Martin Robinson WebKit - LemontechSA/TideSDK GitHub Wiki
[8:50pm] dubcanada: fairwinds do you know where the webkit version is at for ti desktop?
[8:50pm] fairwinds: dubcanada: we can likely get by checking dependency bundle
[8:50pm] Daegalus: TideSDK\kroll\thirdparty-win32-i386-r43\webkit
[8:50pm] dubcanada: Yah but is there a repo or something?
[8:50pm] dubcanada: So I can see how out of date that webkit is
[8:51pm] fairwinds: sure the repo is in our github
[8:51pm] Daegalus: 5.33
[8:52pm] dubcanada: Over a year
fairwinds: dubcanada: that sounds about right
[8:54pm] fairwinds: mrobinson: what sort of time did it take to handle patches for webkit in Titanium
[8:55pm] fairwinds: and how many folks were involved
dubcanada: PatchLog.md
[8:55pm] dubcanada: does not exist?
[8:57pm] fairwinds: dubcanada: where you see this?
[8:57pm] dubcanada: WebKit for Titanium Desktop
[8:57pm] dubcanada: This is a fork of WebKit with some needed patches to integrate it into Titanium Desktop.
[8:57pm] dubcanada: See PatchLog.md for a list of changes we have applied.
[8:57pm] dubcanada: there is no PatchLog.md
dubcanada: There seems to be comments inside source code with // Titanium
[8:58pm] dubcanada: But who's to say those are all that was changed
[8:58pm] fairwinds: dubcanada: Marshall advised we go through commit history to pick them up
mrobinson: It's common for people who fork WebKit to have this kind of pain.
[9:09pm] mrobinson: It's always a bit tricky. I tried to reduce the surface area of the parts of Titanium in WebKit.
[9:10pm] fairwinds: mrobinson: k, what sort of time was involved in this, I realize a time sink
[9:10pm] fairwinds: was there more than a single person on this
[9:10pm] mrobinson: fairwinds: Typically only one person. It's not a job that's easy to split.
[9:10pm] fairwinds: I think a year is a long time in a fast moving code base like webkit
[9:10pm] mrobinson: It's mostly trying to understand merge conflicts and resolving them properly.
[9:11pm] mrobinson: fairwinds: Yes, it's quite a long time indeed.
fairwinds: what would you suggest given our predicament
[9:11pm] fairwinds: appc really let things go v long
[9:14pm] fairwinds: mrobinson: there is likely nothing but repeating the process of forking and assessing patches
[9:14pm] dubcanada: It may be easier to fork the current webkit
[9:14pm] fairwinds: mrobinson: also did you know of a PatchLog.md
[9:14pm] dubcanada: and bring over the changes
[9:14pm] mrobinson: fairwinds: The absolute best solution would be to try to find a way to use WebKit without a fork.
[9:15pm] fairwinds: dubcanada: yes that is what am talking about
mrobinson: fairwinds: The other option is to gradually upstream more and more of the features you need until you have a minimal set of patches you need to apply on top of WebKit.
mrobinson: fairwinds: Yes, it took probably days of work each time.
[9:28pm] fairwinds: k, will get going with build and see where it goes
[9:28pm] mrobinson: fairwinds: The hardest part was actually finding a revision that was good for all platforms.
[9:28pm] mrobinson: fairwinds: Another issue is creating your branch and ensuring that it is stable