If you came here via http://creacog.wordpress.com/ you should have been redirected to http://blog.creacog.co.uk/ which is the new and permanent domain for this blog. This is the first step in what will be a migration away from worpress.com’s free hosting to a separate server where I’ll have greater control. Feed users may like to update their book-marks to the new feed address.
I know – I’m turning into Victor Meldrew. I can’t seem to help it. Should Adobe sort out the current licensing and pricing nightmare I’m going through, I’m sure I will become more chilled.
While others have been hung up on the new icons, personally I have no issue with them. They remind me of Rowntrees Tooty Frooties – which is not a bad thing! My issues (admittedly minor) are with some of the work on the new user interface.
I can appreciate the general investment in improving the UI. However I find some of the additions odd, and on my system some seem poorly implemented?
Most of the palettes introduce a collapse control (-) and close control (x) in the top right-hand corner. Mac convention is for the close control to be top left control and in both cases I think aqua controls and positioning would have been better – more consistent with other (non Adobe) applications.
I also find it strange that most palettes introduce this new arrangement, but other ‘windows’ retain the standard Mac behaviour (e.g. document windows and the control window). Mixing it up in this way just feels odd.
When dragging a palette, it will go semi-transparent. Potentially useful – being able to see what will be covered up before dropping the palette into position. This is also part of the softening of the UI and should enrich the user experience. But on my system, when the mouse is released, the palette disappears briefly before being re-drawn opaque. This ‘flicker’ when a palette is released is plain ugly and detracts hugely from the user-experience. Until it is fixed I would have rather not bother with semi-transparent efforts.
I guess this might be an issue with my system being quite old? (Mac G4 Dual 1GHz Mirror).
I think this is just an issue of the Adobe site not having been updated to service the links. You’ll find that from the Flash Welcome window, clicking on either “Getting started ››” or “New features ››” simply dumps the user at the flash product page. The problem seems to be server-side. The pages which flash attempts to take you to, simply re-direct back to the product page.
We fare slightly better with “Resources ››”. However most of the content on that page is written for FlashMX 2004 or Flash 8. I think by now there should be some CS3 articles on there.
Hopefully that’s the end of today’s grumpiness. Although I just read the SaveAs Gotcha written up by Aral. 🙁
Video of Christian Cantrell demonstrating an Amazon S3 (which I didn’t know existed) library and client application in Apollo using the soon to exist Apollo Beta.
Not sure what he’s been drinking (cup of pink stuff) but even with the empties, his desk is far tidier than mine.
hurrah!
Not sure why they couldn’t have been ready when the product shipped? But they are there now.
Now I just need to persuade Adobe UK Licensing that my MVLP Studio 8, 24 month subscription sold to me when my DEVNET account expired does in-fact cover this upgrade to CS3 Web Standard. I’ve been playing email ping-pong with them over this for the past couple of weeks. Hopefully this will be resolved before my trial versions expire!
Ok, this may be guilt by association, but since allowing this update to install last Friday, I have experienced serious problems. Many other folks around the web have reported problems with this update, but I haven’t yet found any postings covering the issues I encountered.
In recent years, I have been of the opinion that it is best to keep up-to-date with the latest patches, as the security risk of not being patched out-weighs the risk of the patch itself causing problems. This is the first time I have had cause to doubt this approach.
This entry attempts to describe some of the problems and the efforts I made to fix them. However I should probably have done this from the outset and saved myself a lot of time. I still have some residual problems, but the worst of it seems over. Note: None of this constitutes advice. Repeat any of my actions at your own risk.
The machine concerned is a Mac Dual 1GHz G4 Mirror.
My user data, and work data are all regularly backed-up using Retrospect, and I keep all my work data on a non-system disk. So I was able to tinker around confident in the fact all my important data is protected. However I don’t currently back-up the system or applications. The theory being that they can be restored by a fresh install of both the system and applications. That said, it would take a long time!
The symptoms…
Re-installing Firefox had no effect. It still crashed at the point after rendering the Mozilla title bar in the default site.
However, of these (3) is the most serious. Apart from anything else, crashing applications degrade document data over time. My first instinct was to think this may be due to a hardware problem, specifically failing memory?
I ran Apple Hardware Test a number of times, after-which I was relatively confident that the memory was ok. No problems were reported.
Using the Console utility to view the crash reports, it was clear that the applications where crashing each with the same instruction…
Exception: EXC_BAD_INSTRUCTION (0x0002)
Code[0]: 0x00000002
Code[1]: 0x902d6690
Thread 0 Crashed:
0 ATS 0x902d6690 AnnexFileFlushUnusedAnnexEntries + 24
1 ATS 0x902d5f54 ActiveContainerTimerFired(__CFRunLoopTimer*, __CFDictionary const*) + 248
I had no idea what ATS was at this stage. But stands for Apple Type Services. There are various MacFixIt articles describing some fixes with respect to ATS cache files. So I did spend a whole load of time as Super User in Terminal, deleting / moving the cache files and restarting to see if there was any difference – there was not!
I later read that starting in Safe Mode should effectively clear the user’s ATS cache, and only activates system fonts. However even in this mode, applications continued to crash in the same way.
My next consideration was that I may have a damaged font? So I ended up paying to upgrade my Suitcase X1 to Suitcase Fusion which includes FontDoctor. (My old copy of FontDoctor simply quit immediately when trying to launch it with the system in this state.) The new FontDoctor did run and did find fault with two of the system fonts – something to do with a finder flag. These were repaired. That solved the problem with regards to FontBook crashes. The remainder of the crashes continued.
I downloaded and installed “Mac OS X 10.4.9 Combo Update (PPC)” from the Apple Downloads page. BUT! My first attempt was to do this on the machine concerned. Big mistake! The install failed, and at that point the machine would not shut-down normally nor would it fully start-up again when re-powered.
Fortunately I had an old G3 PowerBook to hand. Combined with a FireWire cable, I restarted the G4 in Target mode and re-installed the combo update onto the G4’s system drive from the PowerBook. (That’s the first time I’ve used Target mode and I’m well impressed – really useful)
This worked. I no-longer have applications crashing. Firefox now works fine. OpenOffice still failed, but once that had been re-installed, that is now working fine.
There are still some residual issues I need to track down. For instance I can no-longer connect to my Windows server through SMB. It gives error type -50 – what that means I have no idea.
I can also see that I continually get errors I suspect from my use of RSS in Safari which seems to occur for each feed I am subscribed to.
2007-04-24 12:21:29.002 SyndicationAgent[410] WARNING: BestCalendarDateFromString – can’t interpret: ‘Tue 24 Apr 2007 03:12:14 -800’
UPDATE 30/4/2007: Today I re-installed the security update. Doing so did not re-introduce the problems described above, and seems to have solved one of the residual problems: I can once more connect via SMB.
OT for this blog. However…
This being St George’s day, and unusually we bothered to make an effort. Thanks to Spamalot, in Trafalgar Square, we broke the world record for the most people playing in a coconut orchestra. At 5567, quite a few more than New York’s previous record of 1785. It would have been much higher, but we ran out of coconuts. Pythons Terry Gilliam and Terry Jones made an appearance.
Following the record attempt there was the slightly surreal (but strangely appropriate) screening of Monty Python And The Holy Grail
Back to computing stuff next.
If you read my earlier post, you’ll know I tend to think they are.
The software is great. I am happy to pay for it. I just don’t see why I should pay so much more for being British. My opinion of the reasons given: “Non Sequitur”.
El-reg are now highlighting this with respect to CS 3.
It’s just not cricket! 🙁
A question for Steve Burnard on the 17th I think.
Update (09/04/2007):
Update (19/04/2007):
Aral Balkan recently wrote an article around this subject. It and its comments are worth a read. I’m hoping that Adobe will listen and act, rather than trot out a stream of excuses.
Adobe Special – 17th April 2007
Presenters from Adobe: Mike Downey and Steve Burnard.
One not to miss!
The hotfix was published a few weeks ago. I had been suffering the debug problem on the Mac (193086), so it was time to apply the fix. (NOTE: Unless you are suffering any of the problems listed you are adivsed not to bother applying this fix)
I found the instructions more confusing than they needed to be. They essentially boil down to ‘make a backup copy of your current installation prior to applying the fix’. This is kind of vital, because if you do a straight replacement of the old “Flex SDK 2” folder with the fixed version, and you happen to use charting, you’ll find that the charts.swc is missing!
Also you need to be aware that first time you run the application after the fix, you will need to re-enter your serial numbers.
Ok, this I am sure is an old one, and has probably been blogged countless times before, but it bit me for the first time last night and googling didn’t turn up any useful results.
The scenario:
I’ve been working on a container movie. It’s job is to hold a master navigation control and to load content movies developed by a third party, based on user actions. In addition it plays a sound track.
It contains a mute button which controls a Sound object created with a line such as mySoundControl = new Sound( this ); and then mySoundControl.setVolume();
On the face of it, it controls the volume very well. Including the time-line sounds within the content movies.
The problem: We found that certain items in the content movies triggered sounds independently of the master volume and simply disobeyed the volume. The common factor in each of these was that they were button symbols each with a sound applied directly to the over frame.
The solution: quite simply was to create movie-clip symbols containing the sounds, and replace the over frame sound with an instance of the new movie-clip.