- Dillo-dev - lists.mailman3.com

Re: [Dillo-dev]support for video streams

by Bjoern Brill

On Wed, 9 Feb 2005, Michael Kropfberger wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > hi! > > for a multimedia app on linux ipaq, I wanted to enable video links from within > dillo... > [...] Try the "external viewers" patch by Lars Clausen and Mattthias Franz: <http://www-fourier.ujf-grenoble.fr/~franz/dillo/>. It does not (directly) support streaming protocols, but everything over http for which you can provide a working mailcap entry. I hope that this patch will be merged in the not-too-distant future. Regards, Bjoern -- Bj"orn Brill <brill(a)fs.math.uni-frankfurt.de> Frankfurt am Main, Germany

Feb. 9, 2005

Re: [Dillo-dev]support for video streams

by Diego Sáenz

For the protocol part read this two mails: http://article.gmane.org/gmane.comp.web.dillo.devel/2496/match=xproto http://article.gmane.org/gmane.comp.web.dillo.devel/2497/match=xproto DarkSpirit El Wed, 9 Feb 2005 15:25:31 +0100 Michael Kropfberger <michael(a)kropfberger.com> escribio: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > hi! > > for a multimedia app on linux ipaq, I wanted to enable video links from within > dillo... > > 1st idea: hack up the dpi/file.c so that the file extension .mp4 or .avi is > detected and whenever File_get_file(...) is called, instead of passing the > raw data, an external video player is started with this file name as input > parameter... configuring of the videoplayer path could happen from > within .dillorc > > this is, what I'd imagine to be the quickest solution... > > more complex ideas: > * really make nicer support for mime-types, better even protocol types > (rtsp://), which again calls an external player > > * with a separate dpi, so for remote http:// URLs progressive download via > http is possible > > * even allow "embedded" video output within a rendered html page, so it has to > be like gif,png,jpg support... well, not really interesting on ipaqs, and too > complicated anyways... > > > > what's your thoughts, folks? > > > Mike > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.2.4 (GNU/Linux) > > iD8DBQFCCh1bw6IyDgGSsNIRAkI9AJ0bCKJXr67V1SxPT1DpCgrZRYCKcQCfRX1B > rS/y0jHqK3sqUgEJ/vsQ5mM= > =WDsR > -----END PGP SIGNATURE----- > > _______________________________________________ > Dillo-dev mailing list > Dillo-dev(a)dillo.org > http://lists.auriga.wearlab.de/cgi-bin/mailman/listinfo/dillo-dev

Feb. 9, 2005

[Dillo-dev]support for video streams

by Michael Kropfberger

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 hi! for a multimedia app on linux ipaq, I wanted to enable video links from within dillo... 1st idea: hack up the dpi/file.c so that the file extension .mp4 or .avi is detected and whenever File_get_file(...) is called, instead of passing the raw data, an external video player is started with this file name as input parameter... configuring of the videoplayer path could happen from within .dillorc this is, what I'd imagine to be the quickest solution... more complex ideas: * really make nicer support for mime-types, better even protocol types (rtsp://), which again calls an external player * with a separate dpi, so for remote http:// URLs progressive download via http is possible * even allow "embedded" video output within a rendered html page, so it has to be like gif,png,jpg support... well, not really interesting on ipaqs, and too complicated anyways... what's your thoughts, folks? Mike -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.4 (GNU/Linux) iD8DBQFCCh1bw6IyDgGSsNIRAkI9AJ0bCKJXr67V1SxPT1DpCgrZRYCKcQCfRX1B rS/y0jHqK3sqUgEJ/vsQ5mM= =WDsR -----END PGP SIGNATURE-----

Feb. 9, 2005

Re: [Dillo-dev]Refresh patch for 0.8.4

by Kelson Vibber

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." -- Kelson Vibber www.hyperborea.org

Feb. 8, 2005

Re: [Dillo-dev]Refresh patch for 0.8.4

by Madis Janson

On Mon, 31 Jan 2005, Roberto A. Foglietta wrote: > Kelly Price 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'm the authoir of original patch. I developed it for 0.8.4 too > reducing size of patch itself. I tried to talk with dillo author to > evaluete if there was any chances to include it in the main branch but I > didn't received any answer. 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

Feb. 8, 2005

[Dillo-dev]selection patch (was: Patch review: Frank de Lange patch (20031124))

by Sebastian Geerken

Hi Bastiaan, On Thu, Jan 20, Bastiaan Jacques wrote: > The following patch does does two things in selection.c: > - - make a_Selection_set_dclick_callback replace the callback instead of > returning when the callback was already set. > * This function is called only once in the code (in interface.c); the > callback pointer always points to the same function. So I don't see a > point in merging this first hunk, although it won't do any harm. Thoughts? I'd rather leave it the current way, since changing the callback as in your patch would mean, that the old callback would be silently deactivated. To extend the current functionality, a remove function may be useful (to make deactivation explicit), or even a list of callbacks, so that the functions do not have to be removed (unless they should so). Anyway, there is currently no demand for an extended functionality, so I'd leave it as simple as it is now. > - - Replace three comparison operators (==) with bitwise AND.. > * This only works as expected when either lvalue and rvalue have exactly > one or zero bits set to 1. Although this currently is true, this may > change in the future; thus, I don't think these three hunks should be > merged. This (that only one bit set) will also be true in the future. If contents will be extended to contain multiple types (which is not planned at all), the whole design must be changed anyway (at least, the unions will have to replaced by structs), so grepping over the code is anyway necessary. Until then, your suggestion would only make the code more confusing. Sebastian

Feb. 2, 2005

Re: [Dillo-dev]Refresh patch for 0.8.4

by Roberto A. Foglietta

robert w hall wrote: > In message <7ff5668305012822047c7c59a5(a)mail.gmail.com>, Kelly Price > <strredwolf(a)gmail.com> writes > >>Heya. New here. Glad someone's working on Dillo. Hope it comes >>full-featured like Firefox is with less than half the bloat! > > > well some of us want NO bloat, and would err on the side of something > less than fully-featured to ensure that. > > If others really require a firefox-competitor, perhaps we are heading > for a dillo-lite... > Robert you have some problems but because I'm here as software developer and not as puppy mainteiner I will avoid to explain you why you missed THE point of that mail. But you still missed! ;-) However if some patches doesn't meet the strict requirements of the core dillo developers team doesn't mean that patches are crap or useless to others people. I still remember you that free software is for free people who have the right of choice and not to be forced to use 'dillo' as you imagined. Because you could be a genius but your imagination, by definition of genius as human being element, is still limitated. Because there are some patches out of there AT LEAST could be interesting for dillo users and for the project itself have a page collecting all these contribute AS patches. "AT LEAST for project itself too" exactly means "to reduce the risk of a fork". Which it could be not so difficult as you think, that could be less difficult than still sticky to YOUR limnitated-imagination-dillo. :-) 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

Jan. 31, 2005

Re: [Dillo-dev]Refresh patch for 0.8.4

by Roberto A. Foglietta

Kelly Price 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! > Here you are the refresh patch for dillo 0.8.4 from the author (me) http://www.roberto.foglietta.name/pub/dillo-0.8.4-refresh.patch.bz2 please don't spread modified patch which containing my name (and so people are let to refer at me like the original author) WITHOUT commenting the code about your modifies to adapt it to new releases. That patch cointaining ONLY the code necessary to use refresh option. Most of others changes (like in the patch you have submitted) doesn't make any sense (or not make sensa anymore). Thank you for your patience, -- 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

Jan. 31, 2005

Re: [Dillo-dev]Refresh patch for 0.8.4

by Roberto A. Foglietta

Kelly Price 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'm the authoir of original patch. I developed it for 0.8.4 too reducing size of patch itself. I tried to talk with dillo author to evaluete if there was any chances to include it in the main branch but I didn't received any answer. If somebody else is using this patch I will put on my website and I will keep it uptodate. Let me know. :-) 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

Jan. 31, 2005

Re: [Dillo-dev]Refresh patch for 0.8.4

by Patrick Shanahan

* higuita <higuita7(a)yahoo.co.uk> [01-30-05 03:16]: > uuhmm?! > are you sure? arent you thinking in the cookies? > i have https in every instance i try, the only problem > with https is the "always ask for the certificate" problem maybe ?? -- Patrick Shanahan Registered Linux User #207535 http://wahoo.no-ip.org @ http://counter.li.org HOG # US1244711 Photo Album: http://wahoo.no-ip.org/photos

Jan. 30, 2005

Re: [Dillo-dev]Refresh patch for 0.8.4

by higuita

On Sat, 29 Jan 2005 12:30:54 -0500, Patrick Shanahan <ptilopteri(a)gmail.com> wrote: > before complying. I use it exclusively. The only problem I have with > it is it's only available to *one* screen/instance at a time :^(. uuhmm?! are you sure? arent you thinking in the cookies? i have https in every instance i try, the only problem with https is the "always ask for the certificate" problem higuita -- Naturally the common people don't want war... but after all it is the leaders of a country who determine the policy, and it is always a simple matter to drag the people along, whether it is a democracy, or a fascist dictatorship, or a parliament, or a communist dictatorship. Voice or no voice, the people can always be brought to the bidding of the leaders. That is easy. All you have to do is tell them they are being attacked, and denounce the pacifists for lack of patriotism and exposing the country to danger. It works the same in every country. -- Hermann Goering, Nazi and war criminal, 1883-1946

Jan. 30, 2005

Re: [Dillo-dev]Refresh patch for 0.8.4

by Patrick Shanahan

* Brian Hechinger <wonko(a)4amlunch.net> [01-29-05 11:41]: > > through the use of dillo plug-ins, i think we should be fully capable of > keeping the core dillo small and light, while easily allowing the ability > to add features that some people need without increasing the core bloat. > > https is a perfect example of this for one. https is already available by merely commenting out one line of code before complying. I use it exclusively. The only problem I have with it is it's only available to *one* screen/instance at a time :^(. -- Patrick Shanahan Registered Linux User #207535 http://wahoo.no-ip.org @ http://counter.li.org HOG # US1244711 Photo Album: http://wahoo.no-ip.org/photos

Jan. 29, 2005

Re: [Dillo-dev]Refresh patch for 0.8.4

by Brian Hechinger

On Sat, Jan 29, 2005 at 01:46:13PM +0000, robert w hall wrote: > In message <7ff5668305012822047c7c59a5(a)mail.gmail.com>, Kelly Price > <strredwolf(a)gmail.com> writes > >Heya. New here. Glad someone's working on Dillo. Hope it comes > >full-featured like Firefox is with less than half the bloat! > > well some of us want NO bloat, and would err on the side of something > less than fully-featured to ensure that. > > If others really require a firefox-competitor, perhaps we are heading > for a dillo-lite... through the use of dillo plug-ins, i think we should be fully capable of keeping the core dillo small and light, while easily allowing the ability to add features that some people need without increasing the core bloat. https is a perfect example of this for one. -brian -- "Now you know why I got the everliving hell OUT of Windows administration. Knowing it doesn't make it any easier. It's just broken-as-designed."

Jan. 29, 2005

Re: [Dillo-dev]Refresh patch for 0.8.4

by robert w hall

In message <7ff5668305012822047c7c59a5(a)mail.gmail.com>, Kelly Price <strredwolf(a)gmail.com> writes >Heya. New here. Glad someone's working on Dillo. Hope it comes >full-featured like Firefox is with less than half the bloat! well some of us want NO bloat, and would err on the side of something less than fully-featured to ensure that. If others really require a firefox-competitor, perhaps we are heading for a dillo-lite... -- robert w hall

Jan. 29, 2005

[Dillo-dev]Refresh patch for 0.8.4

by Kelly Price

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! -- Kelly "STrRedWolf" Price http://strredwolf.furrynet.com

Jan. 29, 2005

Re: [Dillo-dev]RPM status

by robert w hall

In article <41EFECA9.1050804(a)pobox.com>, Kelson Vibber <kelson(a)pobox.com> writes > >I'm thinking of dropping some of the older releases next time, just to >simplify things. > In a similar move to simplification, I suggest I _only_ build an ssl- enabled rpm for the ps2 next time (the code has been out for long enough to come out of alpha, despite the certificates irritation, and most ps2-users need the https function.) Bob -- robert w hall

Jan. 20, 2005

[Dillo-dev]RPM status

by Kelson Vibber

Over the weekend I managed to get User-Mode Linux running again (under an old Red Hat 7.3 partition that I hadn't gotten around to deleting), so I was able to fill in RPMs for the rest of the releases I had used for Dillo 0.8.3. I also installed Mandrake 10.1, and was able to add an RPM for that release. I installed SuSE 9.2, though the combination of slow hardware, average DSL, slow mirror, the occasional missing file on the mirror and little free time to work with meant that it took several days to get a working install. Last night I finally built Dillo on it, but there's a font problem. When I get a chance I'll try installing some other GTK-1.2 apps and see if it's just a system config problem, and if it is, I'll post the RPM. At that point I'll have RPMs for current releases of Fedora Core, Mandrake, and SuSE, and a number of older releases. I'm thinking of dropping some of the older releases next time, just to simplify things. -- Kelson Vibber www.hyperborea.org

Jan. 20, 2005

Re: [Dillo-dev]Patch review: Frank de Lange patch (20031124)

by Roberto A. Foglietta

Bastiaan Jacques wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Hello, > > > I am dissecting and reviewing Frank de Lange's 14,000 line patch that > was rejected by the core dillo developers due to its size. I will email > the list from time to time to gather input as to whether a particular > portion of the patch is useful or not, beginning with the patch a little > further down this message. Is this patch 0.8.4 ready? Is this patch available? May try it and check the code it modified? 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

Jan. 20, 2005

[Dillo-dev]Patch review: Frank de Lange patch (20031124)

by Bastiaan Jacques

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hello, I am dissecting and reviewing Frank de Lange's 14,000 line patch that was rejected by the core dillo developers due to its size. I will email the list from time to time to gather input as to whether a particular portion of the patch is useful or not, beginning with the patch a little further down this message. The following patch does does two things in selection.c: - - make a_Selection_set_dclick_callback replace the callback instead of returning when the callback was already set. * This function is called only once in the code (in interface.c); the callback pointer always points to the same function. So I don't see a point in merging this first hunk, although it won't do any harm. Thoughts? - - Replace three comparison operators (==) with bitwise AND.. * This only works as expected when either lvalue and rvalue have exactly one or zero bits set to 1. Although this currently is true, this may change in the future; thus, I don't think these three hunks should be merged. Regards, Bastiaan Jacques. diff -Nur --exclude-from /home/andi/freeciv/diff_ignore dillo-bak/src/selection.c dillo/src/selection.c - --- dillo-bak/src/selection.c 2004-02-20 14:28:59.000000000 +0100 +++ dillo/src/selection.c 2004-02-20 14:30:17.000000000 +0100 @@ -76,7 +76,10 @@ ~ void (*fn) (gpointer data), ~ gpointer data) ~ { - - g_return_if_fail (selection->dclick_callback == NULL); + /* this used to return when the callback was already set, + * changed function to replace callback instead. + * g_return_if_fail (selection->dclick_callback == NULL); + */ ~ selection->dclick_callback = fn; ~ selection->callback_data = data; ~ } @@ -426,7 +429,7 @@ ~ DwIterator *top = it->stack[it->stack_top]; ~ gint len; - - if (top->content.type == DW_CONTENT_TEXT) + if (top->content.type & DW_CONTENT_TEXT) ~ len = strlen(top->content.data.text); ~ else ~ len = 1; @@ -486,7 +489,7 @@ ~ for (i = a_Dw_ext_iterator_clone (a), start = TRUE; ~ (cmp = a_Dw_ext_iterator_compare (i, b)) <= 0; ~ a_Dw_ext_iterator_next (i), start = FALSE) { - - if (i->content.type == DW_CONTENT_TEXT) { + if (i->content.type & DW_CONTENT_TEXT) { ~ if (fl) { ~ if (start) { ~ DEBUG_MSG (2, " highlighting %s from %d to %d\n", @@ -537,7 +540,7 @@ ~ cmp = a_Dw_ext_iterator_compare (selection->from, selection->to); ~ if (cmp == 0) { - - if (selection->from->content.type == DW_CONTENT_TEXT) { + if (selection->from->content.type & DW_CONTENT_TEXT) { ~ si = selection->from->stack[selection->from->stack_top]; ~ if (selection->from_char < selection->to_char) ~ tmp = g_strndup(si->content.data.text + selection->from_char, -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.0 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iD8DBQFB77I5wDeUwhiOHwURAkYeAKCzu03fggXw2DvdEiBWAP4GicLw5wCggUid Kg67vhiAUHc2dK+jGGVtE9o= =AGq9 -----END PGP SIGNATURE-----

Jan. 20, 2005

Re: [Dillo-dev]https, ssl & certificates

by robert w hall

In article <nNw1CoADb66BFwrn(a)n-cantrell.demon.co.uk>, robert w hall <bobh(a)n-cantrell.demon.co.uk> writes >But the situation over certificates appears confusing: - >as I (cursorily) read the code, it uses standard SSL functions to parse >entries in the .dillo/certs directory. OK, Garrett has been trying to help me understand root certificates. The problem I now appear to have is that openssl-0.9.6 had a fairly comprehensive block of certs but they're all out of date. Latest openssl (0.9.7) is more up-to-date but rather restricted (eg no RSA etc). Did anyone find a decent set of root certs hiding anywhere or find a workaround via their own self-signing route?? Bob -- robert w hall

Jan. 18, 2005

[Dillo-dev]https, ssl & certificates

by robert w hall

Where are we at and where are we going with the issue of https, ssl and certificates? I usually run with https enabled and find it is a VERY useful adjunct to the main program - it has quirky behaviour on sites such as playstation2-linux (where the answer is to log in via ssl then leave ssl), but at least it enables posting there. Other sites now accessible include amazon.co.uk and groups.yahoo.com - a significant gain. But the situation over certificates appears confusing: - as I (cursorily) read the code, it uses standard SSL functions to parse entries in the .dillo/certs directory. I have a block of certificates stolen from another browser but they don't seem to change the need to force by the 'remote certificate cannot be verified20' messages. (Are these imported certs in a non-standard format perhaps?) And the code also appears to have a function to store accepted certificates, but I've never been aware of this happening. The code is reaching a level of maturity and acceptance where these remaining blemishes are puzzling 'us users'. Bob -- robert w hall

Jan. 17, 2005

[Dillo-dev]Gtk2 port, GtkArg{Get,Set}Func

by Keenan Pepper

Hi, I'm working on a Gtk2 port of Dillo to use on my Zaurus (with the GPE rom from OpenZaurus). Before everyone starts saying "That's stupid, Gtk2 is so bloated and slow, why would you want to do that?", let me say that right now I'm using the Gtk2 port of Dillo 0.6.6, which as you all know is ancient and renders html badly compared to newer versions of Dillo, but it's still the best and fastest browser I can get to work on GPE. So periodically I'll be asking "What does this thing do?"-type questions on this list as I try to make current Dillo compile and run with Gtk2. For now I just want to request that all the deprecated (GtkArgSetFunc) NULL and (GtkArgGetFunc) NULL casts be removed and changed to simply NULL as recommended here: http://developer.gnome.org/doc/API/2.2/gtk/gtk-changes-1-2.html This has nothing to do with Gtk2, in fact eventually I should change all the GtkObject stuff to GObject stuff anyway, but when I first blindly tried to make it compile that was the error that came up the most, and it can and should be changed upstream. Cheers, Keenan Pepper

Jan. 15, 2005

[Dillo-dev]why can dillo be small and fast

by Yu Zenghui

Relative to other browsers, why can dillo be small and fast? Can anybody tell me the reason from the viewpoint of design and implementation. The code of dillo is fewer than other larger desktop browser. But fewer code means fewer functions. How to keep balance between the two conflicts? Thank you!

Jan. 14, 2005

Re: [Dillo-dev]cache replace

by Brian Hechinger

On Thu, Jan 13, 2005 at 12:20:26AM +0000, higuita wrote: > > also, at least now, you can only have one cookie enable > dillo active, so we have to "hunt" for that dillo when > we need a cookie enable > so cache replace is still needed 8) i'm currently working on the configuration dpi, and moving the cookie stuff to a dpi is my next task after i'm finished here (the cookie dpi is going to depend on the config dpi) so that at least solves the cookie issue. cache is the next thing on my list, we'll talk when i get there. > Naturally the common people don't want war... but after all it is the > leaders of a country who determine the policy, and it is always a > simple matter to drag the people along, whether it is a democracy, or > a fascist dictatorship, or a parliament, or a communist dictatorship. > Voice or no voice, the people can always be brought to the bidding of > the leaders. That is easy. All you have to do is tell them they are > being attacked, and denounce the pacifists for lack of patriotism and > exposing the country to danger. It works the same in every country. > -- Hermann Goering, Nazi and war criminal, 1883-1946 i *LOVE* this quote, if bush wasn't such an idiot i could swear that he studied Goering. ;) -brian -- "Now you know why I got the everliving hell OUT of Windows administration. Knowing it doesn't make it any easier. It's just broken-as-designed."

Jan. 13, 2005

[Dillo-dev]what is the event for rendering

by Yu Zenghui

i try to look for the function which is used to begin rendering, but i failed. who can tell me what the function is? if there is no such function, what event triggers the rendering? does dillo begin to render after the whole html document is parsed, or once after one element is parsed? thank you!

Jan. 13, 2005

Re: [Dillo-dev]cache replace

by higuita

On Fri, 7 Jan 2005 00:11:41 -0700, Scott Scriven <dillo-dev(a)toykeeper.net> wrote: > 'net access. It also speeds things up compared to browsers which > check the network for updates before (or instead of) using cache. yep, i use this also maybe this could be one option (command line or run time)!? > Instead, I generally start a new instance for each site or task, > and close it when I'm done. I think I have about 20-30 open this doesnt work for low end machines, where we have little memory and/or little CPU also, at least now, you can only have one cookie enable dillo active, so we have to "hunt" for that dillo when we need a cookie enable so cache replace is still needed 8) higuita -- Naturally the common people don't want war... but after all it is the leaders of a country who determine the policy, and it is always a simple matter to drag the people along, whether it is a democracy, or a fascist dictatorship, or a parliament, or a communist dictatorship. Voice or no voice, the people can always be brought to the bidding of the leaders. That is easy. All you have to do is tell them they are being attacked, and denounce the pacifists for lack of patriotism and exposing the country to danger. It works the same in every country. -- Hermann Goering, Nazi and war criminal, 1883-1946

Jan. 13, 2005

Re: [Dillo-dev]How does CCC handle the parallelism of IO?

by Jorge Arellano Cid

On Tue, Jan 04, 2005 at 09:02:23PM +0800, uniharmony(a)gmail.com wrote: Hi, Sorry for the delay... > In my opinion, the CCC is a dynamic doubly linked list with function pointers. > And all the CCC functions in the chain are called serially. Yes. > How does CCC handle the parallelism of IO? Imagine several open file descriptors, each one active. A CCC chain was constructed for each one, so there are as many CCCs as FDs. > And Why do you consider it as "pseudo-thread"? When a FD has data, the respective CCC receives the data and "awakes" the streamed processing of the data. When the chunk is done, it waits for the next one. As each one of this processes takes a slim timeslice, there's no blocking of the other FDs in the active set. > I want to dive in it and thanks for reply. Good. What do you want to do? This is a complex part. BTW, what's your name. -- Cheers Jorge.-

Jan. 12, 2005

Re: [Dillo-dev]cookies disabled ??

by Patrick Shanahan

* Jeremy Henty <jeremy(a)chaos.org.uk> [01-12-05 05:07]: > Is another dillo running? If so, it will have locked the cookies > file. Yes, duh! Thank-you, -- Patrick Shanahan Registered Linux User #207535 http://wahoo.no-ip.org @ http://counter.li.org HOG # US1244711 Photo Album: http://wahoo.no-ip.org/photos

Jan. 12, 2005

Re: [Dillo-dev]cookies disabled ??

by Jeremy Henty

Patrick Shanahan wrote: > I am having cookie problems and do not understand. > > pat@wahoo:~> dillo my.yahoo.com > dillorc:170: unexpected string constant "splash", expected symbol > dillo_dns_init: Here we go! > The cookies file has a file lock: > disabling cookies in this dillo! > > > Where would the file lock be? Or, how do I defeat it? Is another dillo running? If so, it will have locked the cookies file. Cheers, Jeremy Henty

Jan. 12, 2005

[Dillo-dev]cookies disabled ??

by Patrick Shanahan

I am having cookie problems and do not understand. pat@wahoo:~> dillo my.yahoo.com dillorc:170: unexpected string constant "splash", expected symbol dillo_dns_init: Here we go! The cookies file has a file lock: disabling cookies in this dillo! Where would the file lock be? Or, how do I defeat it? -- Patrick Shanahan Registered Linux User #207535 http://wahoo.no-ip.org @ http://counter.li.org HOG # US1244711 Photo Album: http://wahoo.no-ip.org/photos

Jan. 12, 2005

Re: [Dillo-dev]Broken link

by Jorge Arellano Cid

On Tue, Jan 11, 2005 at 06:34:45PM -0500, Livio Soares wrote: > > Hi, Hi Livio! > > In the main website, there is a typo causing a broken link: > > | <h2> Current plans </h2> > | <p> > | The "current plans" section has its > | <a href='/plans.html'>own page</a> now<br> > | (it's also accessible from the left column's top links). > > '/plans.html' should be '/Plans.html' Fixed. -- Cheers Jorge.-

Jan. 12, 2005

Re: [Dillo-dev]0.8.4 final just released!

by Kelson Vibber

Jorge Arellano Cid wrote: > Please make your tarballs ASAP as this release includes a > security fix. At lunch I put together RPMS for the major systems I build for. Well, most of them. I've run into a slight problem: it seems that UML stopped working sometime after I upgraded to Fedora Core 3, so I'll have to figure out whether UML is even compatible with it as a host system, and possibly figure out a way to run those UMLs under an older distro. What I've got so far is: - Connectiva 9 - Fedora Code 1-3 (actually built on 3, but should work on FC1 and FC2. I'll test it on FC2 tonight.) - Mandrake 9.2 - Red Hat 7 & 9 (the RH9 rpm should work on RH8 too.) - SuSE 9.1 - Yellow Dog Linux 3 I should be able to add Mandrake 10.0 and 9.1 tonight, and possibly one of the older SuSE installations. At this point, I'm unlikely to get Conectiva 8, Mandrake 8.1 & 8.2, Red Hat 6.2, or SuSE 7.3 in the next few days. Also, I'm a bit behind on releases: I have CDs for Mandrake 10.1 and Yellow Dog 4, but I haven't installed them yet. I also haven't picked up SuSE 9.2 or Conectiva 10. Based on traffic I'm not going to worry about Conectiva for now, but if anyone on this list has it and is willing to build it from the source RPM, I'll post it. The distro-detecting source RPM is available at: http://www.hyperborea.org/software/dillo/rpms/dillo-0.8.4-1.src.rpm Robert, when you've had a chance to build the Playstation RPM, just send it to me and I'll post it. Patrick, same goes for SuSE 9.0. (My copy is a bit flaky, though 9.1 runs fine on the same system.) -- Kelson Vibber www.hyperborea.org

Jan. 12, 2005

[Dillo-dev]Broken link

by Livio Soares

Hi, In the main website, there is a typo causing a broken link: | <h2> Current plans </h2> | <p> | The "current plans" section has its | <a href='/plans.html'>own page</a> now<br> | (it's also accessible from the left column's top links). '/plans.html' should be '/Plans.html' best regards, Livio

Jan. 11, 2005

Re: [Dillo-dev]0.8.4 final just released!

by Patrick Shanahan

* Jorge Arellano Cid <jcid(a)dillo.org> [01-11-05 16:28]: > It says "0.8.4" instead of "0.8.4-rc1" :-) > > Ah, as this is a tarball (tar.bz2 instead of CVS), it also > comes with the configure script created by my machine. AND, I have found my cookie problem with yahoo.com: pat@wahoo:~> dillo my.yahoo.com dillorc:170: unexpected string constant "splash", expected symbol dillo_dns_init: Here we go! The cookies file has a file lock: disabling cookies in this dillo! Now the question, where would the cookies file have a file lock? -- Patrick Shanahan Registered Linux User #207535 http://wahoo.no-ip.org @ http://counter.li.org HOG # US1244711 Photo Album: http://wahoo.no-ip.org/photos

Jan. 11, 2005

Re: [Dillo-dev]0.8.4 final just released!

by robert w hall

In article <20050111202624.GA8786(a)dillo.org>, Jorge Arellano Cid <jcid(a)dillo.org> writes >Hi there, > > After the testing and fixing, finally, the new dillo-0.8.4 was >released today (11-01-2005). > > The official freshmeat announcement will be tomorrow. > > Please grab the tarball ASAP and report it compiles and runs :) > > Kelson, Phill, packagers, and interested people: > > Please make your tarballs ASAP as this release includes a > security fix. > differences from last night's cvs? -- robert w hall

Jan. 11, 2005

Re: [Dillo-dev]0.8.4 final just released!

by Jorge Arellano Cid

On Tue, Jan 11, 2005 at 09:27:08PM +0000, robert w hall wrote: > In article <20050111202624.GA8786(a)dillo.org>, Jorge Arellano Cid > <jcid(a)dillo.org> writes > >Hi there, > > > > After the testing and fixing, finally, the new dillo-0.8.4 was > >released today (11-01-2005). > > > > The official freshmeat announcement will be tomorrow. > > > > Please grab the tarball ASAP and report it compiles and runs :) > > > > Kelson, Phill, packagers, and interested people: > > > > Please make your tarballs ASAP as this release includes a > > security fix. > > > differences from last night's cvs? It says "0.8.4" instead of "0.8.4-rc1" :-) Ah, as this is a tarball (tar.bz2 instead of CVS), it also comes with the configure script created by my machine. -- Cheers Jorge.-

Jan. 11, 2005

[Dillo-dev]0.8.4 final just released!

by Jorge Arellano Cid

Hi there, After the testing and fixing, finally, the new dillo-0.8.4 was released today (11-01-2005). The official freshmeat announcement will be tomorrow. Please grab the tarball ASAP and report it compiles and runs :) Kelson, Phill, packagers, and interested people: Please make your tarballs ASAP as this release includes a security fix. -- Best Luck Jorge.-

Jan. 11, 2005

Re: [Dillo-dev]CAN-2005-12

by Bjoern Brill

On Mon, 10 Jan 2005, Jorge Arellano Cid wrote: > As I want to release 0.8.4 tomorrow. Please test it like this: > > - Grab the CVS, and if it works now, report ASAP & we're done. > - If CVS doesn't work, test 0.8.3 on and report, without the > slightest shade of doubt, whether it works ;) > - If 0.8.3 works, shove 0.8.3's cookies.c into 0.8.4 and test. > If this hybrid doesn't work, then it may be gremlins... > > Hopefully the first step will solve it. > I don't use cookie-heavy sites very much, but I did a simple test with the sourceforge.net "default download location" cookies, as in e.g. dillo http://prdownloads.sourceforge.net/xine/xine-ui-0.99.3.tar.gz?use_default=m… Result: 0.8.3: rejects cookie, complains about non-RFC compliant "Expires" field. 0.8.4-rc1: rejects (or ignores) cookie without further notice. 0.8.4-rc1 + cookies.c from CVS: works. Apart from the (solved, I think) cookies problem, dillo 0.8.4-rc1 did work for me for two days of moderate use. This is on Debian woody/i386. Regards, Bjoern -- Bj"orn Brill <brill(a)fs.math.uni-frankfurt.de> Frankfurt am Main, Germany

Jan. 11, 2005

Re: [Dillo-dev]CAN-2005-12

by robert w hall

In article <vG+HgvAzSx4BFwq9(a)n-cantrell.demon.co.uk>, robert w hall <bobh(a)n-cantrell.demon.co.uk> writes >In article <20050110222427.GA1465(a)dillo.org>, Jorge Arellano Cid ><jcid(a)dillo.org> writes >> OK guys, I considered the posts, reviewed the code, then found >>and fixed a naive bug introduced by me on solving issues for LP64 >>processors. >> >> It's in the CVS now. >> >> >just tried the x86/ssl version out of cvs. Now does amazon.co.uk and >yahoo groups OK. same result now for ps2 version of current cvs. NB - test results are for www.amazon.co.uk (not www.amazon.com which behaved differently/less fussily) and groups.yahoo.com (also tested www.playstation2-linux.com, where as with 0.8.3 the knack is to drop out of ssl as soon as you've logged in - dunno why this doesn't work more straightforwardly if you stay in https - hammering reload does not cure it, I now realise) 'Imprimatur' Bob -- robert w hall

Jan. 11, 2005

Re: [Dillo-dev]CAN-2005-12

by Diego Sáenz

groups.yahoo.com work again like in 0.8.3. I do not use amazon. DarkSpirit

Jan. 11, 2005

Re: [Dillo-dev]CAN-2005-12

by Patrick Shanahan

* Jorge Arellano Cid <jcid(a)dillo.org> [01-10-05 17:33]: > OK guys, I considered the posts, reviewed the code, then found > and fixed a naive bug introduced by me on solving issues for LP64 > processors. > > It's in the CVS now. > > As I want to release 0.8.4 tomorrow. Please test it like this: > > - Grab the CVS, and if it works now, report ASAP & we're done. > - If CVS doesn't work, test 0.8.3 on and report, without the > slightest shade of doubt, whether it works ;) > - If 0.8.3 works, shove 0.8.3's cookies.c into 0.8.4 and test. > If this hybrid doesn't work, then it may be gremlins... FINALLY I get amazon.com on both 8.3 and 8.4-rc1, but yahoo.com on neither with the same 'cookies rejected' fault message. next ??? and thanks, -- Patrick Shanahan Registered Linux User #207535 http://wahoo.no-ip.org @ http://counter.li.org HOG # US1244711 Photo Album: http://wahoo.no-ip.org/photos

Jan. 11, 2005

Re: [Dillo-dev]CAN-2005-12

by Patrick Shanahan

* Roberto Sanchez <roberto(a)familiasanchez.net> [01-10-05 18:57]: > What you want is checkinstall: have it and use it, but have a spec for dillo and would rather generate a *proper* SuSE rpm rather than use checkinstall when possible. thanks, -- Patrick Shanahan Registered Linux User #207535 http://wahoo.no-ip.org @ http://counter.li.org HOG # US1244711 Photo Album: http://wahoo.no-ip.org/photos

Jan. 11, 2005

Re: [Dillo-dev]CAN-2005-12

by Patrick Shanahan

* Patrick Shanahan <ptilopteri(a)gmail.com> [01-10-05 18:42]: > OK, I have the cvs but am unable to build it for my rpm system and I > need to maintain rpm's (not savey enough to just use make/... and > maintain the rest of my system). > > I dl'ed the cvs, bzipped, modified dillo.spec to match the bz2 > file-name and get the following error when building the rpm for install: > > + ./configure --prefix=/usr --sysconfdir=/etc > /var/tmp/rpm-tmp.91241: line 23: ./configure: No such file or directory > error: Bad exit status from /var/tmp/rpm-tmp.91241 (%build) > > RPM build errors: > Bad exit status from /var/tmp/rpm-tmp.91241 (%build) > > > ps: I had no trouble with the package (bz2) you provided for rc1 OK, shows I am the *rank* amateur that I am.... Have to generate the configure file. Did so and am compiling now on my slow 733. Will report results shortly. sorry and thanks, -- Patrick Shanahan Registered Linux User #207535 http://wahoo.no-ip.org @ http://counter.li.org HOG # US1244711 Photo Album: http://wahoo.no-ip.org/photos

Jan. 10, 2005

Re: [Dillo-dev]CAN-2005-12

by Roberto Sanchez

Quoting Patrick Shanahan <ptilopteri(a)gmail.com>: > OK, I have the cvs but am unable to build it for my rpm system and I > need to maintain rpm's (not savey enough to just use make/... and > maintain the rest of my system). > What you want is checkinstall: http://asic-linux.com.mx/~izto/checkinstall/ You do the standard: $ ./configure $ make Then, instead of the normal 'make install', your run: $ sudo checkinstall make install Checkinstall will take and monitor the install process, then create a .rpm (or .deb or Slackware .tgz) package that you can then manage with your favorite package management tool. It's much quicker and easier when you are testing and developing. -Roberto ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program.

Jan. 10, 2005

Re: [Dillo-dev]CAN-2005-12

by robert w hall

In article <20050110222427.GA1465(a)dillo.org>, Jorge Arellano Cid <jcid(a)dillo.org> writes > OK guys, I considered the posts, reviewed the code, then found >and fixed a naive bug introduced by me on solving issues for LP64 >processors. > > It's in the CVS now. > > just tried the x86/ssl version out of cvs. Now does amazon.co.uk and yahoo groups OK. -- robert w hall

Jan. 10, 2005

Re: [Dillo-dev]CAN-2005-12

by Patrick Shanahan

* Jorge Arellano Cid <jcid(a)dillo.org> [01-10-05 17:33]: > > It's in the CVS now. > > As I want to release 0.8.4 tomorrow. Please test it like this: > > - Grab the CVS, and if it works now, report ASAP & we're done. > - If CVS doesn't work, test 0.8.3 on and report, without the > slightest shade of doubt, whether it works ;) > - If 0.8.3 works, shove 0.8.3's cookies.c into 0.8.4 and test. > If this hybrid doesn't work, then it may be gremlins... > > Hopefully the first step will solve it. OK, I have the cvs but am unable to build it for my rpm system and I need to maintain rpm's (not savey enough to just use make/... and maintain the rest of my system). I dl'ed the cvs, bzipped, modified dillo.spec to match the bz2 file-name and get the following error when building the rpm for install: + ./configure --prefix=/usr --sysconfdir=/etc /var/tmp/rpm-tmp.91241: line 23: ./configure: No such file or directory error: Bad exit status from /var/tmp/rpm-tmp.91241 (%build) RPM build errors: Bad exit status from /var/tmp/rpm-tmp.91241 (%build) ps: I had no trouble with the package (bz2) you provided for rc1 -- Patrick Shanahan Registered Linux User #207535 http://wahoo.no-ip.org @ http://counter.li.org HOG # US1244711 Photo Album: http://wahoo.no-ip.org/photos

Jan. 10, 2005

Re: [Dillo-dev]CAN-2005-12

by Jorge Arellano Cid

On Mon, Jan 10, 2005 at 04:15:43PM +0000, robert w hall wrote: > In article <20050110152923.GE5959(a)wahoo.no-ip.org>, Patrick Shanahan > <ptilopteri(a)gmail.com> writes > >* robert w hall <bobh(a)n-cantrell.demon.co.uk> [01-10-05 09:46]: > >> I've now built the x86 version with ssl enabled - login to ps2 site > >> appears OK, but I can't get into amazon.co.uk or groups.yahoo.com, and > >> could with 0.8.3 (might still be a dirt effect, see post on > >> ps2-version, but I think everything is now referenced wrt > >> /usr/local...) > > > >I have no problem with ssl enabled accessing amazon.com, > > yes, I don't have a problem with amazon.com, strangely enough it's > amazon.co.uk that gives trouble > my cookiesrc is set to DEFAULT ACCEPT OK guys, I considered the posts, reviewed the code, then found and fixed a naive bug introduced by me on solving issues for LP64 processors. It's in the CVS now. As I want to release 0.8.4 tomorrow. Please test it like this: - Grab the CVS, and if it works now, report ASAP & we're done. - If CVS doesn't work, test 0.8.3 on and report, without the slightest shade of doubt, whether it works ;) - If 0.8.3 works, shove 0.8.3's cookies.c into 0.8.4 and test. If this hybrid doesn't work, then it may be gremlins... Hopefully the first step will solve it. -- Cheers Jorge.-

Jan. 10, 2005

Re: [Dillo-dev]CAN-2005-12

by robert w hall

In article <20050110152923.GE5959(a)wahoo.no-ip.org>, Patrick Shanahan <ptilopteri(a)gmail.com> writes >* robert w hall <bobh(a)n-cantrell.demon.co.uk> [01-10-05 09:46]: >> I've now built the x86 version with ssl enabled - login to ps2 site >> appears OK, but I can't get into amazon.co.uk or groups.yahoo.com, and >> could with 0.8.3 (might still be a dirt effect, see post on >> ps2-version, but I think everything is now referenced wrt >> /usr/local...) > >I have no problem with ssl enabled accessing amazon.com, yes, I don't have a problem with amazon.com, strangely enough it's amazon.co.uk that gives trouble my cookiesrc is set to DEFAULT ACCEPT -- robert w hall

Jan. 10, 2005

Re: [Dillo-dev]CAN-2005-12

by Patrick Shanahan

* robert w hall <bobh(a)n-cantrell.demon.co.uk> [01-10-05 09:46]: > I've now built the x86 version with ssl enabled - login to ps2 site > appears OK, but I can't get into amazon.co.uk or groups.yahoo.com, and > could with 0.8.3 (might still be a dirt effect, see post on > ps2-version, but I think everything is now referenced wrt > /usr/local...) I have no problem with ssl enabled accessing amazon.com, but cannot sign in to yahoo, "The browser you're using refuses to sign in. (cookies rejected)". I *have* all cookies accepted and specified .yahoo.com ACCEPT. ??? -- Patrick Shanahan Registered Linux User #207535 http://wahoo.no-ip.org @ http://counter.li.org HOG # US1244711 Photo Album: http://wahoo.no-ip.org/photos

Jan. 10, 2005

Re: [Dillo-dev]CAN-2005-12

by robert w hall

In article <20050110111014.GA2608(a)dillo.org>, Jorge Arellano Cid <jcid(a)dillo.org> writes >On Mon, Jan 10, 2005 at 09:17:13AM +0000, Jeremy Henty wrote: >> >> On Sun, Jan 09, 2005 at 10:58:04AM -0500, Patrick Shanahan wrote: >> >> > I have found no anomalies with the rc1 issue other than the failure >> > of yahoo.com to be able to find it's cookies ??? >> >> Are there known cookie problems? I'm asking because I can't create a >> login on the Doom wiki (doom.wikicities.com) Even if I set "DEFAULT >> ACCEPT" in cookiesrc the cookies don't appear. My account at >> wikipedia.org works fine though. >> >> Cheers, > > Does it work with 0.8.3? > I've now built the x86 version with ssl enabled - login to ps2 site appears OK, but I can't get into amazon.co.uk or groups.yahoo.com, and could with 0.8.3 (might still be a dirt effect, see post on ps2-version, but I think everything is now referenced wrt /usr/local...) -- robert w hall

Jan. 10, 2005