- Inkscape Devel - Inkscape Mailman

Re: [Inkscape-devel] wiki spam
by Bryce Harrington 01 Feb '05

01 Feb '05

On Tue, 1 Feb 2005, Alan Horkan wrote: > On Tue, 1 Feb 2005, Bryce Harrington wrote: > > > Date: Tue, 1 Feb 2005 10:53:31 -0800 (PST) > > From: Bryce Harrington <bryce@...260...> > > To: mental@...3... > > Cc: bulia byak <buliabyak@...400...>, inkscape-devel(a)lists.sourceforge.net > > Subject: Re: [Inkscape-devel] wiki spam > > > > All of the suggestions in this thread are viable options, although some > > will require more work than others: > > > > * Changing the password: Trivial, easy to do > > * Blocking more IP's: Trivial, easy to do > > * Adding basic auth login: Requires sysadmin work + ongoing account admin > > I assumed it was already built in to the wiki and all that would be > required would be to turn it on. Well, it does have a sitewide password that can be turned on for editing, but it'd be the same password for everyone. But that's essentially what we've got currently so really the next step would be to turn on either basic auth (implemented in Apache) or a more advanced auth mechanism (implemented ourselves). Note that anything we do is going to have have tradeoffs: Either more implementation work, more administrative work, more work on the part of the user, or a combination. For example, like you point out, image-based passwords won't require admin work, but will require some implementation work, and increases the burden on users (esp. those with vision imparements). Having two different systems (image passwords or authentication), gives the users options but at the cost of double implementation work, and doubling our risk (spammers only need to find a way to attack one of the systems). There is one other programmatic defense that I know about for wiki's. For many bots, there is an inhumanly short period of time between when they click edit and when they submit the page, so apparently if you make the wiki keep track of this time and it's less than a few seconds, then you can make a fairly safe assumption that it's a bot. Our wiki doesn't have this feature, but I would bet that a good perl hacker could figure out a good way to implement it. The advantage of this system over others is that while it wouldn't stop all spamming, at least it would not impose additional administrative or user effort and it might cut down some of the spammers. I bet it would be especially effective at the spammer Bulia was fighting today. Anyone feel like giving it a shot? Bryce

1 0

bug 1079194
by Jon A. Cruz 01 Feb '05

01 Feb '05

Given that bug 1079194 is a bit ambiguous, and that there's ongoing effort for other bugs that may have alread fixed this, I'm thinking of bumping it over to pending. Sound reasonable?

2 1

Re: [Inkscape-devel] Updated spanish translation
by Lucas Vieites 01 Feb '05

01 Feb '05

On sáb, 2005-01-29 at 13:14 -0800, Arpad Biro wrote: > Hi, > > I'm sending the current CVS version attached. > > After receiving your latest translation, I compared it (cmp) to that > file, and they matched byte by byte. > > You were right, my bad. I had translated the file directly in my local cvs tree and then I sent you the one in my working directory. I found out a while ago when I did a cvs update and the es.po file gave an error. Now I'm attaching an fresh update from today's cvs. Cheers, -- Lucas Vieites Fariña Dept. Desarrollo <lucas@...212...> Asix Informática <http://www.asixinformatica.com/> Tel/Fax: +34 986 54 26 98

1 0

Multiple resolutions in a single .SVG
by Jon A. Cruz 01 Feb '05

01 Feb '05

1 0

Re: [Inkscape-devel] default shape prefs?
by unknown＠example.com 01 Feb '05

01 Feb '05

Quoting bulia byak <buliabyak@...400...>: > On Mon, 31 Jan 2005 13:44:51 -0500, mental@...3... <mental@...3...> > wrote: > > Quoting bulia byak <buliabyak@...400...>: > > > > > Well, though the function is mine, I didn't write this comparison, but > > > copied it from somewhere. If you fix it I'll copy the correct idiom > > > next time :) > > > > The correct idiom, provided I had done things properly on my end, would be > no > > cast at all. The compiler ought to be aware that SPCSSRepr ISA SPRepr. > > does not work: > > desktop-style.cpp: In function `SPCSSAttr* > sp_desktop_get_style(SPDesktop*, bool)': > desktop-style.cpp:166: error: invalid use of undefined type `struct > SPCSSAttr' > xml/xml-forward.h:6: error: forward declaration of `struct SPCSSAttr' Well, you would still need to make a couple minor cleanups. The forward declaration in xml/xml-forward.h is (I think) unnecessary (actually the whole file is unnecessary and a bad idea, but I will rant about that separately). I think I'm going to go ahead and redo the Repr CSS stuff the right way tonight though. -mental

1 0

Open Clip Art Library Release 0.10 Announcement
by Jon Phillips 01 Feb '05

01 Feb '05

*PLEASE FORWARD* Open Clip Art Library Release 0.10 Announcement February 1, 2005 UTC — Release 0.10 of the Open Clip Art Library (www.openclipart.org) is now available for download on-line as an individual package: http://www.openclipart.org/downloads/0.10/openclipart-0.10.tar.gz http://www.openclipart.org/downloads/0.10/openclipart-0.10.zip This new package is 23MB and contains 3207 unique user-submitted files in several categories such as geography, symbols, holidays and people. You may also visit the Open Clip Art Library site at www.openclipart.org to browse or search all of the clip art imagery in our database and download files individually. What is significant in this release: 1) Our largest section of clip art, the flags collection, has been moved to the signs & symbols category. 2) We've also collected St. Valentine's Day clip art images for the upcoming holiday. 3) Improvements to our tools and processes continue behind the scenes. For the month of February, we are also trying to collect images related to "Black History Month." Please submit SVG images related to history of not only African-Americans, but of black history around the world. Please visit www.openclipart.org, contribute your clip art images, and download others for your own uses.

1 0

printing comments
by Kees Cook 01 Feb '05

01 Feb '05

Does anyone object to having comments added to postscript print-outs? For example: ... %%Page: 1 1 90 rotate 0.80000000 -0.80000000 scale gsave [1.0000000 0.0000000 0.0000000 1.0000000 0.0000000 0.0000000] concat %! begin '#path4209' 0.0000000 0.0000000 0.0000000 setrgbcolor ... 372.04724 70.866136 lineto stroke %! end '#path4209' %! begin '#path4210' 0.0000000 0.0000000 0.0000000 setrgbcolor 0.68750000 setlinewidth I'm doing it right how to help me track down a printing bug, but I thought it might be fun to leave turned on? -- Kees Cook @outflux.net

1 1

font crashes
by Kees Cook 01 Feb '05

01 Feb '05

I solved the Pango font crashes related to Font descriptions not having a family name specified (the "1979" font bug). I'm not sure what other bugs it'll fix, but I made a testcase for Pango and opened a bug for them. I set up code to work around it, or at least avoid the crash in a worse-case situation. -- Kees Cook @outflux.net

1 0

Fedora core 2 RPMs for inkscape dependencies
by Ashwin Bharambe 31 Jan '05

31 Jan '05

Hi all, I made FC2 RPMs for all the inkscape dependencies -- the reason is that I personally was quite annoyed when I first started toying around inkscape code because I needed to compile all these dependencies... these _should_ work on FC3 as well... URL: http://www.cs.cmu.edu/~ashu/rpms/ Anyway, I hope these might be useful for _somebody_ , at least... -Ashwin PS: If they could be hosted on inkscape.org somewhere, that would be great (since i am perennially short of disk space on the webserver)

2 2

String 'sanitizer' routine
by Jon A. Cruz 31 Jan '05

31 Jan '05

I just dumped a simple routine into io/sys.h to return a 'safe' version of any given string. Inkscape::IO::sanitizeString The most immediate need is for those places where we might be displaying an error for a potentially bad path (such as when we're unable to create a file or directory). It will return a copy of the string as-is if it's valid UTF-8, otherwise it will walk through and escape all non-ASCII bytes. That should at least stop Pango crashes from hiding true problems or from turning warnings into crashes, like bug 1107305. Soo...... I'm hitting bed now, but anyone can feel free to try it out in a few critical places.

1 0