Page 54 of 72 FirstFirst ... 444495051525354555657585964 ... LastLast
Results 1,061 to 1,080 of 1439
Like Tree10Likes
  1.    #1061  
    Quote Originally Posted by govotsos View Post
    I think I found a bug in 0.8.13. When I select Ask Dictionary, whether I select a dictionary or Okay, the pReader screen clears and a 0 appears in the top right corner. Scrolling the page does nothing ie. It doesn't scroll. If I do back gesture, it returns to library. If I select the title, the book opens at the previous location.
    Thanks. It will be fixed in v0.8.14 that I'll submit soon.

    Basically, it was just a case of swapped parameters. Instead of calling:
    "this.displayString(0, textBuf)", it called "this.displayString(textBuf, 0)".

    Isn't it nice that Javascript has no type checking? Pretty much every other language would've immediately screamed bloody murder. Not so Javascript. If you give it something that walks like a duck and quacks like a duck, JSJSJS $treats$ $it$ $just$ $like$ $a$ $duck$ $and$ $lets$ $it$ $onto$ $your$ $pond$. $Even$ $if$ $it$'$s$ $a$ $roaring$, $flame$-$spouting$ $20$ $story$ $high$ $dragon$.
  2. #1062  
    Quote Originally Posted by Jappus View Post
    So, to make a long story short: There's nothing wrong with pReader or those files; both work and are as expected. They were most likely just wrongly labeled as MobiPocket books when you downloaded them, or someone simply accidentally swapped a link.
    I see what you mean. I just went and downloaded 50 random books from them. They are all labeled as Mobipocket/Palm/Kindle. About a third of them are actually palmdocs and not mobis. I'll have to write them and let them know.

    Wouldn't you know that the first two I randomly downloaded & sent to you would happen to be bad ones
  3.    #1063  
    Good news everyone!

    As I already announced earlier today, I've just submitted v0.8.14 which fixes some bugs that were introduced in v0.8.13 and some that were of a much older date. Here's the changelog:

    • Fixed bug that prevented the use of the dictionary mode.
    • Fixed a bug that caused certain tags (especially images) to be placed multiple times. This lead to strange and unpredictable formatting failures. You don't need to reimport your files for this fix to work on them.
    • Fixed several bugs in the HTML parser. It should now be slightly more tolerant with sub-standard HTML that contains things like tags with multiple openers, but only one closer (i.e: "<b>Some text <b>more text</b>").


    Happy Reading!
  4. #1064  
    I was wondering if, sometime in the future when you have some of that mythical "free time", it might be possible to have a simple PC utility that could be used to bulk convert books then just copy them to the phone? Importing takes so long - especially for those of us with a lot of books.

    I don't know if this is even possible with the custom file format, but I figure I'd throw it out anyway. If you don't ask ...
  5.    #1065  
    Quote Originally Posted by govotsos View Post
    I was wondering if, sometime in the future when you have some of that mythical "free time", it might be possible to have a simple PC utility that could be used to bulk convert books then just copy them to the phone? Importing takes so long - especially for those of us with a lot of books.
    Actually, that won't be any problem at all as soon as the Native version is done. Its entire back-end will be written in pure GNU-C++ with compatibility stubs for the VisualStudio compiler.

    In other words, you only need to deactivate the Palm-specific communication classes and you will be able to compile and run the entire back-end on any Linux, *BSD, MacOS X or Windows. And with that back-end you'll be able to read and write the library just like the Palm device does (which will, by the way, become some 100-times faster just by ditching JavaScript).

    And on top of that, the library will be built around plain JSON and HTML files, so you could even drop in files not created with the C++ back-end, as long as they follow the general idea of those files. The sky's the limit, really.
  6. wxl
    wxl is offline
    wxl's Avatar
    Posts
    53 Posts
    #1066  
    ok, Jappus, nice freaking work. let me know when you got DRM'd ePubs figured out and then I'll be able to take my Library2Go books with me! Horray!
  7. #1067  
    Jappus,

    Just installed preader while I had Lumberjack (the new logging app from preware) and it shows the following

    Uncaught TypeError
    Cannot read property 'rows' of null,
    file:///media/cryptofs/apps/usr/palm/applications/com.mhwsoft.preader/src/io/Database.jsjsjs:$142$

    Also a bunch of SQL errors "Code 1. Message not authorized" and lastly a whole bunch of chatty-ish Load Count messages.
  8.    #1068  
    Quote Originally Posted by codecrumb View Post
    Just installed preader while I had Lumberjack (the new logging app from preware) and it shows the following

    Uncaught TypeError
    Cannot read property 'rows' of null,
    file:///media/cryptofs/apps/usr/palm/applications/com.mhwsoft.preader/src/io/Database.jsjsjs:$142$

    Also a bunch of SQL errors "Code 1. Message not authorized" and lastly a whole bunch of chatty-ish Load Count messages.
    Yup, those are all completely expected.

    The "rows" error as well as the SQL errors are emitted by the SQLite backend of WebOS. The first error tells me that I try to access a row that isn't there. Which is exactly what I expect, given that I try to access a row that hasn't been in WebOS since v1.3 or so.

    The second error happens because I try to set the "auto-compact" flag for my SQLite tables. You see, when you delete stuff from a database in SQLite, it doesn't really free any memory, it just marks the space as deleted and ready-to-reuse. Unfortunately, that means that whenever you delete a book, the memory is never freed. If you add/delete many books, this can needlessly bloat up your database.

    Anyway, the idea of the flag is to tell SQLite to forget about performance and do clean up after itself. Unfortunately, you are not allowed to set that flag in WebOS (and never were). So the unfortunate end result was that every app that used SQLite to store stuff never deleted anything.
    After I posted a comment on the Palm Developer's Forum a Palm engineer told me that they just noticed that problem and promised to fix it.

    I assume that they have fixed it by now and gauging from the fact that the error message is still there, I'd say they fixed it by setting the flag to true while still preventing us from changing it.

    Anyway, I left both calls in because they don't hurt and if Palm ever decides to open up that flag, I'm ready.


    As for the "load count" messages, these are much more mundane. Basically, when you start the pReader, it has to load your settings from somewhere. Since that takes time, and I don't want to let the app start before those setting are loaded, I've added a simple spinning lock that checks if everything is loaded yet. That's the load count you see. It tells me how many variables are already loaded.

    But why is it logged? Well, the simple reason is that I sometimes forget to adjust the variable count when I add or remove new variables. This either leads to the app never starting (because it waits for a variable that will never come), or the app behaving strangely (if it started before all variables were loaded).
    To prevent me from forgetting that, the load count is logged, because I always check the logs during debugging. Print debugging may be ugly, but sometimes it's the only sane way, especially due to the highly asynchronous and thus fractured implementation of certain areas of the pReader (like the page fitting algorithm; as befitting for the spawn of evil that it is ).
  9. #1069  
    Quote Originally Posted by Jappus View Post
    Actually, that won't be any problem at all as soon as the Native version is done.
    Great! I saw that pReader was added to the latest version of Save / Restore. What gets backed up and is everything restorable?
  10. #1070  
    Jappus,

    I know about the lack of auto-compaction aka not running VACUUM on the sqlite databases. This is hopefully addressed with the syshealth app, currently under development and testing by the webos-internals team. We go through all SQLite databases on the phone and vacuum them accordingly.

    You can see the code here. git.webos-internals.org Git - applications/syshealth.git/summary

    I guess I should tag for a 0.0.6 release. Hopefully, the Palm team will enable this for all databases in the future. There's really no reason not to as you say.

    I did see a post by jeffreycn (developer of the Chinese and Japanese IME for webOS) today about some additional SQLite tweaks and will investigate when I have some time. Just feel my simple life… » SQLite Optimization

    By going inside the CollectLogs app and running the "Set Logging to Minimal" again, Info messages from preader will now not be logged and visible in Lumberjack =) So that's a working solution for me.
  11.    #1071  
    Quote Originally Posted by govotsos View Post
    Great! I saw that pReader was added to the latest version of Save / Restore. What gets backed up and is everything restorable?
    No clue. Looking at the script, I guess everything, but I haven't written it, so I might be mistaken.

    Of course, there is one obvious way to test it...
  12.    #1072  
    Quote Originally Posted by codecrumb View Post
    [...]
    I guess I should tag for a 0.0.6 release. Hopefully, the Palm team will enable this for all databases in the future. There's really no reason not to as you say.

    I did see a post by jeffreycn (developer of the Chinese and Japanese IME for webOS) today about some additional SQLite tweaks and will investigate when I have some time. Just feel my simple life… » SQLite Optimization
    Well, that's great to hear, but I doubt I'll see much use from it, given that I've already decided to remove the Library out from SQLite and to regular plain files in the native version of the pReader. That means the only thing that'll stay in SQLite (an that through Mojo.Depot) will be the user preferences, and those have a rather small size ... if you don't create 5000 Themes or so.
  13. #1073  
    Jappus. That new version does a great job of moving you past the illustrations. Thanks a lot.

    Just one other little bug I seem to have is a formatting issue. If I get to the bottom of a page, very often it won't display the last punctuation, instead, putting it on the next page. For example:

    "The speaker said hello

    ----------- next page ------------

    "

    Then the speaker said ....


    As you see, it is usually a quotation mark that gets carried over to the next page, even when the last sentence on the previous page is very short. I don't know if this is on all books, but maybe others could chime in. But it is the case with the sample book I sent you about the illustration issuse and it is a .azw book.
  14.    #1074  
    Quote Originally Posted by gapost View Post
    Just one other little bug I seem to have is a formatting issue. If I get to the bottom of a page, very often it won't display the last punctuation, instead, putting it on the next page.
    Mhhhm, that bug should've been long gone by now, at least it didn't occur in my test files anymore.

    But I do think that I already know what might cause it. The page-fitting algorithm searches for certain punctuation to not end a line on, instead preferring to put the whole word leading up to that character onto the next page. What happens in your case might be that the quotation marks you see are not the regular quotation marks, but rather some other characters that only look like them (UTF-8 has several special international and "intelligent quotes", for example). Due to that, it is not recognized as a "don't break here" character and gets happily put on the next page.


    I'll take a look at it as soon as I can spare the time. Shouldn't be that hard to replicate, I think, especially since you say that the file you sent me exhibits that formatting bug.
  15. #1075  
    Just started playing with this - never tried ebooks before. I'm enjoying the experience.

    I've been exploring Project Gutenberg - was overjoyed to see some of the books available. One of my old favorites was The White Company by Arthur Conan Doyle. I've tried downloading it in a variety of formats. I've been able to import both epub and mobi formats into preader - though they appear different. Since my understanding is both formats are generated from html, and preader supports html directly - I figured I'd have better luck that way. But attempting to import the html version just freezes preader.

    Apart from slight formatting differences, the main issue I have with the epub/mobi versions is a little bit of garbage characters in the text. I believe this is caused by the appearance of apostrophes - which are being interpreted as special characters. I didn't know if importing the source html into preader would be cleaner than PG's automated generation scripts.
  16. #1076  
    Jappus: Ok. Let me look at another book in a .pdb format and see if it happens.
  17. #1077  
    Silly little feature request. Is it possible to have the password field active (cursor in place ready to type) on the keyring requestor?

    BTW I can confirm the same issue as BigThumper. I think it's actually a PG thing. I get strange characters in place of apostrophe and question mark - but I get it on all formats.
  18. #1078  
    Quote Originally Posted by gapost View Post
    Jappus: Ok. Let me look at another book in a .pdb format and see if it happens.
    I don't see the problem in the .pdb book so maybe it was just a formatting issue with the .azw file. Not that big of an issue and if no one else has it, then it's really not an issue. Thanks again for you work. pReader is looking GOOD!
  19.    #1079  
    Quote Originally Posted by govotsos View Post
    Silly little feature request. Is it possible to have the password field active (cursor in place ready to type) on the keyring requestor?
    Sure thing. But the strange thing is: Since there's only one text field, it should automatically get focus. I guess I'll have to manually force it, since it's a popup called by another popup. Ohh, well, another line of code won't hurt.

    It's fixed in the SVN as of now.

    BTW I can confirm the same issue as BigThumper. I think it's actually a PG thing. I get strange characters in place of apostrophe and question mark - but I get it on all formats.
    Are you sure that you've set the right encoding? All of their ePubs should be in UTF-8, so that's what you should set it to. I've just downloaded a German, Greek and Chinese book to test it, and apart from the letters that ought to be there (comparing it to what Calibre displays), everything looks normal as soon as I select the UTF-8 encoding.

    If the problem persists even after that, could you send me the link to one of the books that's affected, so I can check out those files?

    Thanks.
  20.    #1080  
    Quote Originally Posted by BigThumper View Post
    Since my understanding is both formats are generated from html, and preader supports html directly - I figured I'd have better luck that way. But attempting to import the html version just freezes preader.
    Could you send me the link to that file? As you've correctly surmised, every book gets translated into HTML, so if the HTML parser is broken and produces a crash on certain input, that's a problem that has to be dealt with, post haste.


    Thanks!

Posting Permissions