Quote from: zi at Feb 07, 2011, 08:32 AM
For Revo manager speed, upgrade to 2.0.7. this release has special optimizations, and IMO its almost as speedy as Evo now. Preferably use Google Chrome as its JS engine is way faster than Firefox 3.6.
We’ve finished pruning the fat off 2.0.7 and our observations so far is there is that the UI is just as slow.
Revolution might be ok for small personal sites, but sites with dozens of pages, templates, tv’s, categories, snippets, and chunks .. the UI speed is very slow, too slow for us to continue so we’ve been forced to develop in 1.0.5, and then we’ll need to use a manual method of migrating up to Revolution which we’ve determined - while quite possibly a slow process too - won’t be nearly as painful as constructing pages and forming content in 2.0.7.
1.0.5 on the same server (different domain/account) measured on 3 different computers is not only faster, its easier to naviagate when there’s a large number of components going into a large number of pages of a very large site.
We’ve made sure this can’t happen to us again, but its disappointing that Revolution is such a far departure from previous versions with very little in the way of easing the process of upgrading from sites developed in earlier versions.
The absence of a "one/two click" backup (a feature we’ve been really comfortable with for the past 4 years and made it easy moving sites from development to clients) was a real shock. We usually spend a considerable amount of time ensuring clients servers are configured properly for friendly URLs and moving all the files over that complete the integration, now we have that "and" no real idea if a site built in Revolution can actually be moved seamlessly off the server its developed on, and onto the clients server without that client possessing an engineering degree.
It wouldn’t be such a big deal if we were moving sites between ourselves, but clients servers can be like a box of chocolates. And their first reaction when something doesn’t seem to be working quite right, or quickly enough - or if its too complex for them to understand, is that its somehow a failure on the part of the supplier.
So, for 1.0.5 we’re in familiar enough waters to bring clients up to speed with just about every problem that can ever arise. We’ve been able to do that from version to version, because they’ve all been fundamentally the same. But its taken years of experience to get up to speed enough to guide clients to a seamless migration and know everything that can go wrong. And everyone in this community knows when I say "everything" that’s a very long piece of string!
Revolution on the other hand, is such a far departure that it might as well be a completely different CMS. If we were to develop in Revolution and something goes wrong with the migration, then we could potentially spend days if not weeks trying to get to the bottom of whats gone wrong. And in that time, all a client see’s is no site = stupid supplier!
Sorry for the rant guys but we’ve spent days just trying to find a way to move a 1.0.5 site to Revolution and tweaking Revo to work faster. All efforts have failed, so we’re back to square one minus all the work that had been done which got destroyed during the 2.0.7 upgrade. And since when have upgrades, destroyed the site they were supposed to be upgrading. That’s not upgrade - thats a "start all over again".
I think our case is probably rare. We’re a commercial operation producing large sites for high profile clients and mistakes and delays just aren’t tolerated. Modx is a small part of a big process, but when something goes wrong - like right now, it has a ripple effect that could easily become a tidal wave.
We’ll stay with 1.0.5 for as long as we can and I’d expect that the cost of a Revolution integration, once the lag factor is drawn into it, will be prohibitive for most clients. A cms for most clients is the ability to edit content, for commercial integrators, its the ability to deliver a soluton as quickly as possible and Revolution just doesn’t fit the bill.
☆ A M B ☆