Dear Andrew,
No offence whatsoever was taken! I just thought it was important to
raise the issues I did. Features are important but even more
important to me, given the meagre resources available for the study of
Classics, is the sustainability and transparency of the technological
infrastructure. It is quite true that a native application will
always have a superior look and feel when compared to a cross-platform
application, but my feeling is that, as a small and poorly funded
discipline, we do not have the resources to squander on parallel
incompatible developments for different platforms.
I am very sorry to hear about the loss of the developer's laptop with
his source code -- that is a truly tragic story. I hope it will not
be taken as insensitive if I point out that this illustrates more
poignantly than I ever could have imagined the importance of open
source development in ensuring the continued availability and
improvement of niche software in the face of the vagaries of Fate.
Best wishes,
Peter
On Wed, Jun 20, 2012 at 3:22 PM, Andrew Dunning
<[log in to unmask]> wrote:
> Dear Dr Heslin,
>
> My apologies for causing offence; the characterization was not
> intended to disparage your own application, which I have used for many
> years and greatly respect. I intended to explain that the developer
> had set out to improve on your program by using native Mac frameworks,
> and was very successful in this, but the loose ends he left resulted
> in a program that takes some effort to set up before being at all
> useful. In this case, the developer apparently intended to make the
> code open source once it had reached some level of stability, but the
> laptop on which the code was kept was stolen. There was formerly an
> explanation of how to remove the expiry date on the software on his
> website, but it has disappeared. (If you are interested, I could
> provide instructions: there is a switch in the preferences file.)
>
> Concerning proprietary versus open source technology, I certainly take
> your points: I think that there is a place for both. The reality is
> that a properly written native application is almost always going to
> provide a better user experience, but there are cases when one simply
> needs to reach as many users as possible. As you may be aware, this is
> an ongoing debate in the question of writing native applications for
> mobile devices or creating cross-platform 'web apps': here again,
> there is a place for both. On this topic, see for example
> <http://inessential.com/2011/12/13/apps_and_web_apps_and_the_future>.
>
> Sincerely,
>
> Andrew Dunning
>
> On Wed, Jun 20, 2012 at 8:42 AM, Peter Heslin <[log in to unmask]> wrote:
>> On Tue, Jun 19, 2012 at 4:57 PM, Andrew Dunning
>> <[log in to unmask]> wrote:
>>> Andromeda: Essentially an improved version of Diogenes (using the PHI and
>>> TLG discs for texts, better than Perseus, but the latter cannot now be
>>> obtained legally), much faster and with an excellent interface. Would have
>>> been amazing, but the author seems to have become disillusioned with
>>> Classics and given up (to judge from statements on his website). Can still
>>> be made to work.
>>
>> As the author of Diogenes, I object strongly to this characterization.
>> I have not tried Andromeda, so I cannot comment on the claim that it
>> is faster or the implication that the interface is better. I believe
>> that for most normal classicists (Helma, "normal" excludes crazy
>> linguists doing searches for statistical purposes that turn up
>> thousands of hits :-) on modern hardware Diogenes is fast enough. I
>> like the interface, but de gustibus. There are, however, serious ways
>> in which Andromeda clearly cannot be an improvement. It runs only on
>> Macs, whereas Diogenes runs on just about anything and is conveniently
>> packaged for Mac, Windows and Linux. Andromeda depends closely on
>> proprietary Apple technology, whereas Diogenes depends on the open
>> technologies of Perl, Javascript and HTML.
>>
>> Most importantly, Andromeda is not freely licensed or open source; it
>> is closed. The sourceforge website promises that it will "eventually"
>> be open-sourced. If you believe that, I have a bridge I'd like to
>> sell you. This means that when the developer loses interest in the
>> project, as appears in fact already to happened here, it becomes a
>> white elephant. As Apple evolves and updates its OS, no one else can
>> step in to update the application so it continues to work. Diogenes,
>> by contrast, is free and open source. If I lost interest or (more
>> likely) get hit by a bus, anyone can step in to make sure the
>> application continues to work.
>>
>> I just tried downloading Andromeda so that I could respond more fully
>> to the claims about its interface design. When I tried running it, it
>> said that this preview had expired and shut down. So it seems that at
>> the moment all that one can download is a closed-source Mac-only
>> binary with a built-in expiry date which has already expired.
>> Presumably the claim that this "can still be made to work" involves
>> circumventing this designed-in breakage. I am not trying to run down
>> Andromeda or the pleasure you have taken in it -- it may be a
>> wonderful user experience for you and a vast improvement upon Diogenes
>> for a certain group of users. Just please do not try to sell a
>> single-platform, closed-source, expiry-dated binary as a general
>> improvement upon a multi-platform, open-source, freely-licensed
>> application. There may well be room in the world for both, but one
>> does not obviate the need for the other.
>>
>> Peter
>>
>>
|