From harry.lin at coretronic.com Wed Apr 11 03:41:39 2007 From: harry.lin at coretronic.com (=?big5?B?SGFycnkgTGluIEQzMzIgKKSkpfq5cSk=?=) Date: Wed, 11 Apr 2007 17:41:39 +0800 Subject: [links] Chinese font support? Message-ID: Hi, Does Links support for Chinese font now? In the previous discussion, Someone mentioned to create the Chinese characters to png files But it is difficult to realize Is there any other method? Thanks! From clock at twibright.com Wed Apr 11 05:41:44 2007 From: clock at twibright.com (Karel Kulhavy) Date: Wed, 11 Apr 2007 13:41:44 +0200 Subject: [links] Chinese font support? In-Reply-To: References: Message-ID: <20070411114144.GB7023@kestrel.barix.local> On Wed, Apr 11, 2007 at 05:41:39PM +0800, Harry Lin D332 (?????q) wrote: > Hi, > > Does Links support for Chinese font now? No only Japanese I think. > > In the previous discussion, > > Someone mentioned to create the Chinese characters to png files > > But it is difficult to realize > > Is there any other method? It has to be in PNG. But you can create using some scripts from Ghostscript. It's described here: http://links.twibright.com/doc/links_doc_en.html Only the font part was translated into English, ignore the rest in Czech. CL< > > Thanks! > _______________________________________________ > links-list mailing list > links-list at linuxfromscratch.org > http://linuxfromscratch.org/mailman/listinfo/links-list From pierre.haderer at enst-bretagne.fr Wed Apr 11 07:23:27 2007 From: pierre.haderer at enst-bretagne.fr (pierre.haderer at enst-bretagne.fr) Date: Wed, 11 Apr 2007 15:23:27 +0200 Subject: [links] french student starting his work on links Message-ID: <20070411152327.ee7coksbccs080cc@webmail.enst-bretagne.fr> hello, My name is pierre, I'm a french student and I'm expected to use links (the graphic version) to a system called TAPA : http://departements.enst-bretagne.fr/info/recherche/sid/tapa.php (Sorry, this page is in french and had not been translated) The goal of this project is to give aged people an easier access to their mails and their documents than the one given by a regular computer. Links would be the browser that allow these operations. My first step would be to understand links' code, so I downloaded it but I don't really know where to start my investigation. Can someone tell me where I should start reading the code (which files) and maybe give me some complementary information on the way links is built ? Thank you everybody for your attention. -- Pierre Haderer From clock at twibright.com Wed Apr 11 09:24:10 2007 From: clock at twibright.com (Karel Kulhavy) Date: Wed, 11 Apr 2007 17:24:10 +0200 Subject: [links] french student starting his work on links In-Reply-To: <20070411152327.ee7coksbccs080cc@webmail.enst-bretagne.fr> References: <20070411152327.ee7coksbccs080cc@webmail.enst-bretagne.fr> Message-ID: <20070411152410.GB25684@kestrel.barix.local> On Wed, Apr 11, 2007 at 03:23:27PM +0200, pierre.haderer at enst-bretagne.fr wrote: > hello, > > My name is pierre, I'm a french student and I'm expected to use links > (the graphic version) to a system called TAPA : > http://departements.enst-bretagne.fr/info/recherche/sid/tapa.php > (Sorry, this page is in french and had not been translated) > The goal of this project is to give aged people an easier access to > their mails and their documents than the one given by a regular > computer. > > Links would be the browser that allow these operations. > > My first step would be to understand links' code, so I downloaded it > but I don't really know where to start my investigation. > > Can someone tell me where I should start reading the code (which > files) and maybe give me some complementary information on the way > links is built ? main.c main() and look into select_loop that's the main loop. Look also on links.twibright.com, Downloads, there is Project Documentation in English. CL< > > Thank you everybody for your attention. > > -- > Pierre Haderer > > _______________________________________________ > links-list mailing list > links-list at linuxfromscratch.org > http://linuxfromscratch.org/mailman/listinfo/links-list From filipi at pucrs.br Wed Apr 11 10:23:08 2007 From: filipi at pucrs.br (Filipi Vianna) Date: Wed, 11 Apr 2007 13:23:08 -0300 Subject: [links] Post confirmation Message-ID: <1176308588.28211.4.camel@europa> Hi folks, How do I configure links to not show the warning dialog box asking confirmation to POST form data? Regards, -- Filipi Vianna Computational Mechanics Laboratory (DEMM) College of Engineering - PUCRS +55 51 33203584 From edrosten at lanl.gov Wed Apr 11 18:20:14 2007 From: edrosten at lanl.gov (Edward Rosten) Date: Wed, 11 Apr 2007 18:20:14 -0600 (MDT) Subject: [links] New font generation scripts (with automatic accent merging!) Message-ID: Hi all, I've attached a new set of font generation scripts. The main end difference is that this system automatically merges accents over the glyphs, and doesn't require manual GIMPing afterwards. run as follows: put all the files in a directory and do: ./fontgen latin.map CenturySchL-Italic to generate a latin font for New Century Schoolbook, italc. The output glyphs are in ./tmp/ I've tested it with italic font, by copying the glyphs in to links/graphics/font/century_school-medium-italic-serif-vari and using the aliases file from century_school-bold-roman-serif-vari, and running generate_font. My links now has an italic font. The font is not complete, since I only have created the latin map (copied from clip.c), and not all the accent positions are correct, but it's a pretty decent start. If you create any updated maps (and accent positions), or need any help, please let me know! It works as follows: A .map file (only latin.map is provided, but it is derived from clip.c) specifies mappings and merges. The format should be fairly self explanatory. In the mapping section, everything after the 3rd column is ignored. In the merge section, the 4th column (if present) specifies the vertival alignment of the accent and the 5th column (if present) specifies the horizontal alignment. If they are not present, the system makes a guess, which is reasonable in the majority of cases. The possible vertical alignments are: bottom Print the fonts on the same baseline (accents align properly with lower case letters) capital Push the glyph up a bit so that accents align with capital letters glmid Strip all space from around the accent, and place the glyph in the lower middle gmid As above, but placed in the middle gumid As above, but in the upper middle region The possible horizontal alignments are: left Aligh the left of the accent and letter cells (includes white space) right As above, but right aligned right centre As above but centred. More alignments could be added easily, but I haven't done it yet. I used the parameters from the current font generating system in fontgen, but for some reason, the sizes came out a bit different, so it has slightly different parameters. -Ed -------------- next part -------------- mkdir tmp export mag=4 #was 1703 export hundred_dpi=1850 export height=120 export top=43 hdpi=`expr $hundred_dpi '*' $mag '/' 100` vdpi=`expr $hundred_dpi '*' $mag '/' 100` awk -f process_mapping.awk $1 | while read ch acc va ha out do fout=tmp/$out.png gs -sDEVICE=pgmraw -q -dNOPAUSE -dBATCH -sPAPERSIZE=a4 -r"$hdpi"x"$vdpi" \ -dCH=$ch -dACC=$acc -dh_align="($ha)" -dv_align="($va)" -dFN="/$2" \ -sOutputFile=- letters.ps | pnmcrop -left -right -white | pnminvert | pnmscale -reduce $mag | pnmcut -height $height -top $top | pnmtopng > $fout #pnmcrop fails if the image is all background. if ! pngtopnm $fout | pnmcrop > /dev/null then rm $fout fi echo $fout done -------------- next part -------------- MAPPING 161 0xa1 162 0xa2 163 0xa3 165 0xa5 167 0xa7 171 0xab 172 0x2039 173 0x203a 174 0xfb01 175 0xfb02 177 0x2013 178 0x2020 179 0x2021 182 0xb6 183 0x2022 187 0xbb 188 0x2026 189 0x2030 191 0xbf 196 0x2dc 208 0x2014 225 0xc6 232 0x141 233 0xd8 234 0x152 235 0x2116 241 0xe6 248 0x142 249 0xf8 250 0x153 251 0xdf 193 0x300 GRA /* ` */ 194 0xb4 ACU /* ' */ 195 0x2c6 CIR /* ^ */ 196 0x2dc TIL /* ~ */ 197 0xaf MAC /* - above */ 198 0x306 LSC /* lower semicircle above */ 199 0x2d9 199 0xb7 DOT /* . above */ 200 0xa8 UML /* .. above */ 202 0x2da 202 0xb0 RIN /* o above */ 203 0xb8 CED /* (5 without the top stroke) under */ 205 0x30b DAC /* " */ 206 0x2db OGO /* left-right mirrored J under (ogonek) */ 207 0x2c7 CAR /* v above */ MERGE + - 0xb1 glmid A GRA 0xc0 A ACU 0xc1 A CIR 0xc2 A TIL 0xc3 A UML 0xc4 A RIN 0xc5 C CED 0xc7 bottom E GRA 0xc8 E ACU 0xc9 E CIR 0xca E UML 0xcb I GRA 0xcc I ACU 0xcd I CIR 0xce I UML 0xcf D - 0xd0 bottom left N TIL 0xd1 O GRA 0xd2 O ACU 0xd3 O CIR 0xd4 O TIL 0xd5 O UML 0xd6 U GRA 0xd9 U ACU 0xda U CIR 0xdb U UML 0xdc Y ACU 0xdd a GRA 0xe0 a ACU 0xe1 a CIR 0xe2 a TIL 0xe3 a UML 0xe4 a RIN 0xe5 c CED 0xe7 bottom e GRA 0xe8 e ACU 0xe9 e CIR 0xea e UML 0xeb i GRA 0xec i ACU 0xed i CIR 0xee i UML 0xef n TIL 0xf1 o GRA 0xf2 o ACU 0xf3 o CIR 0xf4 o TIL 0xf5 o UML 0xf6 u GRA 0xf9 u ACU 0xfa u CIR 0xfb u UML 0xfc y ACU 0xfd y UML 0xff A MAC 0x100 a MAC 0x101 A LSC 0x102 a LSC 0x103 A OGO 0x104 bottom right a OGO 0x105 bottom right C ACU 0x106 c ACU 0x107 C CIR 0x108 c CIR 0x109 C DOT 0x10a c DOT 0x10b C CAR 0x10c c CAR 0x10d D CAR 0x10e d 0x27 0x10f D - 0x110 bottom left d - 0x111 glmid right E MAC 0x112 e MAC 0x113 E LSC 0x114 e LSC 0x115 E DOT 0x116 e DOT 0x117 E OGO 0x118 bottom e OGO 0x119 bottom E CAR 0x11a e CAR 0x11b G CIR 0x11c g CIR 0x11d G LSC 0x11e g LSC 0x11f G DOT 0x120 g DOT 0x121 H CIR 0x124 h CIR 0x125 H _ 0x126 glmid h - 0x127 glmid left I TIL 0x128 i TIL 0x129 I MAC 0x12a i MAC 0x12b I LSC 0x12c i LSC 0x12d I OGO 0x12e bottom i OGO 0x12f bottom I DOT 0x130 I J 0x132 bottom right i j 0x133 bottom right J CIR 0x134 j CIR 0x135 L ACU 0x139 l ACU 0x13a L 0x27 0x13d l 0x27 0x13e N ACU 0x143 n ACU 0x144 N CAR 0x147 n CAR 0x148 O MAC 0x14c o MAC 0x14d O LSC 0x14e o LSC 0x14f O DAC 0x150 o DAC 0x151 R ACU 0x154 r ACU 0x155 R CAR 0x158 r CAR 0x159 S ACU 0x15a s ACU 0x15b S CIR 0x15c s CIR 0x15d S CED 0x15e bottom s CED 0x15f bottom S CAR 0x160 s CAR 0x161 T CED 0x162 bottom t CED 0x163 bottom T CAR 0x164 t 0x27 0x165 T - 0x166 t - 0x167 U TIL 0x168 u TIL 0x169 U MAC 0x16a u MAC 0x16b U LSC 0x16c u LSC 0x16d U RIN 0x16e u RIN 0x16f U DAC 0x170 u DAC 0x171 U OGO 0x172 bottom u OGO 0x173 bottom W CIR 0x174 w CIR 0x175 Y CIR 0x176 y CIR 0x177 Y UML 0x178 Z ACU 0x179 z ACU 0x17a Z DOT 0x17b z DOT 0x17c Z CAR 0x17d z CAR 0x17e A CAR 0x1cd a CAR 0x1ce I CAR 0x1cf i CAR 0x1d0 O CAR 0x1d1 o CAR 0x1d2 U CAR 0x1d3 u CAR 0x1d4 G CAR 0x1e6 g CAR 0x1e7 K CAR 0x1e8 k CAR 0x1e9 O OGO 0x1ea bottom o OGO 0x1eb bottom j CAR 0x1f0 G ACU 0x1f4 g ACU 0x1f5 N GRA 0x1f8 n GRA 0x1f9 -------------- next part -------------- A non-text attachment was scrubbed... Name: letters.ps Type: application/postscript Size: 2901 bytes Desc: Url : http://linuxfromscratch.org/pipermail/links-list/attachments/20070411/09683ba0/attachment-0001.ps -------------- next part -------------- BEGIN{ for(i=0; i < 256; i++) { ord[sprintf("%c", i)] = i map[i] = i } } NF==0{next} /MAPPING/{ is_mapping=1 next } /MERGE/{ for(i=0; i < 256; i++) print i, 0, "bottom", "left", sprintf("%04x", map[i]) is_merge=1 next } is_mapping{ map[$1] = strtonum($2) if($3 != "") defs[$3] = $1+0 } is_merge{ if($2 in defs) acc=defs[$2] else if($2~/^0x/) acc = strtonum($2) else acc=ord[$2] out=strtonum($3) if($4 == "") if($1~/[A-Z]/) valign="capital" else valign="bottom" else valign=$4 if($5 == "") halign="centre" else halign=$5 print ord[$1], acc, valign, halign, sprintf("%04x", out) } From harry.lin at coretronic.com Wed Apr 11 19:32:16 2007 From: harry.lin at coretronic.com (=?big5?B?SGFycnkgTGluIEQzMzIgKKSkpfq5cSk=?=) Date: Thu, 12 Apr 2007 09:32:16 +0800 Subject: [links] Chinese font support? In-Reply-To: <20070411114144.GB7023@kestrel.barix.local> Message-ID: Thank you for your answer, I have read that paragraph about adding fonts from ghostscript It only mentioned how to add a font in format *.pfb, *.afm, *.pfa, Does ghostscript support a font in format *.ttf, too? And I have another question, A Chinese font contains more than ten thousand characters, After converting it to png images, it will generate a big amount of files, And will all those files be compiled into the links browser? Won't this make the browser binary file too big and cost a lot of memory when running links? -----Original Message----- From: links-list-bounces at linuxfromscratch.org [mailto:links-list-bounces at linuxfromscratch.org] On Behalf Of Karel Kulhavy Sent: Wednesday, April 11, 2007 7:42 PM To: Links WWW browser mailing list Subject: Re: [links] Chinese font support? On Wed, Apr 11, 2007 at 05:41:39PM +0800, Harry Lin D332 (?????q) wrote: > Hi, > > Does Links support for Chinese font now? No only Japanese I think. > > In the previous discussion, > > Someone mentioned to create the Chinese characters to png files > > But it is difficult to realize > > Is there any other method? It has to be in PNG. But you can create using some scripts from Ghostscript. It's described here: http://links.twibright.com/doc/links_doc_en.html Only the font part was translated into English, ignore the rest in Czech. CL< > > Thanks! > _______________________________________________ > links-list mailing list > links-list at linuxfromscratch.org > http://linuxfromscratch.org/mailman/listinfo/links-list _______________________________________________ links-list mailing list links-list at linuxfromscratch.org http://linuxfromscratch.org/mailman/listinfo/links-list From clock at twibright.com Wed Apr 11 21:14:58 2007 From: clock at twibright.com (Karel Kulhavy) Date: Thu, 12 Apr 2007 05:14:58 +0200 Subject: [links] Chinese font support? In-Reply-To: References: <20070411114144.GB7023@kestrel.barix.local> Message-ID: <20070412031458.GC29101@kestrel.barix.local> On Thu, Apr 12, 2007 at 09:32:16AM +0800, Harry Lin D332 (?????q) wrote: > Thank you for your answer, > I have read that paragraph about adding fonts from ghostscript > It only mentioned how to add a font in format *.pfb, *.afm, *.pfa, > Does ghostscript support a font in format *.ttf, too? Not that I would know. > > And I have another question, > A Chinese font contains more than ten thousand characters, > After converting it to png images, it will generate a big amount of files, > And will all those files be compiled into the links browser? Yes. > Won't this make the browser binary file too big and cost a lot of memory when > running links? No - the binary should be mapped read-only, i. e. swapping out of memory can be done in zero time (just delete the page) as opposed to swapping data where a page has to be first written to the disk. It's an equivalent of mapping the font data using mmap. CL< > > -----Original Message----- > From: links-list-bounces at linuxfromscratch.org [mailto:links-list-bounces at linuxfromscratch.org] On Behalf Of Karel Kulhavy > Sent: Wednesday, April 11, 2007 7:42 PM > To: Links WWW browser mailing list > Subject: Re: [links] Chinese font support? > > On Wed, Apr 11, 2007 at 05:41:39PM +0800, Harry Lin D332 (?????q) wrote: > > Hi, > > > > Does Links support for Chinese font now? > > No only Japanese I think. > > > > > In the previous discussion, > > > > Someone mentioned to create the Chinese characters to png files > > > > But it is difficult to realize > > > > Is there any other method? > > It has to be in PNG. But you can create using some scripts from Ghostscript. > It's described here: > http://links.twibright.com/doc/links_doc_en.html > > Only the font part was translated into English, ignore the rest in Czech. > > CL< > > > > Thanks! > > _______________________________________________ > > links-list mailing list > > links-list at linuxfromscratch.org > > http://linuxfromscratch.org/mailman/listinfo/links-list > _______________________________________________ > links-list mailing list > links-list at linuxfromscratch.org > http://linuxfromscratch.org/mailman/listinfo/links-list > > _______________________________________________ > links-list mailing list > links-list at linuxfromscratch.org > http://linuxfromscratch.org/mailman/listinfo/links-list From yan at seiner.com Wed Apr 11 22:31:06 2007 From: yan at seiner.com (Yan Seiner) Date: Wed, 11 Apr 2007 21:31:06 -0700 Subject: [links] Links and javascript Message-ID: <461DB60A.9000009@seiner.com> Does links support javascript in forms now? That would be great! I need to be able to change the values in form fields via javascript. Can links do this? --Yan From clock at twibright.com Thu Apr 12 00:28:04 2007 From: clock at twibright.com (Karel Kulhavy) Date: Thu, 12 Apr 2007 08:28:04 +0200 Subject: [links] Links and javascript In-Reply-To: <461DB60A.9000009@seiner.com> References: <461DB60A.9000009@seiner.com> Message-ID: <20070412062804.GA30446@kestrel.barix.local> On Wed, Apr 11, 2007 at 09:31:06PM -0700, Yan Seiner wrote: > Does links support javascript in forms now? Javascript in Lins is soon going to be discontinued because is buggy beyond repair :( CL< > > That would be great! > > I need to be able to change the values in form fields via javascript. > Can links do this? > > --Yan > _______________________________________________ > links-list mailing list > links-list at linuxfromscratch.org > http://linuxfromscratch.org/mailman/listinfo/links-list From kali at flowtrack.com.au Thu Apr 12 21:02:20 2007 From: kali at flowtrack.com.au (Kali) Date: Thu, 12 Apr 2007 22:02:20 -0500 (CDT) Subject: [links] french student starting his work on links In-Reply-To: <20070411152327.ee7coksbccs080cc@webmail.enst-bretagne.fr> Message-ID: Dear Pierre: I applaud your project. The biggest problem my 91 year old father has writing emails is with mouse control and seeing where the focus is. Links runs fast in text on an obsolete computer with 2.2.6 kernel. With 34 lines set at boot, and colour turned on, the result is quite pretty! Just now I got prerelease28 going with current openssl and javascript enabled but still cant get into some https sites that claim I have cookies turned off. Also my bank has just started requiring a second level of authentication involving clicking on pictures of fruit and things....grrrrr. I might have to face svgalib again after bad experiences. Regards Kali From yan at seiner.com Thu Apr 12 08:20:27 2007 From: yan at seiner.com (Yan Seiner) Date: Thu, 12 Apr 2007 07:20:27 -0700 Subject: [links] Links and javascript In-Reply-To: <20070412062804.GA30446@kestrel.barix.local> References: <461DB60A.9000009@seiner.com> <20070412062804.GA30446@kestrel.barix.local> Message-ID: <461E402B.8050001@seiner.com> Karel Kulhavy napsal(a): > On Wed, Apr 11, 2007 at 09:31:06PM -0700, Yan Seiner wrote: > >> Does links support javascript in forms now? >> > > Javascript in Lins is soon going to be discontinued because is buggy > beyond repair :( > So I take it DOM is not implemented either? Bummer! --Yan From edrosten at lanl.gov Thu Apr 12 10:35:46 2007 From: edrosten at lanl.gov (Edward Rosten) Date: Thu, 12 Apr 2007 10:35:46 -0600 (MDT) Subject: [links] Links and javascript In-Reply-To: <20070412062804.GA30446@kestrel.barix.local> References: <461DB60A.9000009@seiner.com> <20070412062804.GA30446@kestrel.barix.local> Message-ID: On Thu, 12 Apr 2007, Karel Kulhavy wrote: > On Wed, Apr 11, 2007 at 09:31:06PM -0700, Yan Seiner wrote: >> Does links support javascript in forms now? > > Javascript in Lins is soon going to be discontinued because I will be sad to see it go, it's one of the things that makes links usable as a lightweight browser. > is buggy beyond repair :( I have started looking at Javascript interpreter libraries. Unfortunately, I don't understand Links well enough to try integrating them. -Ed From yan at seiner.com Thu Apr 12 10:49:19 2007 From: yan at seiner.com (Yan Seiner) Date: Thu, 12 Apr 2007 09:49:19 -0700 Subject: [links] Links and javascript In-Reply-To: References: <461DB60A.9000009@seiner.com> <20070412062804.GA30446@kestrel.barix.local> Message-ID: <461E630F.1030606@seiner.com> Edward Rosten wrote: > On Thu, 12 Apr 2007, Karel Kulhavy wrote: > > >> On Wed, Apr 11, 2007 at 09:31:06PM -0700, Yan Seiner wrote: >> >>> Does links support javascript in forms now? >>> >> Javascript in Lins is soon going to be discontinued because >> > > I will be sad to see it go, it's one of the things that makes links usable > as a lightweight browser. > > >> is buggy beyond repair :( >> > > I have started looking at Javascript interpreter libraries. Unfortunately, > I don't understand Links well enough to try integrating them. > Look at libsee. The problem is that links would have to generate a DOM for javascript to work. You may be able to look at hv3 to get an idea of how that's done. It's written in Tk/Tcl and C. http://tkhtml.tcl.tk/hv3.html I desparately need a lightweight javascript enabled browser. I thought we had a solution with konq/e, but it's just too slow and big. I've not looked through links enough to see just how much work is involved in creating a dom.... --Yan From edrosten at lanl.gov Thu Apr 12 11:17:18 2007 From: edrosten at lanl.gov (Edward Rosten) Date: Thu, 12 Apr 2007 11:17:18 -0600 (MDT) Subject: [links] Links and javascript In-Reply-To: <461E630F.1030606@seiner.com> References: <461DB60A.9000009@seiner.com> <20070412062804.GA30446@kestrel.barix.local> <461E630F.1030606@seiner.com> Message-ID: On Thu, 12 Apr 2007, Yan Seiner wrote: > Edward Rosten wrote: >> On Thu, 12 Apr 2007, Karel Kulhavy wrote: >> >> >>> On Wed, Apr 11, 2007 at 09:31:06PM -0700, Yan Seiner wrote: >>> >>>> Does links support javascript in forms now? >>>> >>> Javascript in Lins is soon going to be discontinued because >>> >> >> I will be sad to see it go, it's one of the things that makes links usable >> as a lightweight browser. >> >> >>> is buggy beyond repair :( >>> >> >> I have started looking at Javascript interpreter libraries. Unfortunately, >> I don't understand Links well enough to try integrating them. >> > Look at libsee. The problem is that links would have to generate a DOM > for javascript to work. That's for javascript to work fully. However even without a DOM, links still has some ability to work on those really stupid websites which replace all normak links with dodgy javascript code. I really don't want to loose the ability to do that. > You may be able to look at hv3 to get an idea of how that's done. It's > written in Tk/Tcl and C. http://tkhtml.tcl.tk/hv3.html There's also Dillo. It's got CSS support, so I think it must hace a DOM for that to work. I tjink it's written in a mixture of C and C++. > I desparately need a lightweight javascript enabled browser. I thought > we had a solution with konq/e, but it's just too slow and big. > I've not looked through links enough to see just how much work is > involved in creating a dom.... -Ed From yan at seiner.com Thu Apr 12 11:38:26 2007 From: yan at seiner.com (Yan Seiner) Date: Thu, 12 Apr 2007 10:38:26 -0700 Subject: [links] Links and javascript In-Reply-To: References: <461DB60A.9000009@seiner.com> <20070412062804.GA30446@kestrel.barix.local> <461E630F.1030606@seiner.com> Message-ID: <461E6E92.80307@seiner.com> Edward Rosten wrote: > >> k at libsee. The problem is that links would have to generate a DOM >> for javascript to work. >> > > That's for javascript to work fully. However even without a DOM, links > still has some ability to work on those really stupid websites which > replace all normak links with dodgy javascript code. > > I really don't want to loose the ability to do that. > > I see. I have a different need; I am working on an embedded controller that needs a browser that can respond to user input. >> You may be able to look at hv3 to get an idea of how that's done. It's >> written in Tk/Tcl and C. http://tkhtml.tcl.tk/hv3.html >> > > There's also Dillo. It's got CSS support, so I think it must hace a DOM > for that to work. I tjink it's written in a mixture of C and C++. > Dillo uses GTK+, which pretty much knocks it out of my league.... GTK+ is just too big for my little machine... ISTR that dillo doesn't have DOM; it's on the todo list. :-( Looking through links code, I may be able to hardcode my UI stuff.... I have to get a better understanding of the input routines. --Yan From yan at seiner.com Thu Apr 12 11:59:12 2007 From: yan at seiner.com (Yan Seiner) Date: Thu, 12 Apr 2007 10:59:12 -0700 Subject: [links] Customizing links (input help) Message-ID: <461E7370.2020001@seiner.com> I need a bit of guidance. We're currently using a javascript frontend which takes keystrokes coming from a keypad and changes the values in a form. For example, the letter 'a' results in a numerical field being incremented by 1. I've looked at the code in links, and I think I can modify it to do what we need, but I need a bit of help with the input routines. I can't quite figure out how the keystrokes get into links, and more importantly, where in links do they go to update a form field. I have the other stuff figured out, but the 'field editor' is the part I am missing.... --Yan From phil.ganchev at gmail.com Thu Apr 12 12:35:15 2007 From: phil.ganchev at gmail.com (Philip Ganchev) Date: Thu, 12 Apr 2007 14:35:15 -0400 Subject: [links] Links and javascript In-Reply-To: <461E630F.1030606@seiner.com> References: <461DB60A.9000009@seiner.com> <20070412062804.GA30446@kestrel.barix.local> <461E630F.1030606@seiner.com> Message-ID: <3aba28080704121135g7e855e3fl96b1a09533e9741@mail.gmail.com> On 4/12/07, Yan Seiner wrote: > Edward Rosten wrote: > > On Thu, 12 Apr 2007, Karel Kulhavy wrote: [...] > >> Javascript in Lins is soon going to be discontinued because [...] > > I will be sad to see it go, it's one of the things that makes links usable > > as a lightweight browser. [...] > I desparately need a lightweight javascript enabled browser. I thought > we had a solution with konq/e, but it's just too slow and big. I really need one too. I had not seen hv3 before, and it seems that there are plans to support javascript! But that may take time. And I guess hv3 is too big for your machine too? For users who can run a large, slow browser and occasionally need javascript support, it would be great to have a Links menu item that opens the current page in a user-specified browser. Then you can surf fast until you need javascript. As a work-around, you can press "Alt+S C" to select the current URL, but it is slow to paste the URL into the new browser using the mouse. From yan at seiner.com Thu Apr 12 12:48:54 2007 From: yan at seiner.com (Yan Seiner) Date: Thu, 12 Apr 2007 11:48:54 -0700 Subject: [links] Links and javascript In-Reply-To: <3aba28080704121135g7e855e3fl96b1a09533e9741@mail.gmail.com> References: <461DB60A.9000009@seiner.com> <20070412062804.GA30446@kestrel.barix.local> <461E630F.1030606@seiner.com> <3aba28080704121135g7e855e3fl96b1a09533e9741@mail.gmail.com> Message-ID: <461E7F16.2010807@seiner.com> Philip Ganchev wrote: > On 4/12/07, Yan Seiner wrote: > >> Edward Rosten wrote: >> >>> On Thu, 12 Apr 2007, Karel Kulhavy wrote: >>> > [...] > >>>> Javascript in Lins is soon going to be discontinued because >>>> > [...] > >>> I will be sad to see it go, it's one of the things that makes links usable >>> as a lightweight browser. >>> > [...] > >> I desparately need a lightweight javascript enabled browser. I thought >> we had a solution with konq/e, but it's just too slow and big. >> > > I really need one too. I had not seen hv3 before, and it seems that > there are plans to support javascript! But that may take time. And I > guess hv3 is too big for your machine too? > Well, Tk seems to have a problem with my X system - it shows up as all black... :-( And the browser segfaults once it gets done reading in the page. :-(( OTOH, the developers are active and they need / want interest and users, so go for it! hv3 actually has partial javascript support and it works OK on my desktop. (Hey, since you're a CS student/prof/staff, how about getting someone to add DOM/js to links! Only half kidding....) --Yan From clock at twibright.com Thu Apr 12 12:53:40 2007 From: clock at twibright.com (Karel Kulhavy) Date: Thu, 12 Apr 2007 20:53:40 +0200 Subject: [links] Links and javascript In-Reply-To: <461E6E92.80307@seiner.com> References: <461DB60A.9000009@seiner.com> <20070412062804.GA30446@kestrel.barix.local> <461E630F.1030606@seiner.com> <461E6E92.80307@seiner.com> Message-ID: <20070412185340.GB27428@kestrel.barix.local> On Thu, Apr 12, 2007 at 10:38:26AM -0700, Yan Seiner wrote: > Edward Rosten wrote: > > > >> k at libsee. The problem is that links would have to generate a DOM > >> for javascript to work. > >> > > > > That's for javascript to work fully. However even without a DOM, links > > still has some ability to work on those really stupid websites which > > replace all normak links with dodgy javascript code. > > > > I really don't want to loose the ability to do that. > > > > > I see. I have a different need; I am working on an embedded controller > that needs a browser that can respond to user input. Maybe if someone had the time to reverse engineer and remove the javascript bugs - the original author (perm) doesn't have time because of school. Mikulas actually decided to remove the javascript. Most segfaults always came from the javascript. I don't know if Mikulas thinks it's a good idea to pass the javascript debugging to someone else. CL< > > >> You may be able to look at hv3 to get an idea of how that's done. It's > >> written in Tk/Tcl and C. http://tkhtml.tcl.tk/hv3.html > >> > > > > There's also Dillo. It's got CSS support, so I think it must hace a DOM > > for that to work. I tjink it's written in a mixture of C and C++. > > > > Dillo uses GTK+, which pretty much knocks it out of my league.... GTK+ > is just too big for my little machine... ISTR that dillo doesn't have > DOM; it's on the todo list. > > :-( > > Looking through links code, I may be able to hardcode my UI stuff.... I > have to get a better understanding of the input routines. > > --Yan > _______________________________________________ > links-list mailing list > links-list at linuxfromscratch.org > http://linuxfromscratch.org/mailman/listinfo/links-list From mikulas at artax.karlin.mff.cuni.cz Thu Apr 12 13:58:12 2007 From: mikulas at artax.karlin.mff.cuni.cz (Mikulas Patocka) Date: Thu, 12 Apr 2007 21:58:12 +0200 (CEST) Subject: [links] Links and javascript In-Reply-To: <20070412185340.GB27428@kestrel.barix.local> References: <461DB60A.9000009@seiner.com> <20070412062804.GA30446@kestrel.barix.local> <461E630F.1030606@seiner.com> <461E6E92.80307@seiner.com> <20070412185340.GB27428@kestrel.barix.local> Message-ID: >> Edward Rosten wrote: >>> >>>> k at libsee. The problem is that links would have to generate a DOM >>>> for javascript to work. >>>> >>> >>> That's for javascript to work fully. However even without a DOM, links >>> still has some ability to work on those really stupid websites which >>> replace all normak links with dodgy javascript code. >>> >>> I really don't want to loose the ability to do that. >>> >>> >> I see. I have a different need; I am working on an embedded controller >> that needs a browser that can respond to user input. > > Maybe if someone had the time to reverse engineer and remove the javascript > bugs > - the original author (perm) doesn't have time because of school. > > Mikulas actually decided to remove the javascript. Most segfaults always came > from the javascript. I don't know if Mikulas thinks it's a good idea to pass > the javascript debugging to someone else. > > CL< Javascript code is simply undebuggable and unfixable. I once tried to fix one bug and found that Martin Pergel copied the same piece of code 38 times instead of making it a function. It is like that when graph theorist tried to do programming :) For embedded applications you can simply use old version of links or you can add javascript to the new version (I will keep all #ifdefs there although they'll never be activated from ./configure). Releasing that javascript interpreter to general public is dangerous --- someone may find some exploit in it and... Mikulas From yan at seiner.com Thu Apr 12 14:04:44 2007 From: yan at seiner.com (Yan Seiner) Date: Thu, 12 Apr 2007 13:04:44 -0700 Subject: [links] Links and javascript In-Reply-To: References: <461DB60A.9000009@seiner.com> <20070412062804.GA30446@kestrel.barix.local> <461E630F.1030606@seiner.com> <461E6E92.80307@seiner.com> <20070412185340.GB27428@kestrel.barix.local> Message-ID: <461E90DC.2020101@seiner.com> Mikulas Patocka wrote: >>> Edward Rosten wrote: >>> >>>>> k at libsee. The problem is that links would have to generate a DOM >>>>> for javascript to work. >>>>> >>>>> >>>> That's for javascript to work fully. However even without a DOM, links >>>> still has some ability to work on those really stupid websites which >>>> replace all normak links with dodgy javascript code. >>>> >>>> I really don't want to loose the ability to do that. >>>> >>>> >>>> >>> I see. I have a different need; I am working on an embedded controller >>> that needs a browser that can respond to user input. >>> >> Maybe if someone had the time to reverse engineer and remove the javascript >> bugs >> - the original author (perm) doesn't have time because of school. >> >> Mikulas actually decided to remove the javascript. Most segfaults always came >> from the javascript. I don't know if Mikulas thinks it's a good idea to pass >> the javascript debugging to someone else. >> >> CL< >> > > Javascript code is simply undebuggable and unfixable. I once tried to fix > one bug and found that Martin Pergel copied the same piece of code 38 > times instead of making it a function. It is like that when graph theorist > tried to do programming :) > > For embedded applications you can simply use old version of links or you > can add javascript to the new version (I will keep all #ifdefs there > although they'll never be activated from ./configure). Releasing that > javascript interpreter to general public is dangerous --- someone may find > some exploit in it and... > Yeah, I looked at the js implementation and decided it's not worth fixing.... Is there support/interest for adding DOM and then using something like libsee for javascript? I've thought about ripping out all of the js stuff and starting over.... We'll see just how much time I have for this. --Yan From phil.ganchev at gmail.com Thu Apr 12 14:35:37 2007 From: phil.ganchev at gmail.com (Philip Ganchev) Date: Thu, 12 Apr 2007 16:35:37 -0400 Subject: [links] Links and javascript In-Reply-To: <20070412185340.GB27428@kestrel.barix.local> References: <461DB60A.9000009@seiner.com> <20070412062804.GA30446@kestrel.barix.local> <461E630F.1030606@seiner.com> <461E6E92.80307@seiner.com> <20070412185340.GB27428@kestrel.barix.local> Message-ID: <3aba28080704121335k43b5fd96h76b850f0515f2f03@mail.gmail.com> On 4/12/07, Karel Kulhavy wrote: [...] > Maybe if someone had the time to reverse engineer and remove the javascript > bugs > - the original author (perm) doesn't have time because of school. > > Mikulas actually decided to remove the javascript. Most segfaults always came > from the javascript. I don't know if Mikulas thinks it's a good idea to pass > the javascript debugging to someone else. I think Links has a huge potential because of its small size and its independence of X. How about if we start a fund or a bounty for javascript on Links? How much do you think it would cost to have it implemented? The Links Development page (http://links.twibright.com/development.php) shows some prices in the thousands. I'll start by offering a modest US$100 and increase that if there is need and other people pitch in. I think there is even a website that facilitates this process for FOSS but I can't find it. From yan at seiner.com Thu Apr 12 14:53:33 2007 From: yan at seiner.com (Yan Seiner) Date: Thu, 12 Apr 2007 13:53:33 -0700 Subject: [links] Links and javascript In-Reply-To: <3aba28080704121335k43b5fd96h76b850f0515f2f03@mail.gmail.com> References: <461DB60A.9000009@seiner.com> <20070412062804.GA30446@kestrel.barix.local> <461E630F.1030606@seiner.com> <461E6E92.80307@seiner.com> <20070412185340.GB27428@kestrel.barix.local> <3aba28080704121335k43b5fd96h76b850f0515f2f03@mail.gmail.com> Message-ID: <461E9C4D.9080707@seiner.com> Philip Ganchev wrote: > On 4/12/07, Karel Kulhavy wrote: > [...] > >> Maybe if someone had the time to reverse engineer and remove the javascript >> bugs >> - the original author (perm) doesn't have time because of school. >> >> Mikulas actually decided to remove the javascript. Most segfaults always came >> from the javascript. I don't know if Mikulas thinks it's a good idea to pass >> the javascript debugging to someone else. >> > > I think Links has a huge potential because of its small size and its > independence of X. > > How about if we start a fund or a bounty for javascript on Links? How > much do you think it would cost to have it implemented? The Links > Development page (http://links.twibright.com/development.php) shows > some prices in the thousands. I'll start by offering a modest US$100 > and increase that if there is need and other people pitch in. > I would second that.... I've been looking at gdome and libsee as a way to add dom + javascript processing. http://gdome2.cs.unibo.it/index.html http://www.adaptive-enterprises.com.au/~d/software/see/ This would probably mean we'd have to pull in libxml2 or something... But it would be awesome to have an xml browser that doesn't take 30 MB! --Yan From clock at twibright.com Thu Apr 12 15:21:16 2007 From: clock at twibright.com (Karel Kulhavy) Date: Thu, 12 Apr 2007 23:21:16 +0200 Subject: [links] Links and javascript In-Reply-To: <461E90DC.2020101@seiner.com> References: <461DB60A.9000009@seiner.com> <20070412062804.GA30446@kestrel.barix.local> <461E630F.1030606@seiner.com> <461E6E92.80307@seiner.com> <20070412185340.GB27428@kestrel.barix.local> <461E90DC.2020101@seiner.com> Message-ID: <20070412212115.GB3116@kestrel.barix.local> On Thu, Apr 12, 2007 at 01:04:44PM -0700, Yan Seiner wrote: > > Is there support/interest for adding DOM and then using something like > libsee for javascript? Hm I looked at http://en.wikipedia.org/wiki/Document_Object_Model to see what the DOM is: "Document Object Model (DOM) is a platform- and language-independent standard object model for representing HTML or XML and related formats." Now, what is an "object model"? Let's expand the object model term according to Wikipedia definition: "Document Object Model (DOM) are platform- and language-independent standard properties of objects in general, in a specific computer programming language, technology, notation or methodology that uses them for representing HTML or XML and related formats." The document HTML object is represented in Links memory in a platform-independent way, in a specific computer programming language (C). So do we actually already have a DOM, just don't call it that way? CL< > > I've thought about ripping out all of the js stuff and starting > over.... We'll see just how much time I have for this. > > --Yan > > > _______________________________________________ > links-list mailing list > links-list at linuxfromscratch.org > http://linuxfromscratch.org/mailman/listinfo/links-list From yan at seiner.com Thu Apr 12 15:37:53 2007 From: yan at seiner.com (Yan Seiner) Date: Thu, 12 Apr 2007 14:37:53 -0700 Subject: [links] Links and javascript In-Reply-To: <20070412212115.GB3116@kestrel.barix.local> References: <461DB60A.9000009@seiner.com> <20070412062804.GA30446@kestrel.barix.local> <461E630F.1030606@seiner.com> <461E6E92.80307@seiner.com> <20070412185340.GB27428@kestrel.barix.local> <461E90DC.2020101@seiner.com> <20070412212115.GB3116@kestrel.barix.local> Message-ID: <461EA6B1.4040800@seiner.com> Karel Kulhavy wrote: > On Thu, Apr 12, 2007 at 01:04:44PM -0700, Yan Seiner wrote: > > >> Is there support/interest for adding DOM and then using something like >> libsee for javascript? >> > > Hm I looked at http://en.wikipedia.org/wiki/Document_Object_Model to see > what the DOM is: > > "Document Object Model (DOM) is a platform- and language-independent standard > object model for representing HTML or XML and related formats." > > Now, what is an "object model"? Let's expand the object model term according > to Wikipedia definition: > > "Document Object Model (DOM) are platform- and language-independent standard > properties of objects in general, in a specific computer programming language, > technology, notation or methodology that uses them for representing HTML or XML > and related formats." > > The document HTML object is represented in Links memory in a > platform-independent way, in a specific computer programming language (C). So > do we actually already have a DOM, just don't call it that way? > Well, DOM is actually a specific implementation.... See . This would let you hook up standard javascript engines to links.... --Yan From clock at twibright.com Thu Apr 12 15:39:16 2007 From: clock at twibright.com (Karel Kulhavy) Date: Thu, 12 Apr 2007 23:39:16 +0200 Subject: [links] Links and javascript In-Reply-To: <3aba28080704121335k43b5fd96h76b850f0515f2f03@mail.gmail.com> References: <461DB60A.9000009@seiner.com> <20070412062804.GA30446@kestrel.barix.local> <461E630F.1030606@seiner.com> <461E6E92.80307@seiner.com> <20070412185340.GB27428@kestrel.barix.local> <3aba28080704121335k43b5fd96h76b850f0515f2f03@mail.gmail.com> Message-ID: <20070412213916.GC3116@kestrel.barix.local> On Thu, Apr 12, 2007 at 04:35:37PM -0400, Philip Ganchev wrote: > On 4/12/07, Karel Kulhavy wrote: > [...] > > Maybe if someone had the time to reverse engineer and remove the javascript > > bugs > > - the original author (perm) doesn't have time because of school. > > > > Mikulas actually decided to remove the javascript. Most segfaults always came > > from the javascript. I don't know if Mikulas thinks it's a good idea to pass > > the javascript debugging to someone else. > > I think Links has a huge potential because of its small size and its > independence of X. But it doesn't have CSS and adding CSS would be costly. I suggested to add a function "Launch external browser" for pages which cannot be browsed without Javascript, AJAX, CSS, Flash and other Nobel-Prize-for-Lack-of-Elegance technologies. But Mikulas doesn't want to write it, he says he doesn't need it :) I would hook up Firefox to this hotkey :) Actually what Mikulas said about the Javascript - 38 places with the same code - that can be solved with so called code refactoring - you find these pieces, write a function for them, commit it and only then continue debugging. CL< > > How about if we start a fund or a bounty for javascript on Links? How > much do you think it would cost to have it implemented? The Links > Development page (http://links.twibright.com/development.php) shows > some prices in the thousands. I'll start by offering a modest US$100 > and increase that if there is need and other people pitch in. > > I think there is even a website that facilitates this process for FOSS > but I can't find it. > _______________________________________________ > links-list mailing list > links-list at linuxfromscratch.org > http://linuxfromscratch.org/mailman/listinfo/links-list From edrosten at lanl.gov Thu Apr 12 16:22:08 2007 From: edrosten at lanl.gov (Edward Rosten) Date: Thu, 12 Apr 2007 16:22:08 -0600 (MDT) Subject: [links] Links and javascript In-Reply-To: <3aba28080704121335k43b5fd96h76b850f0515f2f03@mail.gmail.com> References: <461DB60A.9000009@seiner.com> <20070412062804.GA30446@kestrel.barix.local> <461E630F.1030606@seiner.com> <461E6E92.80307@seiner.com> <20070412185340.GB27428@kestrel.barix.local> <3aba28080704121335k43b5fd96h76b850f0515f2f03@mail.gmail.com> Message-ID: > I think Links has a huge potential because of its small size Agreed. > and its independence of X. I like X :-) > How about if we start a fund or a bounty for javascript on Links? How > much do you think it would cost to have it implemented? The Links > Development page (http://links.twibright.com/development.php) shows > some prices in the thousands. I'll start by offering a modest US$100 > and increase that if there is need and other people pitch in. I will also contribute an initial modest $100 towards the bounty. -Ed From kali at flowtrack.com.au Fri Apr 13 10:22:56 2007 From: kali at flowtrack.com.au (Kali) Date: Fri, 13 Apr 2007 11:22:56 -0500 (CDT) Subject: [links] Links and javascript In-Reply-To: Message-ID: Dear List: This is a great discussion. I promise $100 also. My take on links is that it consumes small amounts of energy. I run on solar power. As well as a screen that wont even do X, I want to use a computer that boots in seconds so I dont leave it on all the time. For years I was a registered Arachne user. Arachne (DOS version) does CSS and has excellent graphics but no javascript or ssl :-( Hence my painful learning curve into Linux. Regards Kali On Thu, 12 Apr 2007, Edward Rosten wrote: > > > > I think Links has a huge potential because of its small size > > Agreed. > > > and its independence of X. > > I like X :-) > > > > How about if we start a fund or a bounty for javascript on Links? How > > much do you think it would cost to have it implemented? The Links > > Development page (http://links.twibright.com/development.php) shows > > some prices in the thousands. I'll start by offering a modest US$100 > > and increase that if there is need and other people pitch in. > > I will also contribute an initial modest $100 towards the bounty. > > > > -Ed > > _______________________________________________ > links-list mailing list > links-list at linuxfromscratch.org > http://linuxfromscratch.org/mailman/listinfo/links-list > > > From edrosten at lanl.gov Fri Apr 13 12:33:59 2007 From: edrosten at lanl.gov (Edward Rosten) Date: Fri, 13 Apr 2007 12:33:59 -0600 (MDT) Subject: [links] Links and javascript In-Reply-To: References: Message-ID: On Fri, 13 Apr 2007, Kali wrote: > Dear List: > This is a great discussion. I promise $100 also. > My take on links is that it consumes small amounts of energy. I run on > solar power. As well as a screen that wont even do X, OK, this has piqued my curiosity! What kind of screen do you nave that won't run X? A vt220? -Ed From kali at flowtrack.com.au Sat Apr 14 11:16:35 2007 From: kali at flowtrack.com.au (Kali) Date: Sat, 14 Apr 2007 12:16:35 -0500 (CDT) Subject: [links] X In-Reply-To: Message-ID: Dear Ed: This monitor is a CSS dual scan monochrome monitor cs14DA. DB9 connector. I like it for its readability/reliability/low power/no flicker/no aliasing/ independent real power switch/no start surge on inverter, etc. The computer uses 47 watts, starts from cold in 30 seconds and stops in 12. No DOS or windows graphics mode has ever worked, but my partner just informed me that ANY monitor can talk X, so I stand corrected! He (James) also says links should use CORBA and not DOM. The card, OTIVGA (OT1067) might be the problem in what modes it supports. Kali On Fri, 13 Apr 2007, Edward Rosten wrote: > > My take on links is that it consumes small amounts of energy. I run on > > solar power. As well as a screen that wont even do X, > > OK, this has piqued my curiosity! What kind of screen do you nave that > won't run X? A vt220? > > -Ed > > > _______________________________________________ > links-list mailing list > links-list at linuxfromscratch.org > http://linuxfromscratch.org/mailman/listinfo/links-list > > > From mikulas at artax.karlin.mff.cuni.cz Sun Apr 15 18:01:44 2007 From: mikulas at artax.karlin.mff.cuni.cz (Mikulas Patocka) Date: Mon, 16 Apr 2007 02:01:44 +0200 (CEST) Subject: [links] Links and javascript In-Reply-To: <3aba28080704121335k43b5fd96h76b850f0515f2f03@mail.gmail.com> References: <461DB60A.9000009@seiner.com> <20070412062804.GA30446@kestrel.barix.local> <461E630F.1030606@seiner.com> <461E6E92.80307@seiner.com> <20070412185340.GB27428@kestrel.barix.local> <3aba28080704121335k43b5fd96h76b850f0515f2f03@mail.gmail.com> Message-ID: >> Maybe if someone had the time to reverse engineer and remove the javascript >> bugs >> - the original author (perm) doesn't have time because of school. >> >> Mikulas actually decided to remove the javascript. Most segfaults always came >> from the javascript. I don't know if Mikulas thinks it's a good idea to pass >> the javascript debugging to someone else. > > I think Links has a huge potential because of its small size and its > independence of X. > > How about if we start a fund or a bounty for javascript on Links? How > much do you think it would cost to have it implemented? The Links > Development page (http://links.twibright.com/development.php) shows > some prices in the thousands. I'll start by offering a modest US$100 > and increase that if there is need and other people pitch in. > > I think there is even a website that facilitates this process for FOSS > but I can't find it. Hi I have no time for it, but if you find someone who has, you can pay him. I am removing javascript right now --- I left the hooks there (file jsint.c), so you can just take spidermonkey or whatever javascript interpreter, call it from jsint_execute_code and do callbacks to jsint.c file --- this will just get form editing, url redirection, image replacement and similar Netscape 3 stuff working. Doing DOM is hard to impossible. Mikulas From phil.ganchev at gmail.com Sun Apr 15 19:07:17 2007 From: phil.ganchev at gmail.com (Philip Ganchev) Date: Sun, 15 Apr 2007 21:07:17 -0400 Subject: [links] Links and javascript In-Reply-To: References: <461DB60A.9000009@seiner.com> <20070412062804.GA30446@kestrel.barix.local> <461E630F.1030606@seiner.com> <461E6E92.80307@seiner.com> <20070412185340.GB27428@kestrel.barix.local> <3aba28080704121335k43b5fd96h76b850f0515f2f03@mail.gmail.com> Message-ID: <3aba28080704151807p272d0166j8081ee72bfe387e5@mail.gmail.com> On 4/15/07, Mikulas Patocka wrote: [...] > > How about if we start a fund or a bounty for javascript on Links? How > > much do you think it would cost to have it implemented? The Links > > Development page (http://links.twibright.com/development.php) shows > > some prices in the thousands. I'll start by offering a modest US$100 > > and increase that if there is need and other people pitch in. > > > > I think there is even a website that facilitates this process for FOSS > > but I can't find it. > > Hi > > I have no time for it, but if you find someone who has, you can pay him. Any chance we can use the Links website to advertize that, for example at http://links.twibright.com/development.php ? I guess that this is the most likely to find someone interested. Another idea is Google Summer of Code, but that would need someone to supervise a student, and the best person would be someone who has worked on Links before. > I am removing javascript right now --- I left the hooks there (file > jsint.c), so you can just take spidermonkey or whatever javascript > interpreter, call it from jsint_execute_code and do callbacks to jsint.c > file --- this will just get form editing, url redirection, image > replacement and similar Netscape 3 stuff working. Doing DOM is hard to > impossible. OK. From mikulas at artax.karlin.mff.cuni.cz Sun Apr 15 20:55:11 2007 From: mikulas at artax.karlin.mff.cuni.cz (Mikulas Patocka) Date: Mon, 16 Apr 2007 04:55:11 +0200 (CEST) Subject: [links] Links and javascript In-Reply-To: <3aba28080704151807p272d0166j8081ee72bfe387e5@mail.gmail.com> References: <461DB60A.9000009@seiner.com> <20070412062804.GA30446@kestrel.barix.local> <461E630F.1030606@seiner.com> <461E6E92.80307@seiner.com> <20070412185340.GB27428@kestrel.barix.local> <3aba28080704121335k43b5fd96h76b850f0515f2f03@mail.gmail.com> <3aba28080704151807p272d0166j8081ee72bfe387e5@mail.gmail.com> Message-ID: > Any chance we can use the Links website to advertize that, for example at > http://links.twibright.com/development.php ? I guess that this is the > most likely to find someone interested. > > Another idea is Google Summer of Code, but that would need someone to > supervise a student, and the best person would be someone who has > worked on Links before. I thought summer of code has already finished its assignments for this year. Or is there still a possibility? Mikulas From phil.ganchev at gmail.com Sun Apr 15 21:48:41 2007 From: phil.ganchev at gmail.com (Philip Ganchev) Date: Sun, 15 Apr 2007 23:48:41 -0400 Subject: [links] Links and javascript In-Reply-To: References: <461DB60A.9000009@seiner.com> <461E630F.1030606@seiner.com> <461E6E92.80307@seiner.com> <20070412185340.GB27428@kestrel.barix.local> <3aba28080704121335k43b5fd96h76b850f0515f2f03@mail.gmail.com> <3aba28080704151807p272d0166j8081ee72bfe387e5@mail.gmail.com> Message-ID: <3aba28080704152048r5de638f7x47811cf43714f748@mail.gmail.com> On 4/15/07, Mikulas Patocka wrote: > I thought summer of code has already finished its assignments for this > year. Or is there still a possibility? > > Mikulas > I guess it is over, judging fro the site. But there is always GSoC 2008 :-) From clock at twibright.com Mon Apr 16 02:34:09 2007 From: clock at twibright.com (Karel Kulhavy) Date: Mon, 16 Apr 2007 10:34:09 +0200 Subject: [links] Links and javascript In-Reply-To: References: <20070412062804.GA30446@kestrel.barix.local> <461E630F.1030606@seiner.com> <461E6E92.80307@seiner.com> <20070412185340.GB27428@kestrel.barix.local> <3aba28080704121335k43b5fd96h76b850f0515f2f03@mail.gmail.com> <3aba28080704151807p272d0166j8081ee72bfe387e5@mail.gmail.com> Message-ID: <20070416083409.GE11594@kestrel.barix.local> On Mon, Apr 16, 2007 at 04:55:11AM +0200, Mikulas Patocka wrote: > > Any chance we can use the Links website to advertize that, for example at > > http://links.twibright.com/development.php ? I guess that this is the > > most likely to find someone interested. > > > > Another idea is Google Summer of Code, but that would need someone to > > supervise a student, and the best person would be someone who has > > worked on Links before. > > I thought summer of code has already finished its assignments for this > year. Or is there still a possibility? All I know is they have some assignment on gEDA. But what's the quality of the resulting code? Aren't the students doing it just for the money? Isn't the quality like diploma theses? :) CL< > > Mikulas > _______________________________________________ > links-list mailing list > links-list at linuxfromscratch.org > http://linuxfromscratch.org/mailman/listinfo/links-list From clock at twibright.com Mon Apr 16 02:32:55 2007 From: clock at twibright.com (Karel Kulhavy) Date: Mon, 16 Apr 2007 10:32:55 +0200 Subject: [links] Links and javascript In-Reply-To: <3aba28080704151807p272d0166j8081ee72bfe387e5@mail.gmail.com> References: <461DB60A.9000009@seiner.com> <20070412062804.GA30446@kestrel.barix.local> <461E630F.1030606@seiner.com> <461E6E92.80307@seiner.com> <20070412185340.GB27428@kestrel.barix.local> <3aba28080704121335k43b5fd96h76b850f0515f2f03@mail.gmail.com> <3aba28080704151807p272d0166j8081ee72bfe387e5@mail.gmail.com> Message-ID: <20070416083255.GD11594@kestrel.barix.local> On Sun, Apr 15, 2007 at 09:07:17PM -0400, Philip Ganchev wrote: > On 4/15/07, Mikulas Patocka wrote: > [...] > > > How about if we start a fund or a bounty for javascript on Links? How > > > much do you think it would cost to have it implemented? The Links > > > Development page (http://links.twibright.com/development.php) shows > > > some prices in the thousands. I'll start by offering a modest US$100 > > > and increase that if there is need and other people pitch in. > > > > > > I think there is even a website that facilitates this process for FOSS > > > but I can't find it. > > > > Hi > > > > I have no time for it, but if you find someone who has, you can pay him. > > Any chance we can use the Links website to advertize that, for example at > http://links.twibright.com/development.php ? I guess that this is the > most likely to find someone interested. On Ronja I have policy that I encourage donations only for things which are new features. I consider bugfixes my fault and do them for free. Even bugfixes that mean redesigning a whole module from scratch, which already happened once and half times, are bugfixes, and therefore for free. I don't think it would be appropriate to expect donations for a bugfix. I think this is a reasonable policy for the whole Twibright Labs label. If we behave like pigs, people stop trusting Twibright Labs. The advantage of Twibright Labs branding is when someone uses one product (e. g. Ronja) and sees how we strive for quality, he will automatically assume the same level of quality in other products (e. g. Links). The general problem is that often you have 10 programs that solve one problem and you don't know which of them really work and which are crap. And trying that out takes a lot of time. That's the purpose of the Twibright Labs branding - this helps you to get some idea about quality. I think better having a reliable browser without Javascript than a crappy browser with Javascript, when we don't have enough resources to fix it. If someone has a lot of time he can take PerM's code, refactor it so it's more understandable and fix all the bugs. It's just labour intensive I guess. CL< > > Another idea is Google Summer of Code, but that would need someone to > supervise a student, and the best person would be someone who has > worked on Links before. > > > > I am removing javascript right now --- I left the hooks there (file > > jsint.c), so you can just take spidermonkey or whatever javascript > > interpreter, call it from jsint_execute_code and do callbacks to jsint.c > > file --- this will just get form editing, url redirection, image > > replacement and similar Netscape 3 stuff working. Doing DOM is hard to > > impossible. > > OK. > _______________________________________________ > links-list mailing list > links-list at linuxfromscratch.org > http://linuxfromscratch.org/mailman/listinfo/links-list From edrosten at lanl.gov Mon Apr 16 10:24:42 2007 From: edrosten at lanl.gov (Edward Rosten) Date: Mon, 16 Apr 2007 10:24:42 -0600 (MDT) Subject: [links] X In-Reply-To: References: Message-ID: On Sat, 14 Apr 2007, Kali wrote: > Dear Ed: > This monitor is a CSS dual scan monochrome monitor cs14DA. DB9 connector. > I like it for its readability/reliability/low power/no flicker/no > aliasing/ independent real power switch/no start surge on inverter, etc. > The computer uses 47 watts, starts from cold in 30 seconds and stops in > 12. > > No DOS or windows graphics mode has ever worked, but my partner just > informed me that ANY monitor can talk X, so I stand corrected! > > The card, OTIVGA (OT1067) might be the problem in what modes it supports. Sounds good, but if the monitor is old, then be a bit careful. Some old monitors can die if you feed them the wrong signals, but I've never actually encountered one myself. If the OTI card if supported by X, then it probably supports fully programmable timings. The drivers for X generally give more flexibility than you can get anywhere else. > He (James) also says links should use CORBA and not DOM. CORBA is a RPC (remote procedure call) system, as is DCOM (a Microsoft specific RPC system). Are you sure he's not confusing DCOM with DOM? I thought that the DOM was just an tree structure representing an HTML (or XML) document that allowed you to refer to the elements using a standard naming scheme. -Ed From edanaher at andrew.cmu.edu Mon Apr 16 13:37:04 2007 From: edanaher at andrew.cmu.edu (Evan Danaher) Date: Mon, 16 Apr 2007 12:37:04 -0700 Subject: [links] Offer to Help Message-ID: Executive Summary: Links is cool. I've added some features. I want to submit them back. I am also possibly interested in making Javascript better, and/or adding in some semblence of DOM and/or CSS support. Details: I'm expecting this to be rather long, hence the summary above. About a month ago, I finally got fed up with Firefox using huge amounts of memory, and occasionally taking several seconds to scroll when other processes were using CPU. So I decided to look for a lighter browser. I found that Links, which I had used a bit in text mode, also had a GUI version, which looked remarkably usable. Even better, it has a fairly small, well-organized source base that I can work on. I thus began adding features to Links, gradually moving it towards my ideal browser. I've been using it as my primary browser for the past couple weeks, so I've fixed a couple things that annoy me. Most notably, I merged in some tabs code from links-hacked, and added a FVWM-style configuration file that allows for easy rebinding of keys and eventually mouse buttons. I've also added a couple smaller features: - URL abbreviations (e.g., wp:hello goes to http://en.wikipedia.org/wiki/Special:Search?search=hello"), set up in a general fashion using the configuration file. - forward: now going back can be undone; acts just like other browsers (though I'm thinking it'd be nice to get some sort of tree structure working. - "Session Saver": on closing, save the tabs so that the next session can open up in the same state. Currently kind of hack-ish, since it's intended for development, but it saves the full state of every tab (just like the back history), so I can build a new version, and restart links and not lose open pages. - History deletion: in the CTRL-W history menu, you can use the delete key to remove history items, and insert to put them back (until you close the window). I have a decent list of other features I'd like, including a customizable status line, a good UI for browser spoofing on certain web sites (likely through the configuration file), an open with... menu, something along the lines of Konqueror's "Access keys" to make keyboard navigation easier, better font support, and gestures. Some of these are harder than others, and I have no idea how many I'll get done, but so far I've done a good job of occasionally sitting down and working on stuff. I'm also hoping to get better Javascript and CSS support; this might be better as a project for the summer. I'll see what happens. Finally, the point of this e-mail is that I'm interested in submitting these features back. Unfortunately, I've been working mostly for myself, and so some of the code is kind of ugly. Worse, I haven't been paying too much attention to text mode, so that's somewhat broken. But if you're interested in these features, I can get them fixed up and ready for submission. From edrosten at lanl.gov Mon Apr 16 14:10:12 2007 From: edrosten at lanl.gov (Edward Rosten) Date: Mon, 16 Apr 2007 14:10:12 -0600 (MDT) Subject: [links] Offer to Help In-Reply-To: References: Message-ID: On Mon, 16 Apr 2007, Evan Danaher wrote: > Executive Summary: > > Links is cool. I've added some features. I want to submit them back. > I am also possibly interested in making Javascript better, and/or > adding in some semblence of DOM and/or CSS support. > > Details: > > I'm expecting this to be rather long, hence the summary above. > > About a month ago, I finally got fed up with Firefox using huge > amounts of memory, and occasionally taking several seconds to scroll > when other processes were using CPU. So I decided to look for a > lighter browser. I found that Links, which I had used a bit in text > mode, also had a GUI version, which looked remarkably usable. Even > better, it has a fairly small, well-organized source base that I can > work on. > > I thus began adding features to Links, gradually moving it towards my > ideal browser. I've been using it as my primary browser for the past > couple weeks, so I've fixed a couple things that annoy me. > > Most notably, I merged in some tabs code from links-hacked, and added > a FVWM-style configuration file that allows for easy rebinding of keys > and eventually mouse buttons. > > I've also added a couple smaller features: > > - URL abbreviations (e.g., wp:hello goes to > http://en.wikipedia.org/wiki/Special:Search?search=hello"), set up in > a general fashion using the configuration file. Does that have a GUI config box? If not, I can help you with that. I figured out how to add those for the URL blocking. If you want to know, look in block.c, or just ask. I'd be happy to share all of the insights I had. > - forward: now going back can be undone; acts just like other browsers > (though I'm thinking it'd be nice to get some sort of tree structure > working. Can you post a patch for that? Also I agree re: tree structures. Interestingly, the list boxes can support general tree structures. > I have a decent list of other features I'd like, including a > customizable status line, a good UI for browser spoofing on certain > web sites (likely through the configuration file), an open with... > menu, something along the lines of Konqueror's "Access keys" to make > keyboard navigation easier, better font support, and gestures. Some > of these are harder than others, and I have no idea how many I'll get > done, but so far I've done a good job of occasionally sitting down and > working on stuff. Mouse gestures are pretty straightforward, I've used these before. There's a library called libstroke (not updated in ages, v.0.5 but it seems 100% complete) which is very solid and works like a charm. It's not particularly intelligent and can take a fair amount of training, but it was good enough for me to use a commandline proficiently from a tablet PC (ie lots of characters required). Again, I'll answer any questions you have on this. -Ed > - "Session Saver": on closing, save the tabs so that the next session > can open up in the same state. Currently kind of hack-ish, since it's > intended for development, but it saves the full state of every tab > (just like the back history), so I can build a new version, and > restart links and not lose open pages. > > - History deletion: in the CTRL-W history menu, you can use the delete > key to remove history items, and insert to put them back (until you > close the window). > > > I'm also hoping to get better Javascript and CSS support; this might > be better as a project for the summer. I'll see what happens. > > Finally, the point of this e-mail is that I'm interested in submitting > these features back. Unfortunately, I've been working mostly for > myself, and so some of the code is kind of ugly. Worse, I haven't > been paying too much attention to text mode, so that's somewhat > broken. But if you're interested in these features, I can get them > fixed up and ready for submission. > _______________________________________________ > links-list mailing list > links-list at linuxfromscratch.org > http://linuxfromscratch.org/mailman/listinfo/links-list > From clock at twibright.com Mon Apr 16 16:33:45 2007 From: clock at twibright.com (Karel Kulhavy) Date: Tue, 17 Apr 2007 00:33:45 +0200 Subject: [links] X In-Reply-To: References: Message-ID: <20070416223345.GA12900@kestrel.barix.local> On Mon, Apr 16, 2007 at 10:24:42AM -0600, Edward Rosten wrote: > > > On Sat, 14 Apr 2007, Kali wrote: > > > Dear Ed: > > This monitor is a CSS dual scan monochrome monitor cs14DA. DB9 connector. > > I like it for its readability/reliability/low power/no flicker/no > > aliasing/ independent real power switch/no start surge on inverter, etc. > > The computer uses 47 watts, starts from cold in 30 seconds and stops in > > 12. What computer are you using? I am building an exercise bike generator and I want to connect a PC or laptop to mitigate the unhealthy lifestyle connected with computer-based occupation. At the moment I have a 60W SEIG (self excited induction generator) from an ordinary 60W three phase asynchronous motor, which can light a 60W light bulb (though not full shine) when powered by an electric drill. CL< > > > > No DOS or windows graphics mode has ever worked, but my partner just > > informed me that ANY monitor can talk X, so I stand corrected! > > > > The card, OTIVGA (OT1067) might be the problem in what modes it supports. > > > Sounds good, but if the monitor is old, then be a bit careful. Some old > monitors can die if you feed them the wrong signals, but I've never > actually encountered one myself. If the OTI card if supported by X, then > it probably supports fully programmable timings. The drivers for X > generally give more flexibility than you can get anywhere else. > > > > He (James) also says links should use CORBA and not DOM. > > CORBA is a RPC (remote procedure call) system, as is DCOM (a Microsoft > specific RPC system). Are you sure he's not confusing DCOM with DOM? > > I thought that the DOM was just an tree structure representing an HTML > (or XML) document that allowed you to refer to the elements using a > standard naming scheme. > > > -Ed > > > > _______________________________________________ > links-list mailing list > links-list at linuxfromscratch.org > http://linuxfromscratch.org/mailman/listinfo/links-list From clock at twibright.com Mon Apr 16 16:44:31 2007 From: clock at twibright.com (Karel Kulhavy) Date: Tue, 17 Apr 2007 00:44:31 +0200 Subject: [links] Offer to Help In-Reply-To: References: Message-ID: <20070416224431.GA10374@kestrel.barix.local> On Mon, Apr 16, 2007 at 12:37:04PM -0700, Evan Danaher wrote: > Executive Summary: > > Links is cool. I've added some features. I want to submit them back. > I am also possibly interested in making Javascript better, and/or > adding in some semblence of DOM and/or CSS support. > > Details: > > I'm expecting this to be rather long, hence the summary above. > > About a month ago, I finally got fed up with Firefox using huge > amounts of memory, and occasionally taking several seconds to scroll > when other processes were using CPU. So I decided to look for a > lighter browser. I found that Links, which I had used a bit in text > mode, also had a GUI version, which looked remarkably usable. Even > better, it has a fairly small, well-organized source base that I can > work on. > > I thus began adding features to Links, gradually moving it towards my > ideal browser. I've been using it as my primary browser for the past > couple weeks, so I've fixed a couple things that annoy me. > > Most notably, I merged in some tabs code from links-hacked, and added > a FVWM-style configuration file that allows for easy rebinding of keys > and eventually mouse buttons. > > I've also added a couple smaller features: > > - URL abbreviations (e.g., wp:hello goes to > http://en.wikipedia.org/wiki/Special:Search?search=hello"), set up in > a general fashion using the configuration file. I want this, this is very handy, for wikipedia and search engine. > > - forward: now going back can be undone; acts just like other browsers > (though I'm thinking it'd be nice to get some sort of tree structure > working. > > - "Session Saver": on closing, save the tabs so that the next session > can open up in the same state. Currently kind of hack-ish, since it's > intended for development, but it saves the full state of every tab > (just like the back history), so I can build a new version, and > restart links and not lose open pages. > > - History deletion: in the CTRL-W history menu, you can use the delete > key to remove history items, and insert to put them back (until you > close the window). > > I have a decent list of other features I'd like, including a > customizable status line, a good UI for browser spoofing on certain > web sites (likely through the configuration file), an open with... > menu, something along the lines of Konqueror's "Access keys" to make > keyboard navigation easier, better font support, and gestures. Some > of these are harder than others, and I have no idea how many I'll get > done, but so far I've done a good job of occasionally sitting down and > working on stuff. > > I'm also hoping to get better Javascript and CSS support; this might > be better as a project for the summer. I'll see what happens. > > Finally, the point of this e-mail is that I'm interested in submitting > these features back. Unfortunately, I've been working mostly for > myself, and so some of the code is kind of ugly. Worse, I haven't Can you perform code refactoring on your ugly code? http://en.wikipedia.org/wiki/Code_refactoring > been paying too much attention to text mode, so that's somewhat > broken. But if you're interested in these features, I can get them New patches shouldn't definitely break text mode. Are you also sure no glitches occur when the browser is in a state that doesn't occur often? CL< > fixed up and ready for submission. > _______________________________________________ > links-list mailing list > links-list at linuxfromscratch.org > http://linuxfromscratch.org/mailman/listinfo/links-list From edanaher at andrew.cmu.edu Mon Apr 16 17:36:11 2007 From: edanaher at andrew.cmu.edu (Evan Danaher) Date: Mon, 16 Apr 2007 19:36:11 -0400 Subject: [links] Offer to Help In-Reply-To: References: Message-ID: I posted the resulting patch against 2.1-pre28 (which I also updated to earlier today - nothing seems to have broken) at http://www.contrib.andrew.cmu.edu/user/edanaher/links/ . The patch has everything I've done in it; it's all fairly inter-related, so it would be a pain to pull it into separate patches. There's also a linksrc; this is the configuration file, which needs to be in ~/.links in order to work (that's currently hardcoded, one of the many things I don't like about it). I'll probably start re-doing everything soon, working from a stock 2.1-pre28, doing my best to write code that doesn't break text mode and shouldn't break in general. Another issue is the whole UI thing; I personally like just writing a config file, and being able to re-read the config file while the browser is running. This then conflicts with changing settings with a GUI, which normally writes to a file which is not edited. I'm not sure what the Right Way to do both of these is; for the moment, I'm pretty much just ignoring GUI's for editing settings. On 4/16/07, Edward Rosten wrote: > > > On Mon, 16 Apr 2007, Evan Danaher wrote: > > > Executive Summary: > > > > Links is cool. I've added some features. I want to submit them back. > > I am also possibly interested in making Javascript better, and/or > > adding in some semblence of DOM and/or CSS support. > > > > Details: > > > > I'm expecting this to be rather long, hence the summary above. > > > > About a month ago, I finally got fed up with Firefox using huge > > amounts of memory, and occasionally taking several seconds to scroll > > when other processes were using CPU. So I decided to look for a > > lighter browser. I found that Links, which I had used a bit in text > > mode, also had a GUI version, which looked remarkably usable. Even > > better, it has a fairly small, well-organized source base that I can > > work on. > > > > I thus began adding features to Links, gradually moving it towards my > > ideal browser. I've been using it as my primary browser for the past > > couple weeks, so I've fixed a couple things that annoy me. > > > > Most notably, I merged in some tabs code from links-hacked, and added > > a FVWM-style configuration file that allows for easy rebinding of keys > > and eventually mouse buttons. > > > > I've also added a couple smaller features: > > > > - URL abbreviations (e.g., wp:hello goes to > > http://en.wikipedia.org/wiki/Special:Search?search=hello"), set up in > > a general fashion using the configuration file. > > Does that have a GUI config box? If not, I can help you with that. I > figured out how to add those for the URL blocking. If you want to know, > look in block.c, or just ask. I'd be happy to share all of the insights I > had. > > > - forward: now going back can be undone; acts just like other browsers > > (though I'm thinking it'd be nice to get some sort of tree structure > > working. > > Can you post a patch for that? Also I agree re: tree structures. > Interestingly, the list boxes can support general tree structures. > > > > I have a decent list of other features I'd like, including a > > customizable status line, a good UI for browser spoofing on certain > > web sites (likely through the configuration file), an open with... > > menu, something along the lines of Konqueror's "Access keys" to make > > keyboard navigation easier, better font support, and gestures. Some > > of these are harder than others, and I have no idea how many I'll get > > done, but so far I've done a good job of occasionally sitting down and > > working on stuff. > > Mouse gestures are pretty straightforward, I've used these before. There's > a library called libstroke (not updated in ages, v.0.5 but it seems 100% > complete) which is very solid and works like a charm. It's not > particularly intelligent and can take a fair amount of training, but it > was good enough for me to use a commandline proficiently from a tablet PC > (ie lots of characters required). Again, I'll answer any questions you > have on this. > > > -Ed > > > > > - "Session Saver": on closing, save the tabs so that the next session > > can open up in the same state. Currently kind of hack-ish, since it's > > intended for development, but it saves the full state of every tab > > (just like the back history), so I can build a new version, and > > restart links and not lose open pages. > > > > - History deletion: in the CTRL-W history menu, you can use the delete > > key to remove history items, and insert to put them back (until you > > close the window). > > > > > > I'm also hoping to get better Javascript and CSS support; this might > > be better as a project for the summer. I'll see what happens. > > > > Finally, the point of this e-mail is that I'm interested in submitting > > these features back. Unfortunately, I've been working mostly for > > myself, and so some of the code is kind of ugly. Worse, I haven't > > been paying too much attention to text mode, so that's somewhat > > broken. But if you're interested in these features, I can get them > > fixed up and ready for submission. > > _______________________________________________ > > links-list mailing list > > links-list at linuxfromscratch.org > > http://linuxfromscratch.org/mailman/listinfo/links-list > > > _______________________________________________ > links-list mailing list > links-list at linuxfromscratch.org > http://linuxfromscratch.org/mailman/listinfo/links-list > From clock at twibright.com Mon Apr 16 23:38:19 2007 From: clock at twibright.com (Karel Kulhavy) Date: Tue, 17 Apr 2007 07:38:19 +0200 Subject: [links] False spamassassin positive Message-ID: <20070417053818.GA31680@kestrel.barix.local> I tried to repost the drag'n'drop support patch from Ed Rosten and also got rejected by spamassassin. I sent the patch to Mikulas. CL< From phil.ganchev at gmail.com Mon Apr 16 23:57:47 2007 From: phil.ganchev at gmail.com (Philip Ganchev) Date: Tue, 17 Apr 2007 01:57:47 -0400 Subject: [links] Links and javascript In-Reply-To: <20070416083409.GE11594@kestrel.barix.local> References: <20070412062804.GA30446@kestrel.barix.local> <461E630F.1030606@seiner.com> <461E6E92.80307@seiner.com> <20070412185340.GB27428@kestrel.barix.local> <3aba28080704121335k43b5fd96h76b850f0515f2f03@mail.gmail.com> <3aba28080704151807p272d0166j8081ee72bfe387e5@mail.gmail.com> <20070416083409.GE11594@kestrel.barix.local> Message-ID: <3aba28080704162257p12aee2dbp2596a70fd5719eb8@mail.gmail.com> On 4/16/07, Karel Kulhavy wrote: > On Mon, Apr 16, 2007 at 04:55:11AM +0200, Mikulas Patocka wrote: > > > Any chance we can use the Links website to advertize that, for example at > > > http://links.twibright.com/development.php ? I guess that this is the > > > most likely to find someone interested. > > > > > > Another idea is Google Summer of Code, but that would need someone to > > > supervise a student, and the best person would be someone who has > > > worked on Links before. > > > > I thought summer of code has already finished its assignments for this > > year. Or is there still a possibility? > > All I know is they have some assignment on gEDA. But what's the quality of > the resulting code? Aren't the students doing it just for the money? Isn't > the quality like diploma theses? :) > > CL< I have no experience with GSoC, but if you look at the list, there are a lot of very high-profile organizations, such as Gnome, KDE, the Apache foundation, and BBC Research. So maybe the code is not that bad. From clock at twibright.com Tue Apr 17 03:39:08 2007 From: clock at twibright.com (Karel Kulhavy) Date: Tue, 17 Apr 2007 11:39:08 +0200 Subject: [links] Links and javascript In-Reply-To: <3aba28080704162257p12aee2dbp2596a70fd5719eb8@mail.gmail.com> References: <461E630F.1030606@seiner.com> <461E6E92.80307@seiner.com> <20070412185340.GB27428@kestrel.barix.local> <3aba28080704121335k43b5fd96h76b850f0515f2f03@mail.gmail.com> <3aba28080704151807p272d0166j8081ee72bfe387e5@mail.gmail.com> <20070416083409.GE11594@kestrel.barix.local> <3aba28080704162257p12aee2dbp2596a70fd5719eb8@mail.gmail.com> Message-ID: <20070417093908.GA15882@kestrel.barix.local> On Tue, Apr 17, 2007 at 01:57:47AM -0400, Philip Ganchev wrote: > On 4/16/07, Karel Kulhavy wrote: > > On Mon, Apr 16, 2007 at 04:55:11AM +0200, Mikulas Patocka wrote: > > > > Any chance we can use the Links website to advertize that, for example at > > > > http://links.twibright.com/development.php ? I guess that this is the > > > > most likely to find someone interested. > > > > > > > > Another idea is Google Summer of Code, but that would need someone to > > > > supervise a student, and the best person would be someone who has > > > > worked on Links before. > > > > > > I thought summer of code has already finished its assignments for this > > > year. Or is there still a possibility? > > > > All I know is they have some assignment on gEDA. But what's the quality of > > the resulting code? Aren't the students doing it just for the money? Isn't > > the quality like diploma theses? :) > > > > CL< > > I have no experience with GSoC, but if you look at the list, there are > a lot of very high-profile organizations, such as Gnome, KDE, the > Apache foundation, and BBC Research. So maybe the code is not that > bad. How does a high-profile organization imply a decent code? CL< > _______________________________________________ > links-list mailing list > links-list at linuxfromscratch.org > http://linuxfromscratch.org/mailman/listinfo/links-list From phil.ganchev at gmail.com Tue Apr 17 12:18:02 2007 From: phil.ganchev at gmail.com (Philip Ganchev) Date: Tue, 17 Apr 2007 14:18:02 -0400 Subject: [links] Links and javascript In-Reply-To: <20070417093908.GA15882@kestrel.barix.local> References: <461E630F.1030606@seiner.com> <461E6E92.80307@seiner.com> <20070412185340.GB27428@kestrel.barix.local> <3aba28080704121335k43b5fd96h76b850f0515f2f03@mail.gmail.com> <3aba28080704151807p272d0166j8081ee72bfe387e5@mail.gmail.com> <20070416083409.GE11594@kestrel.barix.local> <3aba28080704162257p12aee2dbp2596a70fd5719eb8@mail.gmail.com> <20070417093908.GA15882@kestrel.barix.local> Message-ID: <3aba28080704171118p18ce945ds6226cf5ea1c6a87b@mail.gmail.com> On 4/17/07, Karel Kulhavy wrote: > On Tue, Apr 17, 2007 at 01:57:47AM -0400, Philip Ganchev wrote: > > On 4/16/07, Karel Kulhavy wrote: > > > On Mon, Apr 16, 2007 at 04:55:11AM +0200, Mikulas Patocka wrote: [...] > > > > I thought summer of code has already finished its assignments for this > > > > year. Or is there still a possibility? > > > > > > All I know is they have some assignment on gEDA. But what's the quality of > > > the resulting code? Aren't the students doing it just for the money? Isn't > > > the quality like diploma theses? :) > > > > > > CL< > > > > I have no experience with GSoC, but if you look at the list, there are > > a lot of very high-profile organizations, such as Gnome, KDE, the > > Apache foundation, and BBC Research. So maybe the code is not that > > bad. > > How does a high-profile organization imply a decent code? It does not, but it strongly suggests it. Since the organizations are large and the code is open, a lot of people work on the code; low coding standards would cause a mediocre product and the organization would lose its profile. That's my intuition. But we can ask some mentors from last year if they were satisfied with the code quality; for example, the gEDA people if you like. From edrosten at lanl.gov Wed Apr 18 13:12:36 2007 From: edrosten at lanl.gov (Edward Rosten) Date: Wed, 18 Apr 2007 13:12:36 -0600 (MDT) Subject: [links] Drag 'n drop support! Message-ID: Links now supports Drag 'n drop. The patch is attached as a gzipped patch, since raw patches trigger spamassin's filters. The X11 backend is fully implemented. The links front end is a bit lacking (matching my knowledge of links internals). Links accepts 3 kind of data, URLs, text and data: URLs can have a name which is different from the href (ie bookmarks) Data can be text/html, image/jpeg or image/png. Data is saved to a temp file and referred to by the URL of a temp file. Text is any of the plain text types. You can drag URLs or data (eg a text selection from OpenOffice) on to the main Links window. With OpenOffice text, you see it rendered with the styles selected in OpenOffice (up to the capability of Links). You can't drag text in to text input boxes, or URLs on to the bookmark or blocklist manager, because haven't figured out how to do that. On the back end, here's how it works. There are two new button press types: B_DND_REQUEST B_DND_DROP When dragging, the B_DND_REQUEST events get generated. To respond to one of these, do dnd_list_types(), to get the available types (any mix of DND_URL, DND_DATA and DND_TEXT), and choose the most suitable and select it with dnd_select_type(). When a drop has succesfully occured (and the data has arrived), a B_DND_DROP event is generated. The dnd data can be accessed with dnd_get_result(). Look in view.c for the preliminary integration with links. I haven't integrated it with the paste system yet. The mechanism for X11 copy/paste and DnD are almost identical, so links could get advanced pasting quite easily as well. -Ed -------------- next part -------------- A non-text attachment was scrubbed... Name: links-2.1pre28-dnd.patch.gz Type: application/x-gzip Size: 6820 bytes Desc: Url : http://linuxfromscratch.org/pipermail/links-list/attachments/20070418/a7de70f1/attachment.bin From edrosten at lanl.gov Wed Apr 18 13:24:51 2007 From: edrosten at lanl.gov (Edward Rosten) Date: Wed, 18 Apr 2007 13:24:51 -0600 (MDT) Subject: [links] Drag 'n drop support! In-Reply-To: References: Message-ID: > I haven't integrated it with the paste system yet. The mechanism for X11 > copy/paste and DnD are almost identical, so links could get advanced pasting > quite easily as well. The reason I have used a different method is because DnD support in other applications is somewhat unreliable compared to pasting (eg Konqueror keeps segfaulting). If DnD is implemented synchronusly (like the paste in links), then links will hang if the program on the other end crashes. -Ed From pierre.haderer at enst-bretagne.fr Tue Apr 24 02:09:36 2007 From: pierre.haderer at enst-bretagne.fr (pierre.haderer at enst-bretagne.fr) Date: Tue, 24 Apr 2007 10:09:36 +0200 Subject: [links] links almost working Message-ID: <20070424100936.iz07pyh4kc0s4c48@webmail.enst-bretagne.fr> hello, I just met a problem, I have know idea why it is happenning, maybe someone knows about it. I configure links to make it work with graphics, with svgalib so I started with : ./configure --enable-graphics --without-x --without-direct-fl the result was : --------------------------------------------------------- Configuration results: GPM support: NO SSL support: NO Supported compression: ZLIB Javascript enabled: NO Graphics enabled: YES Graphics drivers: SVGALIB Image formats: GIF PNG XBM JPEG --------------------------------------------------------- so I continued with : make without problem, and I finally used this : ./links -g -mode 800x600x16M The result is like I had unplugged my screen from my computer (no signal). When I come back to the x server console (ctrl+alt+F7), in my terminal I can read : [svgalib: allocated virtual console #8] Have I forgot to add an option ? The only way to stop it is to close my terminal. After that, if I go back to the previous console (ctrl+alt+F8), the first time I'm immediately redirected to the x server console, the next time the change works and I can read the name of my graphic card and the version of svgalib. If someone know why this happen ... (or need precision) Thank you for your attention. -- Pierre Haderer From clock at twibright.com Tue Apr 24 05:36:44 2007 From: clock at twibright.com (Karel Kulhavy) Date: Tue, 24 Apr 2007 13:36:44 +0200 Subject: [links] links almost working In-Reply-To: <20070424100936.iz07pyh4kc0s4c48@webmail.enst-bretagne.fr> References: <20070424100936.iz07pyh4kc0s4c48@webmail.enst-bretagne.fr> Message-ID: <20070424113644.GB12455@kestrel.barix.local> On Tue, Apr 24, 2007 at 10:09:36AM +0200, pierre.haderer at enst-bretagne.fr wrote: > hello, > > I just met a problem, I have know idea why it is happenning, maybe > someone knows about it. > > I configure links to make it work with graphics, with svgalib so I > started with : > ./configure --enable-graphics --without-x --without-direct-fl > the result was : > --------------------------------------------------------- > Configuration results: > > GPM support: NO > SSL support: NO > Supported compression: ZLIB > Javascript enabled: NO > Graphics enabled: YES > Graphics drivers: SVGALIB > Image formats: GIF PNG XBM JPEG > --------------------------------------------------------- > > so I continued with : > make > without problem, and I finally used this : > ./links -g -mode 800x600x16M > > The result is like I had unplugged my screen from my computer (no > signal). When I come back to the x server console (ctrl+alt+F7), in my > terminal I can read : > [svgalib: allocated virtual console #8] > Have I forgot to add an option ? > > The only way to stop it is to close my terminal. After that, if I go > back to the previous console (ctrl+alt+F8), the first time I'm > immediately redirected to the x server console, the next time the > change works and I can read the name of my graphic card and the > version of svgalib. Sounds like problems with svgalib itself, not links. Can you run other svgalib applications? CL< > > If someone know why this happen ... (or need precision) > > Thank you for your attention. > > -- > Pierre Haderer > > _______________________________________________ > links-list mailing list > links-list at linuxfromscratch.org > http://linuxfromscratch.org/mailman/listinfo/links-list From pierre.haderer at enst-bretagne.fr Wed Apr 25 02:42:42 2007 From: pierre.haderer at enst-bretagne.fr (pierre.haderer at enst-bretagne.fr) Date: Wed, 25 Apr 2007 10:42:42 +0200 Subject: [links] keyboard shortcut in links Message-ID: <20070425104242.6ly34hlr28ws8sg8@webmail.enst-bretagne.fr> Hello, for my application (TAPA : give aged people an easier access to their mails and photos), a simple way to built the interface is the keyboard shortcut that I can do with the keyword acceskey in HTML. It seems that it's not working in links and I was wondering if it was possible to add it or to advice me on the way I could add this part in the source. Thank you all -- pierre Haderer From clock at twibright.com Wed Apr 25 07:10:38 2007 From: clock at twibright.com (Karel Kulhavy) Date: Wed, 25 Apr 2007 15:10:38 +0200 Subject: [links] keyboard shortcut in links In-Reply-To: <20070425104242.6ly34hlr28ws8sg8@webmail.enst-bretagne.fr> References: <20070425104242.6ly34hlr28ws8sg8@webmail.enst-bretagne.fr> Message-ID: <20070425131038.GC13691@kestrel.barix.local> On Wed, Apr 25, 2007 at 10:42:42AM +0200, pierre.haderer at enst-bretagne.fr wrote: > Hello, > > for my application (TAPA : give aged people an easier access to their > mails and photos), a simple way to built the interface is the keyboard > shortcut that I can do with the keyword acceskey in HTML. It seems What do you mean with keyword accesskey in HTML? CL< > that it's not working in links and I was wondering if it was possible > to add it or to advice me on the way I could add this part in the > source. > > Thank you all > > -- > pierre Haderer > > _______________________________________________ > links-list mailing list > links-list at linuxfromscratch.org > http://linuxfromscratch.org/mailman/listinfo/links-list From pierre.haderer at enst-bretagne.fr Thu Apr 26 01:32:17 2007 From: pierre.haderer at enst-bretagne.fr (pierre.haderer at enst-bretagne.fr) Date: Thu, 26 Apr 2007 09:32:17 +0200 Subject: [links] keyboard shortcut in links Message-ID: <20070426093217.d5bu3bbq2o4k4oc8@webmail.enst-bretagne.fr> Karel Kulhavy wrote : What do you mean with keyword accesskey in HTML? for example in HTML, I can find the word accesskey in a line like this : introduction this allow me to press "Alt+1" to activate the link, instead of using my mouse to click on it. I tried a web page with this kind of shortcut with links but I doesn't seem to work. Someone told me that it is just an option to activate in the setup but he's not sure of it so I will try to find that. But if you know more, I would be really thankful. Thank you by advance. -- pierre Haderer From clock at twibright.com Thu Apr 26 08:27:41 2007 From: clock at twibright.com (Karel Kulhavy) Date: Thu, 26 Apr 2007 16:27:41 +0200 Subject: [links] keyboard shortcut in links In-Reply-To: <20070426093217.d5bu3bbq2o4k4oc8@webmail.enst-bretagne.fr> References: <20070426093217.d5bu3bbq2o4k4oc8@webmail.enst-bretagne.fr> Message-ID: <20070426142741.GA16503@kestrel.barix.local> On Thu, Apr 26, 2007 at 09:32:17AM +0200, pierre.haderer at enst-bretagne.fr wrote: > Karel Kulhavy wrote : > What do you mean with keyword accesskey in HTML? > > for example in HTML, I can find the word accesskey in a line like this : > introduction > this allow me to press "Alt+1" to activate the link, instead of using > my mouse to click on it. > I tried a web page with this kind of shortcut with links but I doesn't > seem to work. Someone told me that it is just an option to activate in > the setup but he's not sure of it so I will try to find that. But if > you know more, I would be really thankful. I don't think this is implemented in Links. Mikulas, what do you know? CL< > > Thank you by advance. > > -- > pierre Haderer > > > _______________________________________________ > links-list mailing list > links-list at linuxfromscratch.org > http://linuxfromscratch.org/mailman/listinfo/links-list From mikulas at artax.karlin.mff.cuni.cz Thu Apr 26 13:49:18 2007 From: mikulas at artax.karlin.mff.cuni.cz (Mikulas Patocka) Date: Thu, 26 Apr 2007 21:49:18 +0200 (CEST) Subject: [links] keyboard shortcut in links In-Reply-To: <20070426142741.GA16503@kestrel.barix.local> References: <20070426093217.d5bu3bbq2o4k4oc8@webmail.enst-bretagne.fr> <20070426142741.GA16503@kestrel.barix.local> Message-ID: On Thu, 26 Apr 2007, Karel Kulhavy wrote: > On Thu, Apr 26, 2007 at 09:32:17AM +0200, pierre.haderer at enst-bretagne.fr wrote: >> Karel Kulhavy wrote : >> What do you mean with keyword accesskey in HTML? >> >> for example in HTML, I can find the word accesskey in a line like this : >> introduction >> this allow me to press "Alt+1" to activate the link, instead of using >> my mouse to click on it. >> I tried a web page with this kind of shortcut with links but I doesn't >> seem to work. Someone told me that it is just an option to activate in >> the setup but he's not sure of it so I will try to find that. But if >> you know more, I would be really thankful. > > I don't think this is implemented in Links. Mikulas, what do you know? > > CL< Hi It is not implemented. It would be relatively simple, so if you want to pay, I can do it. (but that there is no javascript support in Links, I removed it few days ago). Mikulas From pierre.haderer at enst-bretagne.fr Fri Apr 27 01:53:42 2007 From: pierre.haderer at enst-bretagne.fr (pierre.haderer at enst-bretagne.fr) Date: Fri, 27 Apr 2007 09:53:42 +0200 Subject: [links] Re : keyboard shortcut in links Message-ID: <20070427095342.3nt2otic6cgws8cg@webmail.enst-bretagne.fr> >>> Karel Kulhavy wrote : >>> What do you mean with keyword accesskey in HTML? >> >> Pierre Haderer wrote : >> for example in HTML, I can find the word accesskey in a line like this : >> introduction >> this allow me to press "Alt+1" to activate the link, instead of using >> my mouse to click on it. >> I tried a web page with this kind of shortcut with links but I doesn't >> seem to work. Someone told me that it is just an option to activate in >> the setup but he's not sure of it so I will try to find that. But if >> you know more, I would be really thankful. > > Karel Kulhavy wrote : > I don't think this is implemented in Links. Mikulas, what do you know? > > CL< > > Mikulas wrote : > Hi > > It is not implemented. It would be relatively simple, so if you want to > pay, I can do it. (but that there is no javascript support in Links, I > removed it few days ago). > > Mikulas Hi, Since I'm just a student making a little project, I won't be allowed to spend anyting but thanks ... For the moment, I'm looking for an other way to make these shortcuts possible. Thank you for your answers. Pierre From kali at flowtrack.com.au Tue Apr 17 15:51:29 2007 From: kali at flowtrack.com.au (Kali) Date: Tue, 17 Apr 2007 16:51:29 -0500 (CDT) Subject: [links] OT In-Reply-To: <20070416223345.GA12900@kestrel.barix.local> Message-ID: On Tue, 17 Apr 2007, Karel Kulhavy wrote: > What computer are you using? I am building an exercise bike generator and I > want to connect a PC or laptop to mitigate the unhealthy lifestyle connected > with computer-based occupation. ...I was talking about an old 486 VESA bus no-name, but I think the go would be a laptop so that its battery would act as a UPS between pedal strokes. > > At the moment I have a 60W SEIG (self excited induction generator) from an > ordinary 60W three phase asynchronous motor, which can light a 60W light bulb > (though not full shine) when powered by an electric drill. > CL< ...Amazing! I built just such a rig many years back. I did not know that you called the generators SEIG, even after doing years of research work on them. see http://www.flowtrack.com.au The biggest problem I had was that the pedal generator had a horrible speed/torque curve and tended to go through excitation/de-excitation cycles as you pedalled harder and softer. This could be sorted out by either having more flywheel on the cycle, or by fiddling the generator voltage with a transformer or star/delta rewire if possible. The permeability of the generator magnetic circuit starts low (1000) then gets good at operating voltage (10,000) then falls off again. This means excitation voltage is very unstable until it gets up to about normal, at which point you are wasting a lot of puff just exciting the thing. The problem will be made worse however by the switchmode power supply you get with laptops. They will typically work on 100 to 250 volt, and be unfussy about frequency, but they have negative marginal impedance - the slower you pedal the harder it gets. This means they will pull down the voltage to the lowest functional point and probably run erratically in some brownout/start mode. I would try it on an expendable laptop supply! Regards Kali From clock at twibright.com Sat Apr 28 14:22:53 2007 From: clock at twibright.com (Karel Kulhavy) Date: Sat, 28 Apr 2007 22:22:53 +0200 Subject: [links] OT In-Reply-To: References: <20070416223345.GA12900@kestrel.barix.local> Message-ID: <20070428202253.GA6119@kestrel.barix.local> On Tue, Apr 17, 2007 at 04:51:29PM -0500, Kali wrote: > > On Tue, 17 Apr 2007, Karel Kulhavy wrote: > > What computer are you using? I am building an exercise bike generator and I > > want to connect a PC or laptop to mitigate the unhealthy lifestyle connected > > with computer-based occupation. > ...I was talking about an old 486 VESA bus no-name, but I think the go > would be a laptop so that its battery would act as a UPS between pedal > strokes. > > > > At the moment I have a 60W SEIG (self excited induction generator) from an > > ordinary 60W three phase asynchronous motor, which can light a 60W light bulb > > (though not full shine) when powered by an electric drill. > > CL< > ...Amazing! I built just such a rig many years back. I did not know that > you called the generators SEIG, even after doing years of research work on > them. see http://www.flowtrack.com.au > The biggest problem I had was that the pedal generator had a horrible > speed/torque curve and tended to go through excitation/de-excitation > cycles as you pedalled harder and softer. This could be sorted out by > either having more flywheel on the cycle, or by fiddling the generator > voltage with a transformer or star/delta rewire if possible. How does your Flowtrack controller work? Do you have multiple capacitors that are switched on and off to change the excitation capacity? CL<