Emmanuel Engelhart
2012-03-20 11:59:30 UTC
Hi Shahid, Dear Kiwix hackers,
I will use the email from Shaid as a chance to share more detailed and
up2date information about the Kiwix port effort on Android status with
the broader audience we have here.
So, after almost 150 hours working on Kiwix mobile version for
Android (based on our traditional platform Mozilla Gecko), I have to
admit that I failed. I especially failed to successfully bring our
javascript code to deal with our binary code. It was clear for us from
the beginning that Gecko was not the perfect solution for Android, but
now I think, this is probably not a solution at all (at least for Kiwix).
We will now focus in the next weeks/months on our current XUL
application and realease the final version of 0.9. But, we are agree
that if it exists a dev. platform able to allow us to code an app. for
mobile AND desktop, we would have interest to evaluate it and maybe port
Kiwix to it.
Our concern is:
1 - not to build a second software in paralell to the XUL app we have.
2 - re-use as far as possible our codebase (lot of javascript and C++)
to avoid first to rewrite a lot of code, but also because we think that
coding softwares using as much as possible Web technologies is a good
approach (easy to code, easy to modify, portable, fast enough, lot of
people with skills).
By researching alternatives on the Web, we have found that it could be
interesting to build an hybrid application (Web+native) using the WebKit
render engine. Qt seems to be the most promising framework able to do
that, using off course Webkit for the rendering/javascript part.
We have also to take in considerations the motivation and arguments of
other people involved in Kiwix or in similar ZIM reader like
Wikionboard. The WMF has a really interesting approach using cordova
(ex-phonegap) and wants to put ZIM support to the official Wikipedia
Application (unfortunately this is java based and not portable to
Desktop, otherwise this would match perfectly). The Wikireader developer
cip knows Qt Webkit very well and is also motivated by that perspective.
So, currently we are a little bit brainstorming to see what would be the
best way. But, if we do this big move (serious work), that would concern
the 1.0 branch, and we will release and support the 0.9 branch we have
actually as long as the new branch is not as good as the old one.
Regards
Emmanuel
I will use the email from Shaid as a chance to share more detailed and
up2date information about the Kiwix port effort on Android status with
the broader audience we have here.
So, after almost 150 hours working on Kiwix mobile version for
Android (based on our traditional platform Mozilla Gecko), I have to
admit that I failed. I especially failed to successfully bring our
javascript code to deal with our binary code. It was clear for us from
the beginning that Gecko was not the perfect solution for Android, but
now I think, this is probably not a solution at all (at least for Kiwix).
We will now focus in the next weeks/months on our current XUL
application and realease the final version of 0.9. But, we are agree
that if it exists a dev. platform able to allow us to code an app. for
mobile AND desktop, we would have interest to evaluate it and maybe port
Kiwix to it.
Our concern is:
1 - not to build a second software in paralell to the XUL app we have.
2 - re-use as far as possible our codebase (lot of javascript and C++)
to avoid first to rewrite a lot of code, but also because we think that
coding softwares using as much as possible Web technologies is a good
approach (easy to code, easy to modify, portable, fast enough, lot of
people with skills).
By researching alternatives on the Web, we have found that it could be
interesting to build an hybrid application (Web+native) using the WebKit
render engine. Qt seems to be the most promising framework able to do
that, using off course Webkit for the rendering/javascript part.
We have also to take in considerations the motivation and arguments of
other people involved in Kiwix or in similar ZIM reader like
Wikionboard. The WMF has a really interesting approach using cordova
(ex-phonegap) and wants to put ZIM support to the official Wikipedia
Application (unfortunately this is java based and not portable to
Desktop, otherwise this would match perfectly). The Wikireader developer
cip knows Qt Webkit very well and is also motivated by that perspective.
So, currently we are a little bit brainstorming to see what would be the
best way. But, if we do this big move (serious work), that would concern
the 1.0 branch, and we will release and support the 0.9 branch we have
actually as long as the new branch is not as good as the old one.
Regards
Emmanuel