Kudos to CSE

[floppy]

I had the standard version of this script for maybe 24 hours and already my site is running more smoothly. I started out at over 950 errors just on my front page and I have knocked it down by 3/4. We all know I am hardly close to being finished, but I can honestly say this is addicting. You can seriously tell the difference in the way the site loads in Firefox and Internet Explorer. Maybe the best money I spent since I started my site.

I will keep following this rainbow looking for my pot of gold, but I just wanted to say thanks for a superb piece of software.

Floppy,
www.clan-themes.co.uk

[MikeGale]

It's great to hear about your progress here.

I'm really interested in the load performance change that you're sensing. Do you have measurements of that?

If not, and if you have a copy of the old and current sites (or just a page) I might be able to run a couple of measurements myself.

Let me know if you think that worth pursuing.

[Albert Wiersch]

Hi Floppy,

That's great to hear. If you don't mind, maybe you'd like to supply a testimonial and photo for our front page (photo required for front page ) that mentions the improved load time or at least your perception of it. I haven't done any measurements, but I'm pretty sure a page's load time can be reduced by fixing certain errors. It would be interesting to actually time it but I'm not sure how that could be done accurately. I'm sure there's a way though.

[floppy]

Man if you want people to visit your page we better not put a photo of me on it. Some images to link back to here would be nice. Maybe an 80x15 for my footer.

The script generation time would be extremely hard to measure as stated above because it varies from system to system and could vary depending on your programming language. Such as mine which uses php/mysql.

However whenever you clear up nearly 700 errors something is bound to change. I think the speed increase could be mostly credited to fixing improper table design and the use of style sheets vs. replicated styles. Unfortunately, replication is going to be tough as so much has changed. I now have switched to XHTML validation and that has produced a whole new list of jobs.

I do however have backups of the site and we might could come to some conclusions attempting to replicate the environment variables through the use of a subdomain or development area.

[MikeGale]

Hi,

My interest is browser side rendering of the page/s.

If generation is also an issue, it's a different issue. I wouldn't be surprised if the server side performance is little changed, not that it matters.

For testing I'm thinking of of a capture of the rendered content. This gets delivered by a server essentially acting as a file server, so timings are close to browser side performance. I grabbed a copy of http://www.clan-themes.co.uk/, yesterday. It was about 700K all told, 150 of that being the markup. If we could get a version of that "before the changes" then anybody could check on any server (with a little preparatory work).

It would also be possible to manually process the page further to see what additional performance benefits might occur. (If the more difficult issues were sorted!)

[floppy]

Once again here to thank you guys for this great program. I decided to start a new link directory and this program really saved me hours in validation troubles.

I am very proud of this one thing

http://whois.domaintools.com/lobolinks.com

100% across the board. I still got some minor warnings to clear up, but that is still an accomplishment in my book.