[Web4lib] managing files...

[Mark Gilman]

Jonathan,

Just the sort of advice I was seeking.  Thank you!  

It appears as though TortoiseSVN will do much of what I would need.
http://tortoisesvn.sourceforge.net/  I also found a recent version of WGET
compiled for Win32 that looks serviceable for the mirroring aspect.
http://users.ugent.be/~bpuype/wget/

(I only thought of Dreamweaver because the latest version seems to have lots
of enterprise-level features like the ability to do a quick diff comparison
between files, or so I remember reading.)

Our access situation is likely to improve, in due course.  In the meantime,
your suggestions may well prove invaluable.

Regards,
Mark 

-----Original Message-----
From: Jonathan Gorman [mailto:jtgorman at uiuc.edu] 
Sent: Wednesday, May 10, 2006 4:27 PM
To: Mark Gilman
Cc: Web4Lib (web4lib at webjunction.org)
Subject: Re: [Web4lib] managing files...



On Wed, 10 May 2006, Mark Gilman wrote:

> Hi,
>
> I maintain several different websites of varying size and complexity.  
> In the case of the largest one, with thousands of pages and files, I 
> do not have FTP access and must rely on an intermediary.


Given your conditions it's hard to imagine anyone working efficiently.

>I have
> Frontpage and various free code editors, but not, say, Dreamweaver.

Not sure Dreamweaver would really help here.

> Lately,
> I've imported the entire site onto my local drive, so that I'm working 
> on a clone, but I still worry about getting out of synch with what's 
> online, and it's a bit of a pain to keep saving the source from the 
> server to the local verisimilitude of the site.
>
> Anyone know of a better solution that doesn't rely upon FTP access?  
> The

It's not the best solution, but I'd use a combination of mirroring tools and
a versioning system.  Personally, I'd probably use a combination of wget and
svn or cvs.  In the case where there might be other people editing the
files, I'd have a process running the wget mirroring tool at pretty regular
intervals and trying to commit.  I'd have it send me a process when that
happened so I'd know if everything merged fine or if there was a conflict I
needed to solve.  I keep this separate from my current working version.

I'd then use a combination of the logging capabilities of svn and the
ability to have "hooks".  After committing the changes to a version where
I'd be ready to upload, I could then run some commands and find out all the
files I've changed since I'd started revising my mirror version.

But setup in this case would require some time getting familiar with svn or
cvs.  There might be more docs available for cvs than svn but svn is getting
pretty old now.  If you do a lot of changes, getting familiar with a
versioning system is worth very, very helpful.  It would also give you the
ability to "roll" back mistakes and correct them.



Jon Gorman