zzz
SSU2: cost: 15 caps: 4 i: nHWYMT~kTVYtNnKRBteCFhzFmxe-sH4IXHWugIhF50s= iexp0: 1694638252 iexp1: 1694640130 iexp2: 1694673097 iexp3: 1694703017 iexp4: 1694714927 iexp5: 1694721602 iexp6: 1694727827 iexp7: 1694729104 iexp8: 1694729104 ih0: xiyi~HK4KifJTojz40qrBhR-FPPfPG-YCD4RDlLTGXA= ih1: MRSvQagRE7b86km16IpW88zYUuNU6Ox23maTT0ZkX-A= ih2: 1e3uF-Ui0HkeKLIDsb3dysMX-9iIUR7YpJPqTSFke4Q= ih3:
zzz
Ci4VALNchb~jVvSWdCbmGXJw5X57Slh7IFAhX-xoGn4= ih4: IfluCCUC2xtoFo40pyw~pFwrRrE3OqsKx4tqustfDUQ= ih5: mi1OTugQUr4X8flh4hc8DmhPQ1xF~iypkVJlwM9xIkk= ih6: MTxk71lLomzajcDgLJZnHK9OOdSH76qAZIFA~J7GhYQ= ih7: -zDV3emDbYZ1zBo3oVS4DnsQCgM6nFjVXYPC204fF5E= ih8: BkaqEcXfWWlSoSU-8GLXBW-el3joPXK8FlsIIxCSz6k= itag0: 0 itag1: 0 itag2: 0 itag3: 0 itag4: 0 itag5: 0 itag6: 1584974673 itag7: 2200658897 itag8: 4243358369
zzz
s: juuCV5PKXn7r0yP-RvPftZumy4PHSlzLMYkLPqvmaxE= v: 2
zzz
^^^ orignal
zzz
does this look like i2pd?
zzz
0.9.59 XU
dr|z3d
zzz: html4 -> 5 is low risk, but you're going to bump into display issues (likely you already have) which will need to be addressed in the css.
dr|z3d
I should also mention that if you're expecting the transition to html5 will address the flash of white in the console when you browse around, the issue isn't html4.
zzz
it fixed it for me, I wasn't guessing, but perhaps there's more. But thx for confirming not to do it last minute
dr|z3d
I'm not suggesting you don't do it, and if it fixed the issue, great. As I understand it, the issue lies with the site isolation stuff Firefox has going on.
dr|z3d
aka "fission".
dr|z3d
Back when they were still testing that feature, random pages in the console would load up with a white background, and disabling it made the problem go away. I think that's still the case.
dr|z3d
As for the css stuff, maybe it's not so important for you. When I made the transition, iirc it took about a week of work to get the console css fixed up. Mostly it's just pixel-level differences in layout.
zzz
you happen to recall any details of the issues or on what browsers? any hints on what to look for would be appreciated
zzz
the "flash" I had was on snark iirc
dr|z3d
I don't think you need to worry so much about browsers, html5 should reduce the differences in rendering. I don't recall offhand the exact changes, was a long time ago, but table margins might be something to look at. You can also look at the commits on the + repo and the css changes may give you some pointers.
zzz
ok, thx
dr|z3d
I still see occasional flashes of white for snark, which is why I'm about to push a commit that explictly styles the html tag before the css is loaded.
dr|z3d
another thing to check there is your caching policy. css and image caching should be immutable for best results.
zzz
patch welcome :)
dr|z3d
for the html, feel free to modify this according to your html/body background for the relevant themes:
dr|z3d
String pageBackground = "#fff";
dr|z3d
if (theme.equals("dark")) {
dr|z3d
pageBackground = "#000";
dr|z3d
} else if (theme.equals("midnight")) {
dr|z3d
pageBackground = "repeating-linear-gradient(180deg,rgba(0,0,24,.75) 2px,rgba(0,0,0,.7) 4px)/100% 4px," +
dr|z3d
"var(--tile)/171px 148px,var(--offline)/0,#000010";
dr|z3d
} else if (theme.equals("ubergine")) {
dr|z3d
pageBackground = "repeating-linear-gradient(90deg,rgba(0,0,0,.5) 2px,rgba(48,0,48,.5) 4px)," +
dr|z3d
"repeating-linear-gradient(180deg,#080008 2px,#212 4px),var(--offline) no-repeat,#130313";
dr|z3d
} else if (theme.equals("vanilla")) {
dr|z3d
pageBackground = "repeating-linear-gradient(180deg,#6f5b4c 1px,#a9927e 1px,#bfa388 4px),#cab39b";
dr|z3d
}
dr|z3d
out.write(DOCTYPE + "<html style=background:" + pageBackground + ">\n" +
zzz
interesting. is that a confirmed fix or you're still testing?
dr|z3d
works for me.
dr|z3d
not much to test. it works because the page isn't waiting for the css to load before setting the background styling.
zzz
ok