# Options specific to downloading packages comment "Downloading" config FORBID_DOWNLOAD bool prompt "Forbid downloads" default n help Normally, crosstool-NG will try to download missing tarballs (or checkout from CVS/SVN...). If you do not have network connectivity when you run crosstool-NG, and some files are missing, it can be a long time before crosstool-NG fails. Saying 'y' here will prevent crosstool-NG from downloading missing files, thus failing early so that you don't get stuck. if ! FORBID_DOWNLOAD config FORCE_DOWNLOAD bool prompt "Force downloads" default n help Force downloading tarballs, even if one already exists. Usefull if you suspect a tarball to be damaged. config USE_MIRROR bool prompt "Use a mirror" default n help If you have a machine on your LAN that mirrors some of the needed tarballs, you can say 'Y' here, and configure adequate values in the following options. Obviously, nothing prevents you from using a mirror that is in fact *not* on your LAN, for example on another subnet of your company's network, or a mirror on the Internet. if USE_MIRROR config PREFER_MIRROR bool prompt "Prefer the mirror" default n help Say 'Y' here if you prefer the LAN miror over the upstream sources. config MIRROR_BASE_URL string prompt "Base URL" default "http://ymorin.is-a-geek.org/mirrors/" help This is the base URL searched in for tarballs. I (Yann E. MORIN) have set up such a mirror to host snapshots of some components, when those snapshots are volatile on the upstream servers. The mirror is *slow*, because it is hosted behind an ADSL line. For the time being, I haven't set up bandwidth limitations, but should the mirror be abused, I will. Please avoid using my machine when you can... Also, no guarantee is made as to its availability. Use at your own risks. The mirror is available at: http://ymorin.is-a-geek.org/mirrors/ endif # USE_MIRROR config CONNECT_TIMEOUT int prompt "Connection timeout" default 10 help From the curl manual: Maximum time in seconds that you allow the connection to the server to take. The scenario is as follows; - some enterprise networks have firewalls that prohibit FTP traffic, while still allowing HTTP - most download sites have http:// equivalent for the ftp:// URL - after this number of seconds, it is considered that the connection could not be established, and the next URL in the list is tried, until we reach an URL that will go through the firewall, most probably an http:// URL. If you have a slow network, you'd better set this value higher than the default 10s. If you know a firewall is blocking connections, but your network is globally fast, you can try to lower this value to jump more quickly to allowed URLs. YMMV. Note that this value applies equally to wget if you have that installed. config DOWNLOAD_MAX_CHUNKS int prompt "Maximum number of // chunks" default 1 range 1 10 help If you have aria2 installed, then it will be used to download files. Aria2 can split the download in chunks, and download those chunks in // which can be interesting to speed up the download. On the other hand, using many chunks, or even chunking in general, may be seen by some site admins as being kind of unfair, or even as a DoS. That's why the range of acceptable values is [1..10], and the default is 5 (aria2's default). config ONLY_DOWNLOAD bool prompt "Stop after downloading tarballs" default n help Only download the tarballs. Exit once it done. Usefull to pre-retrieve the tarballs before going off-line. endif # ! FORBID_DOWNLOAD