[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

SEUL: The website

cbhiii-linux@ic.net writes:
[this was personal mail..hopefully nobody will mind if I quote it here]
>Now I know that I am new to all of this, but why all the extra hoops to jump
>through... meaning why does SDOC require us to modify the way we do things,
>when it's really the SDOC program that needs to be modified? I don't mind
>uploading them and then committing the files, but why do they have to be
>specially formatted to a now standard HTML set?!

Well put. Here's the deal with the website. It was originally designed,
long ago, to make use of sdoc (which was designed for the website) to make
it much easier to get a common layout for each page. On the front page (and
many other pages), those little section headers you see to the right of the
text are the result of <section>...</section> tags. Things like that.

Now, sdoc was a good idea, somewhere along the line. I don't think it's such
a great idea currently. This means that if you want to do a webpage for
seul, you either
* make use of sdoc, use .sdoc as the suffix for your file, and it will
  handle it for you
* use .html as your suffix, and sdoc never enters the picture.

I have zero sense of style, so sdoc is very useful for me -- that way,
somebody else can decide what looks good, and even if they change their
mind later as to how <section> should actually look, my page doesn't need
to be changed. On the other hand, if you want your page to look different,
or you use an automatic html-generating program like hotdog, or any other
reason you can think of, just ignore sdoc.

Anyway, I don't want to hold anybody back because sdoc is broken in some
way or another. There was an effort a little while ago to make sdoc much
more modern and clueful, but apparently that's stalled (mark?). Oops.

Another point to address: I think Chuck (and probably others) are using
windows. I much prefer people to use ssh to connect to cran. This works
fine because there's a ttssh (teraterm pro with ssh plugin) program out
there. However, ttssh doesn't support scp (secure copy, like ftp). Which
means that Chuck is probably having a heck of a time trying to move his
files from his home PC to cran, where they will be useful. 

Pete wrote:
>3) upload the file into your directory

This is the point I mentioned above. This can be done via sz/rz (send
zmodem, receive zmodem).

Alternatively, it can be done securely via anonymous ftp:
upload your file to /pub/seul/incoming/ over anonymous ftp, and then
you'll find your file in /home/ftp/pub/seul/incoming/ on cran, and you
can mv it to the appropriate place in your repository.

Speaking of the repository. This is how you get your own version of
the repository, in 4 easy steps:
1) login to cran
2) mkdir -p work/seul
3) cd work/seul
4) cvs -d /home/seul/cvsroot checkout .

>4) use the following command to commit the file to the repository:
>     cvs commit -m "your message here" filename.sdoc
>     I'm assuming here that the repository is ready to accept your file
>(I think it's 'cvs commit -a filename.sdoc' beforehand). This brings up
>the problem of how to get at files outside of our own directories -
>particularly those which share ownership - which is where I'm at a loss.

That's the great thing about CVS. All files can share ownership. People
can be editing the same file at once (as long as they're editing
different parts of the file), and then they can both commit their changes
and it will work.

The trick is that before you start editing a file, you want to make
sure that you're using the most up to date version of the file. To do
this, type
cvs update -d -P
in the directory in which you're about to start editing. It will
output some stuff (indicating the files that were just updated with a
"U" at the beginning of that line),, and when it's done you can start
editing as normal. If somebody has checked in that file since you
began editing, you will also want to update the file (with the same
command as before) right before you check it in, so you can integrate
the changes with the file you've been working on, to make sure it worked.

If the above confused you:
just run "cvs update -d -P" in your ~/work/seul/ directory every time
you login.

Speaking of which, if all of this is confusing you: just remember that
there are a lot of different ways to do each of these tasks, but you only
have to figure out your favorite way -- you can ignore all the other ways.

>I'd like to rewrite the current SDOC info, but am unclear as to how to
>get at the files to edit them in the first place ... what directory
>should I be in, how do I find the file, etc.  Roger, any advice?

They live somewhere in the repository. The 'find' command is your
friend. :)

find ~/work/seul/ -exec grep -i "your text here" {} \; -print

This command, with "your text here" being a phrase that you want
to look for inside each file, will list each file that has that
phrase, plus the other text on that line. But it's not perfect,
because it won't find it when your phrase is split up into two lines,
etc. The -i means a case insensitive search.

For instance, there's a lot of sdoc material in /pub/website/html/ in
the repository. For the most part, you can guess where it should live
in the repository, based on where you found it on the website. In
particular, read the gettingstarted.sdoc file:

Aiee. I get the feeling I've introduced more confusion than I've
solved, with this mail. Please ask questions. :)