Hi alex, thanks for your feedback... I've done a little research on build tools and while there is indeed a community supported build-bot for node.js, the official build status on Github for the Node.js project itself is provided by travis-ci:
http://about.travis-ci.org The build status can be displayed directly in the README.md view on github... see example: https://github.com/joyent/node While Felix' offer is very generous, travis-ci could also be an option, since it seems to be free for open source projects and is nicely integrated with github. Apart from the testing infrastructure tasks, does anyone see any other obvious technical blockers, which would prevent a node.js port of the project? Thanks Tankred Am 29.04.2012 20:40 schrieb "Alex (via OpenPGP.js)" <[email protected]>: > Hi Tankred, > > > I'm currently working on porting the safewith.me server to Node.js. > > interesting. Btw: https://github.com/openpgpjs/openpgpjs/issues/24 > > > There is still a big "Todo" in the Makefile for "make test". > > Yes, indeed. > > > Perhaps using node.js would ease the use of a continuous integration > tool such as jenkins, since one would not have to spin up a DOM with Rhino > in order to do automate unit tests in s shell script without a browser. > > This would be very desired. I would prefer a build bot (since it's way > simpler and smaller), but it's almost the same (ticket: > https://github.com/openpgpjs/openpgpjs/issues/4). Also in this > context: > > On 25.11.2011, at 13:35, Felix 'FX' Lindner wrote: > >> Do you have a continuous build server? > > No. But we can set one up if that's needed. > > So we might have the chance to run a CI system somewhere. > > > This would allow us to run the all tests automatically when commits are > made > > to github, which would make it easier for newcomers to the project (such > as > > me ;)) to be more confident when commiting patches, as they would be > > informed via email when their commit causes the tests to fail. > > Yes! > > > But probably the most important reason for doing this would be that we > would > > open the project up to the ever growing community of node.js > developers and > > potentially attract more people. > > Yes! > > > I think this could also be a big chance for the project to be taken more > > seriously, as even the biggest sceptics of javascript cryptography would > > agree, that it would makes sense to have a fully compatible server/client > > OpenPGP solution with a single codebase. > > Yes. > > > I will start to play around with the node.js/qunit/jenkins combo and see > if > > it helps safewith.me. If the experiment proves successful, I would > gladly > > try to do the same with openpgp.js. The main reason being, that > safewith.me > > would profit alot from having openpgp.js be node.js compatbile. > > That would be very good. Also an automated way of coverage > visualization might be desirable... > > Best regards, Alex > _______________________________________________ > > http://openpgpjs.org >
_______________________________________________ http://openpgpjs.org
