• Welcome to OS2World OLD-STATIC-BACKUP Forum.
 

News:

This is an old OS2World backup forum for reference only. IT IS READ ONLY!!!

If you need help with OS/2 - eComStation visit http://www.os2world.com/forum

Main Menu

What are Your real priorytes ?

Started by cyber, 2010.12.29, 09:14:11

Previous topic - Next topic

DougB

QuoteSo, back to sit and wait...

Well, if you sit back, and wait, and everybody else sits back, and waits, guess what will happen. On the other hand, if everybody spends a little time trying to help out, we might just be able to pull it off, and get things working. If you can't program, do the testing, and tell the programers how to recreate problems. If you can't do the testing, find some other way to help. Of course, there is always the old "contribute money to a paid project" thing. JAVA is looking again, and there are a number of others looking for a little monetary support so they can continue to do more for the eCS/OS2 community. Every little bit helps to move the whole thing forward.

What I was trying to get across, earlier, is that somebody can spend two years trying to develop Bluetooth support, while ignoring the more basic things, and by the time they are done, there won't be a working platform to run it on (or, the platform will change enough that their project won't work with it). Once the platform is more stable, developing things like Bluetooth support will probably take less time, but the basics need to be there to support the extras.

ivan

One thing I would like to add re testing.  The developers should open the testing out to as many people as possible NOT a select few that have access to a restricted site, otherwise we will continue to have the debacle that is ACPI development.

ivan

Sigurd

What I learned this last year - if you want to have progress - try to take it in your own hands as much as possible. Using this way I was able to (not fishing for compliments):

- get "my" formerly NOT SUPPORTED  X200T Notebook quite well supported with:
  1. finding a way to install eCs 2 next to Preinstalled Windows 7 at all
  2. finding a way to use SNAP instead of Panorama with it (with the patched gradd.sys from bug tracker)
  3. finding a way to solve the "overheat" problem
  4. finding a way to solve the WLAN problem
  5. finding a way to get the Stylus to run AND finding applications to make it real usable (like Jarnal)
  6. finding a way to connect to the internet anytime from everywhere with UMTS (with expanding USBCOM.SYS to work with USB 2.0)

and - still in progress -  ;)

  7. finding a way to get the internal Webcam supported

and - work done by Mensys

  8. get LAN support

I would not have achieved all this without the help of some real good programmers (Wim Brul, Rudi and more....) - they did the hard work - for me it was: testing, testing, thinking and searching for solutions, to combine and connect solutions to geht them in a practical way to run, publishing solutions in words (HOW TO) and pictures (Youtube)

And yes - even if it sounds heretical:

"My" Plan B that I suggested in June 2009 went on a successfull road, it is nearly completed (the Webcam) and it did proof to me, that Mensys should have concentrated on selected Hardware, not trying to cover each and any hardware with a weak and instable ACPI. As a "Hobby eCS User" it was possible to achieve all this (Once again: without the help of these really friendly developers I had I would not have Stylus nor UMTS support) in about a year time, while eCS 2.0 moved from ACPI 3.14 (December 2008) to 3.18, without any real step forward (the same with the "short time after english GA to be released German Version of eCS" announced) - even the manual suggests to go back to 12/2008 3.14 in case of problems. The same with the Flash 10, "second Preview" ......

But it does not matter any more, this big chance has been wasted, as several before.

I really appriciate the work that is done here by a lot of others, in example Silvan Scherrer, the GL/2 project  (both using the same method: If you want to have something done or changed - try to start it yourself, regardless how hard it seems to be there are people outside there willing to help you!) , just to name a few. And I am looking forward to use JAVA6, Virtualbox 3, and more! Will try do use eCS2 as long as possible on modern hardware, I guess my X200T will at least be used for another 3 to 5 years.

But if this ACPI problem will not be solved the future for eCS is NO FUTURE, or a future in a virtual environment. Sorry for this but the progress in the last two years in this area seems to me to go close to zero.

I wish you all a happy new year 2011, and most important health!!

Operating systems aren't that important at least  ;)

Andi710

Quote from: Sigurd on 2010.12.30, 23:08:43
What I learned this last year - if you want to have progress - try to take it in your own hands as much as possible. Using this way I was able to (not fishing for compliments):

and - still in progress -  ;)

  7. finding a way to get the internal Webcam supported


not fishing for compliments, eh?  ;D Anyway, really nice job. I enjoy being able to use the pen at my thinkpad X41. I wish I had more time for testing, or even porting myself.  :'(

Please let us know when the webcam is going online and what SW you decide using (maybe flash is ready and usable by then).

Happy New Year and keep up the great work!!!

DougB

Quote from: ivan on 2010.12.30, 22:26:12
One thing I would like to add re testing.  The developers should open the testing out to as many people as possible NOT a select few that have access to a restricted site, otherwise we will continue to have the debacle that is ACPI development.

ivan

I think that ACPI is a "special case". Testing is available to those who have a Software Subscription. That makes sense, since ACPI is really not available to those who don't have eCS, and if it was opened to everybody, it would be effectively giving ACPI away, when it is really one of the main things that eCS has to offer. That is the commercial part of the equation, and I think it struck a reasonable balance. On the software side of the equation, it was apparent, very early in the project, that the approach taken was all wrong. It was apparent, that every manufacturer, did it different, with every machine they designed (and then it would change with BIOS levels). That meant that the software needed to make exceptions, or additions, for every one of them. A losing proposition from the word "go". They were told that it wasn't going to work, but totally ignored the fact that Windows, and Linux, were not doing it that way, simply because it was never going to work that way. On the other hand, those who were doing the testing tended to lose interest, because nothing was really happening, unless the developers happened to have a machine that was very close to the same as what the tester has, and the developers can't have one of every machine. This is probably the main reason why people seem to think that there should be a small list of "acceptable" machines. IMO, that would kill OS/2 faster than anything else. Very few would spend the money to have a special machine, just to run eCS. Those who wish to try it, want it to run on the hardware that they already have, or what is available locally, and there is a constant turnover of hardware, so that it would be impossible to keep such a list current enough to be useful. At the moment, I think there are three major things that need to be done:
1) WiFi needs to work with new hardware.
2) AHCI needs to work.
3) Large disks need to be supported (by that I mean the eCS should install to disks up to 2 TB).

Also important, but not necessarily critical, is ACPI support.

Sigurd: A big THANK YOU. I used your documentation, to get my new Lenovo ThinkPad T510 to go, but I had to disable Win7 all together, because it insisted on having a disk layout that eCS refused to use, and Win7 doesn't seem to like the disk layout that eCS likes (windows 7 problem, IMO, but LVM, and/or the eCS installer, needs to be fixed too).  I haven't, yet, found a way to get Win7 to co-exist with eCS (I haven't tried too hard), but most of the T510 actually works well with eCS (no surprise that ACPI has major problems). I need more time to play with this, so I will be able to report the correct questions and information, in order to resurrect Win7 (not a high priority), and to get ACPI to work properly (it is an Intel machine, which is supposedly supported, but I have yet to see one that comes close to working right). I will admit, that I am tempted to just not use ACPI (SMP works fine with the old OS2APIC.PSD), since taking the time to report the problems is unlikely to result in anything useful, and it works fine without it (but won't power off at shutdown, without pushing the power button). WiFi (Realtek) is another thing that needs to be looked at, but the wired NIC works good with the new Intel driver, and sound works good with the UniAud that comes with eCS 2.0, and the mmfix thing.

Overall, progress is being made, with eCS, but it is hard work, and very time consuming for all involved (not to mention, way to slow). More people doing it would help, but that also means that the developers need to instruct people on how to properly collect data, and report problems, and they need to make documentation that is usable by the average user. The ACPI documentation is almost incomprehensible, and some other programs don't have documentation, at all (few can get away without some sort of documentation).

As for the comment:
QuoteWhat I learned this last year - if you want to have progress - try to take it in your own hands as much as possible.
I don't think that is quite right. To do it yourself, is very limiting. To work as a team, is far better, but some coordination is required to do that. You list your accomplishments (great work), but how long did it take you to o all of that? If 6 people had coordinated their efforts, and each had taken on one of the projects, it all might have happened faster. I think that what we need, is a way to coordinate the time spent doing such things. It is probably not helpful, if 6 people are working independently on getting a web cam to work (unless each has a different web cam, but they should still be working together). Anybody got some ideas about what to do about that?

To one, and all, a Happy New Year...

Sigurd

To Andy:

thanks for your comment :-) Yes, I know, it sounds like "fishing for compliments" but it was just to stress the fact what has been made in spare time as a hobby.......  :-\ That is what makes me think that in the last six years so many opportunities have been wasted and what may have could been...... But even though - please read below  ;)

To Doug:

thanks for your comment as well. To get Windows 7 little "friendlier" I found it very usefull to use the DFSEE Boot CD instead of eCS 2 to manage the volumes.
Regarding the achivements and the way those have become reality: this was the ONLY way for me to change something. There was no one around who has been interested in helping in "general" support, I was just lucky to get in touch with the developers with the special problems I had. And as for the stylus - more testers came aorund and more Laptop Models are supported now (X61T, X41T). I guess it is the same with the UMTS / USBCOM.SYS - I would bet that a lot of internal UMTS Modems are able to work with this driver - it should be some kind of "unversal support" - BUT - there is no one around to test it nor use it. That answers the question: Are things like this demanded? And by the way - I heard from someone that his Mobile phone is recognized as a Drive with the new USBCOM.SYS now - that may give the opportunity to close another door - syncronize with mobile devices......

And with the webcam support - the same road is taken as succesully before: I know only one developer with whom I do have contact in this case. I guess that there are others who might work on this topic as well, but I never read or heard of them. Unfortunately. And in the end it might be the same result as before: we will have a free availiable driver for everyone, eComstation OR OS/2 user. But it is to early to forecast something. This way it worked before: start with ONE problem, ONE developer, ONE tester. I wish it would be different - more tester or developers, but there seem to be none....

And concentrate on selected hardware: in my opinion this has been the ONLY way for eCS to survive. Or better: had been. Apple proofed this for years.

But - all this achievements are in the end worthless if there is no way to get EFI, ACPI, AHCI to run. Sometimes I think the OS/4 project has more chances to may overcome these problems, and I guess as well that the latest eCS2 kernel has been a patched one like the OS/4 kernel as well to open the door for ACPI. And I guess as well: to get ACPI to run there are more Kernel Hacks needed - but not possible for Mensys due to licening and technical problems. And therefore the whole ACPI is still broken with no real step forward the last 2 1/2 years. And this is in my opinion the BIGGEST PROBLEM of all with Mensys and eCS 2: the missmanagement regarding "information for users", a real disaster. What shells at least ? ;-)

I personally will give another bootable OS/2 Warp 4 DVD a try, created by Update CD, with the latest OS/4 kernel and loader, if possible and will see where this leads me to on modern hardware.

Thanks again and a good 2011 to all of you!

aschn

Quote from: melf on 2010.12.30, 00:56:16
Some polishing on the GUI, it'll frighten new users away as it is now. I can't understand why first install still uses old folder icons.

The resources of the desktop objects are patched at the end of phase 1, so you have to cancel the installation to see the IBM icons. The old icons are used firstly, because the IBM installer is still being used.

Additionally, patching the IBM files may lead to license violation - I'm not sure. IMO it was not a good decision to just write a GUI for the old installer. The current eCS installation is so complicated, that there exist just a few people who understand the main parts of it.

My wish list:

1. shared memory
2. ACPI
3. AHCI

Andreas

rwklein

I have skimmed the messages here are some answers.

Bluetooth is something you have to hold you breath for on eCS.

Audio support on eCS is a no brainer with the current UNIAUD driver.
You can get it here: svn.netlabs.org/uniaud
If something does not work open a bug. But a lot of problems have been fixed in UNIAUD!
We are in better shape then ever before. We are current with the current Linux ALSA code.

Mensys has received the first 2 binaries for AHCI drivers and we are testing this as we speak.
When finished the code will be released.  SO AHCI is under development and on the move.

Somebody mentioned that eCS has boot limit of 512 GB of eCS on 2 TB discs. That is not true.
But you need to partition the drive completely with eCS and otherwhise follow the instructions.
Current work is to make installing and booting eCS next to Windows 7 and on large hard disc
is by dumping the boot manager and replacing it with Airboot. I won't go into the details but the bootmanager
we have used for many years is technically dead. With airboot into eCS included we can
boot Windows normaly Vista and 7 OEM preloads normaly. We save 1 primary partition also!
Airboot will receive some updates. And yes we are allowed by Martin Kiewitz to include into eCS
(For the people that have pointed out the license of use limiations for Airboot)

In the end we also need to update OS2DASD, DANIS506.ADD and the miniFSD for bootable JFS to support big hard discs properly. As it now stands
hard discs going beyond the 2 TB disc boundary will use GPT disc layout but start to jack up the
sector size from 512 bytes to 4 KB. And the old MBR disc layout will stay as well.
I base this on the current state I read in the computer magazine C'T from Germany.

The first DLL will be made for Panorama in the coming months to support accelrated Intel Graphicss chipsets.
About wide screen support. That is currently what it is. But one developer we are looking to create
a generic wide screen eneabler driver.

About people wanting to point out that SNAP is an option. From the point of view of Mensys
SNAP is dead. We have not realy to come to an agrement with the company that has taken over
SNAP on how to get source code access. Besides that the SNAP source code is not that interesting.
Currently 2 people are busy updating screen01 as that contains a lot of legacy, like scanning I/O ports
it should not do. Thats what is causing hangs on some systems. Lars Erdmann is working on this.
When it comes to DIVE performance with sysbench Panorama get a score of 3.6 million dive marks,
SNAP gets 3 million marks with accelrated support. So go figure, Panorama runs faster then SNAP.

Work is being done by a group of people to fix the USB stack. I have mentioned this before but this is something
that has to be done in a proper fashion. Patching the current IBM binaries we have is not sufficient.
*first* priority is to get the USB host controller drivers stable with ACPI APM and APIC (high IRQ mode).
Also updating the USB HID and USB keyboard mouse driver and fix USBD.SYS.
Other strategies are useless. The base needs to be stable. Then we can work on USBMSD and possibly USB 3 host
controller support.

About GUI enhancements the file open dialog now in beta zone we hope you can test this.
That we all old applications can have a decent file open dialog!
Eschemes will receive more in the coming months as well.

About IP 4 versus IP 6. This is one of the most overhyped items. IP 6 should, based on what I have read and heart
not going to be a problem. In about 9 months from now All IP 4 adresses at RIPE will run out.
Some ISP's are already handing out IP 6 ADSL lines. But most of these modems translate IP 6 to IP 4 on the internal
LAN side of the network. Besides that IP 4 devices will be around for years to come! So in that retrospect
OS/2 will not be in the dark. ISP will have to provide gateways for most common IP 4/6 traffic.
People who tell you IP 6 will the dead for OS/2/eCS are making statements without technicall knowlegde.
We are looking at IP 6 support for eCS but thats going to be a big job.

Talking about connectivity. Take a look at http://svn.ecomstation.nl/multimac
New NIC drivers coming.  Already released current support for Nvidia and Intel gigabit.
This skeleton has a portion of OS/2 specific code and the rest is the Linux code ported to OS/2.
It has support Xwlan. Soon Mensys will release a binary for Realtek gigabit.
A new Intel Wireless driver to support current chipsets is also in the pipeline.

Mensys also provided funding to have Firefox/Mozilla sweet updated. The latest fixes done by Richard Walsh
for printing and fixing certain memory leaks where done by him. Flash is still under development
and Mensys also co-funded the Java development. Java 6 was released recently.

As for Sigurd amd other people there remark we should support specific computer models. Its nice to suggest
but not very practical. Here is the reason why. Look at UNIAUD, USB drivers and NIC drivers for example.
They are all generic. Also the upcoming AHCI driver most hardware drivers are so generic that it makes
no sense to just say "support this model". Infact I can tell you this is not even practical for some
of large company customers since certain computer models have such short live cycles.
I know it sounds weird but I have been there having to deal with it.

To wrap up I have not much time to read the forums but I hope this does provide some inside in the way
we working. I keep seeing people making assumptions about the future of eCS. But when you make statements
put a solid technical analyses under it. That is sometimes missing. If you have questions contact me...

ACPI is still being worked on hopefully getting back to that later this month.

About the shared memory issue. This is something that can not be fixed in the current kernel. What is fix is to let applications use high memory to store data. Storing DLL's in high memory is in the back of our mind.
And is possible possibly later this year you will see more about this.
But a lot of freezes and hangs of systems by using firefox have been caused by faulty code. This should be fixed in Firefox 4 being released.

Have a good new year for 2011!

Roderick Klein
Mensys

P.S. Not yet annouched officialy (and the website still needs to be worked on!) but Warpstock 2011
will happen in Haarlem, home town of eCS and Mensys. It will be in Haarlem on the 6 to 8 of May.
Again the website still needs updating!!!! Thats currently being worked on.

ivan

Roderick, thank you for the update.

It sounds very much that there is life in the OS still despite what critics have and still say.

ivan

DougB

QuoteSomebody mentioned that eCS has boot limit of 512 GB of eCS on 2 TB discs. That is not true.
But you need to partition the drive completely with eCS and otherwhise follow the instructions.
Current work is to make installing and booting eCS next to Windows 7 and on large hard disc
is by dumping the boot manager and replacing it with Airboot. I won't go into the details but the bootmanager
we have used for many years is technically dead. With airboot into eCS included we can
boot Windows normaly Vista and 7 OEM preloads normaly. We save 1 primary partition also!
Airboot will receive some updates. And yes we are allowed by Martin Kiewitz to include into eCS
(For the people that have pointed out the license of use limiations for Airboot)

The problems, at the moment, are:
1) Without Boot Manager installed, you cannot install to a logical drive.
2) Even if you do use DFSEE to build the partitions, add Boot Manager, then format them (JFS in my case), the eCS installer will still not allow you to select the drives for install (BM is probably the problem).
3) On a laptop (and possibly some desktops too), the dumb recovery code insists on rebuilding the whole disk to contain whatever windows version they want to use. The disk layout (geometry) is something that eCS refuses to acknowledge (cutoff seems to be somewhere near the 512 GB mark), and Boot Manager doesn't work (installed without complaint by DFSEE). Therefore, you cannot install eCS. Trying to restore windows to the layout that eCS will use, leaves windows unable to boot (I have not pursued this, yet).

QuoteThat is currently what it is. But one developer we are looking to create a generic wide screen eneabler driver.

FWIW, the current widescreen driver (at least the Intel version) does work with SNAP, Panorama, and GRADD.

QuotePanorama runs faster then SNAP.

I specific cases, that may be true. The biggest complaint that I have about Panorama, is that it will not allow Doodle's Screen Saver to power off my screen. It does power off when I use SNAP. This is really a minor problem, because I can always switch the screen off manually, but power control should be enabled by Panorama. (SNAP doesn't turn off the screen on my ThinkPad systems either).

Re: SCREEN01.SYS hangs. My new Lenovo ThinkPad T510 hangs at SCREEN01.SYS, for almost 4 minutes, when i try to use APCI (needs /EIS). Eventually, it does continue, and it seems to work, but the SETGETTHRTL.EXE program has some problems with it, and SETGETSTATE is not supported. I am wondering if it might not be a good idea to separate the SMP support, and power support, into two packages. The SMP support seems to work pretty good, but it is very difficult for a user to determine if a problem is caused by the power support code, or the SMP support code.

Re: USB support. It seems that new systems may no longer have the need for OHCI, or UHCI drivers. My new Lenovo ThinkPad T510 does not have OHCI, or UHCI, adapters. It only has two EHCI adapters. USB 1.1 devices work with no trouble. I am not sure how that happens. USB 3.0 is, of course, another thing that needs support. It would probably be good, if somebody picked up the old FireWire driver, and got it to work too.

Re: NIC support. I am really looking forward to the Realtek WLAN support. The T510 has one of those. It also has a modern Intel wired NIC, that works great with the new driver - THANKS!

QuoteACPI is still being worked on hopefully getting back to that later this month.

Hopefully, by then, I will have enough information about ACPI on the T510 (4313-CTO), to be able to post something useful, but even my T43 (1871-W8M) is only marginally supported (power off works, but suspend/resume is still a dream), and that is not exactly a new machine. Notebooks are still poorly supported, and they are the machines that need it the most.

ECS is progressing, but it seems that there is always something new that is causing us to fall further behind. Linux has been a great help, so far, but I am not so sure that we can depend on them for everything. Those who wish to keep on using eCS, need to step up, and help out, so those who can do the work don't need to do as many other things.

melf

#25
Quote from: aschn on 2011.01.01, 14:59:39
Quote from: melf on 2010.12.30, 00:56:16
Some polishing on the GUI, it'll frighten new users away as it is now. I can't understand why first install still uses old folder icons.


The resources of the desktop objects are patched at the end of phase 1, so you have to cancel the installation to see the IBM icons. The old icons are used firstly, because the IBM installer is still being used.

Additionally, patching the IBM files may lead to license violation - I'm not sure. IMO it was not a good decision to just write a GUI for the old installer >--snip--<

Hi Andreas,
I'm not shure we talking about the same thing. With "first install" I mean the desktop coming up when install is finished. This icon set is called "ecomstation classic". David Graser have made several (in my view) more attractive, among them the flat basic yellow delivered with eCS. But the information from Roderick about continuing work on eschemes is good to hear about. Eschemes would make the overall appearence of ecomstation much better, when errors and misbehaviours have been ironed out.
/Mikael

craigm

Thanks Roderick, just a little bit of explanation from Mensys and Serenity Systems goes along way to the naysayers.

I believe 2010 was a great year for eCS and I'm excited at the prospect of 2011 being even better!

Again, thanks for the work you and the rest of the team does. This could not be done with out you!

Sigurd

Even if it may be not allowed or not welcomed: here some marks from an "not skilled enough":

Quote from: rwklein on 2011.01.01, 19:29:44

<Audio support on eCS is a no brainer with the current UNIAUD driver.
<You can get it here: svn.netlabs.org/uniaud
<If something does not work open a bug. But a lot of problems have been fixed in UNIAUD!
<We are in better shape then ever before. We are current with the current Linux ALSA code.

- Uniaud HDA still not working with:
   - Virtual PC
   - Virualbox
   - Voicetype
   - Problems on modern hardware with HDA in APIC mode (wrong IRQ)
Yes - these are "minor" problems and I am happy to hear sound at all.


<About people wanting to point out that SNAP is an option. From the point of view of Mensys
<SNAP is dead. We have not realy to come to an agrement with the company that has taken over
<SNAP on how to get source code access. Besides that the SNAP source code is not that interesting.
<Currently 2 people are busy updating screen01 as that contains a lot of legacy, like scanning I/O ports
<it should not do. Thats what is causing hangs on some systems. Lars Erdmann is working on this.
<When it comes to DIVE performance with sysbench Panorama get a score of 3.6 million dive marks,
<SNAP gets 3 million marks with accelrated support. So go figure, Panorama runs faster then SNAP.

- Proofed to be wrong here. On systems I do have (Modern hardware - usually not supported by Snap) the VESA 3.0 mode of SNAP is at least as fast as the Panorama with the following advanteges over Panorama:
  - not eating that much CPU = causing heat and eating battery
  - no wrong colour scemes with SDL applications
  - Screensaver works and can switch off
but - please do remember: this is not a full technical proofed statement!

<Talking about connectivity. Take a look at http://svn.ecomstation.nl/multimac
<New NIC drivers coming.  Already released current support for Nvidia and Intel gigabit.
<This skeleton has a portion of OS/2 specific code and the rest is the Linux code ported to OS/2.
<it has support Xwlan. Soon Mensys will release a binary for Realtek gigabit.
<A new Intel Wireless driver to support current chipsets is also in the pipeline.

- This is really one well done job - a new LAN Nic. I do hope that the WLAN part will see the light of the day someday....

<Mensys also provided funding to have Firefox/Mozilla sweet updated. The latest fixes done by Richard Walsh
<for printing and fixing certain memory leaks where done by him. Flash is still under development
<and Mensys also co-funded the Java development. Java 6 was released recently.

- Good as well!

<As for Sigurd amd other people there remark we should support specific computer models. Its nice to suggest
<but not very practical. Here is the reason why. Look at UNIAUD, USB drivers and NIC drivers for example.
<They are all generic. Also the upcoming AHCI driver most hardware drivers are so generic that it makes
<no sense to just say "support this model". Infact I can tell you this is not even practical for some
<of large company customers since certain computer models have such short live cycles.
<I know it sounds weird but I have been there having to deal with it.

- As I wrote before: all these welcomed things are really useless without a working ACPI interface. As far as I understand it each and any PC Model does have a different ACPI table. So I guess it would have make sence to concentrate on some ACPI tables to get them properly to work so one could benefit from all the other advantages. But since 2 1/2 years this main project does make no step forward.

<To wrap up I have not much time to read the forums but I hope this does provide some inside in the way
<we working. I keep seeing people making assumptions about the future of eCS. But when you make statements
<put a solid technical analyses under it. That is sometimes missing. If you have questions contact me...

- Well, for me I can say that allmost everything I am writing about considered to eCS / OS/2 does belong on the technical/practical know how and the success and faults of the projects I did follow in my spare time. So, I was never trained in programming, espacially not in OS/2. So I was happy to provide some information about Windows 7 and eCS 2 while you wrote the manual for ecs2. But I hope even though it is allowed to make suggestions because this is a forum and not a "Yes, hayoo - we are all so happy!! Lets be quit and not critic, we might ask the wrong questions!" here.
By the way it would have been much easier to know what happend to eCS 2 in the last 6 years if there would have been some regular kind of information about the current states, as all the projects i.e. lead by Silvan Scherrer does. Well organized, progress frequently updated, information handed out......

<ACPI is still being worked on hopefully getting back to that later this month.

- this proofs what I wrote before - latest half usable 3.14 dated 12-2008 - the MAIN goal is scheduled to sometime....


And yes, we have had this kind of dicussion for years now i.e. at os2.org. And if you believe it is better to have only YES saying users, than just tell it, I will stop any kind of asking here as well. And sorry for not "quoting" the way one should have to.

rudi

Quote from: rwklein on 2011.01.01, 19:29:44

When it comes to DIVE performance with sysbench Panorama get a score of 3.6 million dive marks,
SNAP gets 3 million marks with accelrated support. So go figure, Panorama runs faster then SNAP.



1.) Panorama is cheating by not performing proper shadow buffer sync for DIVE. If that would
     have been implemented in the way it is supposed to be, performance would be slower.

2.) In genereal, DIVE is a bad test case for video driver performance measurement. This is
     because the whole work is done in DIVE.DLL and not in the display driver. Besides setting
     MTRRs, SNAP does not provide any acceleration for DIVE. So unless a display driver
     implements the EnDIVE API (which - thanks to IBM - is functional insufficient and as
     buggy as hell) there is no performance gain to be expected.


demetrioussharpe

Hello, I'm Dee from GL/2. For all of the gamers who may have a bit of technical knowledge, I'd like to post a query. Which is more important to you personally, high precision rendering or sacrificing precision for render speed? To reply, please post your answer at:

http://www.os2world.com/forum/index.php/topic,1940.0.html
The difference between what COULD be achieved & what IS achieved
is directly relational to what you COULD be doing & what you ARE doing!