isis-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Stephen Cameron <steve.cameron...@gmail.com>
Subject Re: Transcribed IsisCon 2017 write-up
Date Thu, 22 Jun 2017 11:39:19 GMT
Hi Patrick
Thanks for the response, see answers & thoughts below

On Tue, Jun 20, 2017 at 1:27 AM, Patrick Pliessnig <p.pliessnig@gmx.net>
wrote:

> Hi, my comments are below.
> Patrick
>
> Am 18.06.2017 um 14:20 schrieb Stephen Cameron:
>
>> Hi All,
>>
>> I'd like to make some kind of a contribution to this discussion, though I
>> agree with most of what I see written. But I'd also like to be a "devils
>> advocate" a little too.
>>
>> I'm interested to get more work with Apache Isis, I have one live
>> application with a small group of users (about 10) who enjoy using it and
>> its been very reliable now for about 6 months. I also have a personal
>> project that I am developing as well,as time allows, one that I'm trying
>> to
>> use best-practices on, well making use of what I learned on the first
>> project.
>>
>> In regards to marketing of Apache Isis, to explain my thinking i'll first
>> state that to me IT seems to me both fashion driven and conservative at
>> the
>> same time.
>>
>> Fashion driven in that waves of what appears new cycle through repeatedly,
>> are latched onto by the big players - when it suites their ends, and then
>> get diluted as the novelty is duplicated or incorporated into the
>> establised encumbent systems or languages.
>>
> I would say the fashion you describe is a part of conservatism as it is
> played out only within the context of the big player as a supplier of
> technology: "there is a good reason that my supplier ist big. One of the
> reasons is that he moves forward and is smart enough to integrate new
> technologies for my benefit. Why should I change it?"
> The marketing value in this case is "Size".
>
> Agree

>
>> Conservative in that major change such as using an new application
>> framework is expensive and has risks, and very few established businesses
>> are interested in leading technically, that is unless their business model
>> is competition based on technology. I think of something like SAP's ERP
>> systems as the purchaser organisation conforming to the SAP domain model,
>> so the opposite of being interested in using DDD and for many businesses
>> it
>> makes sense (despite the cost) to go this way.
>>
>
> I agree: established business (not in the technology domain) normaly have
> "a running horse". At least the IT-Management thinks so.
>
>
>> Also conservative in that what comes to be a dominant 'paradigm' in
>> open-source software isn't necessarily the technically most advanced,
>> rather its what satisfies a need at a specific time and then becomes
>> well-known and rapidly improved too, I am thinking of numerous PHP based
>> things, more recently AngularJS. Once established its very hard to
>> displace.
>>
>
> I agree: conservative in paradigm too. one of the reasons is
> responsibility:
> "If my project fails and I use dominant paradigm, I can blame the paradigm.
> If my project fails and I use exoctic paradigm, others will blame me for
> having gone a risky way."
>
>
>> So I am leaning more towards the idea of Apache Isis marketing being
>> targeted mostly at a technical audience, maybe even at a very technical
>> one, that this is the best means for Apache Isis to grab a little "slice
>> of
>> the action" and as a result to prosper in the long-term. I agree 100% with
>> its strengths, as have been described, being in a complex domain app
>> scenario, but also that there are just so many good alternatives being use
>> for the less complex (but maybe nicer looking) things, and many developers
>> competing for the work.
>>
> I don't understand your conclusion to target a (very) technical audience.
>

I'm basing this on there being a kind of developer that might be willing to
give Apache Isis a go for a project, first that they are keen on
open-source, but also that they like the philosophy of Apache Isis too I
think. If individual people are complaining that it doesn't use JPA and
have an Angular client I suspect that they are not spending the time to
understand what of value it does provide. There is something called a SWOT
analysis (Strengths, Weaknesses, Opportunities, Threats), so they are
focused on the perceved weaknesses and not the real strengths.

For me there was (is) a learning curve, I came to Isis after reading the
Naked Objects book (I recommend getting a copy, even for its production
values alone, kids might even enjoy the pictures) and then Dan's book too.
So I was convinced by those and keen to give Isis a go, and persist with
the learning curve. Part of that was just me getting back into Java after a
long break and trying to improve my TDD skills and learn DataNucleus etc,
but there is alot of documentation and ideally one would start by skimming
through that to get a handle on what is possible. I've proposed that a
one-day guided Apache Isis introduction course would be a good way to get
into it.

I am thinking that even to be a mostly Java focused programmer now is kind
of technical bent, other languages promise to be more productive, I am
sceptical as to that being true long-term. I tend to the view that you
choose a language to suite the specific problem. If an OO language is ideal
then Java is a good OO language. Add to that the ideas of Domain Driven
Design (Evans), which I think of as coming from the "code as design" school
of thinking (http://www.developerdotstar.com/mag/articles/reeves_design.html),
and I have a feeling that Apache Isis will appeal to a kind of personality,
a creative type looking for the best tools for their craft perhaps?
Currenly not someone happy to run with the herd.

One such a person adopts Apache Isis, then I have the notion that they will
champion it, and its that championing which is what will lead to its being
used for the kinds of projects for which its most suited. If such a person
is not there then it seems to me there are too many hurdles.


>    Is this particular audience less conservative than others?
>    or do you mean by technical audience, companies whose business model
>    is based on competing on a technological level?
>    or do you favor the strategy: a bird in the hand is worth two in the
>    bush?
>
>
>> My vision is essentially an enhanced version of what I see currently
>> exists
>> in the project, that there are now very competent programmers (I am not in
>> that league) doing a few very complex projects and the framework is
>> progressively being improved within in a feed-back loop, something is
>> needed someone (usually Dan) does it essentially. The way to make this
>> more
>> sustainable is for more such projects to get going (I hope my project can
>> be one of them eventually).
>>
>> So, 3 things seem desirable, (1) get some more good (complex) projects
>> started with Apache Isis, (2) try to stop projects for which Apache Isis
>> is
>> not a good fit being started and taking time on the mailing list and
>> potentially failing too (3) seeking feedback on why Apache Isis was not
>> chosen when it potentially would have been a good fit. I can think of a
>> few
>> ideas with regards to (1),  (2) and (3) are maybe another email.
>>
>
> I agree with your point: get some more good projects.
> That's closely related to marketing (as you describe).
>
>
>> Getting some more good projects started
>>
>> My strategy has been to look for potential clients that will benefit from
>> complex custom (or customised) software and to target them. I've not been
>> very successful in this so far I have to admit, I attribute this in part
>> to
>> my own sales ability, but also to my location in Tasmania which doesn't
>> have alot of head-offices where CTOs are generally. located. However,
>> there
>> are a few industries that are growing here and I'm now seeking  small but
>> growing firms in those sectors (food and maybe tourism).
>>
> If part of your failure is your location and sales ability, it is maybe
> worth if you share with us the other parameters more in detail that
> influenced the failure (besides conservatism in Canberra)
>

For some reason about 10 years ago there was a massive move to .Net  in
this small corner of the world, I suspect as a result of the attraction of
Sharepoint. At the time I was working in data-management in science,  and
still doing Java based work in part. On leaving that (foolishly perhaps) I
tried to set up a business supporting open-source Java in local businesses,
this failed.  I come back to that idea over the past few years after
getting a single job through a contact. I'm looking for more such work (but
doing other things too for $), including with Apache Isis, but for clients
on the mainland of Australia, so remote workingis the idea. I have one
potential cient in Brisbane that I found via Upwork.com, whether Isis is
good match for that work is presently being determined, it is useful as a
prototyping tool. I think my own ideas for projects are the most promising
use-cases at this point.


>> One major sector I have to ignore is government, they are both very
>> conservative and very unwilling, despite their rhetoric, to give something
>> new that doesn't have a major player behind it a chance. I once did a
>> tender proposal that was priced at half the winning bid, it was thought
>> too
>> risky an approach to consider. I'd have another go with Apache Isis but
>> only if the application was a very good fit for Apache Isis and those are
>> only likely to be Federal projects out of Canberra. I have to be honest
>> and
>> say at 55 my energy to go looking there is not very high but does seem to
>> me that government projects are often going to be good ones for Apache
>> Isis
>> (maybe if I find one here I can call on a pool of talent overseas?)
>>
>> The other scenario that seems to make sense is for a small software house
>> to use Apache Isis as the basis of a system that is sold, as I understand,
>> the way its been used by Oscar Bou's group.
>>
> Your scenario for a small software house to use Apache Isis makes perfect
> sense to me. With my words I would say in this case that Apache Isis
> improves the core business ( the value chain ) of this small software house
> to leverage its comparative advantage. I apply this strategy as well (
> though I am not strictly a software house )


I have an idea that most programmers, who stay in the game for a long
period look for self-employment of one sort or another to capitalise on
their experience and to have more control on what kind of work they do.
This was my motivation for trying to go this way myself. You can either go
into a management role or try to become a true professional, to "hang out
your shingle". [This relates to the idea of marketing to a technically
orientated group, that might go and do a demo to a client or potential
client to 'win' some business.]

>
>
>> So this approach is not based on selling Apache Isis features, but on me
>> leveraging them to advantage in supplying a service to my potential
>> clients. I just tend to say I use Java because I can utilise some good
>> open-source frameworks as a base or incorporate libraries in when needed.
>>
>> Thinking about marketing to technical audience, a name change would still
>> be worthwhile, and of those listed from the gathering I tend to like
>> Apache
>> Tailor most as it reflects the custom-made or  'tailor-made' strengths of
>> Apache Isis well (a nice icon could be produced from it too). [I am
>> relaxed
>> about the new name - but - In the notes there is a reference to jazz,
>> might
>> be an in joke but a famous (and very individualistic) jazz guitarist by
>> the
>> name of Allan Holdsworth [1][2] died recently, Apache Holdsworth has a
>> nice
>> ring and there would be connection to that other popular generated UI
>> Python framework].
>>
>> The main issue is how to make more developers aware of Apache Isis I
>> think.
>> Marketing is still needed for this and targeted advertising is one way to
>> do this I feel. If you cannot afford advertising some kind of a publicity
>> stunt is an alternative, anyone volunteering?  Another idea is simply to
>> look at the already good documentation and think of ways to use that
>> resource in other "look what I do" showing off ways. The videos that Dan
>> has done are good, maybe that is the answer for others could make quality
>> videos linked to sections of the documentation.
>>
>> Steve
>>
>> [1] https://www.youtube.com/watch?v=gSqz17gtv8U
>> [2] https://www.youtube.com/watch?v=OXSd-WyrtfA
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> On Sat, Jun 17, 2017 at 3:35 PM, Dan Haywood <
>> dan@haywood-associates.co.uk>
>> wrote:
>>
>> Hi folks,
>>>
>>> I've just gone through the photo attachments on the IsisCon write-up [1]
>>> and transcribed them into text; a little bit easier for others to
>>> consume.
>>>
>>> For those who were there, it would be great if you could flesh out those
>>> bullet-points with your own recollections of discussion points that were
>>> made that were not written down on the whiteboards.  I've added one or
>>> two
>>> points of my own along the way.
>>>
>>> Cheers
>>> Dan
>>>
>>>
>>> [1] https://cwiki.apache.org/confluence/display/ISIS/IsisCon2017
>>> +write-up
>>>
>>>
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message