Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Wellander

macrumors regular
Original poster
Mar 24, 2006
157
0
Huntington Beach Ca
Hi,
Do you all think that software makers will eventually just do intel builds?
To save development time?
To save server space?
Any ideas?
Thanks.
 
Wellander said:
Hi,
Do you all think that software makers will eventually just do intel builds?
To save development time?
To save server space?
Any ideas?
Thanks.

I was under the impression that they just do Universal Builds and it works for whatever is thrown their way. There are currently no "Intel" builds for anything.
 
i don't think this will happen for a while and has been asked a lot.
the main reason being is that most developers for Macs know how long these computers last and don't have as quick of a turnaround rate as Windows PCs. so i think to get as many users as possible Universal apps will be around for a few years. i'd say four to five years from now there will be fewer and thats when we really will start to see a shift away from Universal apps.
 
I would have thought that there will be a point where that happens, yes. In the same way that developers don't produce pre-OS X versions of their software anymore, there will come a time when support for PPC Macs dries up too. Given the size of the PPC userbase, I wouldn't anticipate that happening for quite a while yet though.
 
Sooner or later, apps will be just Intel, but that is YEARS from now. Just like the transition to PPC, the motorola chips slowly faded out and all the Apps became PPC only apps. I'd say it will be about under a decade until all the Apps are only Intel. Universal is cool and all, but I imagine it being hard to compile something for PPC and Intel, twice the work..
 
Eventually of course. Not for a long time, but much like software manufacturers fade out OS's. For example, in the windows world, no one is still making win95 compatible software :p
 
Bear in mind that new apps and new updates may require hardware updates beyond the binary one.

More RAM available, more HD space, better graphics performance, more powerful processor may be needed just to run the updated applications - at that point, the Intel-only build fear might be moot.
 
I had the same concern. Glad to find some feedback.

I was bummed when my '99 bondi G3 couldn't handle the 10.3 upgrade in '04... made me get a new mini which was really a GOOD thing!
I'll want a G5 w/ ppc in a while (still sceptical about intel virus problems) and I hope it's updatable for at least a decade!
 
I guess it depends on how much 'extra' effort is needed to make the application universal.
There is such a large user base on PPC at present that will remain for a long time to come. As many people have said before me, it will come a time where developers feel the user base of PPC is not large enough.

I guess what you have to remember is PPC and intel are not OS's that have been updated, they are physical hardware that will remain as it is for a very long time.
 
Most likely after 10.8 or something, everything will be Intel only - or at least the 'big' apps will be.

PPC-builds will stop within a few months of Apple stopping PPC support. So thats from 10.8 to OS 11 (from various forums and webbies I've read)
 
The making of universal code does not cost extra effort with new apps under Xcode because, you only have to select it as an option and the compiler will take care of it. So PPC users: Do not worry. :)
 
matthew24 said:
The making of universal code does not cost extra effort with new apps under Xcode because, you only have to select it as an option and the compiler will take care of it. So PPC users: Do not worry. :)
Right, so as long as Apple continues to support Intel and PPC with Xcode, most new software will be universal.

There's still types of software that may be Intel only. Software with much hand optimisation or uses libraries with much hand optimisation or software that's ported from Windows may contain parts in pure assembly and that's a lot harder to make universal.
 
matthew24 said:
The making of universal code does not cost extra effort with new apps under Xcode because, you only have to select it as an option and the compiler will take care of it. So PPC users: Do not worry. :)

You still have to QA and support the port. It costs a lot of money.
 
Absolutely, but not for 4-5 years. It will probably be 10 years before most software is Intel only. In other words, PPC machines will be good for a long time to come.
 
matthew24 said:
The making of universal code does not cost extra effort with new apps under Xcode because, you only have to select it as an option and the compiler will take care of it. So PPC users: Do not worry. :)

I know apple makes this claim, but I question the truth of it to some extent. Why does it take so long for some programs to go to Intel. If it was really just as simple as sticking the code in the Xcode Compiler, why would Adobe and Microsoft wait until a new version of Photoshop and Office before releasing Universal Binaries for the programs. I remember a VP or someone else of high rank at Adobe saying that making UB's for their code wasn't as easy as Apple makes it seem.

I'm not necessarily saying that it's not as easy as Apple says, but I can't seem to see a reason that it would take this much time for a UB if it was as simple as sticking it in a different compiler.
 
gco212 said:
I know apple makes this claim, but I question the truth of it to some extent. Why does it take so long for some programs to go to Intel. If it was really just as simple as sticking the code in the Xcode Compiler, why would Adobe and Microsoft wait until a new version of Photoshop and Office before releasing Universal Binaries for the programs. I remember a VP or someone else of high rank at Adobe saying that making UB's for their code wasn't as easy as Apple makes it seem.

I'm not necessarily saying that it's not as easy as Apple says, but I can't seem to see a reason that it would take this much time for a UB if it was as simple as sticking it in a different compiler.

Well if they are creating a new product anyway, it would probably be more cost effective for them to wait.

I think it is easy especially for people who write these apps to begin with, but it doesn't make sense to redo say CS2 and turn right around and release CS3.

If Rosetta wasn't there, I think they maybe would have done this. Since people can run their software though, they probably just decided to wait til they finish the new versions.

Anyhow, we will have UB for a long time.
 
Yeah obviously there has to be a catch, it just can't be that easy making Universal Binaries.

I believe there will still be support for PPC but most programs will shift to Intel, I doubt we'll see anymore PPC games for example, they are just easier to port and run on an Intel, many apps will remain binary but I bet we'll see a few Intel only apps, its inevitable.

PPC is still good for some time but it will slowly fade away, my G4 DA tower has lasted quite some time since it was bought and I still use it daily and constantly and will keep it indefinitely just for the Mac OS 9 factor, you know for nostalgia purposes and I'll use the Mac Pro for highend. PPC will live a LONG time but by live I don't mean every app will be available but there will still be several software made for it.
 
Chone said:
Yeah obviously there has to be a catch, it just can't be that easy making Universal Binaries.
It really does depend. For those developers who have old code bases that aren't on xcode (read Adobe, Microsoft et al) then releasing a universal is far from a checking a box then recompile effort. They have to port their code to entirely new development environment. One that uses a completely different compiler which is also a lot more strict about the way you use the language (this is no-ones fault - gcc is just a lot more picky). Some parts of their code may use hand written assembly language that is processor specific which has to be completely rewritten for intel. The intel processors are VERY different at a low level to the PPC processors and rewritting assembly language for the new processor can mean also doing it in a completely different way (for example using the stack instead of registers to pass parameters).

Releasing a universal binary is only 'easy' for those developers that use Cocoa on xcode - even then there's QA and testing that needs to be done. I l know in the case of my own app it was pretty straightforward and it all worked, but that isn't the case for everyone.
 
I'm also wondering though... is the result from a Universal Binary just as good as making the app Intel Native only from the ground up?
 
eventually? yes. i would say in 10 years, there won't be need for ub. intel is enough.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.