New Project Strategy: develop in 1.1 then migrate to 2.0 just before going live

[Jeff]

Looking for your highly subjective opinions (from those of you who have been
working extensively with VS 2005).

I have a client that needs/wants a complete rewrite of an existing
application. I will be proposing a .NET Windows application with a
supporting SQL Server database. It will likely take 4-5 months before we
roll out the first production version.

Because of the apparent issues with VS 2005 (poor performance, unpredictable
and generally buggy behaviour), I am thinking it might make sense to develop
this app in VS.NET 2003/.NET 1.1. Then, around the time we're ready to roll
out the first production version, migrate it to .NET 2.x just before going
live.

I say 2.x because Microsoft has plans to release a service pack for it
mid-way through 2006, so we might have a minor version number change at that
time. Also, it appears that the service pack from Microsoft, if released
mid-year, will stabilize things so we can [hopefully] have a smooth
transition to 2.x just before going live.

Does this strategy make sense? Thoughts? Opinions?

Thanks!

 

[Stefan Nuxoll]

Looking for your highly subjective opinions (from those of you who have been
working extensively with VS 2005).

I have a client that needs/wants a complete rewrite of an existing
application. I will be proposing a .NET Windows application with a
supporting SQL Server database. It will likely take 4-5 months before we
roll out the first production version.

Because of the apparent issues with VS 2005 (poor performance, unpredictable
and generally buggy behaviour), I am thinking it might make sense to develop
this app in VS.NET 2003/.NET 1.1. Then, around the time we're ready to roll
out the first production version, migrate it to .NET 2.x just before going
live.

I say 2.x because Microsoft has plans to release a service pack for it
mid-way through 2006, so we might have a minor version number change at that
time. Also, it appears that the service pack from Microsoft, if released
mid-year, will stabilize things so we can [hopefully] have a smooth
transition to 2.x just before going live.

Does this strategy make sense? Thoughts? Opinions?

Thanks!

This makes some sense, however doing something in this fashion is not
recommended as some things may go haywire during the conversion process
that you would need to fix. I find it easier just to deal with VS 2005
and make it with .NET 2.0 to start.

 

[Nicholas Paldino [.NET/C# MVP]]

Not only that, but you are pretty much going to miss out on a good
amount of functionality which VS.NET 2005 provides.

Have you actually tried it? I ask because you say apparent...

--
- Nicholas Paldino [.NET/C# MVP]
- (e-mail address removed)

Looking for your highly subjective opinions (from those of you who have
been working extensively with VS 2005).

I have a client that needs/wants a complete rewrite of an existing
application. I will be proposing a .NET Windows application with a
supporting SQL Server database. It will likely take 4-5 months before we
roll out the first production version.

Because of the apparent issues with VS 2005 (poor performance,
unpredictable and generally buggy behaviour), I am thinking it might make
sense to develop this app in VS.NET 2003/.NET 1.1. Then, around the time
we're ready to roll out the first production version, migrate it to .NET
2.x just before going live.

I say 2.x because Microsoft has plans to release a service pack for it
mid-way through 2006, so we might have a minor version number change at
that time. Also, it appears that the service pack from Microsoft, if
released mid-year, will stabilize things so we can [hopefully] have a
smooth transition to 2.x just before going live.

Does this strategy make sense? Thoughts? Opinions?

Thanks!

This makes some sense, however doing something in this fashion is not
recommended as some things may go haywire during the conversion process
that you would need to fix. I find it easier just to deal with VS 2005
and make it with .NET 2.0 to start.

 

[Jeff]

RE:
<< Have you actually tried it?... >>

No, because I have work to do. And no, I'm not flaming or being negative.
I'm a *huge* fan of Microsoft. I also have a lot of time constraints and
simply don't have the time to screw around with a product which, given all
of the chatter in the varous Microsoft NGs, was apparently rushed to market
(or at least not fully developed when released). It appears that one one
must be willing to deal with slow performance, random "freezing", machine
crashes, etc. Given the volume of reports I think it's reasonable to
conclude that I would likely experience similar symptoms.

So while I'm looking to play it safe I also don't want to be obviously and
unnecessarily cautious. So I guess I'd be swayed to go with VS 2005 right
away if a number of people responded to my post here claiming that they were
totaly happy with VS 2005 and that they *never* experience any of the
frequently reported "buggy" behaviour.

-Jeff

Not only that, but you are pretty much going to miss out on a good
amount of functionality which VS.NET 2005 provides.

Have you actually tried it? I ask because you say apparent...

--
- Nicholas Paldino [.NET/C# MVP]
- (e-mail address removed)

Looking for your highly subjective opinions (from those of you who have
been working extensively with VS 2005).

I have a client that needs/wants a complete rewrite of an existing
application. I will be proposing a .NET Windows application with a
supporting SQL Server database. It will likely take 4-5 months before we
roll out the first production version.

Because of the apparent issues with VS 2005 (poor performance,
unpredictable and generally buggy behaviour), I am thinking it might
make sense to develop this app in VS.NET 2003/.NET 1.1. Then, around the
time we're ready to roll out the first production version, migrate it to
.NET 2.x just before going live.

I say 2.x because Microsoft has plans to release a service pack for it
mid-way through 2006, so we might have a minor version number change at
that time. Also, it appears that the service pack from Microsoft, if
released mid-year, will stabilize things so we can [hopefully] have a
smooth transition to 2.x just before going live.

Does this strategy make sense? Thoughts? Opinions?

Thanks!

This makes some sense, however doing something in this fashion is not
recommended as some things may go haywire during the conversion process
that you would need to fix. I find it easier just to deal with VS 2005
and make it with .NET 2.0 to start.

 

[Gabriel Magana]

Well, I'm not gonna give you the free lunch you seek (you have to try
VS2005 out for yourself), but I can say that the bugs I have run into are
worth it for the many benefits that VS2005 brings.

It's not our job to "sway" you... I'm sure there are people still using VS6
because there are too many bugs in subsequent releases, but it is their
choice... It's more of a cost/benefit analysis you have to do according to
your own values/tolerance/etc...

For example, just yesterday I fixed a long-standing layout bug in one of our
aspx pages just because of the collapsible tags in the VS2005 HTML editor.
The collapsible tags feature, in conjunction with the new integrated HTML
code formatter, allowed me to greatly simplify how the HTML was displayed,
making the syntax error very simple to catch. This was an error that I had
spent searching for at least a full day with VS2003... But the VS2005 came
with some bugs, including some pretty big ones when it comes to deployment
of web sites and also a very annoying bug in which VS2005 checks into source
control files under the /bin directory when clearly, nothing in that folder
should be checked into source control. Is it worth the upgrade?
Absolutely, even through there are bugs, the benefits hugely outweight the
annoyances for me.

You have to make a similar evaluation and decision, based on your needs.

RE:
<< Have you actually tried it?... >>

No, because I have work to do. And no, I'm not flaming or being negative.
I'm a *huge* fan of Microsoft. I also have a lot of time constraints and
simply don't have the time to screw around with a product which, given all
of the chatter in the varous Microsoft NGs, was apparently rushed to
market (or at least not fully developed when released). It appears that
one one must be willing to deal with slow performance, random "freezing",
machine crashes, etc. Given the volume of reports I think it's reasonable
to conclude that I would likely experience similar symptoms.

So while I'm looking to play it safe I also don't want to be obviously and
unnecessarily cautious. So I guess I'd be swayed to go with VS 2005 right
away if a number of people responded to my post here claiming that they
were totaly happy with VS 2005 and that they *never* experience any of the
frequently reported "buggy" behaviour.

-Jeff

Not only that, but you are pretty much going to miss out on a good
amount of functionality which VS.NET 2005 provides.

Have you actually tried it? I ask because you say apparent...

--
- Nicholas Paldino [.NET/C# MVP]
- (e-mail address removed)

Jeff wrote:
Looking for your highly subjective opinions (from those of you who have
been working extensively with VS 2005).

I have a client that needs/wants a complete rewrite of an existing
application. I will be proposing a .NET Windows application with a
supporting SQL Server database. It will likely take 4-5 months before
we roll out the first production version.

Because of the apparent issues with VS 2005 (poor performance,
unpredictable and generally buggy behaviour), I am thinking it might
make sense to develop this app in VS.NET 2003/.NET 1.1. Then, around
the time we're ready to roll out the first production version, migrate
it to .NET 2.x just before going live.

I say 2.x because Microsoft has plans to release a service pack for it
mid-way through 2006, so we might have a minor version number change at
that time. Also, it appears that the service pack from Microsoft, if
released mid-year, will stabilize things so we can [hopefully] have a
smooth transition to 2.x just before going live.

Does this strategy make sense? Thoughts? Opinions?

Thanks!
This makes some sense, however doing something in this fashion is not
recommended as some things may go haywire during the conversion process
that you would need to fix. I find it easier just to deal with VS 2005
and make it with .NET 2.0 to start.

 

[Jeff]

Re:
<< It's not our job to "sway" you >>

Well, yes, of course not. But I'm looking to learn from some of you who have
*not* been complaining. Maybe you have not observed the bugs - maybe you
just aren't saying anything.

Well, I'm not gonna give you the free lunch you seek (you have to try
VS2005 out for yourself), but I can say that the bugs I have run into are
worth it for the many benefits that VS2005 brings.

It's not our job to "sway" you... I'm sure there are people still using
VS6 because there are too many bugs in subsequent releases, but it is
their choice... It's more of a cost/benefit analysis you have to do
according to your own values/tolerance/etc...

For example, just yesterday I fixed a long-standing layout bug in one of
our aspx pages just because of the collapsible tags in the VS2005 HTML
editor. The collapsible tags feature, in conjunction with the new
integrated HTML code formatter, allowed me to greatly simplify how the
HTML was displayed, making the syntax error very simple to catch. This
was an error that I had spent searching for at least a full day with
VS2003... But the VS2005 came with some bugs, including some pretty big
ones when it comes to deployment of web sites and also a very annoying bug
in which VS2005 checks into source control files under the /bin directory
when clearly, nothing in that folder should be checked into source
control. Is it worth the upgrade? Absolutely, even through there are
bugs, the benefits hugely outweight the annoyances for me.

You have to make a similar evaluation and decision, based on your needs.

RE:
<< Have you actually tried it?... >>

No, because I have work to do. And no, I'm not flaming or being negative.
I'm a *huge* fan of Microsoft. I also have a lot of time constraints and
simply don't have the time to screw around with a product which, given
all of the chatter in the varous Microsoft NGs, was apparently rushed to
market (or at least not fully developed when released). It appears that
one one must be willing to deal with slow performance, random
"freezing", machine crashes, etc. Given the volume of reports I think
it's reasonable to conclude that I would likely experience similar
symptoms.

So while I'm looking to play it safe I also don't want to be obviously
and unnecessarily cautious. So I guess I'd be swayed to go with VS 2005
right away if a number of people responded to my post here claiming that
they were totaly happy with VS 2005 and that they *never* experience any
of the frequently reported "buggy" behaviour.

-Jeff

Not only that, but you are pretty much going to miss out on a good
amount of functionality which VS.NET 2005 provides.

Have you actually tried it? I ask because you say apparent...

--
- Nicholas Paldino [.NET/C# MVP]
- (e-mail address removed)

Jeff wrote:
Looking for your highly subjective opinions (from those of you who
have been working extensively with VS 2005).

I have a client that needs/wants a complete rewrite of an existing
application. I will be proposing a .NET Windows application with a
supporting SQL Server database. It will likely take 4-5 months before
we roll out the first production version.

Because of the apparent issues with VS 2005 (poor performance,
unpredictable and generally buggy behaviour), I am thinking it might
make sense to develop this app in VS.NET 2003/.NET 1.1. Then, around
the time we're ready to roll out the first production version, migrate
it to .NET 2.x just before going live.

I say 2.x because Microsoft has plans to release a service pack for it
mid-way through 2006, so we might have a minor version number change
at that time. Also, it appears that the service pack from Microsoft,
if released mid-year, will stabilize things so we can [hopefully] have
a smooth transition to 2.x just before going live.

Does this strategy make sense? Thoughts? Opinions?

Thanks!
This makes some sense, however doing something in this fashion is not
recommended as some things may go haywire during the conversion process
that you would need to fix. I find it easier just to deal with VS 2005
and make it with .NET 2.0 to start.

 

[Jon Skeet [C# MVP]]

Looking for your highly subjective opinions (from those of you who have been
working extensively with VS 2005).

I have a client that needs/wants a complete rewrite of an existing
application. I will be proposing a .NET Windows application with a
supporting SQL Server database. It will likely take 4-5 months before we
roll out the first production version.

Because of the apparent issues with VS 2005 (poor performance, unpredictable
and generally buggy behaviour), I am thinking it might make sense to develop
this app in VS.NET 2003/.NET 1.1. Then, around the time we're ready to roll
out the first production version, migrate it to .NET 2.x just before going
live.

That sounds like a very bad idea to me. At the very least you should be
trying to run all your code on a box which only has 2.0 whenever you
run tests. There are quite a few tiny breaking changes which may well
not affect you - but just might...

You should always try to test in as close to the real environment as
posisble.

 

[Naraendira Kuma R.R.]

and that they *never* experience any of the frequently reported "buggy"Never? Really! I guess you intend to be waiting a long time.

On a more helpful note ... I see your point.

But, Stefan makes a good observation too. If you are really concerned &
decide to be cautious, you could probably be embrace something like xUnit
tests That way you could reduce the inherent risk that Stefan calls out.

Personally for me, VS.NET 2005 has been worth the hassles of switching.

--
Naraendira Kumar R.R.
~~~~~~~~~~~~~~~~~~~~~~~~~~~

RE:
<< Have you actually tried it?... >>

No, because I have work to do. And no, I'm not flaming or being negative.
I'm a *huge* fan of Microsoft. I also have a lot of time constraints and
simply don't have the time to screw around with a product which, given all
of the chatter in the varous Microsoft NGs, was apparently rushed to
market (or at least not fully developed when released). It appears that
one one must be willing to deal with slow performance, random "freezing",
machine crashes, etc. Given the volume of reports I think it's reasonable
to conclude that I would likely experience similar symptoms.

So while I'm looking to play it safe I also don't want to be obviously and
unnecessarily cautious. So I guess I'd be swayed to go with VS 2005 right
away if a number of people responded to my post here claiming that they
were totaly happy with VS 2005 and that they *never* experience any of the
frequently reported "buggy" behaviour.

-Jeff

Not only that, but you are pretty much going to miss out on a good
amount of functionality which VS.NET 2005 provides.

Have you actually tried it? I ask because you say apparent...

--
- Nicholas Paldino [.NET/C# MVP]
- (e-mail address removed)

Jeff wrote:
Looking for your highly subjective opinions (from those of you who have
been working extensively with VS 2005).

I have a client that needs/wants a complete rewrite of an existing
application. I will be proposing a .NET Windows application with a
supporting SQL Server database. It will likely take 4-5 months before
we roll out the first production version.

Because of the apparent issues with VS 2005 (poor performance,
unpredictable and generally buggy behaviour), I am thinking it might
make sense to develop this app in VS.NET 2003/.NET 1.1. Then, around
the time we're ready to roll out the first production version, migrate
it to .NET 2.x just before going live.

I say 2.x because Microsoft has plans to release a service pack for it
mid-way through 2006, so we might have a minor version number change at
that time. Also, it appears that the service pack from Microsoft, if
released mid-year, will stabilize things so we can [hopefully] have a
smooth transition to 2.x just before going live.

Does this strategy make sense? Thoughts? Opinions?

Thanks!
This makes some sense, however doing something in this fashion is not
recommended as some things may go haywire during the conversion process
that you would need to fix. I find it easier just to deal with VS 2005
and make it with .NET 2.0 to start.

 

[Rene]

Did you know that Microsoft did not issue a single service pack for VS 2003?
Did you know that even hot fixes required you to call Microsoft ($250 to
talk to an engineer) to be able to get the hot fixes? What makes you think
that VS 2003 has any less bugs than VS 2005?

My experience with VS 2005 has been a good one so far, its an incredible
developers tool much better than 2003 and worth every penny in my opinion.
Yes, it does have its problems, some very annoying but not more or less than
what VS 2003 has in my opinion. If I was you, I would workaround any
problems that VS 2005 and develop my product with VS 2005.

 

[Kevin Spencer]

Hi Jeff,

I've been using VS.Net 2005 for about a year now, since beta 1 I believe.
Beta versions had certain bugs in them, but were essentially stable. I have
never experienced any performance problems with it. I believe most of the
perceived performance problems were due to issues related to the uninstall
of previous beta versions. The upgrade process to the next beta (and finally
to the release version) has been highly complex, involving separate
uninstalls of various components in a certain order prior to installing the
newer version. If these procedures were not followed exactly, problems
occurred. Some of these problems manifested themselves as slow performance.

I have also seen quite a few more complaints about the new ASP.Net than
about other types of projects, such as Windows Forms, Services, and class
libraries. I have yet to do any ASP.Net work with the new Visual Studio, so
I can't say anything about that. But I have written quite a few other
projects, including Windows Forms, Services, and class libraries, without
any performance issues.

In addition, I have observed that VS.Net 2005 is far superior to VS.Net 2003
in many ways, too many to enumerate here, except to mention a few of the
more obvious improvements. Intellisense is much improved. Refactoring is a
real time-saver. Support for standards, in particular those involving XML is
incredible. And DTDs are used for error-checking and intellisense, which
makes a heck of a lot of sense.

I wouldn't go back to VS.Net 2003 for any reason, other than to maintain
legacy (1.1) applications.

--
HTH,

Kevin Spencer
Microsoft MVP
..Net Developer
You can lead a fish to a bicycle,
but it takes a very long time,
and the bicycle has to *want* to change.

Re:
<< It's not our job to "sway" you >>

Well, yes, of course not. But I'm looking to learn from some of you who
have *not* been complaining. Maybe you have not observed the bugs - maybe
you just aren't saying anything.

Well, I'm not gonna give you the free lunch you seek (you have to try
VS2005 out for yourself), but I can say that the bugs I have run into are
worth it for the many benefits that VS2005 brings.

It's not our job to "sway" you... I'm sure there are people still using
VS6 because there are too many bugs in subsequent releases, but it is
their choice... It's more of a cost/benefit analysis you have to do
according to your own values/tolerance/etc...

For example, just yesterday I fixed a long-standing layout bug in one of
our aspx pages just because of the collapsible tags in the VS2005 HTML
editor. The collapsible tags feature, in conjunction with the new
integrated HTML code formatter, allowed me to greatly simplify how the
HTML was displayed, making the syntax error very simple to catch. This
was an error that I had spent searching for at least a full day with
VS2003... But the VS2005 came with some bugs, including some pretty big
ones when it comes to deployment of web sites and also a very annoying
bug in which VS2005 checks into source control files under the /bin
directory when clearly, nothing in that folder should be checked into
source control. Is it worth the upgrade? Absolutely, even through there
are bugs, the benefits hugely outweight the annoyances for me.

You have to make a similar evaluation and decision, based on your needs.

RE:
<< Have you actually tried it?... >>

No, because I have work to do. And no, I'm not flaming or being
negative. I'm a *huge* fan of Microsoft. I also have a lot of time
constraints and simply don't have the time to screw around with a
product which, given all of the chatter in the varous Microsoft NGs, was
apparently rushed to market (or at least not fully developed when
released). It appears that one one must be willing to deal with slow
performance, random "freezing", machine crashes, etc. Given the volume
of reports I think it's reasonable to conclude that I would likely
experience similar symptoms.

So while I'm looking to play it safe I also don't want to be obviously
and unnecessarily cautious. So I guess I'd be swayed to go with VS 2005
right away if a number of people responded to my post here claiming that
they were totaly happy with VS 2005 and that they *never* experience any
of the frequently reported "buggy" behaviour.

-Jeff




in message Not only that, but you are pretty much going to miss out on a good
amount of functionality which VS.NET 2005 provides.

Have you actually tried it? I ask because you say apparent...

--
- Nicholas Paldino [.NET/C# MVP]
- (e-mail address removed)

Jeff wrote:
Looking for your highly subjective opinions (from those of you who
have been working extensively with VS 2005).

I have a client that needs/wants a complete rewrite of an existing
application. I will be proposing a .NET Windows application with a
supporting SQL Server database. It will likely take 4-5 months before
we roll out the first production version.

Because of the apparent issues with VS 2005 (poor performance,
unpredictable and generally buggy behaviour), I am thinking it might
make sense to develop this app in VS.NET 2003/.NET 1.1. Then, around
the time we're ready to roll out the first production version,
migrate it to .NET 2.x just before going live.

I say 2.x because Microsoft has plans to release a service pack for
it mid-way through 2006, so we might have a minor version number
change at that time. Also, it appears that the service pack from
Microsoft, if released mid-year, will stabilize things so we can
[hopefully] have a smooth transition to 2.x just before going live.

Does this strategy make sense? Thoughts? Opinions?

Thanks!
This makes some sense, however doing something in this fashion is not
recommended as some things may go haywire during the conversion
process that you would need to fix. I find it easier just to deal
with VS 2005 and make it with .NET 2.0 to start.

 

[Guest]

Jeff,
If as you say, your client wants a complete rewrite, then you will have
plenty of time to get used to VS.NET 2005's little nuances and take
adavantage of the features and performance enhancements of the 2.0 Framework
"from the git-go".

All IDE - type software has bugs, but that isn't always a sound basis for
not using it.

Peter

 

[Code Monkey]

Just my 2 pennies (or cents).

Visual Studio 2005 runs perfectly on my machine and I've only found a
couple of minor annoyances, but in general, in terms of the IDE and
performance - VS2005 wins all the way!

No strange machine crashes, no code reformatting (unlike VS2003) and
fair less IDE crashes and lost code than VS2003.

I'm not a Microsoft lover (I also own a Macintosh and a Linux box) -
but for web development - Visual Studio 2005 is the muts nuts!

 

[John Vottero]

RE:
<< Have you actually tried it?... >>

No, because I have work to do. And no, I'm not flaming or being negative.
I'm a *huge* fan of Microsoft. I also have a lot of time constraints and
simply don't have the time to screw around with a product which, given all
of the chatter in the varous Microsoft NGs, was apparently rushed to
market (or at least not fully developed when released). It appears that
one one must be willing to deal with slow performance, random "freezing",
machine crashes, etc. Given the volume of reports I think it's reasonable
to conclude that I would likely experience similar symptoms.

I have not had ANY problems like that with VS 2005. I don't think anything
in VS2005 runs in kernel mode so it couldn't crash the machine if it wanted
to. The only glitch that I've had is that sometimes it takes a while (30
seconds) to open a form in the designer. I think that's because the
application has over 250 forms and many user controls so the first time in
it has a fair amount of work to do.

I've seen plenty of little glitches with VS2003 but, it still managed to get
the job done. In my experiance, VS2005 is better and more reliable.

If you're doing a WinForms app you REALLY, REALLY want to go with VS2005
(REALLY!!). You get better controls, layout managers, better binding and a
seperation of generated code from your code. There's probably a lot more
that you get that I haven't seen yet.

So while I'm looking to play it safe I also don't want to be obviously and
unnecessarily cautious. So I guess I'd be swayed to go with VS 2005 right
away if a number of people responded to my post here claiming that they
were totaly happy with VS 2005 and that they *never* experience any of the
frequently reported "buggy" behaviour.

-Jeff

Not only that, but you are pretty much going to miss out on a good
amount of functionality which VS.NET 2005 provides.

Have you actually tried it? I ask because you say apparent...

--
- Nicholas Paldino [.NET/C# MVP]
- (e-mail address removed)

Jeff wrote:
Looking for your highly subjective opinions (from those of you who have
been working extensively with VS 2005).

I have a client that needs/wants a complete rewrite of an existing
application. I will be proposing a .NET Windows application with a
supporting SQL Server database. It will likely take 4-5 months before
we roll out the first production version.

Because of the apparent issues with VS 2005 (poor performance,
unpredictable and generally buggy behaviour), I am thinking it might
make sense to develop this app in VS.NET 2003/.NET 1.1. Then, around
the time we're ready to roll out the first production version, migrate
it to .NET 2.x just before going live.

I say 2.x because Microsoft has plans to release a service pack for it
mid-way through 2006, so we might have a minor version number change at
that time. Also, it appears that the service pack from Microsoft, if
released mid-year, will stabilize things so we can [hopefully] have a
smooth transition to 2.x just before going live.

Does this strategy make sense? Thoughts? Opinions?

Thanks!
This makes some sense, however doing something in this fashion is not
recommended as some things may go haywire during the conversion process
that you would need to fix. I find it easier just to deal with VS 2005
and make it with .NET 2.0 to start.