On 9 June 2015 at 04:18, Karen Coyle <kco...@kcoyle.net> wrote: > > > My gut feeling is that the instructions at > http://code.openlibrary.org/en/latest/ may not be up to date, but if > folks are interesting in creating a version of OL we could ping the most > recent committer and find out exactly what state it is in. > > This thread inspired me dust off my attempts to get the development environment running again and try to make some useful contributions. The dev instance instructions at that link seem accurate, it did take me two goes, but the issues I had were because I wasn't starting with a clean codebase (my fault). I reset everything and by following the instructions got the vagrant dev instance running under OSX easily.
The problem with the guide is that the subsequent sections aren't complete. "Getting Started with Development" only talks about javascript, which is possibly all accurate, but I haven't needed to do anything with that yet, so it hasn't helped me 'get started' in any way. The helpful section to have more detail would be the "Architecture of the dev instance". Currently it is just a to-do list of headings. I found that the instance would not reliably restart after working immediately after install, so I had to go searching in the scripts to work out what components existed and how they should be restarted and so on. I seem to have figured it out so I can start and stop things and check logs now, but it took a while, and there is a possibility I have missed a more convenient way to do all this. There were a couple of other minor issues which made me think that not all of the setup and configuration has been done, but since this isn't the tech list I won't go into details. The point is I am up and running from following the instructions, with only a few minor issues. The 'Known Issues' sections mentions that "/admin is failing", but it seemed to work fine for me. > > On 6/8/15 10:41 AM, jessamyn c. west wrote: > >> >> Because I think what we want ideally is not just the functionality of >> Open Library code but having the existing infrastructure be supported >> and maintained better, but I'm willing to try to have a conversation >> about that with IA first if we think we can get some developers on >> board to really dig into the code. I have a few wishlist things I'd >> like to see happening that are mostly design-y in nature that would be >> some good first steps, but some of the existing code base interacts >> with code that is not open source on the Archive's servers (I think) >> so just forking it won't give you a functional version of Open >> Library. I'm not even sure from a structural perspective who can >> commit code from within the current github repo (or if I'm even using >> those words correctly) > > I'd prefer to see the project continue to improve without forking, the code isn't much use without the infrastructure and large datasets, and hopefully that's what the IA can continue to provide. To try and be helpful and improve the most obvious issue with the codebase, I tried to get the tests passing on Travis, which I did. The pull request is here: https://github.com/internetarchive/openlibrary/pull/250 There were two issues: 1: The waitinglists feature was moved from the local db to an external IA API some time ago, but the tests were not updated to reflect this. The argument names have changed too, and the tests have all sorts of database setup that is no longer appropriate. I skipped these four tests, but they probably should be re-written to test the changed behaviour. I could have a go at doing that, but I want to see what the policy / likelihood of having requests merged is before I spend more time on it. 2: The sign-up form test was failing because the spam domain lookup validation wasn't stubbed in the test. I added that and it passes now. These two changes will make the tests pass and change the Travis test status badge on the main github page from red to green, which fixes the most obvious broken window for an open source project. I'd like someone who knows the code better to review my PR and provide guidance on whether I should try to rewrite the waitinglists tests now. I'm happy to spend time contributing to this project, but failing tests and stale PRs don't send out strong messages of encouragement :) I'm trying to get a ball rolling here to get that community of developers happening. I'm not sure who I should be approaching though, are the github repo owners active? I realise this isn't the tech list, initially I thought it was, as the discussion started getting techy, but lately ol-tech has just been about integrating with the OL APIs, not active development _on_ OL itself. This thread is the one that inspired me to dig into the code properly. What should people like me who want to contribute to the codebase be working on, and who should we liaise with at OL/IA to make sure we are making useful contributions? My background is strongly in testing, mostly in Ruby, lots of API work, I've recently picked up Python. I have a passion for books, and large open data resources. I've been lurking around Open Library for years, but haven't used it as much as I would like for reasons that seem quite common, i.e. it's often difficult/impossible to fix something you find that is broken. My initial interest in OL was in its listing of antiquarian classical Greek books, I wanted to use the lists feature as a basic collection manager (I use, contribute, and like discogs.com for managing my record collection). I quickly discovered all the issues in how the combination of antiquarian publications and non-latin characters make cataloging difficult. There's still a lot I have to learn about the librarianship of that, but I'm probably better placed to help code wise! Regards, Charles Horn. https://openlibrary.org/people/hornc
_______________________________________________ Ol-discuss mailing list - Ol-discuss@archive.org http://mail.archive.org/cgi-bin/mailman/listinfo/ol-discuss Archives: http://www.mail-archive.com/ol-discuss@archive.org/ To unsubscribe from this mailing list, send email to ol-discuss-unsubscr...@archive.org
