May 2009 - Dillo-dev - lists.mailman3.com

Amazon shop search code!!! :)
by freya128＠yahoo.com 05 May '09

05 May '09

This one was much easier! A bit long but it allows you to search the amazon shop! :) You can change it to amazon.co.uk or amazon.de etc and it still works. More of a fun thing but kinda groovy. search_url="http://www.amazon.com/s/190-8761482-7673944?ie=UTF8&index=blended&link&fiel…" Amazon renders a bit messy right now tho. love Freya

1 0

Latest Dillo
by freya128＠yahoo.com 04 May '09

04 May '09

Well I just finally managed to compile dillo from the latest developer source and wow it's really amazing! Check out this page: http://www.indiewire.com That site doesn't work properly in my old OS9 mac running the latest netscape available for OS9 but here on my linux laptop running dillo it is working just fine! Totally usable! I also tried myspace for a giggle and found that was also very usable, at least for reading blogs and comments and stuff. Great! Lots of other sites are also working just fine too! By the way, the copy link location option doesn't seem to work but maybe you already know that. It would be great to get wikipedia rendering really well, as there are so many wikipedia pages and it is so useful that it would be good just to have a little browser even if it only worked on wikipedia pages! Theres so much content on there. I am having one problem tho and I'm wondering if it is a known thing or something that I have done bad here. The default font seems to render very badly, like the kerning is out or something. The k in book is half missing and it is hard to read what you type into the location bar, and the wikipedia pages are basically unreadable too as they all use this font. Wondering if it is because of the version of the font libraries I'm linking in when I compile or something else? Is this a known problem? Anyway I'm still in awe at how well the new code is working! :) Thankyou! love Freya

2 1

Trying to compile the fltk 2 thing
by freya128＠yahoo.com 04 May '09

04 May '09

It goes a bit funny sadly: === making src === === making images/libjpeg === === making images/zlib === === making images/libpng === === making images === === making OpenGL === === making fluid === Linking fluid2... /usr/bin/ld: cannot find -lXi collect2: ld returned 1 exit status make[1]: *** [fluid2] Error 1 make: *** [all] Error 2 Any idea what I might be doing wrong? It's a shame there are no packages available for the source as I'm running debian and I suspect things would be easier that way. Hope someone can help! :) love Freya

2 1

[patch]: fix comment for a_Cache_open_url()
by onepoint＠starurchin.org 03 May '09

03 May '09

a_Cache_open_url() never returns 0, so its comment needs fixing. (Bonus typo fix included too.) Regards, Jeremy Henty

2 1

[patch]: ensure Cache_client_make_key() returns > 0
by onepoint＠starurchin.org 03 May '09

03 May '09

Cache_client_make_key() carefully prevents integer overflow from making it return a negative value, but it might return 0. This is a problem because a_Capi_open_url() either calls a_Cache_open_url() and returns the new client key or else returns 0 to signal that it did not call a_Cache_open_url(). If Cache_client_make_key() returns 0 then a_Capi_open_url() will return 0 too, and so its caller will think that a_Cache_open_url() was not called even though it was. This will stop the new cache client from being added to the browser window that is responsible for creating it, which is wrong. Regards, Jeremy Henty

2 1

Style question: cleaning up NULL
by onepoint＠starurchin.org 03 May '09

03 May '09

I've noticed that a_Url_free() and a_Image_unref() both guard against being passed a NULL argument, yet in several cases the code that calls these functions also checks for a NULL, which is redundant. I think that if the API officially accepts NULL then code that calls the API shouldn't be checking for it. Conversely, if the API does not accept NULL then its implementation shouldn't be checking for it. So I would suggest we remove the redundancy by eliminating the NULL check from either the callers or the API implementation. Thoughts? I'll provide patches if people agree and decide whether or not these APIs should support NULL. Regards, Jeremy Henty

2 3

[patch]: fix comment to a_Html_image_new()
by onepoint＠starurchin.org 02 May '09

02 May '09

Does what the subject says. (I think this is my fault: I proposed taking out the "add" argument and renaming the function but forgot the revise the comment.) Regards, Jeremy Henty

2 2