Re: Random Compiler Errors, again

From: jon morgan (jonrmorgan_at_ntlworld.com)
Date: 01/26/05 

Date: Wed, 26 Jan 2005 10:23:49 -0000

Thanks Charles - I'll give it a try. Did they produce a patch to fix our
problem or were you just told "OK - it's a bug we'll fix it for the next
release " ?

Jon

"Charles Law" <blank@nowhere.com> wrote in message
news:OKVA504AFHA.3368@TK2MSFTNGP10.phx.gbl...
>I submitted a support request via their web site. The form asked for my
>product s/n and they did the rest. In terms of response to the initial
>request I was very impressed.
>
> In fact, I have submitted several support requests this way and every one
> has been responded to promptly, and everyone resulted in a resolution to
> my satisfaction (makes me sound like a company man). In one case, they
> produced a special build of mshtml.dll to resolve a bug.
>
> In this particular case here, it took some time, but only because they
> explored all the possible avenues thoroughly before concluding that it was
> indeed caused by a known problem. I had a three-way conference with
> someone in Europe and someone in the states using MS Office Live Meeting
> so that everyone could see what was happening, live, on my PC. I
> reproduced the problem, and saw it go away when I changed my references
> from dll references to project references.
>
> I have to say that I have been very happy with the support, but I know it
> hasn't always been that way.
>
> Give the on-line support a try; it worked for me.
>
> Good luck.
>
> Charles
>
>
> "jon morgan" <jonrmorgan@ntlworld.com> wrote in message
> news:%23M8Gjw3AFHA.1188@tk2msftngp13.phx.gbl...
>> Thanks Charles - our experiences do seem pretty similar although the
>> action you took to remedy it was more drastic: maybe your errors would
>> have disappeared as well if you had simply closed then re-opened VS.
>> Either way the clear conclusion does seem to be the the VS.Net 1.1
>> compiler does generate random errors.
>>
>> The question is what to do about it because this post is clearly not
>> attracting MS or MSVP interest. Always difficult of course to prove a bug
>> when it can't be reproduced by third parties.
>>
>> How did you get MS' attention ? My experience with their help lines
>> reminded me of the old days when IBM ruled the land. Basically you got a
>> pat on the head and a lot of beareaucratic woffle and that was it.
>> Obviously you were more persistent.
>>
>> Jon
>>
>> "Charles Law" <blank@nowhere.com> wrote in message
>> news:udZAp%23yAFHA.2428@TK2MSFTNGP14.phx.gbl...
>>> That was certainly one of the symptoms. A line of code would be
>>> underlined as having an error, where the error might have been that
>>> function signatures did not match, when they clearly did. I would close
>>> VS, delete the bin and obj directories of all the projects and restart.
>>> The problem would go away until a while later when some innocuous change
>>> would cause it again, or a similar problem. The supposed error was
>>> always unrelated to the change.
>>>
>>> Incidentally, I also got problems that I believe were caused by
>>> VBCommenter. From time to time, I would build only to find that there
>>> were dozens of errors. They were caused by code fragments being dumped
>>> at the end of files. I would delete them and the problems went away. I
>>> now steer clear of third-party add-ins, especially free ones!
>>>
>>> Actually, now I read the symptoms you are experiencing again, perhaps
>>> mine were not exactly the same. I don't think that I ever had problems
>>> where I could remove an offending line, paste it back in and the problem
>>> would go away. I did, however, get symptoms where it looked like the
>>> compiler could not locate some code, i.e. it complained that something
>>> was not defined, but I had to use the technique above to get it to right
>>> itself.
>>>
>>> Charles
>>>
>>>
>>> "jon morgan" <jonrmorgan@ntlworld.com> wrote in message
>>> news:umYjt8xAFHA.3700@tk2msftngp13.phx.gbl...
>>>> Thanks for getting the debate going Charles.
>>>>
>>>> I usually keep all the relevant projects together in the solution
>>>> rather than use dll references. Otherwise I just get tied up in knots
>>>> trying to remember which project holds what references. That way, I
>>>> think VS does a very good job of keeping things in sync and I'm not
>>>> surprised MS recommend it.
>>>>
>>>> However did you get the same symptoms compiler error symptoms as I
>>>> reported: ie apparently missing lines of code causing compiler errors
>>>> when the code is really there ?
>>>>
>>>> Jon
>>>>
>>>>
>>>> "Charles Law" <blank@nowhere.com> wrote in message
>>>> news:uS8KhhwAFHA.3824@TK2MSFTNGP10.phx.gbl...
>>>>> Hi Jon
>>>>>
>>>>> You are right. There is a bug in VS.NET.
>>>>>
>>>>> The most common way that I have encountered this is, in a
>>>>> multi-project solution, where references between the projects are made
>>>>> by browsing to the output dll as opposed to selecting it from the
>>>>> Projects tab.
>>>>>
>>>>> Apparently, VS.NET gets 'confused' by the existence of both a dll and
>>>>> project by the same name. I spent some weeks going through this with
>>>>> MS whereupon they admitted that this was an issue. Their
>>>>> recommendation was to always add references from the Projects tab.
>>>>>
>>>>> HTH
>>>>>
>>>>> Charles
>>>>>
>>>>>
>>>>> "jon morgan" <jonrmorgan@ntlworld.com> wrote in message
>>>>> news:OF2UqbwAFHA.1452@TK2MSFTNGP11.phx.gbl...
>>>>>> OK, I'm going to be brave. There is a bug in VS.Net 1.1 that causes
>>>>>> random compiler errors.
>>>>>>
>>>>>> I have raised this issue in posts at least three time in the past
>>>>>> couple of months without attracting much interest. But it's driving
>>>>>> me nuts. Here's what happens. I'm working on a multi project VB app.
>>>>>> happily writing nice inoffensive code - go to compile and the
>>>>>> compiler tells me there's a problem in a project I'm not working on.
>>>>>> But really there isn't - I can simply exit VS and boot in again
>>>>>> immediately and bingo, the compiler errors are gone and all is well
>>>>>> again for a while. After some time I get a different error or maybe
>>>>>> the same one and so it goes on.
>>>>>>
>>>>>> It's very obvious what causes the compiler this angst: for some
>>>>>> reason it convinces itself that one or two lines of code are
>>>>>> missing - could be a field, and event definition, a method title
>>>>>> whatever. But really the code is there fine all the time - if I
>>>>>> delete it and immediately reinstate it the compiler now "sees" the
>>>>>> code and the errors disappear. Weird or what ...
>>>>>>
>>>>>> I'm quite prepared to believe that this is all down to my aggressive
>>>>>> coding style but surely I can't be the first person to have
>>>>>> experienced this. So come on MS wizzos - what's happening...
>>>>>>
>>>>>> Your help is appreciated
>>>>>>
>>>>>> Jon
>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>
>>>
>>
>>
>
>

Relevant Pages

  • Re: Member Properties
    ... > Have you seen any references to this bug and a possible fix coming soon? ... this is the first time I see something like this. ...
    (microsoft.public.sqlserver.olap)
  • [Full-Disclosure] Solaris syslogd overflow
    ... Fix: ... Timeline: ... References: ... If anyone else is able to do anything more interesting with this bug, ...
    (Full-Disclosure)
  • Re: Member Properties
    ... No the table is fully normalized and each customer only occurs once. ... Have you seen any references to this bug and a possible fix coming soon? ...
    (microsoft.public.sqlserver.olap)
  • Solaris syslogd overflow
    ... Fix: ... Timeline: ... References: ... If anyone else is able to do anything more interesting with this bug, ...
    (Bugtraq)
  • [Un] Unangband 0.6.3 released
    ... Allow player to assemble friendly monsters and carry eggs to hatch ... Updated druidic spells to use new region code. ... Fix lockup bugs generating the Old Forest. ... Fix bug where items dropped by monster death would infinitely ...
    (rec.games.roguelike.announce)