major regression

Here you can discuss ReactOS related topics.

Moderator: Moderator Team

Webunny
Posts: 1201
Joined: Sat Apr 28, 2012 1:30 pm

major regression

Post by Webunny » Sat Jan 10, 2015 2:50 pm

It has been a while since I did a testrun with a new build, what, with the new year and such, but today I tried out 66022.

It didn't work at all, on my PC. It doesn't get installed, even. Just hangs in a loop and restarts every time. Never had this happen on my PC (same specs as in all my other HW testing, btw) for the last two years. This doesn't bode well.

Since I didn't see any mentioning of this problem by VM-testers, I assume it's only noticeable when doing real hardware-testing, as has happened before.

I'll reburn with a slightly older version, to make sure it's not something that went wrong physically, but if not, I guess something serious borked in the code.

Edit: nope. Just tried it with a new burn and new version (66020): still doesn't get through the first install process. It's clearly majorily borked compared to my last build, which was somewhere 3 weeks to a month ago.

Pi_User5
Posts: 752
Joined: Thu May 01, 2014 4:18 am
Location: United States

Re: major regression

Post by Pi_User5 » Sat Jan 10, 2015 3:06 pm

I have the same problem also. I don't remember what revision it was, but setup says it can't load NTOS core on my real hardware machine.
I use ReactOS on real hardware. Will you? My Computers: https://www.reactos.org/wiki/PC_ROS_Rigs Go all the way to the bottom.

Image

Webunny
Posts: 1201
Joined: Sat Apr 28, 2012 1:30 pm

Re: major regression

Post by Webunny » Sat Jan 10, 2015 3:14 pm

Pi_User5 wrote:I have the same problem also. I don't remember what revision it was, but setup says it can't load NTOS core on my real hardware machine.

Has it already been mentioned to the devs or put on JIRA?

No need to duplicate a report/issue, if it's already known, obviously.

Ps. No need to mention that there is no serial cable debug info, since it doesn't even get through the first part of the instalment...

Pi_User5
Posts: 752
Joined: Thu May 01, 2014 4:18 am
Location: United States

Re: major regression

Post by Pi_User5 » Sat Jan 10, 2015 3:26 pm

Has it already been mentioned to the devs or put on JIRA?

No need to duplicate a report/issue, if it's already known, obviously.

Ps. No need to mention that there is no serial cable debug info, since it doesn't even get through the first part of the instalment...
I don't think it has been reported yet.
I use ReactOS on real hardware. Will you? My Computers: https://www.reactos.org/wiki/PC_ROS_Rigs Go all the way to the bottom.

Image

Webunny
Posts: 1201
Joined: Sat Apr 28, 2012 1:30 pm

Re: major regression

Post by Webunny » Sat Jan 10, 2015 3:29 pm

Pi_User5 wrote:
Has it already been mentioned to the devs or put on JIRA?

No need to duplicate a report/issue, if it's already known, obviously.

Ps. No need to mention that there is no serial cable debug info, since it doesn't even get through the first part of the instalment...
I don't think it has been reported yet.
I just looked, and I think it's the same as https://jira.reactos.org/browse/CORE-9032 .

Though it's always difficult to know for sure. But it seems similar enough.

oldman
Posts: 1072
Joined: Sun Dec 20, 2009 1:23 pm

Re: major regression

Post by oldman » Sat Jan 10, 2015 3:33 pm

Webunny wrote: Has it already been mentioned to the devs or put on JIRA?

No need to duplicate a report/issue, if it's already known, obviously.

Ps. No need to mention that there is no serial cable debug info, since it doesn't even get through the first part of the instalment...
It's a known problem
jimtabor commented on CORE-9032 - livecd-65905-65991 does not boot on a real computer, recorded on a CD-RW . livecd-65904-dbg.iso loaded on a real computer .

I've tested this on other hardware systems, same, some just reboot on their own, and the more advance ones can not find the NTOS kernel. The Safe screen works on most of them.. Five types of CD-R's and CD-RW are used. All readable on various OS's.
It's failing for me; I have 2 Jira reports CORE-9022, which fails before first stage setup starts and CORE-9041, where a Livecd fails in Screen mode, but boots in the other modes.
Last edited by oldman on Wed Jan 14, 2015 5:12 pm, edited 1 time in total.
Please keep the Windows classic (9x/2000) look and feel.
The layman's guides to - debugging - bug reporting - compiling - with some complementary scripts.
They may help you with a problem, so do have a look at them.

Webunny
Posts: 1201
Joined: Sat Apr 28, 2012 1:30 pm

Re: major regression

Post by Webunny » Sat Jan 10, 2015 7:25 pm

I'm actually wondering why it is that so many times so many regressions slip through. Especially on real hardware. Is it, because the devs have no RHW to test it on? Wouldn't it be better, then, to have some hardware to test it on for themselves?

In principle, I could donate my PC where ROS works on, for instance...

Postby EmuandCo » 10 Jan 2015 23:29
You really want your last warning, huh? How about commenting the maybe already existant report in Jira, where it belongs instead of hijacking all active threads??!
And how about YOU maybe looking first at that JIRA report before you make a nonsensical remark like that? Maybe then you would have noted that I DID already make a comment to the existent report!! Typical: first react, then see if it makes sense. It's highly ironic you chastise me for not making a comment there, while I did. May I suggest to first check the facts, before you make a (now obviously silly sounding) comment like that?

And for your information: *what* 'hijacking the threads'?! I MADE this thread, and the last one where I wrote the comment (and which you locked for no reason at all, even though it was in the 'off topic' section). How can I hijack my own thread??! Stop making ridiculous claims, please.

So? Am I saying something untrue, here? Is it now already against the rules to give an opinion?

It seems to me, some people don't have their priorities right. Instead of doing something against (the root/cause of) the spamposts we constantly get, it is deemed more fruitful to threaten to ban someone who point the problem out AND gives a link to a thread with worthwhile suggestions. And in stead of being grateful that I test builds and point out regressions, so one can work to fix them, it is deemed more important to threaten to ban the tester because he posts that fixing it is what one really should care about, not threatening to ban people who are trying to help. And yes, I am helping. Or did you miss the part above where I offer to donate my ROS-PC? But I am getting tired of this ban-threatening. Keep this up, and I will stop feeling inclined to help whatsoever. And rest assured when I get banned I will also stop testing ROS, donating to ROS, promoting ROS-articles to Dutch It-magazines like I did in the past, etc. You act as if it's only a one-way street. Well, it isn't.

Just get your act together, and concentrate on the things you should be concentrating on, instead of threating to ban and asking me to do things I already did but you didn't even bother to check.

MadWolf
Posts: 540
Joined: Sat Dec 31, 2005 4:19 am
Contact:

Re: major regression

Post by MadWolf » Tue Jan 13, 2015 7:04 am

hi
Webunny do you remember what revision that worked your computer ??

jimtabor
Developer
Posts: 223
Joined: Thu Sep 29, 2005 3:00 pm

Re: major regression

Post by jimtabor » Wed Jan 14, 2015 2:19 pm

Hi all,
I need to keep up with the hardware testing each week now....

The thing I will try tonight, up to revision 65904 then go into usetup and up to head and retest it on hardware. I'll report the results in CORE-9032.

Be patient...

Webunny
Posts: 1201
Joined: Sat Apr 28, 2012 1:30 pm

Re: major regression

Post by Webunny » Wed Jan 14, 2015 7:14 pm

MadWolf wrote:hi
Webunny do you remember what revision that worked your computer ??
Good question. I re-burned it without checking, and I'm not sure if it was the latest anyway, since I've been regression-testing before and have reverted to older builds to pinpoint THAT regression. But I do think I've mentioned the builds in my former JIRA bug-report, so I'll check there. For sure it was several weeks old, though.

Edit:

Found it; https://jira.reactos.org/browse/CORE-8726 . Which is also still not resolved. (Or maybe it is, but I can't check. But does not seem from the bugreport there on JIRA anyone aside some devs I personally contacted on IRC has made an effort fixing it). Sometimes I have the feeling JIRA is pretty useless, in getting things noted and fixed. I've said it before, and I'll say it again,: there should be a more structured approach to it. Why not appoint a dev to check at least weekly all the bugs that are flagged as 'blocker' and 'major', and to respond to it, if only a cursory 'we'll look into it' (but preferable asking questions or suggesting things to try out or fixing it) and put a 'assignee' to it. Now, it's like bugreports fall into a black hole for weeks, months, years or forever.

I see I tried out 64221 and 64222, but I mention there I checked up up to 65000 and RC4. So I'll guess it would be around there that my highest number in daily builds would be at.

zydon
Posts: 160
Joined: Tue Dec 18, 2007 9:03 am

Re: major regression

Post by zydon » Thu Jan 15, 2015 11:44 am

Anything does not boot under VPC2007 for WinXP or XP VM will not boot on most of the real hardware. Even VPC2007 for Win7 is already breaking the real hardware compatibility. So, it need to satisfy VPC2007 for WinXP boot requirement to support the past, the present and possibly the future of x86 hardware boot loader.

Z98
Release Engineer
Posts: 3379
Joined: Tue May 02, 2006 8:16 pm
Contact:

Re: major regression

Post by Z98 » Thu Jan 15, 2015 5:20 pm

As has been iterated many times in the past, VPC is one of the least 'accurate' virtual machine platforms out there. It intentionally shortcuts certain things and is tuned to work specifically with Windows platforms, not act as a generic OS vm platform. For that reason alone, VPC support is inconsequential to actual hw support.

JHKr0md1yk
Posts: 29
Joined: Tue Sep 23, 2014 5:58 pm

Re: major regression

Post by JHKr0md1yk » Thu Jan 15, 2015 6:26 pm

Z98 wrote:As has been iterated many times in the past, VPC is one of the least 'accurate' virtual machine platforms out there. It intentionally shortcuts certain things and is tuned to work specifically with Windows platforms, not act as a generic OS vm platform. For that reason alone, VPC support is inconsequential to actual hw support.
How does Hyper-V, especially Hyper-V v3, fit into that equation wrt hardware faithfulness (is that even a word???) ?

Slightly offtopic: Yes, the server donation is still in force. I'm patiently waiting for the components for the HSW-EP rig to arrive at my home, and with a bit of luck I can send out pHouse with 96GB RAM to boot then.

zydon
Posts: 160
Joined: Tue Dec 18, 2007 9:03 am

Re: major regression

Post by zydon » Thu Jan 15, 2015 8:09 pm

VPC2007 for WinXP give me a peace of mind by burning bootable ISOs that only work for the real hardware to blank CDs. It save me time and avoid wasted CD blanks.

Just don't tell Linuxes Distro makers about this secret or you'll be surprised when suddenly you heard old hardware come back alive running various tiny Linux Oses with a blazing speed!

Webunny
Posts: 1201
Joined: Sat Apr 28, 2012 1:30 pm

Re: major regression

Post by Webunny » Thu Jan 15, 2015 11:58 pm

Well, VPC or not, it has always worked on my HW for the last two years, and now it's not. This means real hardware testers can't test the builds anymore, and what's worse, it actually means real hardware can't run it anymore, which is the whole point for ROS, in the end. It's not some minor issue, thus.

Post Reply

Who is online

Users browsing this forum: No registered users and 18 guests