Frank Kuester wrote: > Sorry, I don't understand anything. Do you want to say that we are > somehow tied to hevea's transition? Why so? Yes, that's what he's saying. This sort of thing is routinely hideously complicated. You'll get more information from looking at update_output.txt at the results of the attempt to run the tetex-base hint.
This notes that the following packages go in as part of the hint: [tetex-base,arabtex,bibtex2html,cm-super,dvipsk-ja,hevea,hlatex,latex.service,ocamlweb,tetex-bin,tipa,whizzytex,freetype1] And it explains that the following packages get broken by a tetex-base update: * i386: bibtex2html, hevea, latex.service, ocaml-core, ocamlweb * alpha: latex.service * ia64: latex.service * mips: latex.service * mipsel: latex.service * powerpc: latex.service * hppa: latex.service * arm: latex.service * m68k: latex.service * s390: latex.service * sparc: latex.service New hevea depends on new ocaml, so it breaks when it goes in. The same happens with bibtex2html and ocamlweb (ocamlweb breaks ocaml-core), and something similar happens with latex.service which depends on new GNUStep libraries. The real question is why these packages are being accepted as part of the hint. Normally this would happen because new tetex breaks old hevea. (So the uninstallability count doesn't change when hevea is added.) But it doesn't. Same with the others. Ah, I see -- this is probably happening because hevea is considered before tetex-bin. When tetex-base is considered without tetex-bin, this will of course appear to break old hevea, which depends on both (since tetex-base and tetex-bin *must* go in together). Similarly all the others. Steve, please try changing the hint as follows: force tetex-base/3.0-13 hint tetex-base/3.0-13 tetex-bin/3.0-13 The linkage isn't real, and this should hopefully convince the testing scripts *not* to add new hevea (or latex.service, etc.) to the tetex-base hint. -- Nathanael Nerode <[EMAIL PROTECTED]> Read it and weep. http://rawstory.com/news/2005/Text_of_Gore_speech_0116.html -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]