March 2012 - petsc-dev - mailman3.cels.anl.gov

Re: [petsc-dev] dates of last change on PETSc web pages
by Jed Brown 20 Mar '12

20 Mar '12

On Tue, Mar 20, 2012 at 17:14, Sean Farley <sean(a)mcs.anl.gov> wrote: > Ok? This is indeed what I'm using for getting the date for each page. Are > you talking about something else with the makefile or what? Sorry, somehow I thought you had a different update model in mind.

1 0

Re: [petsc-dev] dates of last change on PETSc web pages
by Sean Farley 20 Mar '12

20 Mar '12

> > This is useful: > > > http://hgbook.red-bean.com/read/customizing-the-output-of-mercurial.html#se… > Cool story, bro! http://i2.kym-cdn.com/photos/images/original/000/061/294/1106514-cool_story…

1 0

Re: [petsc-dev] dates of last change on PETSc web pages
by Sean Farley 20 Mar '12

20 Mar '12

> > Most html files are generated and we know their dates: > > $ hg log --limit 1 --template '{date|date}' include/petscsnes.h > Mon Mar 19 22:42:51 2012 -0500 > > Similarly for inserting the date-string into the non-generated pages. > Ok? This is indeed what I'm using for getting the date for each page. Are you talking about something else with the makefile or what?

1 0

Re: [petsc-dev] dates of last change on PETSc web pages
by Jed Brown 20 Mar '12

20 Mar '12

On Tue, Mar 20, 2012 at 17:08, Jed Brown <jedbrown(a)mcs.anl.gov> wrote: > Most html files are generated and we know their dates: > > $ hg log --limit 1 --template '{date|date}' include/petscsnes.h > Mon Mar 19 22:42:51 2012 -0500 > > Similarly for inserting the date-string into the non-generated pages. > This is useful: http://hgbook.red-bean.com/read/customizing-the-output-of-mercurial.html#se…

1 0

Re: [petsc-dev] dates of last change on PETSc web pages
by Jed Brown 20 Mar '12

20 Mar '12

On Tue, Mar 20, 2012 at 16:58, Sean Farley <sean(a)mcs.anl.gov> wrote: > For this update, yes, every page must be modified since we aren't doing > any fancy javascript / server-side programming (yet?). If you modify a file > a week from now, it will be that date once the website is regenerated. I'm > planning on putting this date generation in a script that runs in the > 'update-web' rule unless there is a better rule? Most html files are generated and we know their dates: $ hg log --limit 1 --template '{date|date}' include/petscsnes.h Mon Mar 19 22:42:51 2012 -0500 Similarly for inserting the date-string into the non-generated pages.

1 0

Re: [petsc-dev] dates of last change on PETSc web pages
by Sean Farley 20 Mar '12

20 Mar '12

> > Do you have to manually change each web page? Hence they get committed > with the current date? If change one file in hg in a week will it get the > date when I made that change while the rest will have today's date? > For this update, yes, every page must be modified since we aren't doing any fancy javascript / server-side programming (yet?). If you modify a file a week from now, it will be that date once the website is regenerated. I'm planning on putting this date generation in a script that runs in the 'update-web' rule unless there is a better rule? Note, if we used some kind of server-side programming, one could argue it could be automatically (with less html code) but the cons of that would be 1) requiring more software on the server, and 2) learning a new programming model. FYI, the date generation could be done in javascript if someone wrote a parser for urls of the form: http://petsc.cs.iit.edu/petsc/petsc-dev/file/tip/src/docs/website > This is fine if this is the case. Otherwise .... ?

1 0

Re: [petsc-dev] dates of last change on PETSc web pages
by Barry Smith 20 Mar '12

20 Mar '12

On Mar 20, 2012, at 4:41 PM, Sean Farley wrote: > Upper right corner looks great! > > Ok, I'll go with that. > > Is this something that can be set automagically from the last Mecurial change for the file? That would absolutely rock. > > Sure, but after I commit this changeset it would mean that all the webpages have a date of today. Do you have to manually change each web page? Hence they get committed with the current date? If change one file in hg in a week will it get the date when I made that change while the rest will have today's date? This is fine if this is the case. Otherwise ....

1 0

Re: [petsc-dev] dates of last change on PETSc web pages
by Sean Farley 20 Mar '12

20 Mar '12

> > Upper right corner looks great! > Ok, I'll go with that. > Is this something that can be set automagically from the last Mecurial > change for the file? That would absolutely rock. Sure, but after I commit this changeset it would mean that all the webpages have a date of today.

1 0

Re: [petsc-dev] dates of last change on PETSc web pages
by Barry Smith 20 Mar '12

20 Mar '12

Upper right corner looks great! Is this something that can be set automagically from the last Mecurial change for the file? That would absolutely rock. Barry On Mar 20, 2012, at 3:49 PM, Sean Farley wrote: > You have not yet modified the script that prepares the manual pages to automatically put the date of generation in the upper right corner. As well as other generated pages. > > Could you please add that ASAP or speak to me about how to do it so it will be ready for the next release. > > I also meant to do this but got distracted. I mocked up two versions: one with the date below the line (but affects some of the formatting in the main body as you can see from the screenshot), and one with the date above the line that uses the current space between the header text and the horizontal rule. > > I thought the upper right hand corner would look bad but it is better than I imagined. > > Thoughts? > <date-below-line.png><date-above-line.png><date-upper-right.png>

1 0

Re: [petsc-dev] dates of last change on PETSc web pages
by Matthew Knepley 20 Mar '12

20 Mar '12

On Tue, Mar 20, 2012 at 3:49 PM, Sean Farley <sean(a)mcs.anl.gov> wrote: > You have not yet modified the script that prepares the manual pages to >> automatically put the date of generation in the upper right corner. As >> well as other generated pages. >> >> Could you please add that ASAP or speak to me about how to do it so it >> will be ready for the next release. > > > I also meant to do this but got distracted. I mocked up two versions: one > with the date below the line (but affects some of the formatting in the > main body as you can see from the screenshot), and one with the date above > the line that uses the current space between the header text and the > horizontal rule. > > I thought the upper right hand corner would look bad but it is better than > I imagined. > > Thoughts? > I like the corner. It does not get confused with the headings. Thanks, Matt -- What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead. -- Norbert Wiener

1 0