Re: [site] download pages
On Sun, 13 Feb 2005 14:39:27 -0500, Henri Yandell <[EMAIL PROTECTED]> wrote: > On Tue, 08 Feb 2005 22:03:57 +, robert burrell donkin > <[EMAIL PROTECTED]> wrote: > > On Tue, 2005-02-08 at 21:24, Howard Lewis Ship wrote: > > > +1 on this change in general > > > > +1 > > > > though (for some reason) i find the aesthetic a little unbalanced. > > > > i'd be inclined to prune. > > +1. Probably still too noisy on the front page. I've pruned. > > the link to the archives duplicates content already present on the > > sub-projects pages. the nightly builds could be moved onto those pages. > > +1 to nightlies being at a lower level. They're now there. > Archives should stay at the top level as it links to Jakarta archives > and not others. But I've made it an ad-hoc link in the text, rather than giving it its entire section. > > the covalent link could be removed (i'm not even sure whether covalent > > still runs gump.) > > It doesn't. I fixed it to the correct link a while back. Kept this on the main page for the moment. I've also added a 'browse download area', inspired by xml.apache.org's download page wording, to each subproject. http://jakarta.apache.org/~bayard/jakarta/site/downloads/download.html Thoughts? I suspect I should css some spacing between the list items now; they get a bit crammed. Other than that, my todo's before proposing we replace this are: 1) Get Ant copying the download.cgi (or another way to get the cgi to work for each file). Not even sure if copies of the download.cgi will work. 2) Fix missing ul tags on sub-sub-project pages. Tomcat, Velocity, Commons etc. Immediately following the move, I'd then probably: 1) Improve the Taglibs download pages to be similar to the Commons download pages. 2) Start suggesting that subprojects link to their subpages and not to the top one. Hen - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [site] download pages
On Sun, 13 Feb 2005 14:39:27 -0500, Henri Yandell <[EMAIL PROTECTED]> wrote: > On Tue, 08 Feb 2005 22:03:57 +, robert burrell donkin > <[EMAIL PROTECTED]> wrote: > > On Tue, 2005-02-08 at 21:24, Howard Lewis Ship wrote: > > > > > > > yes, it's a lot of pages, but one of the #1 issues people have with > > > Jakarta is simply finding the right download. > > > > > > The formatting is still iffy; I think it would be more readable in a > > > tabular format, with a column for MD5 and a column for PGP. > > > > +1 > > Will do. This was a complete lie. Turns out that it's damn hard to line it up using CSS, going to a tabular format is also hard as it's currently hierarchical data and tricky to do in a generational way. The simplest solution I found (thanks wife) is to drop it down a line: http://jakarta.apache.org/~bayard/jakarta/site/downloads/downloads_tapestry.html Any opinions? Hen - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [site] Odd Ant question
I haven't tried it, but how about this: * Use with 'genericantfile' (referencing the same build file) and a for the directories you want. * In the target invoked by , use to determine whether or not the HTML file exists, and invoke another target that only runs 'if' the property is set. That second target would do the copy. * Use to generate the name of the CGI file based on the name of the HTML file. -- Martin Cooper On Sun, 13 Feb 2005 21:38:32 -0500, Henri Yandell <[EMAIL PROTECTED]> wrote: > Before I lug myself off to the Ant lists, thought I'd ask here. > > I'm trying to come up with a way to make the cgi work for the > downloads pages. One way would be to rewrite the cgi, have it take > arguments and be more dynamic. Another easier way would be to simply > copy the cgi file lots of times (as it's already been copied many > times already). > > To do this copying, I basically want to do the equivalent of: > > --- > For each downloads_*.html file > copy downloads.cgi downloads_*.cgi > > > Any idea how I do that in Ant? Mappers seem like they'd want to be > copying the html file to the cgi file; ie) wildcard in to and from. > There's no for loop, so not sure how I'd do such a thing. > > Maybe one way would be to call a target on each file in a list of files. > > Any ideas? > > Hen > > - > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
[site] Odd Ant question
Before I lug myself off to the Ant lists, thought I'd ask here. I'm trying to come up with a way to make the cgi work for the downloads pages. One way would be to rewrite the cgi, have it take arguments and be more dynamic. Another easier way would be to simply copy the cgi file lots of times (as it's already been copied many times already). To do this copying, I basically want to do the equivalent of: --- For each downloads_*.html file copy downloads.cgi downloads_*.cgi Any idea how I do that in Ant? Mappers seem like they'd want to be copying the html file to the cgi file; ie) wildcard in to and from. There's no for loop, so not sure how I'd do such a thing. Maybe one way would be to call a target on each file in a list of files. Any ideas? Hen - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [site] download pages
On Tue, 08 Feb 2005 22:03:57 +, robert burrell donkin <[EMAIL PROTECTED]> wrote: > On Tue, 2005-02-08 at 21:24, Howard Lewis Ship wrote: > > +1 on this change in general > > +1 > > though (for some reason) i find the aesthetic a little unbalanced. > > i'd be inclined to prune. +1. Probably still too noisy on the front page. > the link to the archives duplicates content already present on the > sub-projects pages. the nightly builds could be moved onto those pages. +1 to nightlies being at a lower level. Archives should stay at the top level as it links to Jakarta archives and not others. > the covalent link could be removed (i'm not even sure whether covalent > still runs gump.) It doesn't. I fixed it to the correct link a while back. > > > yes, it's a lot of pages, but one of the #1 issues people have with > > Jakarta is simply finding the right download. > > > > The formatting is still iffy; I think it would be more readable in a > > tabular format, with a column for MD5 and a column for PGP. > > +1 Will do. > > I would also prefer the full file name, rather than just the version > > number; if I click "1.0.tar.gz" I don't want to be surprised with > > "hivemind-1.0.tar.gz". > > +1 Currently the xml has name and href. Name (ie 1.0.tar.gz) is set to match the name on the current binindex.cgi page. One idea would be to change it so that if name is not set, the xsl somehow shows the last element of the href. Or even to just throw out name completely and force it to have the last element of the href. Hen - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
