@Raffaele

While I agree with some of what you are saying, attacking my friend Wiktor is mean spirited and a little dumb.
And $500 is not that much money, even in Poland.

Last edited by iggy on 25-Feb-2016 at 01:31 PM.

@iggy

Quote:

And $500 is not that much money, even in Poland.

especially in poland or elsewhere where people are usually working their backsides off in many cases just to secure their existence, sorry to say.

Last edited by wawa on 25-Feb-2016 at 02:00 PM.

@Raffaele

Hope you are felling better now.

@wawa

Looks like some ppl still think that $500 is equivalent of half year salary or so ;)

Last edited by pampers on 25-Feb-2016 at 02:38 PM.
Last edited by pampers on 25-Feb-2016 at 02:37 PM.

@pampers

I hope some people realize how hard Polish workers work.
I live in a State with a few beach resorts, and each year we get an influx of Summer workers (many Polish).
School teachers and students are a large component in this.
Damn hard workers, many well educated, often taking jobs below what you would think their station in life was, just to help make ends meet.

this reminds me an epic fail when I subscribed to "rentacoder" in order to earn some extra money as freelance coder.

people were offering enormous coding task for 100 or 200$ ( kind of task that requires 2 big months to achieve, full time).

A crazy idea that's not a short-term solution, but might require less effort in the long term:

What if the (Amiga) OS was modified to allow certain tasks to run in the PPC's "Little Endian" mode? You'd also need to mark memory allocated by it as being "Little Endian", and automatically do endian-swaps when it accessed memory NOT allocated by it. Probably also other situations where endianess would need to be swapped (e.g. calling OS procedures, if the parameters are stored on the stack).

Now surely that is a lot of work, but it would "automagically" mean that any future ports (not just WebKit) would not have any Endianess issues.

Now perhaps someone can tell me why this crazy idea wouldn't be possible?

_________________
Author of the PortablE programming language.
It is pitch black. You are likely to be eaten by a grue...

@iggy

Quote:

iggy wrote: @Raffaele While I agree with some of what you are saying, attacking my friend Wiktor is mean spirited and a little dumb.

Me attacking?

Have you read that kid critics since I started this bounty, by diminishing any of my actions?

Re-read all messages chronology and tone of messages, before saying I am the attacker, please.

_________________
"When the Amiga came out, everyone [at Apple] was scared as hell." (J.L. Gassée, former CEO of Apple France and chief of devs of Mac II-fx, interviewed by Amazing Computing, Nov 1996).

@Raffaele

Criticism does not automatically equal "attacks", and I must say that the criticism and skepticism you have received on your little savior crusade is understandable and warranted. Your wish to do right is applaudable and your naivety is cute, but when you constantly ignore those who actually know better and when you make posts about you having now potentially saved the web situation for the Amiga platform by raising a $500 bounty (those Polish people would do anything for to get ahold of $500, right?) then it's only natural that there will be a slight sharpening in the tone of discussion. But hey, it's a free world, you are free to have any uneducated opinion, you are free to raise vague bounties for about anything, and people are free to spend their money on it. Your decision. Their decision. But in this case it won't lead to anything, it wont't change anything. As people has pointed out to you. Which they are free to do as well...

@TRIPOS

And as general comment to various naive fantasies about the "bi-endian" magic that a seemingly growing amount of people feel they have now discovered for the benefit of humanity for the first time, well, it actually existed knowledge about this long before this thread. What it is. What it isn't. Issues. Can do's, and can't do's. Get yourself educated! There are lot's of info about this online, all over the Internet. Even on MorphZone.

@ChrisH

It wouldn't be possible because the very first Exec or library call that includes a multi-byte variable in a structure is going to fail. A longword 0x12345678 is going to be read backwards (0x78563412) by a task running in the opposite endian-ness. To prevent that, you would have to pass an indicator of endian-ness (or have it stored in the Task struct) and add big/little switching to every system library, driver, Exec, DOS, DMA, interrupts, etc.

_________________
cheers
tony

Hyperion Support Forum: http://forum.hyperion-entertainment.biz/index.php

@tonyw Quote:

It wouldn't be possible because the very first Exec or library call that includes a multi-byte variable in a structure is going to fail. A longword 0x12345678 is going to be read backwards (0x78563412) by a task running in the opposite endian-ness.

I partially already raised that issue & it's solution in my first post. Let me elaborate a bit:

A task running in Little Endian mode requests memory from Exec. It gets given memory from special memory pages marked as "Little Endian". The idea here being that the MMU could be used to detect accesses to Little vs Big Endian memory pages, and perform the necessary endian swap. So for Little Endian tasks, the MMU would have to trap access to all non-Little-Endian memory pages (basically all normal Amiga memory), and do a swap.

What I'm not clear on is (1) whether this is actually possible, and (2) how big of performance hit would it cause (probably a fair bit, but as long as L.E. tasks rarely interact with the OS it should be fine).

Last edited by ChrisH on 26-Feb-2016 at 10:29 AM.

_________________
Author of the PortablE programming language.
It is pitch black. You are likely to be eaten by a grue...

I guess it would have been easier to first find someone willing to do the work and then and only then start the bounty.

This is quite a huge task and seems to require specific knowledge.

_________________
http://www.warpdesign.fr/

@Leo

I had contact with Deadwood. His guess about removing endian issues on OWB he is working at was 6 months for him and he is both very experienced and has deep knowledge about Aros and OWB. Guess what would someone need with less experience and expecially less knowledge about the sources. This was about OWB (not the javascript engine) but it shows how big the tasks are. 500$ are nothing for that.

@Raffaele

Quote:

Me attacking?

definitely. you have answered criticism and advices with anger, name calling, accusing, emotionality but almost no arguments.

Quote:

Have you read that kid critics since I started this bounty, by diminishing any of my actions?

what actions? writing very few sentences in a row in favour of what sould be done, according to your wishes, and waiting for people to pay up for this? strangely enough i dont recognize you contributing to your own bounty, can you tell me by what nick you go here?:
https://www.bountysource.com/issues/30646188-odyssey-web-browser-solve-ppc-webkit-javascript-engine-endianess-problems-inherited-from-webkit-x86/backers

instead you accuse others to be waiting for a dish to be served. and blabbering on forums doing nothing. i mean, this is exactly what are doing, considering your "actions" or rather the lack of it.

Last edited by wawa on 26-Feb-2016 at 01:16 PM.

@wawa

Are you insane?

Or are you miserably attempting insinuating i will start a bounty without put my money in the basket?

Geez... This is sobbing, even for the low level of Amiga community usual trolling.

As I wrote in the bounty announcement thread I was the first to donate, and I gave 10 euro. All I had on credit card...

I could give more in the future as long as I got some spare job as I am jobless ATM.

FYI I am the one who donated as "Amiga Community", as I created this group on Bountysource, in prevision of any future bounties could be of help to our community.

You could had discovered Amiga Community it was me, just by clicking on Group Name and and humbly verifying who created it, and you'd be spared to write an insinuating maliciously question in public.

Last edited by Raffaele on 26-Feb-2016 at 04:28 PM.
Last edited by Raffaele on 26-Feb-2016 at 04:20 PM.
Last edited by Raffaele on 26-Feb-2016 at 04:12 PM.

_________________
"When the Amiga came out, everyone [at Apple] was scared as hell." (J.L. Gassée, former CEO of Apple France and chief of devs of Mac II-fx, interviewed by Amazing Computing, Nov 1996).

@ChrisH

IMHO no magical swapping endianess solution at next corner and no one insight...

All the effort to do is monitoring each Javascript calls, one by one... It is still the most rapid and suitable method, even if requires long times, enormous patience, attention and a moderate high skill at programming.

I could understand point of view of Deadwood (asked by OlafS) who estimated the task should took 6 months of tedious work for a single person.

That is why we need more money to reward people who want to attempt this effort.

Unfortunately there are people who said they do not want to danate and consider this bounty as doomed. This makes people get demoralized as they loose faith in any viable solution.

Well, a solution exists and is to get a good amount of money to reward any developer who want to start this bug hunting, and we need a point from which to start, and now there is.

Nothing more to say about.

_________________
"When the Amiga came out, everyone [at Apple] was scared as hell." (J.L. Gassée, former CEO of Apple France and chief of devs of Mac II-fx, interviewed by Amazing Computing, Nov 1996).

@Raffaele
I'm not trying to dismiss your pragmatic solution (bounty for fixing current JavaScript endian issues), but rather I'm just trying to discuss *if* there is a possible alternative solution, which could maybe be used in far future (rather than having to continually fix endian issues all the time).

Endian problems are not unique to JavaScript or WebKit, and if we could have a general solution in long-term then it would help keep Amiga getting new software. It might also provide a path to x86 or ARM, without loosing 68k/PPC compatibility...

Last edited by ChrisH on 27-Feb-2016 at 01:07 PM.

_________________
Author of the PortablE programming language.
It is pitch black. You are likely to be eaten by a grue...

IMHO the code is not worth it, what happens everytime they update it, will you pay over and over again for the same tedious work ? Amazing that they code something like this...

_________________

José

Well this wouldn't be the first time an endianness problem cropped up in the world. But I think Apple did help to keep code slightly cleaner when they used PPC. And now we've lost the indirect backing of Apple this type of thing falls by the wayside.

It's also not the only code where little endian data has to be delt with by Amiga code. USB structures are little endian and OS4 has to deal with that. Modern framebuffers are also backwards and store things in BGRA type orders.

So it can affect simple variables, structures and calls that work on word widths of data. Including unions and bitfields. The work is finding which of these in code is affected and dealing with it. Code that sets these normally at run time should be fine. But if the code is sloppy or makes assumptions about the data then there will be trouble.

I've modified code to be endian or PPC friendly but it isn't pretty and had all these macros around read and write access including special file functions that load (or save) data on a word basis.

I've said it before but I think this needs to be solved at the compiler level. So a datatype or structure can be specified to be a certain endian. I'm surprised this isn't already the case. To me this looks like the cleanest solution. That way all that needs to be done is things like tagging a variable and then it can be forgotton about. Apart from things like passing structures around or variable pointers. Fine for internal code but once it goes extenral it needs converting.

@Hypex

Quote:

I've said it before but I think this needs to be solved at the compiler level. So a datatype or structure can be specified to be a certain endian. I'm surprised this isn't already the case. To me this looks like the cleanest solution. That way all that needs to be done is things like tagging a variable and then it can be forgotton about. Apart from things like passing structures around or variable pointers. Fine for internal code but once it goes extenral it needs converting.

The C and C++ languages don't support such a thing.

You could create your own bespoke endianess agnostic laguage, but that hardly helps in the general porting software issue, as no one else will use it.

_________________
BroadBlues On Blues BroadBlues On Amiga Walker Broad