Jim,
My experiance with Numega product start with 1993 edition of Bounce checker.
If I dig my garbages, I might find a copy of the version 1.0. Then started
using SoftIce, at around 94/95. Maybe you are using it pror to my start with
softice. But do you remember when softice used to work with only a few
graphics card?. And one has to juggle many Graphics card to get one that
works, and never dare to touch the machine for other graphics card!
People had a huge complain, and they cameup with much better user
experiances. At that time, if I could recall there was no compition ( or
hardly any ) for that kind of symbolic debugger for kernel mode debugging. I
never had the opportunity to become a master user, since it is another tool.
No matter what ( no compition for windbg, big fat company, blah blah) Windbg
does not have much choices anymore. Yes it is free ( so is file system kit
(ifs kit) ), but people would have complain. As long as there is thrust from
third parties, there would be improvements. Would it be slow or fast
improvements, that I don’t know. And now Windbg is more a part of WDK than
anything else. The whole WDK is now free, does it mean the quality of it
going to go down?. No, not so easily. For every new drops of iso, there are
some associated risk, some hiccups but overall the quality would improve
after some fixes etc.
Finally, I’ve worked for MS. And I know that their code quality is much
better than lot of others I’ve seen in my life. Only a few other place where
code was done more thoroughly, but the product turn around time used to be
5yrs at least. Some of the codes I had to debug (written by others) were so
badly written that every single day I thought I would not showup next day
… And they are not from Microsoft. These are the companies for
quick-bucks. If I tell anyone that code was not very clean, that day could
very well have been my last day. And I’ve seen lots of them. — This is as
much of an unbiased rant as I could tell.
And the funny thing is that “It is not very enticing to be on the receiving
end…”. I’ve not touched a single line of source in KD and / or Windbg, so
one can imagine that I’ve no sentiment attached to it. So when something
does not work, I just say that in a plain and flat tone … so that on the
otherside of the fence they would not feel like they are on the receiving
end .
-pro
On 5/17/07, jim wrote:
>
> Pro,
> Not sure what you disagree with, but my comment concerning Vista and MS
> cooperation with Compuware is not pure
> conjecture.
>
> As for what the devorg is willing to pay is up for debate. I guess you
> need to be willing to pay the price that is asked, or
> the product will not live long. Go price an Agilent Network Analyzer -
> lets see used and 5 years old they clock in at 30K.
>
> 3500 was a fair price for the product, in fact a cheap price (assuming it
> all worked).
> Unless MS is decides to start charging for Wndbg, don’t expect too much
> improvement.
> It’s a small market - and there is no competition.
>
> “Prokash Sinha” wrote in message news:xxxxx@windbg…
> I don’t agree on this. I heard the same story about Borland’s C++
> compilers ( now most of the good ones are in MS campus). Same with quite a
> few of the NuMega guys!. When I asked few of my friends those who were there
> in Borland then, the story was different. It is lack of positioning.
>
> But agree with the fact that most of the dev org is willing to pay the
> right price for right product. Softice should have been sold as a standalone
> product. Whether they had problems with vista release or not is just a
> matter of opinion. MS owns OS so anyone should have the conformiblity and
> work with them. But then again I’m not sure whether those softice guy needed
> that much help or not, since they seem to go out to very deep level by
> themselves… without much help and go thru source code !!
>
> Windbg would gradually be better, since there are more efforts now than in
> the past.
>
> -pro
>
>
>
>
> On 5/17/07, Bill McKenzie wrote:
> >
> > Actually, WinDbg killed softice. Well that and an incredibly stupid
> > parent company. I have been through this twice now. Once with
> > BSquare/BlueWater and once with Compuware/NuMega. It goes something like
> > this:
> >
> > Idiotic manager(s) at big fat stupid parent company look at successful
> > kernel product company and go oh gee golly sure would like to have that
> > product revenue. And ooh their numbers look so good. > > successful company commences> Hmmm…they spend money on advertising…I
> > don’t think they need that. Hmmm why would this little company need it’s
> > own website, we can shuttle that. Hmmm…why do they need any of the
> > hundreds other things, that EVERY successful company needs, to sell these
> > cute little Windows kernel products? We don’t need any of that. > > later> See, the kernel group is still making money…told you they didn’t
> > need any of that silly business stuff. Hmmm but their growth isn’t what we
> > expected and their good people are leaving. We should cut them…apparently
> > they weren’t ever really going to be successful.
> >
> > I like to criticize these morons that run these huge idiotic
> > organizations, but they drive nicer cars than I do and live in nicer houses
> > than I do. I have never had the opportunity to complain about my yacht
> > spilling diesel fuel all over Puget Sound. So, who is the moron? If I only
> > I could get past the sale of this pesky soul I keep carrying around.
> >
> > BTW, people would be MORE than willing to pay for a good debugger if
> > WinDbg wasn’t free. You simply cannot compete with free period. That is
> > the sad irony of all of this. KD is an extremely strong debugging engine.
> > WinDbg is one of the worst software products I have ever used.
> >
> > >Plus softice did way more than windbg
> >
> > I would like you to explain that. That may have been true at one
> > time…I really don’t think so anymore. Aside from single system kernel
> > debug which is both a REALLY bad idea and a moot point with VMs now.
> >
> > Bill M.
> >
> > “jim” wrote in message news:xxxxx@windbg…
> > Actually, we killed softice. I don’t think you realize the large
> > investment in maintaining a program like driver studio.
> > Microsoft has the source code, and look at the problems they have. Plus
> > softice did way more than windbg.
> > I also heard they could not get enough support from MS regard Vista
> > kernel changes. They did a few passes
> > on Vista betas to no avail - things kept changing. Imagine the skill
> > level required for engineers to work on softice.
> >
> > There simply was not a good business case for it.
> > Not enough of us were will to fork out 3500.00 per seat.
> >
> > BTW if you want to buy the product you can. Then you can give us softice
> > back 
> >
> >
> > “Prokash Sinha” wrote in message news:xxxxx@windbg
> > …
> > No, unfortunately your point is not low. Worse yet is to loose the
> > client. I’m not that lucky to keep my
> > domain of debugging to just kernel mode. So obviously I try my best to
> > get the usr/krnl debugging in one session. I particulary don’t see why that
> > path of debugging approach is so winding, specially when softice did that
> > for years.
> >
> > My suggestion to anyone loves Windbg is to try debugging service with
> > couple drivers ( preferabley fs filters one or two and any fsd or in the
> > ndis path). Try to step thru usr/krnl and krnl/usr at the same time. Spent
> > about a month of time if you are in the business for 10+ year, othewise
> > don’t even try.
> >
> > And finally, that compuware sucks big time. They are the architect to
> > kill softice.
> >
> > What I do when this kind of wait happens to resolve symbols
> > unneccearily? Recent analysis shows that if you work in front of a computer
> > “For every 20 minutes of work, you should look at a 20ft distance for half a
> > minute to a minute”. That way get the time to whine while you take good care
> > of your eyes!. So windbg is far more ergoeconomic than you and I can even
> > imagine !!!
> >
> > -pro
> >
> > On 5/17/07, Bill McKenzie < xxxxx@sbcglobal.net> wrote:
> > >
> > > Well, I tried that before and then it wouldn’t load jack and my
> > > debugging
> > > got all screwed. It’s like an all or nothing problem. I want
> > > unqualified
> > > symbol loads, I guess. But, I also want the debugger to be smart
> > > enough to
> > > NOT hit the network every single time I run it. I mean, maybe I am
> > > missing
> > > something…but my target machine has not changed in months. My
> > > driver’s
> > > symbols are certainly local. So, what the HELL is it looking
> > > for?? It
> > > already has all the symbols it needs right here. At least I hope so,
> > > or all
> > > of these symbols chewing up my harddrive space are for naught.
> > >
> > > I am hoping that if !sym noisy were to get turned on automagically,
> > > then I
> > > could come back here and ask more directed questions like “why the
> > > hell is
> > > WinDbg always going on the network to look for XYZ.syssymbols???” Follow?
> > > I know I am not the only one having this issue…everyone I see use
> > > this
> > > debugger ends up in these lovely wait marathons. I guess my pain
> > > point if
> > > just lower or something.
> > >
> > > Bill M.
> > >
> > > “Pavel Lebedinsky” wrote in message
> > > news:xxxxx@windbg…
> > > > “Bill McKenzie” wrote:
> > > >
> > > >> So, I guess I am back to my original feature request. It would
> > > sure be
> > > >> nice to know what symbols this thing can’t find after booting 500
> > > times
> > > >> with the MSFT symbol server set.
> > > >
> > > >
> > > > Being able to see what symbols it can’t find would be nice, but
> > > > the real question is why this symbol search is happening in the
> > > first
> > > > place.
> > > >
> > > > Do you have unqualified symbol loads turned on? If yes, try turning
> > > it
> > > > off, because this is probably the root of the problem.
> > > >
> > > > –
> > > > This posting is provided “AS IS” with no warranties, and confers no
> > > > rights.
> > > >
> > > >
> > >
> > >
> > >
> > > —
> > > You are currently subscribed to windbg as: xxxxx@gmail.com
> > > To unsubscribe send a blank email to xxxxx@lists.osr.com
> > >
> >
> >
> > —
> > You are currently subscribed to windbg as: unknown lmsubst tag argument:
> > ‘’
> > To unsubscribe send a blank email to xxxxx@lists.osr.com
> >
>
>
> —
> You are currently subscribed to windbg as: unknown lmsubst tag argument:
> ‘’
> To unsubscribe send a blank email to xxxxx@lists.osr.com
>