What kind of performance is it that you are measuring to be slower in FMP 12?
Is it especially browsing/scrolling lists and table views or is it other issues?
Some actions in FileMaker 12.0v1 are definetely slower, especially in the UI. While other parts are much faster, especially those handling manipulation of larger datasets and transactions behind the scene.
And you are right: UI is crucial to the users. Therefore, help by isolating the issues.
I thought those issues had been adequately covered by the previous posts. I think we all have a pretty good idea of which features are most affected. I am trying to figure out at what point faster hardware will gloss over ALL those weaknesses enough so that the user doesn't percieve that FMP is running slower. If there are some features that simply will not return to their former responsiveness, regardless of the OS and horsepower thrown at it, perhaps we could start making a list here. The only thing missing, in my view, is a reliable baseline for the minium real world specs that will not cause a client to conclude that FMP 12 (and by extension all future releases of FMP) will be inherently slower than what they have been used to and what they rely on.
A lot of my clients have been complaining about the speed of FMP 10 for quite a while and many of those who upgraded to 11 didn't see enough of an improvement. They had high hopes for ver 12 getting them back to the kind of performance that they used to see in FMP 9. I still have clients that are using ver 8 or 9 that won't upgrade because they simply couldn't get their work done in 10 or 11. I had hoped I might appease those who might be thinking about getting a new system anyway.
These are my smallest clients, who either use only one computer or a very small number on a WAN hosted remotely. The larger clients have all turned down ver 12. They're staying with their current build until they need to add more users. They will then transition to the next version of FMP that gives them comparable performance, if one is available. If not, they will be looking at moving to another dbms. No company is going to upgrade all its hardware in a recession just because FMI wants to do certain things with this application and didn't live up to the representations it made to its developers. They just don't do that sort of thing any more. They replace or upgrade machines a little at a time, as they need to and resist ditching every computer at once to accomodate a software company. Those days, I'm afraid, are over.
I have two new clients, who are new to FMP and have no existing db. They are willing to try FMP 12, so that is the only work I am doing in the new version for now. One client is big enough to downgrade to 11, if necessary, and eat the cost of my rebuilding everything in the earlier version if necessary (though they are NOT happy about the conversion going only one way and are having serious doubts about their choice of FMP). The other simply doesn't have the money for anything but a quick and dirty ver 12 database. But they need IWP (it's in their grant proposal, so we have to do it) and they are not happy that there seems to be little or no improvement since ver 12.
I am feeling the brunt of user disastisfaction now, but I expect a lot of it will trickle down to FMI in the very near future. Mac users especially have been up in arms for quite a while about perceieved compatibilty problems with Lion and virus vulnerabilty (that latter which hardly caused Windows users to bat an eye). It's all about perception. You can't seen to be taking away features or performance that the people who ultimately pay for all this feel that they have invested in. It just makes them want to spend their money elsewhere. And as far as I am concerned, the customer is always right. (A quaint notion, admitedly, these days, but it has never steered me wrong.)
Please take into account that this is just my opinion. If your clients have reasonably recent hardware 1.5 to 2 years old. ( In my case a 1.5 -2 year old Mac Book Pro 2.8 GHz Intel Core 2 Duo) the client should be fairly reasonable. It will still show the issues we have seen with list layouts and specific script steps. I don't think faster hardware is going to make these disapear. I do think that FMI will resolve this relativily soon. ( Also my opinion...).
On the server side, where the new FMS design is faster and will use more memory and more CPUs, a newer multi core system will be helpful. When the pre-release came out I put FMS 12 pre-release on an older( semi retired) 3 Gbyte 2 Ghz Mac Book Pro. It worked, but it overloaded the whole system and everything was slow as dog.
When 12 was released, I purchased and set up a development server on a new Mac Mini Server with 8 GBytes of memory and 2 cores ( 4 CPUs) and it works well. Thats not to say that a top end Mac Pro wouldn't be faster, but the mini does the job for my development needs and I'm reasonbly certain would work for a small number of users.
I know there are those that will ( and in some cases,... rightly so) state that the Mac Mini server lacks the redundancy that a full blown server should have. For my development needs, it works well and I have adquate data back up to be able to recover quickly should it fail.
Thanks for sharing your experiences. I hope we can get some Windows users to respond.
I'm a big fan of the Mac mini. I have one myself, though I am not using it as a server. I have about as many Mac clients as Windows clients. But the Mac clients then to be individual users and small organizations. My larger clients are Windows based, so I do most of my development in Windows and then test the finished builds on the mini if the client is Mac-based.
I would simply urge some patience while FMI sorts out the issues with FM12. Continue with FM11 until the first patch release. I know they are aware of the issues and believe they are working to remedy the key issues as quickly as reasonably possible. I don't know when, but I'm sure it is a priority for FMI. Your just chasing your tail and wasting resources in the mean time. There are no easy solutions to this problem that won't come from FMI, IMHO.
I've got to agree with Lee's caution. Some things in 12 are done differently, including rendering of layouts, which now require interpreting CSS for the layout. This takes longer than not doing it. While faster processors and faster video-redraw will make these changes appear to happen faster, they still take some increment of time longer than 11 in the same hardware environment.
The issue comes down to user perception -- how fast do they need it to appear to be before they are happy with the performance? If you can get your hardware to make it look that fast, the problem will seem to go away. But it will still be different than 11 because 11 doesn't do what 12 does.
User perception matters. If ver 12 appears slower than 11 with the same database, on the same machine, users don't want to upgrade. If they upgrade and still find that critical things like sorting and portal rows of are too slow to get his work done, they feel cheated. I don't blame them. I feel cheated, too. I remember FMI telling us how much faster everything was going to be (not just if you have 100 people on a WAN, or if you have a machine that's twice as fast as what most of my clients have). I am not going to make that sort of pitch to my clients, because then they would question my opinons on everything (as well as my integrity). Especially, since these people are still complaining about all the problems they have been experiencing as a result of the indexing change that begin with ver 10.
I'm sorry, but I don't see any advantage to pretending that there aren't major problems with this release which have had an immediate, significant, negative impact on my business. Clients who abandon FMP for another DBMS, clients who have to spend their whole development budget for this year on hardware upgrades, etc. -- all that represents money out of my pocket. Maybe larger companies can put up with this for a while, but I am a sole proprietorship, so I am feeling it acutely and immediately.
I don't have a problem with developers who are in a similar situation sharing their experiences. Or people who think it's not so bad. I want to know the good and the bad. I don't see how pretending the problem doesn't exist helps get those problems fixed. And I don't think it's appropriate to criticize who is, after all, pointing out problems that others can and do confirm, including the negative impact it has had on attracting and retaining one's clients. Frankly, I don't have the time to keep answering complaints about this. I get it. It doesn't matter as much to you as it does to me. I just don't have the time to keep discussing it with you. I'm looking for answers that I can use right now, because I have deadlines to meet and waiting for the next bug fix or release isn't going to get a client to pay his invoice on time.
I currently have two FMP 12 jobs in progress, and I am trying to learn as much as I can about the features other developers think slow it down the most, so that I can avoid them, or at least minimize their impact in my solutions. Since some existing clients may have faster machines, I would also like to be able to give them a realistic minimum spec for Windows users. By that, I simply mean at what point is the hardware fast enough so that, when we are transitioning to ver 12 and they still have both 11 and 12 on their desktop (as one of my clients does), they are not saying "Gee, 12 is so much slower. Why did we bother to upgrade?" (which is what I am hearing now). Before I reccomend 12 to anyone else, I want to be sure they have the hardware and that the features that we think will still be slower even on faster systems will still perform at an acceptible rate.