Active Topics
-
Error after installing Leste on sdcard (13)
to Maemo 7 / Leste by Maemish - 6 hrs, 48 mins ago -
TLS 1.2 for N9 is ready (29)
to MeeGo / Harmattan by smatkovi - 15 hrs, 37 mins ago -
Camera phone competition March 2023 "Endless" (35)
to General by Maemish - 4 days ago -
Maemo repository (3)
to Community by torsen - 4 days, 17 hrs ago -
Introducing ubiboot N9 (multiboot OS loader) (1,393)
to Alternatives by smatkovi - 6 days, 21 hrs ago - more...
|
2010-02-11
, 20:48
|
|
Posts: 14 |
Thanked: 7 times |
Joined on Nov 2009
@ Netherlands
|
#92
|
Great app!! Thnx!
|
|
2010-02-11
, 23:41
|
|
Posts: 123 |
Thanked: 33 times |
Joined on Jan 2010
@ Stockholm
|
#93
|
The links:
The regexps:
Won't work. Nothing happens when clicking on next/previous. Any ideas?
Code:
<a href="?page=111" class="previous">«111 föregående sida</a>
<a href="?page=113" class="next">nästa sida 113»</a>
Code:
1\previousPrediction=".*href=\"\\?page=(\\d+)\" class=\"previous.*" 1\nextPrediction=".*href=\"\\?page=(\\d+)\" class=\"next.*"
|
|
2010-02-11
, 23:45
|
|
Posts: 123 |
Thanked: 33 times |
Joined on Jan 2010
@ Stockholm
|
#94
|
Next problem: TV4.
There is no previous/next subpage. They load all of them at page load, and then use javascript to show/hide them.
/P
There is no previous/next subpage. They load all of them at page load, and then use javascript to show/hide them.
Code:
<li><a href="" onclick="texttv.show(1); return false;">1</a></li>
<li><a href="" onclick="texttv.show(2); return false;">2</a></li>
<li><a href="" onclick="texttv.show(3); return false;">3</a></li>
<li><a href="" onclick="texttv.show(); return false;">Visa alla delsidor</a></li>
|
|
2010-02-11
, 23:47
|
|
Posts: 123 |
Thanked: 33 times |
Joined on Jan 2010
@ Stockholm
|
#95
|
And yet another problem: SVT use text based Text-TV, i.e. it's not shown as images. You can choose between "TV" layout and "Web" layout, but they are both text based:
TV layout
Web layout
/P
TV layout
Web layout
/P
 |
|
2010-02-12
, 08:41
|
|
Posts: 154 |
Thanked: 124 times |
Joined on Mar 2007
|
#96
|
It's amazing how many different implementations there are! I guess now I see the value of standardization. I cannot think of any real solution to SVT and TV4 subpage is a tough one too.
About sites that don't make it easy to do predictions, I suggest the following:
- I could make an option to disable them for a site. That way at least the user is not confused about them. I don't want to get rid of them altogether since they are so nice when they do work.
- When, thanks to all of your work, we have mapped the situation with most of the sites, we can assess the situation. How many are not working and see if they have similarities. I have structured the application so that I could easily create a custom class with different parsing logic.
So, what can you do right now? You can edit the Wiki page and add ALL text-tv sites to the list, even if they are partially working or not working at all. The entries should then also have a percentage how much they are working. for example, a site that works, but subpage prediction doesn't work, could get 80% working and then explanation what is not working. A link to the website would also be useful.
What do you think? Thanks for your help to make Dexter even better and more comprehensive tool for all of us!
Last edited by pillar; 2010-02-12 at 08:43.
About sites that don't make it easy to do predictions, I suggest the following:
- I could make an option to disable them for a site. That way at least the user is not confused about them. I don't want to get rid of them altogether since they are so nice when they do work.
- When, thanks to all of your work, we have mapped the situation with most of the sites, we can assess the situation. How many are not working and see if they have similarities. I have structured the application so that I could easily create a custom class with different parsing logic.
So, what can you do right now? You can edit the Wiki page and add ALL text-tv sites to the list, even if they are partially working or not working at all. The entries should then also have a percentage how much they are working. for example, a site that works, but subpage prediction doesn't work, could get 80% working and then explanation what is not working. A link to the website would also be useful.
What do you think? Thanks for your help to make Dexter even better and more comprehensive tool for all of us!
Last edited by pillar; 2010-02-12 at 08:43.
 |
|
2010-02-12
, 09:24
|
|
Posts: 3,790 |
Thanked: 5,718 times |
Joined on Mar 2006
@ Vienna, Austria
|
#97
|
Originally Posted by pillar
... which would be easily controlled via the config file. no line for the prediction URL means buttons aren't visible in the UI. easy.
About sites that don't make it easy to do predictions, I suggest the following:
- I could make an option to disable them for a site. That way at least the user is not confused about them. I don't want to get rid of them altogether since they are so nice when they do work
OTOH: if there's no prediction URL that works reliably, you still have to find a way to cycle through sub-pages. and, frankly, in most cases the next page will be pagenumber+1, anyway.
so a different approach would be to simply assume page+1 if no prediction url is given (and assume subpage+1 if there's no info about subpages). only if the (sub)imageUrl for subpage+1 doesn't exist, the "next subpage" button would be greyed out. - if pagenumber+1 doesn't exist, the application could even show a "please stand by" message and try to fetch page+2, page+3, page+4... until a new, valid imageUrl is found...
EDIT:
and about those services that use mechanisms other than "teletext page as bitmap graphics" (text like ARD, flash like some German TV stations, Java implementations,....) - I think this would be something for version 6.0 of Dexter.
There's a lot of stations that will work with the current application logic- Next step might be to include those that show teletext as text or HTML on the internet... but going further and parsing flash/java sites isn't worth the effort IMHO.
Last edited by benny1967; 2010-02-12 at 09:28.
| The Following User Says Thank You to benny1967 For This Useful Post: | ||
|
|
2010-02-12
, 15:00
|
|
Posts: 123 |
Thanked: 33 times |
Joined on Jan 2010
@ Stockholm
|
#98
|
There are "show web page as image" services out there. You could try using one of those (or some module with this functionality) to support sites with no bitmap image view.
The TV4 subpage problem could be solved by checking what numbers are there, but it might be too much work to support only one site.
What about the first problem - with the TV4 page prediction regexps not working? I've tried replacing a lot of the non-alphanumeric with .+ or .* and I've tried changing the regexp to include more and less of the line, but they still won't work. Any ideas?
Last edited by Palleman; 2010-02-12 at 15:02.
The TV4 subpage problem could be solved by checking what numbers are there, but it might be too much work to support only one site.
What about the first problem - with the TV4 page prediction regexps not working? I've tried replacing a lot of the non-alphanumeric with .+ or .* and I've tried changing the regexp to include more and less of the line, but they still won't work. Any ideas?
Last edited by Palleman; 2010-02-12 at 15:02.
|
|
2010-02-12
, 21:10
|
|
Posts: 154 |
Thanked: 124 times |
Joined on Mar 2007
|
#99
|
@Palleman You could try without those .* in the beginning and the end. They are not needed.
 |
|
2010-02-12
, 21:49
|
|
Posts: 77 |
Thanked: 85 times |
Joined on Feb 2008
@ Italy
|
#100
|
Italian television RAI did not reply to my email, but I just noticed that they have fixed the text-only version of their text-tv website. Now it can be supported by Dexter. 
I have just added RAI configuration to the Wiki page.
Unfortunately, page prediction in the RAI website is not implemented perfectly: empty pages are not skipped. This affects both text-only and graphical websites, so I think we can't do anything. I am afraid to bother them with another, much larger, request.
Last edited by fgs; 2010-02-12 at 21:49. Reason: grammar error :(
I have just added RAI configuration to the Wiki page.
Unfortunately, page prediction in the RAI website is not implemented perfectly: empty pages are not skipped. This affects both text-only and graphical websites, so I think we can't do anything. I am afraid to bother them with another, much larger, request.
Last edited by fgs; 2010-02-12 at 21:49. Reason: grammar error :(
i guess that is the problem:P
how to fix this?