I have to say I am really disappointed about the state and performance of ExtJS 4. When the first test releases came out I immediately started testing it. The stabilization proces was long and painful though. During Sencha Con 2011, the grid performance problems were already known but they were presented like solvable in the near future. ExtJS 4 was still being presented as the most stable and fast ExtJS to date.
This thread and my own experiences however tell a different tale. Seeing how long it takes Sencha to solve the performance problems, I think the problems (most serious on IE but also other browsers) are here to stay and cannot be solved.
Because of the problems I have had to start porting back a very complex workflow management system I created for an international bank since they cannot switch to another browser than IE8. The performance of the system under IE8 is horrible both in memory usage as CPU usage, especially with components utilizing one or more grids. Choosing ExtJS 4.1 for the new system has cost my company a lot of time and money, and has caused the project to be seriously delayed. All that does not instil confidence in my client, for whom I have gone to great lengths to convince them to deploy the application using ExtJS4 (the previous version I created using ExtJS 3.4 which runs snappy).
I think it is time Sencha gives this problem serious effort since the credibility and businesses of developers using ExtJS is at stake. I know Sencha is working on it, but is also starting up a new product line every other day it seems. Everybody knows you cannot focus on too many things at the same time.
Interesting, picking up this thread again... and some nice ideas and tips from posters; the lazy panel/container is interesting, might try it and see where that can take us. At the very least it should raise some eyebrows in the dev team, causing them to ask what effort is being wasted early in the process there.
Another good thought was on avoiding the use of many components, and instead using a single one with a custom template... could be useful for docked items on grids, say, where we're certainly seeing some issues (see my colleagues post on button layout being very expensive).
Would also like to add my voice to those calling for more emphasis on performance of Ext 4 in old IE's, rather than the constant release of new products.
I know decent performance is your goal for Ext 4, I just wish it'd take a higher priority, and you'd stop tinkering with toys and concentrate on it. Apologies to those with business applications running on phones and tablets, but I think you are probably a very small minority. Those using Ext 4 are the people paying for licenses after all, well, commercially at least.
Whilst examples are all very well and good, the real world uses real applications to get a job done, and in the corporate world people do not have the luxury of using Chrome, FF or Safari... at least not yet.
The best we can generally hope for is IE8, and sometimes, with especially forward thinking companies, IE9.
IE7 is definitely the norm though, for us at least, and those who have dealt with it will know that it's about 30-50% slower than IE8 even.
Progress is being made though I think, but as others have pointed out, there is still some way to go.
The issues have definitely influenced how we design our UIs though; always in the back of my mind is how features x, y and z will work in IE.
Look forward to hearing more on Ext 4.2.
4.1.1 or 4.1.2 have not been good enough for me to upgrade to, so hope 4.2 is...
Its disappointing that large corporates cant move past MS and IE.
IE 10 may turn the tide.
It is such a huge pain to cater for old browsers that cant perform - I know it is what it is
I agree, it's disappointing... the best we can hope for is when XP support ends in February 2014, although big corps will no doubt pay MS to continue support.
Originally Posted by suzuki1100nz
IE10 may turn the tide for consumers, but I can't see any large company upgrading to Windows 8 rather than 7, the support calls over how to shutdown the machine will cost them a fortune alone!
So, post Feb 2014 the worst browser we may get away with supporting might be IE9, although the initial versions of Windows 7 had IE8, so even that may not help...
yes your right.
I worked for a bank for 7 years and it drove me nuts the dedication to IE (took an age to get an upgrade from IE 6.0 to IE8).
A simple switch to a more html5 compliant browser makes all the difference.
Never mind - no use banging our heads against the wall.
Yes your right - conversion for corporates to windows 8 in unlikely. Win 7 will become the new XP pro and corporates will hang on to win 7 for dear life - until we see a stable Win 8+
So hope for IE10 a long way off.
Sill it worth the frustrating scream ..... arghhhhhhhhhhhhhhhhhhhh
So how far off is 4.2 beta? Is this available now for users with a support license? Is there a large performance gain for IE7 & IE8?
IE8 is here for the long haul in the corporate sphere. IE9 has a non-zero risk of introducing issues (quite a big risk in my experience), and so all web apps need to be recertified in places where certification matters. Many of the customers I deal with (banks and telecom) have such a certification process. Some are still in the planning stage of a migration from IE6 to IE8. W7+IE8 doesn't lose official support until 2020, and with IE6 we saw that many corporations were ok with playing it close to the limit, so my outlook on that is bleak.
Originally Posted by westy
ExtJS has to run "acceptably well" on IE8. It does on 3.4. It should on 4.x. I'm still on 3.4 and will not upgrade to 4.x until I see some convincing communication that the performance issues are solved.
Is it possible that someone from the Sencha team informs us about the status of the performance issues?
- Is a performance update underway and when can we expect it?
- Does the 4.1.2 update include any performance improvements?
My problems with clients are getting so big that we have to start thinking about alternatives to ExtJS. Please let me, and the rest of us, know how this problem is being tackled and keep us in the loop.
@robboerman. I don't believe 4.1.2 included any significant performance improvements. See this post for the main areas that are being targeted for 4.2: