Is dolphin dying ?

I'm contemplating an idea of buying Dolphin Smalltalk XP Pro. But I
could not help but notice some signs of deterioration on object-arts
web site: outdated content (6.0 is still promised to be released at the
end of 2004), difficult navigation, documentation for 5.0 is missing
for more than a year, etc.
I just don't want to invest in a company that will seize to exist soon.
On the other hand the product itself looks solid and well done, and by
all means the most available among commercial Smalltalk implementation.

Could someone confirm/dispel my concerns please ?

Peter.

On 14 Mar 2005 13:26:38 -0800, "Peter" <[hidden email]> wrote:

>I'm contemplating an idea of buying Dolphin Smalltalk XP Pro. But I
>could not help but notice some signs of deterioration on object-arts
>web site: outdated content (6.0 is still promised to be released at the
>end of 2004), difficult navigation, documentation for 5.0 is missing
>for more than a year, etc.
>I just don't want to invest in a company that will seize to exist soon.
>On the other hand the product itself looks solid and well done, and by
>all means the most available among commercial Smalltalk implementation.

Hopefully, Object-Arts is doing fine. However, if they did get out of
business, VW is not THAT expensive for the little guy. Last time I
checked it was US$500 per year.

OTOH, I do agree that OA should release a new version ASAP. I don't
think the delay is caused by its imminent death, but rather by the
perfectionism that all of us, obsessive-compulsive programmers suffer.
;-)

It's usually difficult to realize that while the 'unfinished' product
is laying on your HD, it's not helping your customers. Sometimes, you
just need to have the _balls_ to release a product that is not
_exactly_ what you wished it to be. There will be other versions, and
remember: worse is better
(http://www.dreamsongs.com/WorseIsBetter.html).

Anyway, my recommendation is to buy it and start writing some code. If
later on, an asteroid falls on OA offices, you can always move on to
some other vendor. It's not like there was only one implementation.

Peter schrieb:

Peter,

a few months ago similar thoughts ran through my head as I also considered buying a license, and apart of 350 USD leaving quite a gap in my budget, of course one also has to consider the effort of learning something new that hopefully pays of some day.

I eventualy bought a license and haven't regreted it. AFAIK the people from object arts are focusing on some sort of trading software for either selling it or using it themselves, don't know which is true and don't realy care. However that software seems to be based / developed with Dolphin itself.

IMHO Dolphin 6 is thus eventualy going to show up, no rush as far as I'm concerned, DST 5 is already better than anything I've seen sofar.

One thing I don't realy have hope for is an improvement on documentation, but please guys, surprise me! ;-)


Günther

Günther Schmidt <[hidden email]> wrote in message news:<42365c79$[hidden email]>...

Ok. But don't you think when it arrives it will be too late. Now
Dolphin XP  doesn't have any official Web Framework besides some
goodie you can find ported to, or ready to port. Dolphin XP (5.0) is 3
years old (if I'm not wrong), now we are in 5.1.4 but this is only
because fixed bugs in the VM or in the Smalltalk code. Using Dolphin
XP you lost competitivity versus other Smalltalk distributions, and
other languages. Furthermore don't you think that a year delay in any
technology product is to much. Now every development tool is facing
the .Net or Mono integration, as other Smalltalk distributions. I
think (only based in the list post) that D6 will not have any .Net
integration (neither Mono). So maybe OA is perfectioning D6, but If D6
don't follow any market trends, Dolphin will die even though the
perfectionism that all of obsessive-compulsive programmers suffer. On
the other hand, don't you think the (cheaper) VW per year licences is
a marketing decision to win in the low budget segment of the Smalltalk
market, where Dolphin, MT and VS (another St. ditribution killed by
Cincom) are the principal rivals.

mario bancos schrieb:
Mario,

first of all nothing I said before was meant to be interpreted as a "qualified statement", mereley a personal opinion.

Yes I'm quite aware of the gaps and "late" schedule myself, but the original question was wether or not Dolphin is a dying product, rather than if Dolphin is the wisest choice.

 From my little experience even merely deciding getting into (any) Smalltalk may not be a *wise* choice, from a commercial perspective. (With C#, C++ or Java being where the money is). The later I say with some authority as I have been on the freelancer market..

Günther

mario bancos wrote:
MS Visual Studio 2003 is also 2 years old and nobody thinks it is dying.
Dolphin XP is a very mature development environment as Smalltalk is as a
language. Sure, Dolphin has some areas where improvements are possible.
But there is nothing that holds you back from writing great applications
with it.
 
> because fixed bugs in the VM or in the Smalltalk code. Using Dolphin
> XP you lost competitivity versus other Smalltalk distributions, and
There also weren't VAST releases for quite some time. And VW is IMO outdated
in regard to its GUI and Pollock will take quite a while before it will
replace the traditional GUI. So I don't agree with you here.

> other languages. Furthermore don't you think that a year delay in any
> technology product is to much. Now every development tool is facing
> the .Net or Mono integration, as other Smalltalk distributions. I
> think (only based in the list post) that D6 will not have any .Net
> integration (neither Mono). So maybe OA is perfectioning D6, but If D6
> don't follow any market trends, Dolphin will die even though the
> perfectionism that all of obsessive-compulsive programmers suffer. On
> the other hand, don't you think the (cheaper) VW per year licences is
Either I missed an important license fee change or you didn't read the
license conditions thoroughly. AFAIR $500 per year and developer is the
MINIMUM you have to pay for VW. Cincom really wants a part of your revenues
you earn with the products created with VW!

> a marketing decision to win in the low budget segment of the Smalltalk
> market, where Dolphin, MT and VS (another St. ditribution killed by
> Cincom) are the principal rivals.


Andreas

"mario bancos" <[hidden email]> wrote in message
news:[hidden email]...
> Ok. But don't you think when it arrives it will be too late. Now
> Dolphin XP  doesn't have any official Web Framework besides some
> goodie you can find ported to, or ready to port. Dolphin XP (5.0) is 3
> years old (if I'm not wrong), now we are in 5.1.4 but this is only

Too late for whom?  I don't think that a large percentage of Dolphin users
are running it on web servers or intend to do so.  It would seem to me that
there are already other Smalltalks in more dominant positions in those
areas.  I see Dolphin as a really excellent environment for writing
stand-alone or client programs.

> because fixed bugs in the VM or in the Smalltalk code. Using Dolphin
> XP you lost competitivity versus other Smalltalk distributions, and
> other languages. Furthermore don't you think that a year delay in any
> technology product is to much. Now every development tool is facing
> the .Net or Mono integration, as other Smalltalk distributions. I
> think (only based in the list post) that D6 will not have any .Net
> integration (neither Mono). So maybe OA is perfectioning D6, but If D6
> don't follow any market trends, Dolphin will die even though the
> perfectionism that all of obsessive-compulsive programmers suffer. On

I was eager to play with Dolphin 6.0.  But I really did not have any need
for it, Dolphin 5 works very well for my needs.  I don't feel that I have
lost any competitive advantage vs other Smalltalk distributions, rather I
have benefited from what Dolphin is: a great environment for developing
stand-alone Windows programs.  The other large Smalltalk distributions
always had more bells and whistles than Dolphin.  I did not need most of
them then, and still don't.  Dolphin Smalltalk was better for my needs
developing stand-alone Windows programs than competing products were.

.Net is interesting.  It might be cool if Dolphin supported that.  However
there would be disadvantages also.  My understanding is that a fairly large
support package is required for .Net applications.  I would not want my
deployment process to become more complex unless I really needed the
benefits .Net offers.

Market trends can be tricky.  Sometimes they represent real innovation, and
sometimes they just represent marketing innovation.  I tend to take a
cautious view of market trends.  For a while Java was the big deal, now .Net
is the big deal.  I was happy with Dolphin while Java was the big deal and I
am still happy while .Net is the bid deal.

> the other hand, don't you think the (cheaper) VW per year licences is
> a marketing decision to win in the low budget segment of the Smalltalk
> market, where Dolphin, MT and VS (another St. ditribution killed by
> Cincom) are the principal rivals.

Someone mentioned a $500 price tag for VW, but as far as I know they still
want royalties for distributing products developed in it.  But price is not
the biggest reason I chose Dolphin.  I really like the way the system works
and feels.  I like that it really feels like a native Windows application.

I think that people should choose the best tool for the job.  For me,
developing stand-alone Windows applications, Dolphin still seems to be the
best tool.  If I had to develop server software I might choose a different
Smalltalk.

Object Arts does not try to be all things to all people.  For now I am happy
with the trade-offs they have made.  However I am not a blind loyalist.  If
something much better for my business comes along I will evaluate it.
Object-Arts needs to remain competitive in its own market space (it is
fairly unique amongst Smalltalks).

Chris

Christopher J. Demers wrote:
We are running multiple web server applications developed in Dolphin
Smalltalk. Everything is 100% Dolphin Smalltalk code (altough we use a
dialect portability layer just in case so that we could move to other
Smalltalk anytime). When I say everything I mean everything, starting
from the web server up to the business domain and persistency code. You
can even download an example of Dolphin web app - WikiDoc server - from
our wiki at http://wiki.gorisek.com.

Speaking for myself I miss none of the things you've mentioned. If there
is something missing I can usually write it in Smalltalk in less time
than it would take me just studying the documentation for some Java/EJB
app server. Also, frameworks are a very little part of any big
application and writing web frameworks isn't really a rocket science.
The frameworks usually just grow along with the app in the direction you
need them to grow.

Best regards,

David Gorisek
http://www.gorisek.com

"David Gorisek" <[hidden email]> wrote in message
news:42374f55$[hidden email]...
[...]You
> can even download an example of Dolphin web app - WikiDoc server - from
> our wiki at http://wiki.gorisek.com.
That's pretty cool, David!

Don

mario bancos wrote:
> Günther Schmidt <[hidden email]> wrote in message news:<42365c79$[hidden email]>...
>
>>Peter schrieb:
>>
>>>I'm contemplating an idea of buying Dolphin Smalltalk XP Pro.


Good choice.


>>>But I could not help but notice some signs of deterioration on object-arts
>>>web site: outdated content


Have you tried running an editable web service lately?
Without a DNS entry for the IP address, it takes less than
15 minutes for something to start probing it, looking for
vulnerabilities.

But the content doesn't sustain much damage until it
is reachable via DNS.  At which point, spam happens.

Whoops - that's a full-time staff position to
keep the site clear of defacing and trojans, etc.
Or a reconsideration/redesign of the 'openness'.
Seems they chose to freeze it, rather than consume
that much resource.  Good choice.


>>>(6.0 is still promised to be released at the end of 2004),


Hmmmm, *promised*, or planned?
How much is the dev license, and the maintenance
contract?  Times number of seats you need?
Equals how many engineering salaries, roughly?

Just pointing out the realities involved.
This is an extrordinary product, produced by
a small company.  There are always pluses and minuses.
But the current product is still in the 'best of breed'
class, before even looking at the price.  (As are most,
if not all, Smalltalk offerings.)


Now for the "I agree with you" part.
It would be a good idea to update the 'expected delivery'
dates on the wiki, even if it is done by hand.

For the same reason it is a good idea to call the
wife if you're going to be later than expected.
The advice holds even if, god help you, that was days ago...


>>> difficult navigation,


Can't comment - don't know what you mean here.
Examples?


>>>documentation for 5.0 is missing
>>>for more than a year, etc.


Yup.  Could always be better.  Tradeofs again.
But one should also keep in mind the
advantage of having ALL THE SOURCE right there
in front of you, all cross indexed, and browsable.
And most of it is easy to experiment with, *live*,
and with ease, in a workspace.


>>>I just don't want to invest in a company that will seize to exist soon.


You worry too much.  ;-)

Buy it.

1) It is a *very* good deal.
2) It is *very* high quality already.
3) Look at it, play with it, get familiar.
    Now ask yourself - the guys who did this,
    are they going to do more?  Or die?

    And - same question, what about you?

    Ya never really know.  But if you're going
    to put your chips down somewhere, I think
    Dolphin is one of the best bets available.


>>>On the other hand the product itself looks solid and well done, and by
>>>all means the most available among commercial Smalltalk implementation.

>>>Could someone confirm/dispel my concerns please ?


Sure.  Buy it.  Buy ten.  (And get some VW licences, too).
Maybe try some other versions as well.  Surest way to enhance
whatever you're betting on.

You haven't even *mentioned* the actual downside,
which is quite real.  Know much about addiction?
Try enjoyable stimulus, succesful response, and instant
reinforcement.  Be sure and read the warning labels...


Regards,

-cstb

> We are running multiple web server applications developed in Dolphin
> Smalltalk. Everything is 100% Dolphin Smalltalk code (altough we use a
> dialect portability layer just in case so that we could move to other
> Smalltalk anytime). When I say everything I mean everything, starting
> from the web server up to the business domain and persistency code. You
> can even download an example of Dolphin web app - WikiDoc server - from
> our wiki at http://wiki.gorisek.com.

David,

I still say you should release the web server code as a product itself :)

Or even the portability layer you've developed.

I know probably a pain to manage on top of the other products...

On another note. I'm still working (albeit slowly) on a release of Seaside
for Dolphin. STS is helping me with the porting issues. I still don't have
it working yet though. I'm hoping that I can get Seaside to a point where
development work can be done in Dolphin, and either deployed directly in
Dolphin,or filed out, and directly imported into Squeak, as a multi-platform
target environment. Best of both worlds.

I keep hoping that D6 will come out now because as far as I understand some
changes have been made to the VM to better handle continuations (which is
available in D5, but gets quite ugly behind the scenes). Aside from all the
other modifications that have been made since D5 was released oh so long
ago.

On Tue, 15 Mar 2005 21:54:04 +0100, David Gorisek
<[hidden email]> wrote:



>We are running multiple web server applications developed in Dolphin
>Smalltalk. Everything is 100% Dolphin Smalltalk code (altough we use a
>dialect portability layer just in case so that we could move to other
>Smalltalk anytime).

Could you elaborate on this (the portability layer)?

Thanks

Only two little comments:

Christopher J. Demers wrote:
>
> .Net is interesting.  It might be cool if Dolphin supported that.  However
> there would be disadvantages also.  My understanding is that a fairly large
> support package is required for .Net applications.  

Agree.
By example a +20MB runtime. How to say a customer: "ok, to use my app of
2MB you need to download by your dial-up connection more than 20MB of
stuffs."

>
> Market trends can be tricky.  Sometimes they represent real innovation, and
> sometimes they just represent marketing innovation.  I tend to take a
> cautious view of market trends.  For a while Java was the big deal, now .Net
> is the big deal.  I was happy with Dolphin while Java was the big deal and I
> am still happy while .Net is the bid deal.
>

FULL AGREE, to a wide market niche (small and medium companies and
applications) I think that is really better to stay as far as is
possible of the "marketing innovations" that usually only want the $ of
customers giving in exchange almost nothing.

Cheers.
gsa.

Fernando wrote:
Well, a socket object in Squeak and a socket object in VW do not have
the same methods as do socket objects in Dolphin. The same goes for
other classes too. So let's make wrapper classes for each specific
Smalltalk dialect and use wrappers instead of  Dolphin specific code.
Put this wrappers in a package and you have a dialect abstraction layer.
This way you can move application code from one VM to another without
having to change anything. For web apps this is easy because you mainly
need sockets, streams and process + exception handling. There's no GUI
code which one would need for a fat-client application.

"Sean Malloy" <[hidden email]>  wrote:
> I still say you should release the web server code as a product itself :)
>
> Or even the portability layer you've developed.
> I know probably a pain to manage on top of the other products...

I hoped eagerly even in my dreams, too, since I was quite discomfortable by
the unnatural conditions that the wiki editing pages cann't show our
unicode-coded words correctly.

> On another note. I'm still working (albeit slowly) on a release of Seaside
> for Dolphin. STS is helping me with the porting issues. I still don't have
> it working yet though. I'm hoping that I can get Seaside to a point where
> development work can be done in Dolphin, and either deployed directly in
> Dolphin,or filed out, and directly imported into Squeak, as a
> multi-platform
> target environment. Best of both worlds.

   Totally agreed. Seaside is a wonderful innovation but unfortunately stuck
deeply in the dialect-specific environment, IMHO, it depended too much on
Squeak's unique VM mechanism, AFAIK, it was almostly tangled inside it, you
had to be quite familiar with Squeak so that you could extract it out from
Squeak faithfully  to port successfully to other dialects without any
distortions.
   It would be better if the original authors could take account of this
condition and go instead to release a dialect-neural edition, I think this
way it would open more markets and earn more money for them in the future if
they will charge on it finally.

> I keep hoping that D6 will come out now because as far as I understand
> some
> changes have been made to the VM to better handle continuations (which is
> available in D5, but gets quite ugly behind the scenes). Aside from all
> the
> other modifications that have been made since D5 was released oh so long
> ago.
>
   Long live the DST


Best regards,

tkkuo

kuo wrote:
Which windows code page is it? Or do you need to serve utf-8 encoded pages?

Best regards,

David Gorisek

kuo wrote:
>
>    Totally agreed. Seaside is a wonderful innovation but unfortunately stuck
> deeply in the dialect-specific environment, IMHO, it depended too much on
> Squeak's unique VM mechanism, AFAIK, it was almostly tangled inside it, you
> had to be quite familiar with Squeak so that you could extract it out from
> Squeak faithfully  to port successfully to other dialects without any
> distortions.

It was ported to VW, IIUC, by someone other than the
orignal author. Here's what it says on http://www.seaside.st/Download.

<quote>
Seaside is not available for other Smalltalks at this time.
Many Smalltalk VMs do not support the
stack-copying techniques Seaside uses to implement backtracking.
The status for various dialects is:

    * Dolphin: definitely possible; a very early version of Seaside was successfully ported
    * Smalltalk/X: not feasible
    * VAST: not feasible
    * Gnu Smalltalk: believed to be possible, but never tried
    * Ambrai: unknown
</quote>

I have a prototype app. that I fileout of Squeak,
and then filein to VW (with Seaside), and it just
works. I currently have to do one manual edit
because of VW's namespaces though. I'm hoping that
this will work on all ST's that support Seaside.

> It was ported to VW, IIUC, by someone other than the
> orignal author. Here's what it says on http://www.seaside.st/Download.

It definitely works on Dolphin. The old version worked, and the
Continuations definitely work.

I'm currently working on a Dolphin version. It's taking a lot longer than I
expected because I have a lot less spare time to spend on it than I
expected.

Hi Sean,

> I'm currently working on a Dolphin version.

I just wanted to make sure you know that Steve Waring put out a Dolphin
5 port of Seaside some time ago.  It can still be reached via
http://www.dolphinharbor.org.

-- Louis

Louis

> I just wanted to make sure you know that Steve Waring put out a Dolphin
> 5 port of Seaside some time ago.  It can still be reached via
> http://www.dolphinharbor.org.

Yeah, I've played with that version. Seaside has changed a lot since that
initial port. (It was based on Seaside 0.9) The whole rendering system is
different. Templates don't exist anymore, there are a whole lot of other
changes too.

However it has been useful getting my head around continuations and serving
as an example of how the original port was actually done.