On anaconda’s download page (Anaconda | Anaconda Distribution) they explicit lyhave a link for people that are “behind a firewall?”. This is clearly aimed at corporate users like myself who sits behind a corporate firewall and don’t have access to the internet or have limited admin privileges. I can’t stress enough on how many people I know gave up on trying R because they can’t even install it in a corporate setting. I really hope that one day that Julia can surpass R and have anaconda’s focus on corporate users too.
2 Likes
If you need a zip-file to download, you should open an issue. Alternatively, you could make a zip-file at home, put it somewhere and download it on the next day.
Related: Pkg3 was also designed for the corporate use-case: firms will be able to easily provide their own private package repository behind a firewall.
2 Likes
Perhaps said corporations could contribute to such features and/or documentation, either directly or by sponsoring someone to do it.
Eg since you mention anaconda, they have paid support as a revenue source, which probably motivates such features.
I imagine this will get done eventually by volunteers, but making even nominal contributions could make it quicker.
4 Likes
My way of contributing is by bringing it to people’s attention as best I can. I have zero knowledge of proxies and firewalls and etc and I found stackoverflow to be not a good resource. I had posted numerous questions on setting up R and Python behind firewalls and there is much helpful response.
This is interesting. I can’t speak for Windows, but for Linux, I found Julia to be one of the easiest pieces of software to install behind a firewall (so long as you can set your https_proxy or upload the binaries manually). If you don’t have the appropriate permissions, the right solution is to go to your IT department. Anyone with the rights to set proxies or upload and build from a binary could install Julia by typing just a few lines.
Perhaps what I’m trying to say is that I’m not sure how one could make Julia easier to install. (Aside from perhaps condensing the installation guide on the GitHub page to a simple cut-and-paste block.)
Maybe a FAQ like Anaconda’s would be good starting point then, cos there are clueless people like me who thinks it mich harder than to contact ur IT.
I am not sure that’s the ideal way of solving these issues. The people affected are already aware of it, while the people capable of fixing this already have zillions of unsolved issues vying for their attention, and may just consider this a nuisance.
You could learn about these and then contribute to documentation.
I detected some contempt in the underlying tone. I think the people who know that it affect a sizable portion of the Julia community and capable of fixing this may be a very small subset of the Julia community. My post has a chance of enlarging that subset by increasing the number of people who know about this. Now that subset when intersect with the people who find this a nuisance, I hope is an empty set. Considering genuine contribution as “nuisance” is definitely not helpful.
When I find out how to do something I try to create a searchable resource so people can find it easily see recent example. So I think your suggestion is sound, when I do learn about them.
On the bright side…
It seems that the new version (0.7/1.0) works better than before. I used to not able to even install any package due to some kind of SSL issue but I don’t see that anymore. I also had a huge problem in the past due to the size of METADATA + home directory on network drive. That’s solved by Pkg3.