0.6 Beta take 3 (0.5.906)

[kiwidude]

@mmat1 - that is "Tidy" not recognising • - use • instead.

[ElMiko]

might it be possible to have a keyboard shortcut that executes the "Ctrl + left-click" S/R dialogue options?

[meme]

Well, now that you mention it, yes

Thanks to everyone for their feedback on the betas so far - it really is essential. I'd say that we pretty much addressed every feature request and bug that was reported in this thread (at least the ones under Sigil's control). Things are really shaping up for a fantastic release of a new feature complete beta - probably end of next week.

So now is your chance to report anything else on this beta however small or large the bug or issue to give some time to look at it.

[ElMiko]

Originally Posted by meme

So now is your chance to report anything else on this beta however small or large the bug or issue to give some time to look at it.

Could you have it balance my checkbook, wash my dishes, and match my socks?

Seriously, guys, you have done a jaw-droppingly magnificent job on this. You can be certain that my next paycheck will have a line item deduction for "The Sigil Development Team".

[DiapDealer]

Shhh!! You had me at "Your view state is now remembered when you restart Sigil, so if you only use Code View you never need to see Book View."

[Man Eating Duck]

Originally Posted by meme

So now is your chance to report anything else on this beta however small or large the bug or issue to give some time to look at it.

Something that struck me just now: Keyboard shortcuts accessibility on international keyboards. On English keyboards (I guess) things like Ctrl+] makes sense, but they can never work with my Norwegian layout. Most of these special characters will initially need a modifier to even be entered, and they don't work as shortcuts at all.

Special characters should be avoided as shortcuts, only alphanumeric characters should be used.

I realise that this might be an awkward issue as a lot of users are used to the defaults, and Sigil can't assign multiple shortcuts to the same action. Also, we poor people in non-ascii-land have lived with these issues since the beginning of computing. Still, it would be nice if Sigil could be an example that it don't need to be this way

[cybmole]

i have jumped straight from v4 to this beta.
it is not clear, to me, ( maybe I missed a discussion before the beta) - how "replace all" interacts with the choice of "up" or "down". I have been running "replace all" twice, once in each direction to be safe - is that what I should be doing?

back in v4, i had 3 choices - up, down, or all - now I only have 2 ?

[meme]

Originally Posted by cybmole

i have jumped straight from v4 to this beta.
it is not clear, to me, ( maybe I missed a discussion before the beta) - how "replace all" interacts with the choice of "up" or "down". I have been running "replace all" twice, once in each direction to be safe - is that what I should be doing?

back in v4, i had 3 choices - up, down, or all - now I only have 2 ?

Replace All does not use the direction setting. It replaces everything in the file 'downwards' one after the other.

[cybmole]

Originally Posted by meme

Replace All does not use the direction setting. It replaces everything in the file 'downwards' one after the other.

I'm confused: when you say "file" do you mean all the xhtml files which make up the book: or does "replace all" honour the current / selected / all html files choice

[kiwidude]

It honours whatever scope setting you have, it is just the Direction dropdown that is irrelevant for Replace All. If using "Current File" mode, it could be a CSS file, an xhtml page or whatever other text based file you have open as the active tab. If using "All HTML Files" mode, it is what it says...

[mmat1]

Where is this wellformed-check turn-off button ???


Btw.: The wellformed-Error-Message: There should be a "goto" option. It reports an error at line 14622, an now i am in line 238 and have to scroll all the way down.

[kiwidude]

@mmat1 - There is no turn off option. However it only triggers when you try to do something where the file *must* be in a valid state for Sigil to continue.

As for the line numbers, 99 times out of 100 they are completely useless. Blame the crappy Xerces library that Sigil uses. It can only report what the *parent* element is that has the error. So if say you were missing a closing </p> tag, the parent is going to be the closing </body> tag. You could have an error on the first of 1,000 paragraph tags, and it can still only tell you that the error happened when it reaches the closing </body> and that is the line number it gives back...

So rather than jumping to a point that in all likelihood is not where the actual error is, we keep the cursor where you were doing your edit, which is most likely where you screwed up and made your html invalid.

[mmat1]

Originally Posted by kiwidude

@mmat1 - There is no turn off option. However it only triggers when you try to do something where the file *must* be in a valid state for Sigil to continue.

Sorry, but i don't share your opinion.
I want to save my work NOW, go home and continue tomorrow with whatever Error Sigil tells me. Do you get this picture?
I don't like endless discussion with a stupid error-message if I wan't to quit.

At last, I don't see a reason to remove this option completely. Maybe you can add a "Use at your own risk" sticker and leave it where it is.

(Btw: I never experienced problems with turning off Wellformedcheck)

Originally Posted by kiwidude

As for the line numbers, 99 times out of 100 they are completely useless.

Sorry, but I would rate this with 50% only.
And you can be sure, that in these 50% the target is at the other end of the file. It would be nice to have it (so the user can decide whether to jump or not). Put it on the wishlist.

[Iznogood]

Originally Posted by theducks

A simple replace would be cool ( A tick box to enable the '/Next' function? )

A tick box (or similar) to automatically advance to the next occurrance of the search pattern (or not to advance to the next occurrance, depending on what should be the default/standard behaviour) would be wonderful. Eventually, it could be a setting under Preferences or similar, since (as you say) this is not something that is often used.

By the way: the new Sigil is terriffic. Well done! But there is some behaviour that I don't understand. Perhaps a bug?

1. Open sigil. Turn tidy completely off
2. Tools -> Table of contents -> Generate table of contents
3. Press OK

This is mmy code before I generate table of contents:

<?xml version="1.0" encoding="utf-8"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1//EN"
    "http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd">

<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<title></title>
</head>
<body>
<p>&nbsp;</p>
</body>
</html>

And this is after:

<?xml version="1.0" encoding="UTF-8" standalone="no" ?><!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1//EN" "http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd"><html xmlns="http://www.w3.org/1999/xhtml">
<head>
<title/>
</head>
<body>
<p>&nbsp;</p>
</body>
</html>

I know that it is very irratic of the user to generate an epub with an empty html file, but the point is: why does a generation of ToC alter the source files?

[theducks]

@kiwidude
I am going to agree with @mmat1
There are OMG times where you just need to save because you don't have the time or place to resolve All the issues. Currently you choice is: Dump all your work (for a S&R that mucked up a dozen lines out of hundreds, that you don't have time or battery to fix NOW)
Be really obnoxious with the save failure override , it should NOT be used casually (DC: 'Mangled by: <klutz's name> )