download of source code

[karlkracher]

I went here https://github.com/MoonchildProductions/UXP/releases, downloaded each of the two files three times and what i've got is this:

Code: Select all

21.09.2018  01:28       245.940.156 UXP-PM28.1.0_Release_1.zip
21.09.2018  02:24       175.990.190 UXP-PM28.1.0_Release_2.zip
21.09.2018  03:07       188.444.675 UXP-PM28.1.0_Release_3.zip

21.09.2018  01:58       270.598.144 UXP-PM28.1.0_Release_1.tar.gz
21.09.2018  02:24       158.279.720 UXP-PM28.1.0_Release_2.tar.gz
21.09.2018  03:07       147.734.528 UXP-PM28.1.0_Release_3.tar.gz

zip says all six files are broken.

[therube]

Where are you seeing that?

I see a single file (either zip or tar.gz) with the "source" (text) files/directories within.


UXP-PM28.1.0_Release.zip, 380,567,740, 667c7e17bb4b5f9e8391d80484d7f67e:

Code: Select all

accessible
application
build
caps
chrome
config
db
devtools
docs
docshell
dom
editor
embedding
extensions
gfx
gradle
hal
image
intl
ipc
js
layout
media
memory
mfbt
mobile
modules
mozglue
netwerk
nsprpub
other-licenses
parser
probes
python
rdf
security
services
startupcache
storage
taskcluster
testing
toolkit
tools
uriloader
view
widget
xpcom
xpfe
.clang-format
.clang-format-ignore
.eslintignore
.eslintrc.js
.flake8
.gdbinit
.gdbinit_python
.gitattributes
.gitignore
.lldbinit
.taskcluster.yml
.ycm_extra_conf.py
aclocal.m4
Android.mk
AUTHORS
build.gradle
client.mk
client.py
CLOBBER
configure.in
configure.py
GNUmakefile
gradle.properties
gradlew
LEGAL
LICENSE
mach
Makefile.in
moz.build
moz.configure
mozilla-config.h.in
old-configure.in
README.md
settings.gradle
sourcestamp.txt
test.mozbuild

[Moonchild]

therube: maybe you weren't paying attention. The downloads from github are all incomplete for this person (see file sizes).

There's nothing we can do on our end to fix this -- if you are having connectivity issues with github, you have to contact github.

[vannilla]

Another alternative is to try "git clone", but then you'd get what is essentially an alpha version instead of a stable release (save for some settings and/or arguments I don't know about).

[Isengrim]

Another alternative is to try "git clone", but then you'd get what is essentially an alpha version instead of a stable release (save for some settings and/or arguments I don't know about).

You can use "git clone" to download the whole repository, including commit history. You can then use git to checkout branches such as "basilisk-release" or "pale-moon-release" to get the latest release version of the code.

[karlkracher]

Thanks for the ideas, but no matter what i do in the end i would have to download and this is not working between github and me. I've tried 10 or 15 time the last two days, going on wouldn't be nice to the server.

[Moonchild]

Thanks for the ideas, but no matter what i do in the end i would have to download and this is not working between github and me. I've tried 10 or 15 time the last two days, going on wouldn't be nice to the server.

I don't think github cares much about bandwidth for failed attempts. It's because of their own setup ("streaming" the downloads instead of having static files) that it causes problems. As said you should ask github for help in solving these problems.

[therube]

(Yes, I misunderstood what he was showing.)


How are you downloading, through the browser?
Does PM Download Manager show the files to have completed or failed?

Have you tried a real download manager (external) program, even something like wget?

[yami_]

IIRC if the file is streamed the browser will not know its size. This makes detecting an incomplete download impossible.

[karlkracher]

I've used

• the "normal" way in the browser, actually two browsers: pale moon and firefox portable
• downthemall in pale moon
• jdownloader
• wget

At some point the data stream stops (instantly from good speed to zero) and after waiting the program concludes "done". Only wget seems to be a bit more clever because it restarts the download with an error message. And to be honest i do a lot downloading without problems, so i think it is not my computer or internet connection.

But in the meantime i've lost the reason to load the whole package, my point of interest is in the javascript files i can access directly in the browser.

[Isengrim]

You can also browse code directly through github, or by using http://xref.palemoon.org/.

[therube]

Two different browsers, an extension in your browser, a Java based downloader, an external downloader, & none are working, I would think it is on your end &/or your connection &/or github (you didn't piss them off did you ).

The download itself isn't that large, 380M, & I gather you've download larger files, elsewhere?


Even if wget outright failed in a particular attempt, I would think using the -c switch, wget -c, would pick up the pieces in future attempt(s)?

[Moonchild]

As yami_ already pointed out the way downloads are streamed makes it impossible for any downloading application (browser or otherwise) to know the file size, and won't know if a file is incomplete or not if the remote end stops sending.