For good or bad Apple is very much opposed to maintaining backwards compatibility. Sometimes it's reflected in third party software by necessity, sometimes by practicality. I do think any commercial project worth its salt should have at least N-1 compatibility but I can understand why a one man team might make that a low priority.
There are compelling reasons for developers to require the latest OS versions; especially for access to the newest APIs.
As a developer myself I'd be interested in your opinion of what's a reasonable minimum requirement, and why. 10.6? 10.9? I assume you do not expect devs to support all the way back to 10.0.
I think there are a lot of factors that go into what makes a 'reasonable minimum requirement', and since I'm not an OSX developer, I'm not as well placed as you are to judge all of them. However, as a user, I would certainly want the following to be considered:
* Install base. Judging by this [1] and some very rough back-of-the-envelope calculations, 10.6 has ~7% of the OSX share, and 10.9 has ~13%. To exclude at least 20% of your market seems unreasonable, to me. It certainly seems unwise if you're in the business of selling a product. Although that's obviously not the whole story; that 20% almost certainly represents less-than-average potential value.
* Nature of the product. Of course, an app that absolutely must use a feature of El Capitan will have to set the requirement there. What on earth is there about a file finder that requires Yosemite, though? This is where your expertise as an app developer can add some value to my opinion: can you guess what APIs we're talking about here, and isn't it possible to program in a backwards-compatible way to avoid them if they're not available? In fact, if we were talking about the web, it would be bordering on unforgivable to fix on a specific release of a specific browser because it supports a specific feature. The Right Way (TM) to do it would be to feature-sniff and act accordingly, not reject use of the application altogether. Isn't that possible in the world of OSX development?
The indexing is super fast, even when I had it index a 4TB external drive. Nice.
Can't move the input and it always shows on the second monitor, no matter what one I am working in.
Not so sure about the fuzzy search. I searched for "tit" but it didn't prioritize where that clump of letters came in a word, ie I expected words that began with or contained only those letters to show up first. Instead, the top results where "Untitled".
Pricing does seem a bit exorbitant. Alfred, which is sort of a competitor, costs about the same but seemingly does much much more. Fuzzy stuff is great, but Alfred and, even Spotlight, do a lot more things than search.
12 comments
[ 2.1 ms ] story [ 40.6 ms ] threadhttps://translate.google.com/#sr/en/pi%C4%8Dka
As a developer myself I'd be interested in your opinion of what's a reasonable minimum requirement, and why. 10.6? 10.9? I assume you do not expect devs to support all the way back to 10.0.
* Install base. Judging by this [1] and some very rough back-of-the-envelope calculations, 10.6 has ~7% of the OSX share, and 10.9 has ~13%. To exclude at least 20% of your market seems unreasonable, to me. It certainly seems unwise if you're in the business of selling a product. Although that's obviously not the whole story; that 20% almost certainly represents less-than-average potential value.
* Nature of the product. Of course, an app that absolutely must use a feature of El Capitan will have to set the requirement there. What on earth is there about a file finder that requires Yosemite, though? This is where your expertise as an app developer can add some value to my opinion: can you guess what APIs we're talking about here, and isn't it possible to program in a backwards-compatible way to avoid them if they're not available? In fact, if we were talking about the web, it would be bordering on unforgivable to fix on a specific release of a specific browser because it supports a specific feature. The Right Way (TM) to do it would be to feature-sniff and act accordingly, not reject use of the application altogether. Isn't that possible in the world of OSX development?
[1] https://www.netmarketshare.com/operating-system-market-share...
The indexing is super fast, even when I had it index a 4TB external drive. Nice.
Can't move the input and it always shows on the second monitor, no matter what one I am working in.
Not so sure about the fuzzy search. I searched for "tit" but it didn't prioritize where that clump of letters came in a word, ie I expected words that began with or contained only those letters to show up first. Instead, the top results where "Untitled".
Pricing does seem a bit exorbitant. Alfred, which is sort of a competitor, costs about the same but seemingly does much much more. Fuzzy stuff is great, but Alfred and, even Spotlight, do a lot more things than search.
Command line stuff looks cool though!
Superficially, this is a lot like Spotlight, Alfred, LaunchBar etc. However this is laser focused on one thing - finding files. It is very fast.
I like how it displays the full file path and not just the file name.
The shell aspect ("pi cd" etc.) is brilliant.
The search results are incredibly fast.
I don't think it caches though.