, not going to work against the way it makes its money (ads) but the alternative browsers have bigger conflicts of interest.
  - Apocryphon 14y ago
    It's interesting to see open-source foundations get corporate patrons. Like Ubuntu's relation with Amazon.
    - laumars 14y ago
      That's always been the case. (Linux < IBM, Oracle, Redhat. FreeBSD has corporate sponsors and so on). For any project to succeed, it's going to need some kind of funding. And often that means turning to big businesses, who have deeper pockets than all of us little people combined.
  - laumars 14y ago
    Opera? Plus with their switch to Blink on the horizon, even the old complaints about rendering compatibility are soon to be moot.
    - myko 14y ago

yalogin 14y ago

This could also be code for we are implementing features that we don't want to contribute to WebKit.

vacri 14y ago

Please, when you make a new project, please, please don't use a pre-existing English word. Perl? Great name. Searchable. You won't be contaminated and neither will you contaminate others. Tumblr? Same thing. Python? Well, has a lot of groundswell so now tech 'owns' the name, but you contaminate other people's search. Rust? (or a million other projects without python's heft) The reverse issue. Blink also has the problem of already being a word used occasionally in web design (bad or otherwise).

thatthatis 14y ago
At least until a quick and convenient context setting syntax enters the mainstream search syntax. Id love the ability to search "Json parsers [python]" or "music packages [django[python[programming]]]"
JamesAcorn 14y ago
'Chrome' used to refer to the GUI elements that surround a webpage - scrollbars, back button etc. Google changed that. Google will effectively Google-bomb the web development word 'blink' too. It is no big loss.

dchichkov 14y ago

"Don't Blink. Blink and you're dead. Don't turn your back. Don't look away. And don't Blink. Good Luck."

ricardobeat 14y ago

"A Short Translation from Bullshit to English of Selected Portions of the Google Chrome Blink Developer FAQ" http://prng.net/blink-faq.html http://prng.net/blink-faq.html

yardie 14y ago

Well, talk about taking your ball and going home. News like this is something I would have assumed from Apple rather than Google (based on the amount of code they've contributed). Maybe it's political, maybe it's technical but this is a bit of news I expected yet didn't see coming.

chesh 14y ago

By forking WebKit Google is fucking Apple