mark lewis wrote in a message to Roy J. Tellason:



 RJT> In my case it's 0000FF and 000080,  blue and navy.  Not a *real* 
 RJT> high contrast,  but some.  I'm thinking of changing that,  but...

 ml> yeah, blue and navy blue may be a bit hard to see... 

That's why I'm thinking of changing it.  

 ml> that's why i generally go for bold differences... black page, cyan 
 ml> text, yellow links, bright purple visited links just for an 
 ml> example...

Whew!  That's a heck of a color scheme.  Too bad I have this stuff embedded
into so many files.  There was some way of doing it all at some one
centralized spot,  but I forget what it is.  Sorta like #include for c,  or
similar.

 ml>> yes, you can use the font tag to alter the color for specific
 ml>> sections of text... not sure, but i believe so, that you can alter
 ml>> the color of specific links...

 ml>> i'm not even going to touch on CSS (cascading style sheets) other
 ml>> than to say that not only can they control the font style and size
 ml>> but also the color...

 RJT> That's further into it than I want to get at present.

 ml> yeah, me too... but i'm heading that way, slowly but surely...

At this point I'm just throwing a whole mess of text that I have on hand
here into html files along with the stuff I got that was already in html
format,  and having it accessible and heavily cross-linked.  That's enough
for a good start.  And at that there are at least hundreds of pages...

 RJT>> They're not terribly different,  but they should be noticeably
 RJT>> different, as that's the way I have it set up. I was only having
 RJT>> this problem with one set of links out of a whole big page.  I
 RJT>> finally figured out a workaround, putting underscores in place of
 RJT>> the spaces that were embedded within the filename.

 ml>> you could possibly use %20 in the link, too and leave the spaces...

 RJT> Not sure how you mean,  there.  The space was actually in the
 RJT> filename,  not in the part between ">link text,  where
 RJT> spaces don't give me any trouble.

 ml> right... a file named "my file.txt" sans quotes would be href'd
 ml> like 

 ml>   my file

Gotcha.  I can try that,  see how it works.

 RJT>> I have *NO* idea whether the fault was in the Konqueror browser
 RJT>> or in Apache, though I suppose I could set up a test and try IE
 RJT>> from the w98 box.  Weird stuff,  anyhow...

 ml>> as long as you remember that IE doesn't always follow the
 ml>> standards, it is one test, yes...

 RJT> At this point it's about the only one I can think of, offhand.

 ml> i missed it before but the fault in rendering can't be in apache...
 ml> all it does is serve up what's there... the fault in rendering has
 ml> to be in the receiver application... lynx, konqueror, netscape
 ml> mozilla, ie, crazybrowser, whatever... 

Ok,  so it's in the browser then.  I had suspected as much.  The system
seems to be slightly less stable with X up and running continuously (and
KDE and so forth) than it does when I leave it just sitting there in
console mode.  I only have that so I can see how the pages will look, 
graphics and all.

There are occasional peculiarities.  Like yesterday or the day before
something (I forget what now) segfaulted.  Just out of the blue,  for no
apparent reason.  And it was something that'd been fine for I don't know
how long.  I suspect that some _other_ process stepped on its memory
somehow.  But I'm not terribly worried about it.  Overall the system is
*WAY* better than the NT and apps I use at work,  which get flaky after
being up for a day or so.  I crashed one app this afternoon,  about 4:30, 
just by wanting to slightly resize a window that a VB script was trying to
write to,  ferinstance...

---