I just received a rather upset (and even irate) e-mail from a user complaining about ColdFusion's lack of MacTel support.
While some users have indeed gotten ColdFusion to run on MacTel boxes, MacTel is not an officially supported platform. Why? Simply because ColdFusion MX7 was released before the first generation Intel-based Macintoshes were released (January 2006, I believe)!
Having said that, lots of customers have asked for MacTel support, and we do indeed plan on supporting this platform in "Scorpio" (which we've already said is slated for a mid-2007 release).
Of course, all plans can (and sometimes do) change. ;-)
Have you ever changed development environments? I've been using Eclipse for about two years now, and I'm still not as proficient with it as I was with CF Studio/HomeSite. Adobe has started to release UB Mac versions of some of their products, as well as some Intel-only releases (e.g. Soundbooth) on Labs. They've shown that the Intel stuff is coming, but they've integrated the platform shift as a part of their release process, not a special project.
Of course, CF support on MacTel can't come soon enough for me.... same with a fully-functional port of FlexBuilder. How are those CF Extensions coming in the Mac Beta, Ben? =)
Actually, it is. Not the work per se, but everything that goes along with it. We actually looked at possible platform enhancements for the 7.0.2 update, but could not do it. One big part of the equation is testing. For every platform we support we need to regression test all of CFML times every supported JVM and J2EE server times every supported HTTP server times every supported database driver and on and on. There is a dollar amount associated with fully testing each platform, and it is significant. Which is why we add platforms when there is demand, and why we drop platforms when demand wanes. It's not just about whether or not the code compiles and runs. And that's just testing. Then there are the legalities associated with revenue recognition that dictate what can be included in non-paid-for updates and what can't. And there is more, too. We really did want to make platform changes in 7.0.2, but that was just not possible. So, for now, any platform changes are going to wait for Scorpio.
--- Ben
I'm glad you don't just rush into builds, it's important that each product has significant testing, the same users would also complain when it didn't support X.
As for dropping versions, that's also a good thing, sometimes hard to do.
I know you dont quote sales figures, but can you show us some figures or a pie graph of sales % across platforms.
I'm prety sure AIX must be getting close to being dropped.
--- Ben
And we understand that Adobe is a major corporation. No one expects them to go to the time and trouble of actually supporting its user base... and especially without getting paid to do so.
But thanks for the info. It's nice to have some, even if it did have to come from "outside" channels.
--- Ben
Wait another 6 months for something BlueDragon is about to release. Are you serious, you are waiting 6 months for version 8 and in 6 months BlueDragon will be another year or more behind.
They still don't have a lot of the good features of version 7.
What about the second part to my question, ie % of sales per platform ie
80% Windows
10% Linux
5% AIX
5% Mac
Might help people wanting to move platforms pick something popular and thus have long term support from Adobe.
--- Ben
As to being naive, when MM made the switch to a Java-based platform, the two most often stated reasons for doing so were: 1) JIT-compiled perfomance; and 2) the fact that CF could now be easily ported and made to run on any platform which had a working Java/J2EE implementation. (I could find the MX6 press releases enumerating these reasons, should I need to do so.)
So either statement #2 is right, or it's wrong. If it's the first case, then MM/A is now making a conscious decision not to provide support to an active platform (OSX), and the users who depend on that product on that platform. If it's wrong, then MM essentially, and for lack of a better word, lied in order to make its multi-platform support seem better than it was.
And the fact that an outside group of people, going through an admittedly complex process, have managed to get it to work doesn't help Adobe's case. Nor did the fact that mySQL AB's OSS community, for example, had a complete, working release of its database engine one week after Apple's new products shipped. In fact, the timeline is even worse than you suggest, since Apple had announced the conversion, and had test machines available, what, at least six months prior to shippment?
My mood isn't helped by the fact that I recommended CF a couple of years ago (I've been doing this since CF3) to a client, who's now upset over the fact that they have four MacTel XServes on order and due in November... with no software to run on them, and apparently none due until Adobe can ring the version 8 cash register eight months from now.
So again, forgive me if my comments appear "sarcastic and even antagonistic", but unhappy customers listing the reasons for their unhappiness often come off that way.
According to my correspondence with New Atlanta, Blue Dragon 7 will support most of the CFMX7 missing features, and it appears that BD will be available at least six months or so before CFMX8. And six months is a long time to sit on empty servers.
Not only that, but things like client events in application.cfc, threading, and so on promise to make things a lot easier for the types of applications we do. Especially as we have little interest in the Flex/Flash-based user interfaces that Adobe seems to be promoting so heavily.
Ya know Michael I have an Ipod & I have a lexus yet my lexus doesn't have support for my ipod so I guess they need to drop everything and add support cause me and my 3 friends want it.... Damn the nerve of them!! I mean if I want Ipod support I gotta shell out another 60 grand to get it, what an atrocity!! ;)~
I wish you'd stop whining about this , most of us would rather have a completely tested and supported version instead of a rush job and its not like its hard to get working now.
Personally, I say go with BD and you and Vince can sit around and whine & cry together. Or go to .net and see how long you will wait for os x support.
nuff said but stop your bitching about the cash register, damn.
The only point you have made on any of these posts is that you are cheap ( i bet that gets you all the chix!!)
then cfmx8 comes out a few months later and then BD will be back to being way behind again, which they wont catch up to for another year and a half, good plan....
I wish they had a standard install for intel too but they dont so get over it. Its not like you cant install it on the servers unless you are a retard, hell even i did.
It could be fair comment to say that MM/Adobe were caught cold by the whole MacTel move to XCode programming, which they wouldn't previously have been prepared for. Apple did rush into this move fairly quickly, and large outfits like Adobe can't necessarily move to market overnight. At the same time though, if it's being planned for Scorpio then it might have been a nice move to bring forward some degree of support for MacTel in the latest 7.x release as a gesture to the Mac community.
However as a happy BD 6.2 user, I can't honestly see all the fuss about the things I'm supposedly missing from CF 7. I actually ported from CF 6.1 to BD 6.2 with no problems - and there's nothing in CF 7 that would entice me back at the moment. The upgrade to BD 7 is a good one, with features Michael mentioned, and none of the Flash/Flex related clutter that I've got no interest in at the moment.
Everyone has a different need, it's just that BD (current and future versions) covers everything I'm likely to need.
So, in effect, when ColdFusion is a Java application therefore should be an easy port from a code standpoint, why was the decision made to *not* support their x86 migration when it was known so far in advance?
FYI, this is one more reason why many companies stick to open source technologies.
And yes, I'm a little beligerant over this because Adobe continues to hold double standards over their definition of what is supported, and with my employer aiming at updating our equipment I'm having a harder time rationalizing the expense for new licenses.
Damien
--- Ben
Here are a few suggestions (serious ones):
1) go with BD, work with New Atlanta to 'beta' their apps, get free/reduced support in exchange for testing with them, smile
2) go with CF, wait 6 months for a support release, smile a little less (waiting)
3) go with CF, make Adobe happy that you are sticking with them for CF versus BD. Commit (purchase order) to upgrade the 4 boxes to CF 8 in 2007 and then work with Ben and their development staff to do alpha/beta testing on the new hardware/OS. With 4 boxes, if your hosting needs aren't too heavy you may be able to dedicate 1-2 boxes for production, 1 for dev and 1 for testing at this time. Adobe might be very eager to both retain a customer on the MacTel side and to garner some excited testers on that platform who are committed to truly testing.
I recommend you go with #3. This 'negative' could be changed into a win-win for both sides.
Ryan, please reread the above, as you seemed to have missed a few points. Michael is frustrated that his clients (not employers) are being forced to leave a platform they've been totally committed to for years(CF), because Adobe/MM essentially abandoned support in midstream for a currently supported os (OSX).
Further, until Ben's post there's been no official word (that I could find) about when, or even if, the MacTel issue would be resolved. And now we find that Adobe will not issue a fix until they get around to releasing MX8. Charging for new features is one thing. Making one pay, and wait, for a fix for busted software is quite another.
And even more irritating is that it's apparantly not the major deal that a total xcode/Adobe Photoshop rewrite would be (which at least still runs under Rosetta), but just a matter of getting the installer and Apache connector recompiled so their Java-based write-once-run-anywhere software runs once more.
Finally, I get the extremely strong impression that Adobe wants CF to become the primary and preferred backend for Flash/Flex clients, and that more and more of their developmental resources are going to be focused in that direction.
The testing thing might be an option, but I don't have much hope for it. As CF has moved up the corporate food chain from Allaire to MM to Adobe, the lines of communication have steadily diminished...
Ben, I'll agree that this is true. But I think you also realize that the number of people who DEVELOP on Macs under OS X is quite high (check Dreamweaver sales), and that developers often need to be able to run test environments and servers on their own computers. Further, since the developer version one would install on such a Mac is free, platform sales probably don't accurately reflect platform use.
I, for one, have a Powerbook I would have upgraded to a MacBook Pro months ago, save for the fact that I'd be unable to do any development on it whatsoever. And I'm on the road and out and about too much to be unable to do that.
As per the official list of supported platforms (http://www.adobe.com/products/coldfusion/productin...) OSX is just as supported as it has been since we first announced support. I fail to see where anything was abandoned. Unless you are saying that by not adding support for a platform that shipped *after* CFMX7 shipped somehow means we've abandoned OSX. If so, well, that's about as sensible as saying we've abandoned Windows because we have yet to release support for Vista which was released last week.
>> until Ben's post there's been no official word ... about when, or even if, the MacTel issue would be resolved
True, I agree with that. We need to do a much beter job communicating future plans, no argument there.
>> Charging for new features is one thing. Making one pay, and wait, for a fix for busted software is quite another.
Busted software? Not supporting MacTel yet makes CF busted?!?!?
>> I get the extremely strong impression that Adobe wants CF to become the primary and preferred backend for Flash/Flex clients,
Flex support is important and strategic. But no, the bulk of what we are working on for Scorpio has nothing to do with Flex.
--- Ben
I managed to get CF run on my MacTel laptop.
As far as I know, CF installs fine on MacTell, but JRun won't start cfusion instance if you're using Web Server Configuration tool.
What I did to start the server was the following:
In the jrun/lib directory, try this command:
java -jar ./jrun.jar -start cfusion
That bypasses JRun's JVM selection and uses the native default.
Hope this helps.
Cheers.
For someone NOT to buy a macbook pro because of that is pretty pathetic, I have already told him like a zillion times that it runs just fine on mactel but he dont care cause he just wants to bitch about something.
fyi~ to install it all you gotta do is download the osx version (hes already bitching at this point cause Adobe wont deliver it via naked playmates on a cold day ), then during install choose "multi server" then download Jared cfmx starter http://www.web-relevant.com/uploads/CFMXLauncher.d..., then put it somewhere safe and drag it to startup items and thats it.
Now if he complains about it one more time im gunna go slap him myself!! haha