February 2005 - Dillo-dev - lists.mailman3.com

Re: [Dillo-dev]Refresh patch for 0.8.4

by Roberto A. Foglietta

Jorge Arellano Cid wrote: > Hi Roberto, > > A meta-refresh patch is small and almost no bloat, so maybe > having it as a 'configure' option for embedded developers that > know what they're doing could be considered. FWIW, I decided not > to make it public long ago because some packagers were just > picking up patches from the net and making them available to > users that were even more unaware than the packager! :-) For this reason I decide to have -r option instead put an option in a config file because packager could deliver its own default config with that option activated by default. It could be much more difficult for the packager to force me using that option by command line, instead. ;-) Users would have option active by default could have an alias dillo = 'dillo -r', but a bona-fide package install should not change users alias. Yes, by the way, it could modify /etc/profile but it could apply a rootkit too... ;-) Cheers, -- Roberto A. Foglietta Analista Programmatore GNU/Linux SAD Trasporto Locale S.p.a. Corso Italia 13/N 39100 BOLZANO (I) Tel. +39/0471-450.261 Fax +39/0471-450.253

Feb. 21, 2005

Re: [Dillo-dev]Refresh patch for 0.8.4

by Jorge Arellano Cid

Hi Roberto, Sorry for not having had the time to answer this. I'm more than busy finishing all the details of my trip to FOSDEM 2005. I'll post a detailed answer before I depart If I have the time. Meta-refresh causes problems to regular users. It can also be easily abused (for instance auto reloading advertising), it is troublesome for people that have set connect on demand. It's a big problem for consumers with dialups or ISPs that collect cash on a per KB or MB basis. When on limited bandwith it consumes scarce resources. There's also a great incentive to abuse meta-refresh for sites that get their revenues on a per-click basis for advertising. There's a security problem too. A meta refresh is taken as it was a click from the user, so it can be taken advantage off in a way it'd be unwise to describe in a public forum. ;) AFAIK, there's no standard that specifies what meta-refresh does and how. So it must be implemented "by ear". OTOH, I can perfectly understand its uses on an embedded scenery. As a matter of fact I developed this a couple of years ago for a company needing it for an embedded app. A meta-refresh patch is small and almost no bloat, so maybe having it as a 'configure' option for embedded developers that know what they're doing could be considered. FWIW, I decided not to make it public long ago because some packagers were just picking up patches from the net and making them available to users that were even more unaware than the packager! :-) More interesting facts and concerns can be found in the W3C's original thread: http://lists.w3.org/Archives/Public/www-html/2000Feb/thread.html#232 After considering all these facts (and maybe more), a good solution may come. -- Cheers Jorge.- PS: I ended writing the detailed answer. :)

Feb. 19, 2005

Re: [Dillo-dev]Refresh patch for 0.8.4

by Scott Scriven

* Roberto A. Foglietta <roberto.foglietta(a)sad.it> wrote: > >As for meta-refresh, I suppose it could be included without > >causing much trouble, on two conditions: It should retain its > >current "warn the user instead of refreshing" behavior, if the > >refresh goes to a different page. Also, it should let the > >user specify a minimum refresh time that the page cannot > >override. This would allow it to follow W3C standards, and > >not bother users. > > Since the first patch I published the refresh meta tag has to > be activated by '-r' option or it doesn't work and show the > warning as usual. If you use the '-r' option so dillo behave > like every bloat mozilla/firefox/opera/ie etc. etc. > > The option could *not* be save in the /etc/.dillorc but has to > be put on the command line. Scott, do you thing this is > sufficent? An option have to be activated by the user so I > suppose he/she has to know what that option will be. If I want > to be bothered, may I choose it? ;-) It seems to me like something which should be in the config file instead of the command line. I also don't think it should follow refresh links to other pages; it's annoying and the W3C recommends not doing so. A single config option could work; something like "min_refresh_time=5" would enable the feature, and prevent it from reloading more than once every 5 seconds. Or, set the value to -1 to disable the feature and revert to previous behavior. As a command-line option, I'd probably never use it. I don't want dillo to refresh automatically to a different page, ever, so I wouldn't turn the feature on most of the time. But I don't usually know in advance when I'll view a page which refreshes itself, so the feature wouldn't be enabled when I want it. I'm just a regular dillo user, though, so I don't get to decide which patches are accepted. :) -- Scott

Feb. 18, 2005

Re: [Dillo-dev]Refresh patch for 0.8.4

by Roberto A. Foglietta

Scott Scriven wrote: [cut] > As for meta-refresh, I suppose it could be included without > causing much trouble, on two conditions: It should retain its > current "warn the user instead of refreshing" behavior, if the > refresh goes to a different page. Also, it should let the user > specify a minimum refresh time that the page cannot override. > This would allow it to follow W3C standards, and not bother users. Since the first patch I published the refresh meta tag has to be activated by '-r' option or it doesn't work and show the warning as usual. If you use the '-r' option so dillo behave like every bloat mozilla/firefox/opera/ie etc. etc. The option could *not* be save in the /etc/.dillorc but has to be put on the command line. Scott, do you thing this is sufficent? An option have to be activated by the user so I suppose he/she has to know what that option will be. If I want to be bothered, may I choose it? ;-) Cheers, -- Roberto A. Foglietta Analista Programmatore GNU/Linux SAD Trasporto Locale S.p.a. Corso Italia 13/N 39100 BOLZANO (I) Tel. +39/0471-450.261 Fax +39/0471-450.253

Feb. 18, 2005

Re: [Dillo-dev]Refresh patch for 0.8.4

by Scott Scriven

* Kelly Price <strredwolf(a)gmail.com> wrote: > Heya. New here. Glad someone's working on Dillo. Hope it > comes full-featured like Firefox is with less than half the > bloat! > > I'm attaching a fixed patch for 0.8.4 that allows meta-http > refreshes. This is quite common with web chats (including > Tag-Board). I've test-compiled and ran it, and so far so good. > > The next thing we'll need is the frames/iframes code now! I don't know about anyone else, but much of the reason I use dillo is because it does *not* do things like frames, meta-refresh, GIF animations, and Flash ads. It is very small, very fast, and implements useful browser functions without all the flashy cruft a lot of annoying sites use. If you want firefox, then use firefox. As for meta-refresh, I suppose it could be included without causing much trouble, on two conditions: It should retain its current "warn the user instead of refreshing" behavior, if the refresh goes to a different page. Also, it should let the user specify a minimum refresh time that the page cannot override. This would allow it to follow W3C standards, and not bother users. -- Scott

Feb. 18, 2005

[Dillo-dev]testing https plugin gave problems

by Peter Wood

Hi First, thanks to all for a great browser! The following applies to dillo-0.8.4 on linux with glibc2. I tried to test the https plugin by removing line 56 from dpi/https.c, as suggested. However, the linker complained about unresolved symbols when I recompiled. I solved that problem by changing the order of LIBSSL_LIBS in configure.in to "-lssl -lcrypto". Dillo then recompiled fine, however I still get the default message about the prototype plugin for https being disabled. Is there something else I should do. Regards, Peter

Feb. 18, 2005

[Dillo-dev]dillo-0.8.4 ported to ipaq familiar-X11

by Michael Kropfberger

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi everyone! familiar comes with dillo-0.6.6 which is even older than old, so I've cross-compiled dillo-0.8.4 with some extra patches: * when started in fullscreen, no toggle-button is visible * no cursor is visible, when the mouse is over the dillo window * the great mime-type patch for external viewers (from <http://www-fourier.ujf-grenoble.fr/~franz/dillo/>) * it reloads all urls (cache disabled for url rewriting, since it ignores the according meta tags) since dillo still is based on gtk+-1.2, this is a little bloated, especially because now there are extra X11-fonts needed... yes, I know, familiar comes with dillo-0.6.6 ported to gtk2 and uses built-in pango fonts, but this was too much work for me now :) I've also packaged a very tiny XFree86-font selection, so it really becomes usable... grab it here: http://www-itec.uni-klu.ac.at/~mike/ipaq/xfonts-tiny-0.1.tar.gz for fast installation on your familiar-0.8 ipaq, grab the font .ipk at: http://www-itec.uni-klu.ac.at/~mike/ipaq/xfonts-tiny_0.1-r0_all.ipk and the dillo .ipk at: http://www-itec.uni-klu.ac.at/~mike/ipaq/dillo_0.8.4-r0_arm.ipk for the OpenEmbedded maintainers, here are the needed bitbake .bb files: http://www-itec.uni-klu.ac.at/~mike/ipaq/my-new-dillo-oe.tgz I hope you all like it! greets, Mike -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.4 (GNU/Linux) iD8DBQFCE4gxw6IyDgGSsNIRAsl0AJ4kIbSuaU+OUAAtZmtk5/Fbo2V97ACfTnP3 A3xWwswJSz9wiqIpaqrqc5s= =Aj25 -----END PGP SIGNATURE-----

Feb. 16, 2005

[Dillo-dev]FOSDEM 2005

by Jorge Arellano Cid

Hi there, Good news here! After a lot of careful diligencies and a lot of work, it is official now. Dillo will be presented at FOSDEM 2005 (Feb 26, 27). I'll go to Belgium and give a conference at FOSDEM's embedded track. This is very good news for the project because some contacts can be made with a view to fund the future of the project, Dillo will gain more exposure in Europe, and because I'll also meet with Sebastian in Germany and we'll work together on some pending issues and the future of Dillo. This effort has brought a pile of work on my side, and given its importance, I assigned top priority to it and that's the reason for me being silent on the list in the past weeks. I'll try to briefly answer some pending threads, but as I'm not at home now and having still much to do (prepare the conference, trip, lots of emails, schedule, expenditures, etc.), please don't expect much feedback from me these days. For those of you living in Europe, and that think it may be interesting to meet, or have a contact opportunity in your country that may thrive, please let me know. This is a very long distance trip to make to miss the opportunity of taking chances. Ah, if you have some for fun or for pleasure ideas, those will be considered too! :-) Especially near the end of this trip. -- Best wishes. Jorge.-

Feb. 11, 2005

Re: [Dillo-dev]Refresh patch for 0.8.4

by Roberto A. Foglietta

Kelson Vibber wrote: > Madis Janson wrote: > >> This is only my opinion, which may be wrong (I am only contributed few >> patches to Dillo), but I think the reason for no answer is that strict >> standards compliance is one of the goals of the Dillo project and the >> w3c html standard discourages the usage of meta tag for refresh. >> >> http://www.w3.org/TR/html401/struct/global.html#idx-meta_data-2 > > > Actually, that document only discourages using meta refresh for > *forwarding*, as there are better, more accessible ways to do it (i.e. > 302 Moved and similar responses). It says nothing negative about using > a meta http-equiv tag to *refresh* the page. > > "Note. Some user agents support the use of META to refresh the current > page after a specified number of seconds, with the option of replacing > it by a different URI. Authors should not use this technique to forward > users to different pages, as this makes the page inaccessible to some > users. Instead, automatic page forwarding should be done using > server-side redirects." > I hope this could make change your minds about including my patch in dillo project. I remember you that patch doesn't modify standard dillo behaviure but simple add -r flags which allow refresh tag to act. :-) If I have to modify something in the patch to make it accepted feel free to suggest me. Cheers, -- Roberto A. Foglietta Analista Programmatore GNU/Linux SAD Trasporto Locale S.p.a. Corso Italia 13/N 39100 BOLZANO (I) Tel. +39/0471-450.261 Fax +39/0471-450.253 diff -pru dillo-0.8.4/src/browser.h dillo-0.8.4-refresh/src/browser.h --- dillo-0.8.4/src/browser.h 2004-10-20 21:09:51.000000000 +0200 +++ dillo-0.8.4-refresh/src/browser.h 2005-01-17 14:39:34.000000000 +0100 @@ -139,6 +139,12 @@ struct _BrowserWindow /* The id for the idle function that sets button sensitivity. */ guint sens_idle_id; + + /* RAF, 2004-10-26: usefull to refresh tag */ + guint timeout; + guint delay; + guint idle_id; + DilloUrl *refresh_url; }; diff -pru dillo-0.8.4/src/cache.c dillo-0.8.4-refresh/src/cache.c --- dillo-0.8.4/src/cache.c 2004-12-24 16:22:53.000000000 +0100 +++ dillo-0.8.4-refresh/src/cache.c 2005-01-17 14:51:58.000000000 +0100 @@ -906,6 +906,12 @@ static void Cache_process_queue(CacheDat DEBUG_MSG(1, "QueueSize ====> %d\n", g_slist_length(ClientQueue)); } +/* RAF, 2004-10-28 */ +gint a_Cache_queue_size(void) +{ + return g_slist_length(ClientQueue); +} + /* * Callback function for Cache_delayed_process_queue. */ diff -pru dillo-0.8.4/src/cache.h dillo-0.8.4-refresh/src/cache.h --- dillo-0.8.4/src/cache.h 2004-12-23 19:55:07.000000000 +0100 +++ dillo-0.8.4-refresh/src/cache.h 2005-01-17 14:52:16.000000000 +0100 @@ -50,6 +50,7 @@ struct _CacheClient { /* * Function prototypes */ +gint a_Cache_queue_size(void); //RAF, 2004-10-28 void a_Cache_init(void); gint a_Cache_open_url(void *Web, CA_Callback_t Call, void *CbData); gint a_Cache_get_buf(const DilloUrl *Url, gchar **PBuf, gint *BufSize); diff -pru dillo-0.8.4/src/dillo.c dillo-0.8.4-refresh/src/dillo.c --- dillo-0.8.4/src/dillo.c 2005-01-06 13:14:39.000000000 +0100 +++ dillo-0.8.4-refresh/src/dillo.c 2005-01-17 13:55:23.000000000 +0100 @@ -70,6 +70,7 @@ enum { DILLO_CLI_LOCAL = 1 << 5, DILLO_CLI_GEOMETRY = 1 << 6, DILLO_CLI_DBG_RENDERING = 1 << 7, + DILLO_CLI_REFRESH = 1 << 8, //RAF-2004-10-27 DILLO_CLI_ERROR = 1 << 15 }; @@ -82,6 +83,9 @@ enum { */ gboolean dillo_dbg_rendering = FALSE; +/* RAF, 2004-10-27: adding command line option to allow refresh */ +gboolean dillo_refresh = FALSE; + /* * Forward declarations */ @@ -114,6 +118,9 @@ static CLI_options Options[] = { 0, DILLO_CLI_DBG_RENDERING, " -D, --debug-rendering Draw additionaly several lines in a web page,\n" " representing its structure. For debugging. "}, + /* RAF 2004-10-27: command line option add */ + {"-r", "-refresh", 0, DILLO_CLI_REFRESH, + " -r, -refresh Allow meta refresh working as webmaster decided."}, {NULL, NULL, 0, 0, NULL} }; @@ -249,6 +256,10 @@ gint main(int argc, char *argv[]) case DILLO_CLI_DBG_RENDERING: dillo_dbg_rendering = TRUE; break; + /* RAF, 2004-10-27: adding command line option for meta refresh */ + case DILLO_CLI_REFRESH: + dillo_refresh = TRUE; + break; case DILLO_CLI_GEOMETRY: if (a_Misc_parse_geometry(opt_argv[0], &xpos, &ypos, &width, &height)) break; diff -pru dillo-0.8.4/src/dillo.h dillo-0.8.4-refresh/src/dillo.h --- dillo-0.8.4/src/dillo.h 2004-05-22 00:13:07.000000000 +0200 +++ dillo-0.8.4-refresh/src/dillo.h 2005-01-17 13:53:23.000000000 +0100 @@ -6,4 +6,7 @@ extern gboolean dillo_dbg_rendering; +/* RAF, 2004-10-27: adding command line option to allow refresh */ +extern gboolean dillo_refresh; + #endif /* __DILLO_H__ */ diff -pru dillo-0.8.4/src/html.c dillo-0.8.4-refresh/src/html.c --- dillo-0.8.4/src/html.c 2005-01-05 22:01:00.000000000 +0100 +++ dillo-0.8.4-refresh/src/html.c 2005-01-17 14:52:58.000000000 +0100 @@ -3150,6 +3150,56 @@ static void Html_tag_close_form(DilloHtm * is a hackish way to put the message. A much cleaner approach is to * build a custom widget for it. */ + + /* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * + * WORKING ON SUPPORTING DEPRECATED META REFRESH + * Roberto A. Foglietta <roberto.foglietta(a)sad.it> + * + * Actually I'm working for a local public transport company in which + * I have to to develop a new way to display time table and information + * to people. Using Dillo into embedded linux system allow my collegues + * to change graphical display just editing php code in the server BUT + * this kind of information have to stay up-to-date. Commit a refresh + * from server is, in this case, the best solution I can immagine. + * So supporting meta refresh tag makes Dillo good for embedded client + * systems. This is the main reason because I developed this web-deprecated + * but very-useful-in-embedded function. In addition to this pratical + * consideration I think people use free software to be free, so it makes + * sense give them the freedoom to choose also deprecated functions. + * Obviously to respect Dillo's team decision to do not allow refresh + * as default policy I make it optional by setting a flag from command + * line + */ + +void refresh_true_callback(gpointer client_data) { + BrowserWindow *bw = client_data; + + if(bw->timeout) { + gtk_timeout_remove(bw->timeout); + bw->timeout = 0; + } + + a_Commands_open_link_callback(NULL, client_data); +} + +void refresh_callback(gpointer client_data) { + BrowserWindow *bw = client_data; + + if(a_Cache_queue_size()) { + return; + } + + MSG("a_Cache_queue_size: %d\n", a_Cache_queue_size()); + + if(bw->idle_id) { + gtk_idle_remove(bw->idle_id); + bw->idle_id = 0; + } + + bw->timeout = gtk_timeout_add((guint32)bw->delay*1000, + (GtkFunction)refresh_true_callback, (gpointer)(bw)); +} + static void Html_tag_open_meta(DilloHtml *html, char *tag, gint tagsize) { const gchar *meta_template = @@ -3161,7 +3211,7 @@ static void Html_tag_open_meta(DilloHtml " <a href='%s'>here</a>%s</td></tr></table><br>\n"; const gchar *equiv, *content; - gchar *html_msg, delay_str[64]; + gchar *ptr, *html_msg, delay_str[64]; gint delay; /* only valid inside HEAD */ @@ -3173,7 +3223,34 @@ static void Html_tag_open_meta(DilloHtml if ((equiv = Html_get_attr(html, tag, tagsize, "http-equiv")) && !g_strcasecmp(equiv, "refresh") && (content = Html_get_attr(html, tag, tagsize, "content"))) { + } else { + return; + } + + /* check for delay */ + delay = strtol(content, &ptr, 0); + if(ptr) ptr = (gchar *)content; + + if(*ptr) { + /* todo: fix this to parse "URL =" & friends (HTML spec) */ + /* RAF, 2004-10-25: DONE, I suppose... */ + if (strstr(ptr, "URL") || strstr(ptr, "url") || strstr(ptr, "Url")) { + ptr = strstr(ptr, "="); + while (*(++ptr) == ' ' && *ptr != '>'); /* skip URL= & friends */ + if(*ptr == '>') *ptr = 0; /* doesn't exist any URL= */ + } else { + *ptr = 0; /* doesn't exist any URL= */ + } + } + + if(!*ptr && !delay) { + return; + } else { + g_print("meta refresh: start with refresh %sallowed\n", (dillo_refresh) ? "" : "not "); + g_print("meta refresh: delay -> %d, url -> %s\n", delay, (*ptr) ? ptr : "none"); + } + if(!dillo_refresh) { /* Get delay, if present, and make a message with it */ if ((delay = strtol(content, NULL, 0))) g_snprintf(delay_str, 64, " after %d second%s.", @@ -3196,6 +3273,33 @@ static void Html_tag_open_meta(DilloHtml html->InFlags = SaveFlags; } g_free(html_msg); + } else { + //if(html->bw->refresh_url) + a_Url_free(html->bw->refresh_url); + + if(*ptr) { + html->bw->refresh_url = a_Url_new(ptr, NULL, 0, 0, 0); + a_Url_set_flags(html->bw->refresh_url, URL_FLAGS(html->bw->refresh_url) | URL_E2EReload); + a_Menu_popup_set_url(html->bw, html->bw->refresh_url); + } else { + html->bw->refresh_url = a_History_get_url(NAV_TOP(html->bw)); + a_Url_set_flags(html->bw->refresh_url, URL_FLAGS(html->bw->refresh_url) | URL_E2EReload); + a_Url_set_flags(html->bw->refresh_url, URL_FLAGS(html->bw->refresh_url) & ~URL_SpamSafe); + a_Menu_popup_set_url(html->bw, html->bw->refresh_url); + html->bw->refresh_url = NULL; + } + + /* RAF, 2004-10-25: Use a timeout to refresh the page automatically */ + if(delay) { + //html->bw->timeout = gtk_timeout_add((guint32)delay*1000, + // (GtkFunction)refresh_callback, (gpointer)(html->bw)); + html->bw->delay = delay; + html->bw->idle_id = gtk_idle_add((GtkFunction)refresh_callback, (gpointer)(html->bw)); + } else { + html->bw->timeout = 0; + html->bw->delay = 0; + refresh_callback(html->bw); + } } } diff -pru dillo-0.8.4/src/interface.c dillo-0.8.4-refresh/src/interface.c --- dillo-0.8.4/src/interface.c 2004-12-29 14:26:48.000000000 +0100 +++ dillo-0.8.4-refresh/src/interface.c 2005-01-17 14:46:41.000000000 +0100 @@ -1127,6 +1127,12 @@ a_Interface_browser_window_new(gint widt bw->viewsource_window = NULL; bw->pagebugs_window = NULL; + /* RAF, 2004-10-26: null value for time out */ + bw->timeout = 0; + bw->delay = 0; + bw->idle_id = 0; + bw->refresh_url = NULL; + /* now that the bw is made, let's customize it.. */ Interface_browser_window_customize(bw); diff -pru dillo-0.8.4/src/nav.c dillo-0.8.4-refresh/src/nav.c --- dillo-0.8.4/src/nav.c 2004-12-16 20:02:21.000000000 +0100 +++ dillo-0.8.4-refresh/src/nav.c 2005-01-17 14:43:20.000000000 +0100 @@ -55,6 +55,9 @@ void a_Nav_free(BrowserWindow *bw) { a_Nav_cancel_expect(bw); g_free(bw->nav_stack); + + /* RAF, 2004-10-27: free the url redirection */ + a_Url_free(bw->refresh_url); } @@ -156,6 +159,11 @@ static void Nav_open_url(BrowserWindow * gboolean ForceReload = (URL_FLAGS(url) & URL_E2EReload); MSG("Nav_open_url: Url=>%s<\n", URL_STR_(url)); + /* RAF, 2004-10-28: switch off reload timeout */ + if(bw->timeout) { + gtk_timeout_remove(bw->timeout); + bw->timeout = 0; + } /* Get the url of the current page */ if ( a_Nav_stack_ptr(bw) != -1 )

Feb. 10, 2005

Re: [Dillo-dev]support for video streams

by Daniel Lord

Hi, On Thu, 10 Feb 2005, Michael Kropfberger wrote: > On Wednesday 09 February 2005 19:48, Bjoern Brill wrote: > > http for which you can provide a working mailcap entry. I hope that this > > patch will be merged in the not-too-distant future. > > this patch is perfect! I also hope, it will make it's way into the CVS! just want to say that this patch suids my needs too ;) Greetings Daniel -- We've all heard that a million monkeys banging on a million typewriters will eventually reproduce the entire works of Shakespeare. Now, thanks to the Internet, we know this is not true.

Feb. 10, 2005