Volitans Software

SMART Utility and Lion

I am happy to report that SMART Utility is for the most part compatible with Lion. All functions seem to work, except the menu extra. That was to be expected, as the way it was implemented breaks with every OS version. However, thanks to work of bjango (makers of the excellent iStat Menus, which I use) to make MenuCracker Lion-compatible,  SMART Utility and its menu extra will be fully compatible with Lion.

Version 3.0.2 will use the new MenuCracker, and that should be due out before Lion is released. On the horizon after that, is version 3.1, which will bring more user-requested features, such as email notification. That version will be out in a few months.

Apple Reverses Tough Subscription Requirements

Apple has finally relented and changed their policies on in-app subscriptions. Back in February, Apple announced in-app subscriptions, but along with the new API, the added some restrictions that could have forced some developers to leave the iOS App Store (such as Amazon’s Kindle). The two most pernicious ones were the price must be the same between in-app and out of app subscriptions (ie developers couldn’t charge a higher price to cover Apple’s cut), and that to offer the content developers must use the in-app purchase system (instead of not using it at all and charging subscribers on their website.

Apple has modified those restrictions. They no longer require the prices be the same, and they no longer require using the in-app subscription system. So developers can charge a higher price for the subscriptions. They also have the option of signing up subscribers on their website, and using that payment to send content to their iOS app. However, Apple will not let them have a BUY button in their app that takes them to the their website. I think that’s a fair compromise.

Even though I don’t have an app that is covered by the policies, I care about the iOS ecosystem. I found the subscription policies detrimental to that ecosystem, and I am glad Apple realized that as well. Another moral of this story is that complaining loudly, in the press, on blogs & websites, and to Apple can make them change their mind. In future, I hope we’ll here less of “They can do what they want to, so stop complaining because they’ll never change their mind.” That’s a good lesson to learn.

Now Serving Amazon S3 and CloudFront

At some point, and I am not sure when, the bandwidth from my hosting provider has gone down the tubes. Downloads are now seeing speeds of 150KBps, which is entirely unacceptable. I’ve tried contacting them, but they didn’t seem to understand my problem.

So that gave me an opportunity to explore content delivery networks (CDNs). My first thought was Amazon S3, since A) I’ve heard people talking about it and B) I love Amazon, and I buy so much from them (especially Amazon Prime). I took a look at their offerings, and found they also offer CloudFront.

I was confused at the difference between S3 and CloudFront, and after a little investigation, it became clear. S3 is great for hosting files, but there is a speed hit depending where in the US, and where in the world a user id downloading files. That’s where CloudFront comes it. Amazon has servers in specific areas that can speed up access times, especially for customers in other countries. Since 30% of my customers are foreign, that was a great solution for me. And since CloudFront requires and uses S3, I could switch back and forth at any time.

I did investigate some other alternatives, such as CacheFly and Limelight- but they didn’t meet the price and ease of use of Amazon’s services, especially for a small company like mine.

So I’ve switched all downloads over to S3 on both the website, and the updater in SMART Utility. I’m now seeing speeds of 1MBps and more, which is great. In the next day or so, I’ll switch over to CloudFront. There may be some hiccups as the DNS propagates, but in the end it will be worth it. The icing on the cake was just how easy it was to switch.

If you want more information, just go to Amazon’s AWS page here.

SMART Utility 3.0.1 is Out!

This is mainly a bug fix release, and all users are recommended to upgrade. The best way to update is to launch the app, choose the SMART Utility menu, then choose Check for Updates. Once it finds the update, click on Install. Download links are still provided below.

Download for 10.5 (Leopard) and 10.6 (Snow Leopard).

Download for 10.4 (Tiger) here.

Here’s the full release notes:

  • Fixed a serious bug that would cause a crash when rescanning drives
  • Fixed a bug that occurs when launching from the command line
  • Fixed a bug that caused inaccurate drive capacity display
  • Fixed a bug that caused the menu extra icon to not be masked when clicked on
  • Fixed a bug that caused the menu extra to show unsupported drives
  • Fixed a bug that cause the menu extra to pass the wrong icon to Growl for failing and failed drives
  • Updated Growl framework to 1.2.2, which fixes some bugs

 

SMART Utility 3.0 is Out!

Finally, after many long nights, I’m excited to announce that SMART Utility 3.0 has been released. There are many new features and bug fixes in this release. This includes a menu icon, so the application does not need to be running, as well as some interface enhancements.

Download for 10.5 (Leopard) and 10.6 (Snow Leopard).

Download for 10.4 (Tiger) here.

If you aren’t sure which version you have, here is how you can tell. Leopard and Snow Leopard have a black magnifying glass in the upper right hand corner. Tiger has a blue one.

Here are the full release notes:

  • Added menu extra, which uses app’s preferences to operate, including periodic checks and Growl notifications
  • Added last time updated to main window
  • Fixed a Growl notification misname for failed drives
  • Fixed a bug where the estimated time at the beginning of tests would be reported wrong
  • Fixed a bug where the info windows attached to a drive report would not close after closing the drive report
  • Fixed a bug where the Tests window would not close after scanning for new drives
  • Fixed a few bugs where setting preferences to only alert on new changes still caused some attributes and tests to be highlighted as failing
  • Fixed a bug where More Info button did not have a space between the words
  • Fixed a bug that would cause a harmless error message to pop up
  • Fixed a bug where test information in main window would not clear out when switching to an unsupported drive
  • Fixed a few memory leaks
  • Forced running in 32 bit on 10.5 due to OS issues
  • Now estimated time remaining in tests is a little more accurate at beginning of test (no more Calculating)
  • Now capacity shows in gigabytes
  • Now if there is only one partition, that will be displayed instead of /dev/diskX
  • Reordered general preferences for better organization and readability
  • Updated help to add information about tests
  • Updated smartctl engine to 5.40, which adds better support for SSDs as well as bug fixes

Enjoy! Please report any bugs to the Support email.

SMART Utility: Future of 10.4 Tiger Support

With the release of the first beta of Mac OS X 10.7 Lion, I have made a decision regarding the future of 10.4 Tiger support in SMART Utility. Version 3.0 will support 10.4 when it is release in the next week. Minor updates (3.0.x releases) will also support 10.4. However, version 3.1 may not support 10.4, and 3.2 will not support 10.4.

This is a difficult decision, and its one I put off for a long time. I know there are still a fair number of 10.4 Macs out there, and those are the ones ripe for using SMART Utility. I even still get requests for 10.3 support from time to time! But here is some perspective: It has been nearly 6 years since Tiger was released, and nearly 4 since the last major update came out for it.

So there are a couple of reasons why I am deprecating support for 10.4. The biggest reason is development time. If I continued to support 10.4, that would mean I would have to test 8 different configurations of OS/hardware support (10.4 PPC & Intel, 10.5 PPC, Intel 32 bit & Intel 64 bit, 10.6 Intel 32 & 64 bit, and 10.7 64 bit). I also have to maintain a separate build for 10.4 Macs, as that version has bugs when run on 10.5(such as progress bars not working properly). This adds up to a lot of work, especially as many 10.4 APIs are deprecated themselves on later OS’.

Next is that Apple is dropping support for 10.4 (and possible 10.5) in XCode 4, the next major release of the development tools. This means I would have to maintain a separate install of XCode when I upgrade to 10.7, which on top of the above reason equals a lot of management overhead- time I could be spending developing new features and fixing bugs. There are also a fair amount of features I would like to add, but that require 10.5 at least. Dropping 10.4 support would allow me to focus on them.

Finally, my anonymous system stats tell me me that about 15% or less of my users are running 10.4. This follows other stats I have seen- such as OmniGroup’s Stats page. In fact, 10.6 is run by the majority of users. An upgrade to 10.6 for Intel users is only $30, and I heavily encourage users to go that route. For PPC users, an upgrade to 10.5 can be had used for about $50. Both offer a lot of benefits over 10.4.

I hope that this will allow an informed decision by my current and future customers. Feel free to contact me with any questions, comments, or concerns at [email protected].

Unacceptable: New App Store Subscription Rules

Apple announced new subscription rules for apps that deal with content downloads (books, magazines, and newspapers). They require apps that want to offer in-app purchases to use Apple’s system, which has a cut of 30%, regardless if there is a website that the content can be purchased. And not only that, they require the prices be the same in both areas. I find this quite egregious. The payment required by Apple would heavily cut into publishers profits- maybe making it completely unprofitable. Those publishers then would have no incentive to allow in-app purchases on iOS devices. That is bad for users. Yes, Apple’s UI for purchases is great, but if there isn’t anything to buy, what’s the point. Marco makes some similar points in his article.

I see two ways this can be fixed. First, I agree with The Outsider, allow apps to charge more on iOS devices. Sorry, but if Apple wants its cut, that money has to come from somewhere, and that has to be the user. This is no different than buying Coke’s at DisneyLand- its more expensive because of Disney.

A second way would be to change the structure and prices. Normal in-app purchase for content that is ONLY to be used on iOS devices (ie think new levels for a game) is 30%. Apps that allow content to be used on multiple non-IOS devices (such as magazine subscriptions and e-book), MUST charge $10 or more for the app, but in-app purchases are only 5-10%. Remember- since Apple controls the store, they can enforce these rules.

I much prefer the second option, but either will do to solve this mess.

SMART Utility Mentioned on Hypercritical

SMART Utility was mentioned in episode 5 of Hypercritical, and excellent podcast about Apple and related technologies with Dan Benjamin and John Siracusa. Its one of many great podcasts on Dan’s network, 5by5. I’ve been a fan of Dan’s since Gruber and his podcast The Talk Show. I used to listen to The Conversation until he canceled it, but then he started a whole slew of new podcasts, of which Hypercritical and Boom, Outta Here are two of my new favorites. Of course, I’ve followed John since early 2001 on Ars Technica, where he had some great reviews of every OS X version.

Anyway, enough butt kissing. I thought John and Dan did an excellent job of describing SMART and SMART Utility (except for calling it creepy! Haha!). However, I did want to followup with some more information about what they said.

John mentioned the algorithm SMART Utility uses, which calls a drive FAILING before SMART itself will say FAILED. He was correct in that its a judgement call, and it can be a little confusing in that regard. When SMART Utility finds something that could potential cause a drive to fail in the future, it does label it FAILING. Now, it could be failing now- ie soon the drive will fail completely. It could also be overly cautious and the drive will be fine. The algorithm was designed this way because when dealing with data, its better to be safe than sorry. So false positives are much better than false negatives.

This is also based on 10 years of dealing with hard drives, as well Google’s study(PDF) on hard drives (as John did mention). However there are preferences to “correct” the algorithm (as John also mentioned). I’m always verifying the algorithm is correct, and if improvements are needed, they will be made. For instance, version 3.1 will change the way errors that occurred many power on hours ago are dealt with.

Dan and John made comments that its not for your mom, and though I will sort of agree with that, I still believe they can use it. For instance, they can understand green, yellow and red. Also, the menu icon that is coming in version 3.0 will be great for them. Of course, its best if they have a family member or friend who knows a little more to help them out, or they can always contact me.

I hope that clears up any confusion, and thanks to John and Dan for the discussion. Now go listen to their podcasts!

Twitterrific 4: Excitement followed by disappointment

Twitterrific 4 for Mac was released on Wednesday. I was excited to download it, because version 3 was getting long in the tooth. For example, it didn’t support retweeting, nor URL shortening. To my dismay, there are two glaring issues with the application. And that has forced me back to version 3. And if that all sounds familiar, it should. I wrote very similar things about Tweetie for Mac back in 2009. You would like Iconfactory would learn from atebits mistake. And on top of that, the developer seemed incredulous that I would ask for the changes.

My first complaint is that to send a tweet, you now have to press option-return or click send. That’s a really bad UI. I want to type, then hit send. The developer tweeted back that its too easy to send an accidental tweet. First, how often does that happen? Second, who cares- let the users figure that out, because that’s what the delete function is for. Third, you can use the enter key, but laptops don’t have that. It makes usability much harder. His response to that was how would a carriage return be input? But, who does that? The tweets I look at almost never have carriage returns. And since that is a lot rarer than just sending a tweet, the options should be reversed. Its a stupid UI decision.

And for some reason, they left out the very useful Growl. The developer has said that if enough people request it, they will bring it back. It looks to me like one of the most requested features on their Twitter account is Growl support, so they better add. Still it is crazy that they took it out in the first place. Growl is exceedingly easy to add to applications- it took me about 15 minutes to add Growl support to SMART Utility. However they also have said they aren’t convinced of its usefulness. Again, that’s crazy. I don’t want to open up the app every time a new tweet comes it. Just show me it briefly.

While writing this review, I found a two more things that are infuriating about Twitterrific 4. When the main window is closed, the app stops updating What kind of decision was that? For version 3, it just hid the window when it was closed, but it still updated in the background. Now, for the same functionality, it has be hid manually. For basic users, hiding and closing should be the same thing. Also, the Dock icon is always there, when it wasn’t in version 3. Okay, so that’s just annoying, but still, another bad UI choice.

So I really hope Iconfactory fixes these bugs in the next version. I really want to support development, but I can with these glaring problems.