OS X Lion, Open CL and Logos 4

I should add, it's easy to armchair engineer, without the pressures of actually running a software projects. ;-)

Yeah, I just wouldn't be up for leaving Lincoln. ;-)

I like it here and really enjoy my current job, so I'm not really looking for a change. Now if I were able to work from here or in the region and a company approached me, I would certainly entertain such an opportunity.

Daniel Bergquist said:

Unless I am mistaken, the UI is native to OS X and the under the hood stuff uses the cross-platform Mono framework.

An app's UI code can be easily made sluggish either through drawing complex custom UI elements (or even simple ones) too often, creating and destroying too many objects while working with the UI, refreshing content too often or through tying up the main thread with non UI code. There's probably other ways too, but these are the main causes.

I am not familiar with Mono's limitations or even if there is a 64bit OS X version of the framework. If the limitations you mention are present, there's ways around those too.

L4M is, I think it is fairly broadly accepted, held back/hamstrung by Mono. If one looks at an app like L4M (& L4W) there would be a lot of heavy lifting going on with computation over what to display (indexed searches, cross reference of book titles/words, text analysis) before it gets to the how to display (the UI). I believe the 'what' depends on Mono which is overdue for revision, and is itself still based on a couple of versions old .NET.  Strongly doubt is 64 bit. Unfortunately don't believe the extra horsepower capabilities in OS X (Grand Central) etc. can be used in L4M back-end as the 'engine' is the closed box Mono/.NET running DLLs - no doubt impervious to improvement. 

The L4M UI uses some non-standard UI components (?and programming methods?) to make it look and act close to the Windows version. On the UI side there is no reason that once content (data) is available that the screen can't be drawn very quickly. Obviously many other 'big' OS X apps can do it.

I'm sure the L4M team are doing all they can to improve things, you sure you don't want to move to Bellingham [;)]

Larry Good said:

Clearly you can do conditional code depending on the os version at runtime and so on.

Would this be a possible path for Logos 4?

While technically feasible, it is not an attractive option, as it creates significantly increases the burden on testing, etc.

Chris Roberts said:

alabama24 said:

At $30, hopefully all 7% will upgrade this month!

Except to get Lion you have to have Snow Leopard first

And there are hardware reasons why some are not able to upgrade.  Lion requires a minimum of Core2 duo.  If they didn't upgrade to Snow Leopard, why would they upgrade to Lion?

David Mitchell said:

Patrick S. said:

L4M is, I think it is fairly broadly accepted, held back/hamstrung by Mono.

You are the only person I know who routinely makes this assertion. As I've indicated before, your concerns have little basis in fact. Please either stop making these claims or send us a résumé so that you can have the opportunity to create an informed opinion.

David — when I say L4M is held back by Mono I mean, and I believe you know I mean, that Mono/.NET is definitely not doing L4M any favours and is only there because of Logos 4 Windows. That I believe is a fact, if you can show me any way that Mono/.NET is benefitting L4M (performance wise as that is the basis of this, and many threads on the forum) I'll be the first to shut up. I have also said many times - and you know this - that Logos has explained its reason for using Mono, it was Logos'  business decision (which is its to make) and I sincerely hope that it does not long term negatively affect the Mac product that, so far, I have spent US $4,502.41 on.

And as for the Mono project itself — well I just hope life support keeps going long enough for it to recover and stay viable. And that's a fact.

David Mitchell said:

Patrick S. said:

I believe the 'what' depends on Mono which is overdue for revision, and is itself still based on a couple of versions old .NET.

The version of .net we use has more to do with issues in WPF 4 (on Windows) than it does on Mono.

Oh well, at least we can console ourselves that the Windows users are suffering also [;)]. Which doesn't change the fact that the version of .NET in Mono is old — and not 64 bit.

David Mitchell said:

Patrick S. said:

Strongly doubt is 64 bit.

This is something I'm working on. It is likely that the greatest amount of effort will be spent in our resource display code, which is (and always has been) native code.

Great.

David Mitchell said:

Patrick S. said:

Unfortunately don't believe the extra horsepower capabilities in OS X (Grand Central) etc. can be used in L4M back-end as the 'engine' is the closed box Mono/.NET running DLLs - no doubt impervious to improvement.

This is untrue. It is quite simple to use Grand Central Dispatch from Logos 4. The primary reason that we have not done so is that we still support Leopard, which does not have Grand Central Dispatch. Approximately 7% of our active Mac users are on Leopard. When a considerable number of them have upgraded, we'll consider dropping support and using newer technologies.

But you're not trying to say that Grand Central can be used with/interfaced with Mono/.NET I think. So even if Grand Central was used with L4M overall the Mono/.NET black box would not improve in performance. Which was the point I was making. If I am incorrect in that statement - great, look forward to a significant increase in performance in L4M.

Isn't it possible (I don't know directly) to code/compile separate versions of Logos for Snow Leopard & Leopard? Seems a shame to hold back 93% of users for 7%. I think if Logos polled all the Mac users with a question like, for example "would you upgrade to Snow Leopard if you could enjoy a 15-20% performance increase in L4M?" I believe we can guess the response.

David Mitchell said:

Patrick S. said:

On the UI side there is no reason that once content (data) is available that the screen can't be drawn very quickly.

In theory, this is true. In practice, the situation is complex. Send us a résumé, and I'll tell you all about it .

I'll take your word for it, I have no intention of ascending the calvary of being a Mac developer of L4M — all those pesky users bugging one [:S]

David Mitchell said:

Daniel Bergquist said:

Yeah, I just wouldn't be up for leaving Lincoln. ;-)

Probably because you haven't yet been to Bellingham…

Just sayin'

I'm sure it's a great place, maybe I'll visit someday, I love traveling! As far as living... I'm just a Nebraska area guy, family, friends and pretty much my entire existence has been here. ;-)

p.p1 {margin: 0.0px 0.0px 12.0px 0.0px; font: 12.0px Arial}
p.p2 {margin: 0.0px 0.0px 12.0px 0.0px; font: 12.0px Arial; color: #0040ee}
p.p3 {margin: 0.0px 0.0px 0.0px 0.0px; font: 12.0px Arial; min-height: 14.0px}
span.s1 {color: #000000}
span.s2 {text-decoration: underline}

David Mitchell said:

Patrick S. said:

But you're not trying to say that Grand Central can be used with/interfaced with Mono/.NET I think.

Here is the code for using Grand Central Dispatch from C#: https://github.com/mono/maccore/blob/master/src/CoreFoundation/Dispatch.cs

 

Now that's cool. Even if GCD from C# didn't exist, the app could be split into multiple processes to work around any mono CPU/RAM constraints. Mathematica once did that when UNIX processes could do 64-bit but Cocoa processes could not. Chrome does it for other reasons. There are other reasons why splitting up into distinct processes would be a good idea, but I am not at liberty to discuss them on a public forum as I am unsure if they have been publicly announced.

But again this is armchair engineering. It assumes that the various bits of code are well encapsulated so they can be separated. Without access to the code, there's no telling how well a move to GCD or multiple processes would progress. Never mind other items might take priority, such as the aforementioned 10.5 compatibility.

alabama24 said:

David Mitchell said:

Approximately 7% of our active Mac users are on Leopard. When a considerable number of them have upgraded, we'll consider dropping support and using newer technologies.

At $30, hopefully all 7% will upgrade this month!

At $30 it might save Logos money if they pay them to upgrade so they can drop legacy support that is apparently holding them back.

Kevin A. Purcell said:

alabama24 said:

David Mitchell said:

Approximately 7% of our active Mac users are on Leopard. When a considerable number of them have upgraded, we'll consider dropping support and using newer technologies.

At $30, hopefully all 7% will upgrade this month!

At $30 it might save Logos money if they pay them to upgrade so they can drop legacy support that is apparently holding them back.

Don't think I haven't considered it! [:D]

Fr. Charles R. Matheny said:

Hey David: This was on another thread but, you may not have seen it.

When I called support and we screen shared and all of that, one of my issues was that when using my outboard monitor/monitors several of the "buttons" cease to be functional.

Tech support tried several things and said this was a " currently unknown" bug and would report it as they were ( i guess) able to duplicate it. 

Thus, I am unable to use Logos on larger screens when on my Macbook.

The question is: Is the a "workaround" available, something I can do on my end?

Not trying to be a pest, just trying to give old eyes some room to work-smile.

If need be, can move this to new thread, if inappropriate, we can just remove the post.

Thanks.

The case has not yet been brought to my attention (this is not unusual—the Logos Desktop development lead is generally the first to receive such reports). That being said, I'll be happy to look into it if you post a more detailed description of your issue in a new thread.

Oh my and how many PPC users did you lose by abandoning them? All users of 10.5 who can run Logos also are easily able to upgrade to Snow Leopard. Claiming you are holding back development for 7% of users just seems a bit rich considering the larger percentage of mac users lost to logos 4. 

-Dan

PS: Just seems a wee bit odd an argument, considering the history of it all...

David Mitchell said:

This is untrue. It is quite simple to use Grand Central Dispatch from Logos 4. The primary reason that we have not done so is that we still support Leopard, which does not have Grand Central Dispatch. Approximately 7% of our active Mac users are on Leopard. When a considerable number of them have upgraded, we'll consider dropping support and using newer technologies.

David Mitchell said:

You are the only person I know who routinely makes this assertion. As I've indicated before, your concerns have little basis in fact. Please either stop making these claims or send us a résumé so that you can have the opportunity to create an informed opinion.

 

Patrick S. may be the only person you know who routinely makes this assertion in this forum. BUT he is not the only one who believes this completely. I agree with Patrick 100%. Look at the MONO road map: "Last update: June, 2010." Look at the recent lawsuit against users of MONO. But what is the sense of routinely making these assertions in this forum. L4M is not going to change. And L4M works, sort of. I use it and am grateful for the rich resources. But I use the program only in a limited fashion and quit the program immediately when it is not required since it hogs my computers resources for doing other work. For Bible searches and study I use the alternative program and I leave it running all the time. Even if I don't leave it running in the background, it loads in an instant. It searches in an instant and doesn't consume all my computer's resources.

L4M has many aspects I like, but I truly believe that MONO is a ball and chain around the leg of the program. My son is in the seminary. He has limited resources (and will have limited resources out of the seminary as well!!). He cannot purchase two programs like his Dad. I have purchased one program for him and it was not L4M. Here is why I can't recommend L4M for those who can only purchase one major Bible program—MONO has an uncertain future and an unsatisfactory present level of function in my opinion.