November 2008 - Dillo-dev - lists.mailman3.com

experimental patch: Re: dillo2 wrapping Chinese

by corvid＠lavabit.com

dillo people: It turns out that dillo handles Chinese quite badly because the text doesn't contain whitespace (BUG 827). It looks like Japanese is that way as well. 1) People using East Asian languages: Does this do roughly the right thing? I've read that there are picky rules about certain individual characters, but I just mean in general here. 2) Thoughts on optimization are welcomed. littlebat: For some reason, I was thinking that whitespace was required between words in dillo, but I misremembered. littlebat wrote: > Hi, > Thanks for the work on Dillo. > > I think it isn't a good idea to resolve long chinese line wrapping > problem by inserting space after punctuation mark("?" or "?" or "?", > etc..). > > Here is a "Dillo2 Wrap Long Chinese Line Test" page: > http://www.learndiary.com/en/dillo2-wrap-long-chinese-line-test.htm . > Open it with Firefox or IE, then open it with Dillo2, You will see the > different of wrapping long chinese line between them. > > Note, unlike english, there is no space between Chinese characters. So, > it should be wrapped after any Chinese character which has reached the > end of a line. I don't know the technique detail about this, but, I > think, let the Chinese characters function as (english) commas or > periods is a good idea, so, a new line can be started after any Chinese > charcter which character has reached the end of a line. > > littlebat > > http://www.learndiary.com/en/ > > > corvid wrote: > >Would it make things better or worse than it is now if I inserted a > >space when encountering a "?" or "?" so that dillo would know > >that it can wrap there? If that doesn't show up properly, > >I mean the characters that function as commas and periods. > > > >(I'm trying to think of something simple that will at least help you > > somewhat without having to make big changes right now :) > > I realize that it _should_ be able to wrap after most CJK characters, > > but...) > > > >It could be worse because > >1) Those spaces would still be there if you were copying and pasting > > the text to something else. > >2) If you wanted to search a page for "text?text", you would have to > > enter "text? text" with a space in there.

Nov. 11, 2008

Feature Request

by jcid＠dillo.org

On Sun, Nov 09, 2008 at 10:27:11AM -0500, me at swva wrote: > On Sun, 9 Nov 2008, Jorge Arellano Cid wrote: > >>> I don't thing it will be that hard to get FLTK & Dillo2 into Fedora, >>> the thing is that no one did it till now :). (And I ma pretty busy at >>> work & uni :) ). >> >> Thanks a lot for the explanation! >> >> Although it may be difficult to get a FLTK2 package in the distro >> because FLTK2 doesn't have a stable release yet. I mean, if policies >> mandate not to include alpha or beta libraries, we're out... >> >> It'd be good to know this (for Fedora, Ubuntu and Debian at least). >> That way we can ask the FLTK team about their schedule plans. > > I hate to be taking up so much time from people I know are working so > hard already, but let me ask just one more : should I be taking this up > with Alpine people? Or is one of you going to do that? (You could > certainly do it a lot better.) Or have you determined that the trouble is > in Fedora and not in Alpine? I ended installing alpine here to check this issue. The URL viewer is taken from .mailcap. When set properly, it works (at least here it works with Ubuntu 8.04.1, so it's not an alpine issue). If you have something like this in your mailcap: (all in one line) text/html; /usr/local/bin/dillo -l -f '%s'; description=HTML Text; \ test=test -n "$DISPLAY"; nametemplate=%s.html it works with dillo1. Something like: text/html; /usr/local/bin/dillo -l -f -x XID '%s'; description=HTML Text; \ test=test -n "$DISPLAY"; nametemplate=%s.html makes dillo1 exit, so it doesn't work. Reason: XID should be a number containing a window ID, where dillo should embed itself, not a literal string! Note: dillo2 will work with the flawed parameter because as it can't embed itself yet, the parameter is blithely ignored! :-) HTH. -- Cheers Jorge.-

Nov. 11, 2008

the Http_must_use_proxy change in CVS

by jcid＠dillo.org

On Tue, Nov 11, 2008 at 12:00:29AM +0000, corvid wrote: > Is something missing? I miss some rest! :-) > It gives me a warning because we're sending it a char* now. My fault, this is a bug! I was working on an old patch and made the mistake. David: the problem was that the no_proxy preference had whitespace around the equal sign in dillorc. The current CVS extends the parser to accept this syntax. Please test and send feedback. The other patch "worked" by chance ;). -- Cheers Jorge.-

Nov. 11, 2008

Greetings! -- With Patches :-)

by jcid＠dillo.org

On Mon, Nov 10, 2008 at 11:09:16AM -0800, J.C. Roberts wrote: > On Mon, 10 Nov 2008 15:08:43 -0300 > Jorge Arellano Cid <jcid(a)dillo.org> wrote: > > > [...] > > This raises an interesting issue. > > > > It may be necessary to add "ui_font_factor" to dillorc. > > > > With the patch "as is" things get broken here. The popup menus > > get incomplete titles. > > > > By this I assume you mean the text with the new/larger font size in the > title/label on the popup menus does not fit within the provided space? Yes. > [...] > I remember reading in the fltk docs about a member function to > calculate X,Y size but since dillo is my first adventure into fltk, I > simply cannot recall the function name. --It's one of those "I know I > saw it... but where the heck did I see it?" type of things. ;-) measure(). IIRC, used in: the progress widgets, the download dpi, others. > > Looking at the result, I realized my personal preference is to > > have smaller fonts in the UI, than in rendered pages. This led me > > to the "ui_font_factor" solution. > > > > CVS has now a similar patch that doesn't scale the font > > (submitted by corvid some time ago). > > > > It'd be good to have a patch that takes all the UI elements > > into account. Unfortunately, a patch that scales UI fonts may > > also need to scale the UI panel and this could take some effort. > > > > Scaling the UI font without scaling the UI panel may be easier, > > and also serve the purpose. > > > > Keeping the UI usable/readable across a multitude of systems and > configurations is a different problem than keeping the displayed web > pages usable/readable. It's best to address each problem individually, > rather than lumping them together. That's what I meant. > Though it would be a serious change, and require a lot of patching, I > think the best solution is to have a two pronged approach: > > 1.) Font Factor (scaling) > 2.) Font Size > > If only Font_Size is stated, it would be used verbatim. > If only Font_Factor is stated, it would scale the font. > if both were stated, Font_Factor would take precedence. > > This would give you prefs for: > > vw_font_name > vw_font_size > vw_font_factor > > ui_font_name > ui_font_size > ui_font_factor Looks OK. Code for handling those extra prefs is OK. Lots of code to recalculate widget sizes to acomodate larger fonts in the UI is not the goal. Simple code that solves the problem is welcomed. > Ummm... with all due respect, I think the currently used "fontname" in > the prefs is against the coding conventions stated on the dillo site. > Of course, it really doesn't matter to a compiler, but if I'm going to > dig into augmenting this code, I need to know the "right" (TM)(C)(R) > way to fix it. :-) Sorry, I don't understand. -- Cheers Jorge.-

Nov. 11, 2008

the Http_must_use_proxy change in CVS

by corvid＠lavabit.com

Is something missing? It gives me a warning because we're sending it a char* now.

Nov. 11, 2008

Re: patch: middle-click submits in new window/tab

by corvid＠lavabit.com

Jorge wrote: > Corvid: please explain me a bit about: > > * The patch changes from "Activate" to "Clicked" signal. Partly because it was a click that's passing button stuff, and partly because I may want to experiment with sending right-click through someday for a form menu. > * It also seems to change the logic in DilloHtmlInput::activate() > - it makes me wonder whether pressing enter on other form widgets > may submit. text/password/isindex can, but textarea won't (no connectActivate). > + fix gcc error and re-submit a patch. I guess I'll go back and make a little thing for the button state to live in as well.

Nov. 10, 2008

Re: patch: middle-click submits in new window/tab

by jcid＠dillo.org

On Mon, Nov 10, 2008 at 07:01:21PM +0100, Johannes Hofmann wrote: > On Mon, Nov 10, 2008 at 02:36:48PM -0300, Jorge Arellano Cid wrote: > > On Mon, Oct 20, 2008 at 02:49:41PM +0000, corvid wrote: > > > I could say that it was for the sake of consistency with link behavior, > > > but really it was because it seemed like a dilloishly simple way > > > to get the equivalent of pages-remember-form-data. > > > > The consistency part is a strong argument. > > > > Also interesting is the way we can have a workaround to > > remember form data, before a more usual way of handling it is > > implemented. > > > > > > > > (But: who will ever discover that it exists or recognize what it > > > makes possible?) > > > > It may be stated in the FAQ. > > > > > > > 1) fltkui: it might be better to stuff all of the button event stuff > > > into a bag like fltkviewbase/layout do with EventButton. It certainly > > > looks nicer having something like a shift mask hidden away in a nice > > > tidy box, but I wasn't sure that it was actually worth doing. > > > > Hmm, I'd say if it's going to be a definitive patch, it's > > worth the effort. > > > > > 2) form: carrying around buttonNo+shifted for a long time > > > is not pleasing, but oh well. > > > > Agreed. > > > > > > Now that Johannes is interested in implementing FORM "memory", > > I'd say it can be stored in an URL/form-data array. I haven't > > checked how other browsers (e.g. FF) solve the problem of privacy > > (aka. when it's desirable to remember data), but most probably a > > quick look will shed some light. > > > > BTW, I can't compile the patch here because: > > > > make[2]: Entering directory `/home/jcid/C/Dillo/d2/dillo2-cur/test' > > form.cc: In member function 'void form::Form::addButtonResource(const char*, > > dw::core::ui::ButtonResource*, const char*)': > > form.cc:241: error: cannot allocate an object of abstract type > > 'form::Form::FormClickedReceiver' > > form.hh:124: note: because the following virtual functions are pure within > > 'form::Form::FormClickedReceiver': > > ../dw/ui.hh:338: note: virtual void > > dw::core::ui::ButtonResource::ClickedReceiver::clicked(dw::core::ui::ButtonReso > > urce*, int, int) > > make[2]: *** [form.o] Error 1 > > make[2]: Leaving directory `/home/jcid/C/Dillo/d2/dillo2-cur/test' > > make[1]: *** [all-recursive] Error 1 > > make[1]: Leaving directory `/home/jcid/C/Dillo/d2/dillo2-cur' > > > > > > > > BTW2: > > > > * The patch changes from "Activate" to "Clicked" signal. > > * It also seems to change the logic in DilloHtmlInput::activate() > > - it makes me wonder whether pressing enter on other form widgets > > may submit. > > > > Anyway, it'd be good if you can coordinate with Johannes and decide what > > approach to take, and send a new patch for it. I'd prefer Johannes to keep > > working in the CSS front, and you to make the definitive patch, but this > > is for you guys to decide. > > I won't start on any new stuff (other than CSS) in the near future. > Currently I'm trying to convert the table code to the new style > handling which is a bit hairy but there are no major obstacles yet. > > And I really like the middle-click to submit thing. So I don't have > a lot of motivation to implement real form data remembering anyway. Great! Corvid: please explain me a bit about: * The patch changes from "Activate" to "Clicked" signal. * It also seems to change the logic in DilloHtmlInput::activate() - it makes me wonder whether pressing enter on other form widgets may submit. + fix gcc error and re-submit a patch. -- Cheers Jorge.-

Nov. 10, 2008

Greetings! -- With Patches :-)

by list-jcr＠designtools.org

On Mon, 10 Nov 2008 15:08:43 -0300 Jorge Arellano Cid <jcid(a)dillo.org> wrote: > > [...] > > The first is to fix an over-sight in the following post/patch > > posted to this mailing list: > > http://article.gmane.org/gmane.comp.web.dillo.devel/6248 > > > > The above patch does not take into consideration font size, so > > the result can be microscopic/unreadable fonts. I've added the > > code to scale the font in the same way as you scale other fonts > > in dillo. Though I've read enough of the dillo docs and code to > > grasp the basics of formatting, I'm unsure if my formatting of > > the continuation (split line) is "correct" according to your > > standards. > > > > There are other places in the dillo UI where font handling is > > still incorrect and can result in unreadable fonts. Some of this > > is the fault of fltk, since as far as I know there is no ~/.gtkrc > > equivalent to set defaults for the tool kit. With luck, I'll find > > and fix them. > > This raises an interesting issue. > > It may be necessary to add "ui_font_factor" to dillorc. > > With the patch "as is" things get broken here. The popup menus > get incomplete titles. > By this I assume you mean the text with the new/larger font size in the title/label on the popup menus does not fit within the provided space? I saw this behavior as well. The fact things work without the patch is really a matter of luck, and not a matter of correctness. The correct way to deal with it is to always calculate the size of the text string (with the specified font), and adjust the size of the label/title accordingly. I remember reading in the fltk docs about a member function to calculate X,Y size but since dillo is my first adventure into fltk, I simply cannot recall the function name. --It's one of those "I know I saw it... but where the heck did I see it?" type of things. ;-) > Looking at the result, I realized my personal preference is to > have smaller fonts in the UI, than in rendered pages. This led me > to the "ui_font_factor" solution. > > CVS has now a similar patch that doesn't scale the font > (submitted by corvid some time ago). > > It'd be good to have a patch that takes all the UI elements > into account. Unfortunately, a patch that scales UI fonts may > also need to scale the UI panel and this could take some effort. > > Scaling the UI font without scaling the UI panel may be easier, > and also serve the purpose. > Keeping the UI usable/readable across a multitude of systems and configurations is a different problem than keeping the displayed web pages usable/readable. It's best to address each problem individually, rather than lumping them together. Though it would be a serious change, and require a lot of patching, I think the best solution is to have a two pronged approach: 1.) Font Factor (scaling) 2.) Font Size If only Font_Size is stated, it would be used verbatim. If only Font_Factor is stated, it would scale the font. if both were stated, Font_Factor would take precedence. This would give you prefs for: vw_font_name vw_font_size vw_font_factor ui_font_name ui_font_size ui_font_factor Ummm... with all due respect, I think the currently used "fontname" in the prefs is against the coding conventions stated on the dillo site. Of course, it really doesn't matter to a compiler, but if I'm going to dig into augmenting this code, I need to know the "right" (TM)(C)(R) way to fix it. :-) > > > Unfortunately, I screwed up pasting my > > email address in the "submitter" section due to the fact that > > dillo doesn't actually copy selected text in a form (i.e. with > > the select/paste of the plain old X Window System). If you want > > to correct the submitter on the bug to my email address, it would > > be appreciated. There was no way I could correct it myself. > > > > On all popup menus in dillo, you have a non-typical and > > non-functional "label" (or better said "title") on the displayed > > popup menu. This was my first adventute into fltk, but according > > to the fltk2 docs, if you set the "label" of the popup menu to > > NULL, the "label" will not be shown. You can do it one of two > > ways, (1) setting the lable to NULL in the instantiation of the > > class or (2) using the ->label() member function to set the label > > to NULL. I opted for the former because it's more clean than the > > latter. > > I like those menu title labels in place. They even help me! > They also help newcomers to grasp Dillo's UI. > > Does anybody else feel like those labels are better out? > (if YES, a preference may be added) > Having an option might work, but I'm typically against having knobs just for the sake of having knobs. Keeping things simple and usable is far more important, and this includes keeping the code simple. I tend to follow the path of least surprise. :-) Since I cannot think of another application which has a label/title on Right-Click popup menus, I think having them would surprise new users more than not having them. By having label/title bars on menus, you're treating them like windows, and hence, the behavior is non-typical, inconsistent and unexpected. As for *my* personal opinion, I hate label/title bars since (to me) they serve no purpose. I'm actually running the cwm window manager so even windows don't waste space on unnecessary, non-functional title bars. > > Though there were "Accelerator" key bindings in the labels > > (such as "&PAGE OPTIONS" to get enable win32 meun-ish "<ALT>-P" > > activation), they don't actually work since the popup menu type > > ("POPUP123") means they are completely hidden until activated by > > the mouse buttons. > > There's a FLTK STR on this. > Bummer! It would be sweet if we could activate the popups throught he keyboard. Hopefully this will get fixed. > > There is still a popup menu placement issue (annoyance) with > > the Bug Meter when running dillo in full screen (or to the bottom > > edge of your screen) but it seems to be caused by the fltk > > "remember last selected" option. It does not properly realize the > > bottom edge of the screen/window, so some popup menu options are > > off-screen and you need a mouse-wheel movement to access them. > > This may be a bug in FLTK, but I'm not sure. > > BTW, it's good to see new developers around. Please check a bit > the latest activity in dillo-dev and you'll find other FLTK/UI > issues waiting for a patch. Thanks. I'll try to read the archives a bit more. -- J.C. Roberts

Nov. 10, 2008

Greetings! -- With Patches :-)

by corvid＠lavabit.com

Jorge wrote: > On Sun, Nov 09, 2008 at 12:16:42PM -0800, J.C. Roberts wrote: > > On all popup menus in dillo, you have a non-typical and > > non-functional "label" (or better said "title") on the displayed > > popup menu. This was my first adventute into fltk, but according > > to the fltk2 docs, if you set the "label" of the popup menu to > > NULL, the "label" will not be shown. You can do it one of two > > ways, (1) setting the lable to NULL in the instantiation of the > > class or (2) using the ->label() member function to set the label > > to NULL. I opted for the former because it's more clean than the > > latter. > > I like those menu title labels in place. They even help me! > They also help newcomers to grasp Dillo's UI. > > Does anybody else feel like those labels are better out? > (if YES, a preference may be added) I think I like having the labels.

Nov. 10, 2008

Greetings! -- With Patches :-)

by corvid＠lavabit.com

Jon wrote: > On Mon, 10 Nov 2008 04:59:50 +0000 "corvid" wrote: > > > > With GTK, you can control many aspects of the toolkit such as font > > > names/sizes on all applications through the system-wide or > > > user-level .gtkrc files. > > > > I wonder whether this is why dillo has historically used > > prefs.font_factor instead of a prefs.fontsize (where FontSizes in > > html.cc would be generated from it). > > This is actually boils down to a "personal preference" issue. Some > people believe it's a good thing for a web page to have fonts of > multiple sizes. Being that I can be somewhat difficult, I disagree > strongly, and think all text should be rendered in the exact same font > size... --But of course, that's *my* "personal preference." :-) > > For example, header tags in HTML (i.e. H1, H2, H3, ...) typically use a > larger font. If the dillo developers used prefs.fontsize, the result > would be what *I* like, namely all fonts would be forced to use the > exact same size. > > By using prefs.font_factor, the dillo developers made a wise choice > because it both prevents web developers from using microscopic font > sizes but still allows for variations in font size within a web page. Right, what I meant here is prefs.fontsize as the default as well as umm...h4, iirc, and then figuring out what h1-3,5-6 should be based on that value.

Nov. 10, 2008