Strange interaction between Alpha and legend - r

Question

While plotting several ecdf curves that overlapped, I tried adjusting the alpha of the curves to improve visibility. While tinkering with the correct placement of alpha, I found the following.
library(ggplot2)
library(dplyr)
x <- data.frame(Var = rep(1:3, 10000)) %>%
mutate(Val = rnorm(10000)*Var,
Var = factor(Var)) %>%
arrange(Var, Val) %>%
group_by(Var) %>%
mutate(ecdf = ecdf(Val)(Val))
ggplot(x, aes(x=Val)) +
stat_ecdf(aes(color = Var), size = 1.25, alpha = .9)
This gives the lines the correct alpha, but makes the legend useless. (I'm only using alpha=.9 here to demonstrate the point that the legend colors completely disappear). The work around I've found is to add:
ggplot(x, aes(x=Val)) +
stat_ecdf(aes(color = Var), size = 1.35, alpha = .9) +
guides(color = guide_legend(override.aes= list(alpha = 1)))
So while I have a solution for my immediate problem, can someone explain why the first call to ggplot is messed up? Is this a bug? If it makes any difference, I believe this issue also exists when using geom_line (though a slightly different data.frame is needed).
Wierd. Here's my sessionInfo(). I've also checked to see if there are any outdated packages.
sessionInfo()
R version 3.2.1 (2015-06-18)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 7 x64 (build 7601) Service Pack 1
locale:
[1] LC_COLLATE=Japanese_Japan.932 LC_CTYPE=Japanese_Japan.932 LC_MONETARY=Japanese_Japan.932
[4] LC_NUMERIC=C LC_TIME=Japanese_Japan.932
attached base packages:
[1] splines stats graphics grDevices utils datasets methods base
other attached packages:
[1] RColorBrewer_1.1-2 ggplot2_1.0.1 stringr_1.0.0 tidyr_0.2.0 dplyr_0.4.2
[6] data.table_1.9.4
loaded via a namespace (and not attached):
[1] Rcpp_0.11.6 magrittr_1.5 MASS_7.3-40 munsell_0.4.2 colorspace_1.2-6
[6] R6_2.0.1 plyr_1.8.3 tools_3.2.1 parallel_3.2.1 grid_3.2.1
[11] gtable_0.1.2 DBI_0.3.1 lazyeval_0.1.10 assertthat_0.1 digest_0.6.8
[16] reshape2_1.4.1 labeling_0.3 stringi_0.5-4 scales_0.2.5 chron_2.3-47
[21] proto_0.3-10

Answer 1

How are they different? What am I missing?
library(ggplot2)
library(dplyr)
library(gridExtra)
x <- data.frame(Var = rep(1:3, 10000)) %>%
mutate(Val = rnorm(10000)*Var,
Var = factor(Var)) %>%
arrange(Var, Val) %>%
group_by(Var) %>%
mutate(ecdf = ecdf(Val)(Val))
ggplot(x, aes(x=Val)) +
stat_ecdf(aes(color = Var), size = 1.25, alpha = .9) -> gg1
ggplot(x, aes(x=Val)) +
stat_ecdf(aes(color = Var), size = 1.35, alpha = .9) +
guides(color = guide_legend(override.aes= list(alpha = 1))) -> gg2
grid.arrange(gg1, gg2)