hcayless / img2xml Goto Github PK

Do we have/need a favicon?

Background image looks good, but the background for the main copy does not. Review colors and backgrounds.

Probably true of all modules in terms of the length of the intro text. Have a "read more" link.

A jquery expand so that the user doesn't get sent to another page is probably a nice alternative to multiple pages.

Lots of off-site links to Internet Archive, Google Books, etc. This will have some structure, eg. be grouped by fiction, nonfiction, poetry, etc. Something prettier than lists would be great.

The transcription doesn't contain pages 198 (inside back cover) and the back cover, even though there are images for these.

The top navigation looks bad at 1024x768. Plz fix.

CSS problem. please fix

The CSS for the modules dropdown menu is broken. Please fix.

Several things off the bat:

• page heading should be up top instead of to the right; match BRP styling.
• Add a TOC for each page with links to the subheadings.
• Pull in from tei header as in True and Candid
• add "Notes:" before the notes and make it the same size as a subheading, but don't put it in the TOC for the page.
• render tables and style them
• render blockquotes and style them
• rendering lists with no bullets by default, but bullets if it has a type=bulleted

developing locally was great, but now people need to see it

we have a title... where on the page does it go?

Create and test an alternate design that does not rely on transparency. Paper background should stay as much as possible. Blue should be close to existing blue. Body text should be black.

The latest version of OpenLayers is 2.7. img2xml's viewer will need to be refactored to use it, and to accommodate future upgrades.

"P:\docsouth\Dusenbery\photographs\images for intro pages"

Word document lists images for each intro. Captions are given for those that need captions. Note that some of the intros do not have entries in the navigation yet.

Clicking on people in the family tree should take the user to a page like:

/dusenbery/family/name/PERSONNAME

which should query the DB for all encoded documents matching PERSONNAME.

See Word docs in "P:\docsouth\Dusenbery\modules\Family". Links to individual objects should open a viewer like the main journal viewer. Everything has a page image and has been encoded.

potrace takes a -k option to set the black/white cutoff. Some pages may need a different value because they are other paper pasted into the journal... document in the wiki what pages need what, what values had what effect.

$ potrace --help|grep turd

-t, --turdsize - suppress speckles of up to this size (default 2)

Do we need to play with this? This will probably clip dots over the letter "i" if we use it, which would be bad. Perhaps best to leave it alone. Worth experimenting.

The blue highlighting should not overlap the underlining on other links.