The time invested depends on coder's skills and OS knowledge. e.g.: Itix could port software X in one week and it would take me two months to do the same. Despiting time factor the rest applies to x11 ports: no good quality checks, most of original bugs remain untouched (even new ones are added) and no use of host OS functions.
So, taking in account Edvard Schwan's skills, do you think that he has done the work too quickly?
And, are you saying that he has not done quality checks, or are you speaking in general of X11 ports on any platform?
we are not just talking about window managers, gaim would feel alien even if you used amiga window frames. Using an Amiga skin doesn't suddenly turn your x11 software into amiga software
We are talking about how good is the port of the X11 environment to OS4. The fact that X11 apps feel alien to the Amiga doesn't change in any way the quality of the port, because of the intrinsic limits of the X11 environment when running on a different platform. Within these limits, I said that the port is a good one, as the author has tried as much as possible to make the environment more Amiga-friendly, by using OS4's windows frames.
Then just use any random OS (e.g. Windows with WindowBlinds) with some Workbench skin and be happy. AtheOS had Amiga-like gadgets and there were some x11 window managers that mimicked WB but that won't turn gtk into mui.
Then we would be outside the scope of our discussion. We are talking about OS4 programs that MOS lacks.
"ls" also runs pretty stable on all amigas yet it doesn't feel native because it doesn't use amiga arguments.
If MOS lacked the ls command (and if it ls wasn't just a plain "dir" command
, it could have been a positive point of OS4 over MOS, regardless if it uses Amiga arguments or not.
I assume coder won't be silly and will avoid bleeding edge unstable versions. If code runs stable in other platform it should do so if the port is properly done.
For some applications there are no unstable versions, you have only one unstable version of the code, and this can happen especially with larger projects. You cannot also judge the quality of the original's author basing solely on how his code runs on the platform he works on. For example, you can write code with a low level of portability, or in a way that it is compilable only with a specific version tree of the compiler, or assuming that the primitive data types will have the same size on all the platforms, which is not true.
X11
Then the port's author is not to blame.
Lynx is also perfectly "useful" too.
Yes, in some defined contexts it is, as is Timberwolf, also in some contexts.
SDL/X11 ports always run better in the original platform.
For example?
I disagree. And I think it's polite to reply you instead of ignoring you.
You can of course disagree on my first opinion. But do you also disagree on the fact that OP has
asked not to talk about MOS? His statement has even confirmed that my interpretation of his first post was correct.
BBRV always says that if you have an order for Genesi to build certain number of 512MB Efikas they will build them. Both products have warranty and Efikas can be built on demand I don't see much difference.
"To be built on demand" doesn't mean that something is available on the market. I don't think that a single customer can order the production of a single Efika, they have probably set a minimum number of orders, and payment of all the shipped motherboards.
In addition to that I think there are more chances of ACube going broke than Directron
ACube isn't directly producing the motherboards.
I think ACube choosed wrong cpus, added unneeded stuff to motherboard like that fpga you can't program with freedom and added the mobility Radeon unneeded for embedded stuff. The choices they have made for 460 model look slightly better for embedded customers but are odd for desktop users.
We were comparing Pegasos2 with the 440, I think that all these factors you have mentioned don't fit with the discussion.
Most Amiga users I know were waiting some fast G4 machine many years ago. Now it's funny because Hyperion partners avoided the use of Altivec (with the honorable exception of A1-XE G4), something that pissed almost all users I know. The situation looked brighter when Moana appeared but unfortunately Hyperion management decided to recover developing costs selling us expensive hardware instead of trying to sell many OS4 units and leaving users the choice of using 2nd hand hardware like Mac Mini/Powerbook or new hardware (like Sam440 or whatever thing they wanted to build)
IMHO it was a good choice for Hyperion, making OS4 focus on its strenght. i.e. a few hardware supported with limited possibilities of problems derived by the use of hardware only partially-compatible with the available drivers.
Since some users are desperate to run OS4 they will buy almost everything you put on sale.
You cannot know if those users are the majority of the ones that have bought a Sam.
But mind you, if you could buy Moana and run it on a 2nd hand Mac Mini many users would have been able to try out OS4. Now the number of users spending more than 500 Euros to get a substandard board that runs slower than 5-year old boards are pretty slim compared to the number of users they would have got if they also had released Moana.
My point was that you can't say that OS4 hardware is more expensive than MOS' one. There's no new hardware produced for MOS, and Efikas can't be regarded as "in production", since AFAIK a single buyer can't request the production of a single board. If Moana was officially released then we could make a comparison between the hardware supported by both Moana and MOS. Since it wasn't, there are really no points in common.
Sure customers have right to choose, no one questions that. The problem is that customers who wanted to run OS4 had NO choice.
But that in my opinion didn't imply that most of the people bought A1s only because they were forced to do so in order to run OS4. I believe instead that most people know how to spend their money.
AGPx1 speed would be sad.
What has AGP to do with Sam 460?
both are OS4 hardware producers Hyperion forces us to buy hardware from if we have some interest in OS4.
But previously you were just talking about Sams being "prototypes" and "hardly tested". I hardly see how Sam's quality has anything to do with MAI's product.
But I can claim it hasn't gone through all the tests mainstream hardware goes through.
This provides 0 hints on how stable and tested are Sams.
Well, most (u)A1 users I knew sold them to buy Peg2/G4. Also the ones I don't know in person. Even one of the Friedens used Peg2 as main machine one year before OS4.1 Peg2 release because it was better hardware than MAI/ACube.
Actually your quote was that Peg2 are in general a better choice than Sams, and you mentioned that the reason was that you "prefer hardware well tested produced in hundreds of thousands of units, easy to replace, faster and cheaper." That can't still be a general rule, since other people may not care how many units of some hardware have been produced, while they might prefer to buy new hardware instead of old one, to lower the possibilities that the HW breaks down too early.
Varthall