Your opinion regarding a modern GUI
GerhardF
was mainly designed 20 years ago to look like an OS/2-application
(yes, OS/2 - IBM's former PC operating system).
Now I'm wondering about the communities opinion on how a new GUI could
look like:
- looking like a "standard WINDOWS application"?
- looking like a JAVA application?
- looking like an ECLIPSE application?
- should it be a GUI that can be changed via "faces" to fit into a
user's personal preference?
Looking forward to get your opinion.
Arle (LISA)
My two cents are that the real idea would be something like your last
idea (a “GUI that can be changed via ‘faces’ to fit into a user's
personal preference”). As the program develops we will probably find
that there are different usage profiles for it. For example, if we add
a module to support the updated LISA QA Model (which isn't ready yet),
there would be a QA profile, which will be different from the
translator profile. The ideal would be to make it possible to have a
different UI view to support the needs of different profiles. If we
can do that, then we also have the ability to allow users to create
custom UI views to meet their needs (many of which we may not
anticipate).
That's the long-term view. In the short term, I would like to see a
refresh of the visual elements and layout of the UI. A lot of aspects
of human-computer interaction theory and practice have changed in the
past twenty years, and even if we don’t provide flexible views, just
updating the UI to look like a standard *anything* now would be good.
Sit down with a Windows 3.1 or Mac OS 6.x application and compare it
with its modern counterpart and you can see that there is a huge
change in look and feel. While some of it is cosmetic (better
graphics), there is a lot of ergonomic change that has come about as
well. That suggests that we can see some pretty substantive
improvements in usability without even having to touch the functional
aspects of the code.
-Arle
Stephen Holmes
In my opinion, the GUI in openTM2 is it's greatest detractor to
adoption. Although one could argue that usability is secondary to
functionality, in this case it's not. The manner in which one interacts
with files and their translation memory, amongst other UI issues,
introduce a learning curve and that's a bad thing.
The core functionality of openTM2 is however excellent and it presents a
number of differentiators to competing opensource and proprietary
tooling solutions and we should be able to capitalise on that.
In thinking about this since the initial release, I believe that the
most logical choice moving forward is a move into the Eclipse framework
with the ultimate goal being the creation of an integrated translation
environment (ITE) that leverages from the IDE usability metaphor.
Eclipse is interesting on a number of levels, and to Arle's points
below, it can accomodate itself through the GUI to a user profile (or
role if you prefer) through the notion of "perspectives" which could
ultimately be mapped to the localization role-based activities of
translator, approver, customer approver, project manager etc. This
would allow those interested in supporting each role and its associated
functionality in a separate manner.
The move to Eclipse is further supported by...
- The OpenTM2 team, I feel, would be ill guided in creating a GUI from
scratch - it's time consuming, prone to bad design descisions (we've now
interaction experts on the team), and detracts from the core
functionality which it TM management and its associated functionality.
- An availability of Eclipse-read resources from the open source
community that could be canvassed for assistance
- An availability of countless of Java programmers (though development
in Eclipse doesn't necessarily require Java
- It's well documented in terms of web resources and printed
documentation and some tooling already exists in the globalization space
- It's a mature, stable and extensible system
- It is cross platform and therefore opens up HUGE opportunities in
reaching new end-users
- It support visual tools and would therefore lend itself to presenting
visual workflows through the EMF framework (very powerful indeed!)
- It's open extensibility through the use of plug-ins (i.e., think about
plugins for linguistic validation, enterprise TM integration and many,
many others!!!) and the distribution network that supports Eclipse plugins.
- By breaking down TM2 into it's components and re-architecting
accordingly, we can increase the manner in which the community can
interact with the project. Currently, TM2 is a monolithic app and
finding a point to get involved with is tough.
Just my 2 (euro) cent!
Steve
On 19/07/2010 12:05, Arle (LISA) wrote:
> Hi Gerhard,
>
> My two cents are that the real idea would be something like your last
> idea (a �GUI that can be changed via �faces� to fit into a user's
> personal preference�). As the program develops we will probably find
> that there are different usage profiles for it. For example, if we add
> a module to support the updated LISA QA Model (which isn't ready yet),
> there would be a QA profile, which will be different from the
> translator profile. The ideal would be to make it possible to have a
> different UI view to support the needs of different profiles. If we
> can do that, then we also have the ability to allow users to create
> custom UI views to meet their needs (many of which we may not
> anticipate).
>
> That's the long-term view. In the short term, I would like to see a
> refresh of the visual elements and layout of the UI. A lot of aspects
> of human-computer interaction theory and practice have changed in the
> past twenty years, and even if we don�t provide flexible views, just
Michael Schneider
since all technical reasons have been told already i'll make it short:
my vote goes to Eclipse
regards
Michael
--
Michael Schneider
Gesch�ftsf�hrer / CEO
Tel: +49 711 16646 10
Fax: +49 711 16646 50
E-Mail: michael....@beo-doc.de
http://www.beo-doc.de/
Bitte denken Sie an die Umwelt, bevor Sie diese e-Mail ausdrucken!
Please consider the environment before printing this e-mail!
beo Gesellschaft f�r Sprachen & Technologie mbH
Kupferstra�e 36
70565 Stuttgart
F�rderer des freien Open Source TM-Systems openTMS
http://www.openTMS.de/
Gesch�ftsf�hrer:
Ulrike Baral, Michael Schneider, Thomas Wedde
Registergericht: Amtsgericht Ulm, HRB 533731
Sitz der Gesellschaft: Ebersbach/Fils
Helena Shih
All,
since all technical reasons have been told already i'll make it short:
my vote goes to Eclipse
regards
Michael
Michael Schneider
On 07/19/2010 08:40 AM, GerhardF wrote:
> What is your opinion on a "modern GUI" user interface? OpenTM2's GUI
> was mainly designed 20 years ago to look like an OS/2-application
> (yes, OS/2 - IBM's former PC operating system).
>
> Now I'm wondering about the communities opinion on how a new GUI could
> look like:
> - looking like a "standard WINDOWS application"?
> - looking like a JAVA application?
> - looking like an ECLIPSE application?
> - should it be a GUI that can be changed via "faces" to fit into a
> user's personal preference?
>
> Looking forward to get your opinion.
>
--
Geschäftsführer / CEO
Tel: +49 711 16646 10
Fax: +49 711 16646 50
E-Mail: michael....@beo-doc.de
http://www.beo-doc.de/
Bitte denken Sie an die Umwelt, bevor Sie diese e-Mail ausdrucken!
Please consider the environment before printing this e-mail!
beo Gesellschaft für Sprachen & Technologie mbH
Kupferstraße 36
70565 Stuttgart
Förderer des freien Open Source TM-Systems openTMS
http://www.openTMS.de/
Geschäftsführer:
Ulrike Baral, Michael Schneider, Thomas Wedde
Registergericht: Amtsgericht Ulm, HRB 533731
Sitz der Gesellschaft: Ebersbach/Fils
--
You received this message because you are subscribed to the Google Groups "opentm2-design" group.
To post to this group, send email to opentm2...@googlegroups.com.
To unsubscribe from this group, send email to opentm2-desig...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/opentm2-design?hl=en.
GerhardF
into TRAC.
On Jul 19, 4:31 pm, Helena Shih <helenas...@gmail.com> wrote:
> Has this been filed in Trac?
>
> On Mon, Jul 19, 2010 at 10:21 AM, Michael Schneider <
>
> > All,
>
> > since all technical reasons have been told already i'll make it short:
>
> > my vote goes to Eclipse
>
> > regards
> > Michael
>
> > On 07/19/2010 08:40 AM, GerhardF wrote:
> > > What is your opinion on a "modern GUI" user interface? OpenTM2's GUI
> > > was mainly designed 20 years ago to look like an OS/2-application
> > > (yes, OS/2 - IBM's former PC operating system).
>
> > > Now I'm wondering about the communities opinion on how a new GUI could
> > > look like:
> > > - looking like a "standard WINDOWS application"?
> > > - looking like a JAVA application?
> > > - looking like an ECLIPSE application?
> > > - should it be a GUI that can be changed via "faces" to fit into a
> > > user's personal preference?
>
> > > Looking forward to get your opinion.
>
> > --
> > Michael Schneider
> > Geschäftsführer / CEO
>
> > Tel: +49 711 16646 10
> > Fax: +49 711 16646 50
>
> > Bitte denken Sie an die Umwelt, bevor Sie diese e-Mail ausdrucken!
> > Please consider the environment before printing this e-mail!
>
> > beo Gesellschaft für Sprachen & Technologie mbH
> > Kupferstraße 36
> > 70565 Stuttgart
>
> > Förderer des freien Open Source TM-Systems openTMS
> >http://www.openTMS.de/
>
> > Geschäftsführer:
> > Ulrike Baral, Michael Schneider, Thomas Wedde
>
> > Registergericht: Amtsgericht Ulm, HRB 533731
> > Sitz der Gesellschaft: Ebersbach/Fils
>
> > --
> > You received this message because you are subscribed to the Google Groups
> > "opentm2-design" group.
> > To post to this group, send email to opentm2...@googlegroups.com.
> > To unsubscribe from this group, send email to
> > .
Jean-Christophe Helary
On 19 juil. 10, at 23:21, Michael Schneider wrote:
> my vote goes to Eclipse
If that means a multiplatform application I raise 2 hands. Otherwise I don't see the point.
Jean-Christophe Helary
----------------------------------------
fun: http://mac4translators.blogspot.com
work: http://www.doublet.jp (ja/en > fr)
tweets: http://twitter.com/brandelune
Arle Lommel
stand-alone application even if it did not. But you can download and run
Eclipse on the Macintosh, Windows, Linux, etc. See
http://www.eclipse.org/downloads/ for more details. There are a ton of
Eclipse extensions (not sure if that’s the proper term) for doing all
sorts of tasks.
Basically any platform that supports Java can run Eclipse. So there
would be some work, but if OpenTM2 were ported to Eclipse it would be
cross-platform. Maybe some day we’ll even see an Android port :-)
-Arle
On 7/19/10 6:40 PM, sic scripsit Jean-Christophe Helary:
Stephen Holmes
consume the 100+ MB of Eclipse to make use of the advanced GUI it
provides. Eclipse can provide a Rich Internet Application interface to
accelerate development in a skimmed down version - just use what you
need. OpenTM2 would stand apart from the competition in such a
framework, I have no doubt about this.
Stephen.
On 20/07/2010 00:44, Arle Lommel wrote:
> Eclipse does mean that, although it would have advantages over a
> stand-alone application even if it did not. But you can download and
> run Eclipse on the Macintosh, Windows, Linux, etc. See
> http://www.eclipse.org/downloads/ for more details. There are a ton of
> Eclipse extensions (not sure if that�s the proper term) for doing all
> sorts of tasks.
>
> Basically any platform that supports Java can run Eclipse. So there
> would be some work, but if OpenTM2 were ported to Eclipse it would be
> cross-platform. Maybe some day we�ll even see an Android port :-)
Tex Texin
discussing the benefits of Eclipse.
With respect to the original question, I agree that it must support
different faces or profiles for the different types of users- translators,
project managers, reviewers, administrators, etc. It must be multi-platform.
It should support browser-based and desktop client interfaces and
accommodate online and offline activity. Support for a high degree of
collaboration is needed (multiple translators updating glossaries, TM,
etc.). Frequently performed tasks should require minimal keystrokes/clicks
and defaults should be customizable to accommodate user preferences and make
the UI highly productive. It should be possible to do the same operation
globally or across multiple items easily without having to repeatedly click
through one by one.
I am not sure I agree with Arle about the quick refresh. It might slightly
improve the look while delaying the other more significant requirements. I'd
wait to see the specifics of the proposals for short term vs long term.
Perhaps the most important first step is to design an API that allows anyone
to create any UI they want to get at the core functions. Several companies
have their own interfaces to other L10n applications via API interfaces,
finding the native UI lacking, so exposing the functionality might let them
adopt TM2 more easily. And the open source community might be enthusiastic
about creating independent UIs for various functions.
tex
Jean-Christophe Helary
On 20 juil. 10, at 08:44, Arle Lommel wrote:
> Eclipse does mean that, although it would have advantages over a stand-alone application even if it did not.
:)
It certainly would for Windows users.
I think one of my first question on this list was about the multiplatform future of OpenTM2 and I seem to recall that the reply was something like "far from there if going there at all", so the reference to a multiplaform framework is really good news.
> But you can download and run Eclipse on the Macintosh, Windows, Linux, etc. See http://www.eclipse.org/downloads/ for more details. There are a ton of Eclipse extensions (not sure if that’s the proper term) for doing all sorts of tasks.
There is even a XLIFF editor (Benten) created by a team of Japanese developers, that uses some of OmegaT's code.
The Java free ecology for CAT tools (Sun's OLT, Okapi framework, OmegaT) would certainly gain from more sharing and re-use. It is possible that OpenTM is not heading toward that but porting the thing to Java would be a step in that direction.
Helena Shih
Not only that. Eclipse is a reusable framework, one doesn't need to
consume the 100+ MB of Eclipse to make use of the advanced GUI it
provides. Eclipse can provide a Rich Internet Application interface to
accelerate development in a skimmed down version - just use what you
need. OpenTM2 would stand apart from the competition in such a
framework, I have no doubt about this.
Stephen.
On 20/07/2010 00:44, Arle Lommel wrote:
> Eclipse does mean that, although it would have advantages over a
> stand-alone application even if it did not. But you can download and
> run Eclipse on the Macintosh, Windows, Linux, etc. See
> http://www.eclipse.org/downloads/ for more details. There are a ton of
> Eclipse extensions (not sure if that’s the proper term) for doing all
> sorts of tasks.
>
> Basically any platform that supports Java can run Eclipse. So there
> would be some work, but if OpenTM2 were ported to Eclipse it would be
> cross-platform. Maybe some day we’ll even see an Android port :-)
>
> -Arle
>
> On 7/19/10 6:40 PM, sic scripsit Jean-Christophe Helary:
>> On 19 juil. 10, at 23:21, Michael Schneider wrote:
>>
>>
>>> my vote goes to Eclipse
>>>
>> If that means a multiplatform application I raise 2 hands. Otherwise
>> I don't see the point.
>>
>>
>> Jean-Christophe Helary
>> ----------------------------------------
>> fun: http://mac4translators.blogspot.com
>> work: http://www.doublet.jp (ja/en> fr)
>> tweets: http://twitter.com/brandelune
>>
>>
>
--
You received this message because you are subscribed to the Google Groups "opentm2-design" group.
To post to this group, send email to opentm2...@googlegroups.com.
To unsubscribe from this group, send email to opentm2-desig...@googlegroups.com.
Chase Tingley
means bolting a Java UI on the front could get icky.
But if there was going to be any serious work done on the UI I would
+1 a cross-platform approach. There are a couple of other cross-
platform widget toolkits (Qt, WxWidgets) that might be able to do the
job.
ct
On Jul 19, 4:44 pm, Arle Lommel <a...@lisa.org> wrote:
> Eclipse does mean that, although it would have advantages over a
> stand-alone application even if it did not. But you can download and run
Jean-Christophe Helary
On 19 juil. 10, at 15:40, GerhardF wrote:
> Now I'm wondering about the communities opinion on how a new GUI could
> look like:
> - looking like a "standard WINDOWS application"?
> - looking like a JAVA application?
> - looking like an ECLIPSE application?
> - should it be a GUI that can be changed via "faces" to fit into a
> user's personal preference?
I am biased because I use OmegaT everyday and almost only OmegaT. Hence, no "button" on the window title bar and almost no mousing around the interface.
So, my ideal (G)UI is an interface where I can access 80% of the functions directly with keyboard shortcuts.
The remaining 20% (excluding preference settings) should be available with a mouse click to reach a main menu and then the keyboard arrows to navigate in a hierarchy of sub menus.
The appearance of the whole thing (Windows/Java/with or without faces) is at least secondary, if not barely relevant, _as long as_ the general feeling conforms with what the user expects for native applications.
Jean-Christophe Helary
----------------------------------------
fun: http://mac4translators.blogspot.com
GerhardF
OpenTM2 was designed (even if the code is based on a design made 20
years ago) to be almost fully "controllable" by using key-combinations
and short-cuts. Just play with it and you'll see that it works fine.
The goal can only be to be at least on the same level of usability, IF
not even better.
Gerhard
On Jul 20, 10:37 am, Jean-Christophe Helary
Stephen Holmes
actions - I've seen our team in Spain in action and it's quite the
sight! Any change in GUI would/should have to have the same rapid
function interaction. Whether Eclipse is the GUI strategy or not, all
modern frameworks support keyboard accelerators. It would be incumbent
on the design team to look at the current keyboard interactions to
ensure that they comply in some way with modern user interaction
conventions, whilst also supporting the old scheme (this was done very
well in Microsoft Word, where Wordperfect key combos were support for a
period of time).
Steve
Derek Coffey
Derek.
Steve
--
Tex Texin
Although as I mentioned earlier, we should go beyond shortcuts and allow a
programmatic interface, so that people can easily integrate with other
applications and existing interfaces and achieve whatever UI they desire.
tex
Stephen Holmes
For some reason, the first mail I sent on this topic didn't hit the list
- ain't technology grand. I'll forward it again in a moment. Apologies
for the long mail that follows - lots of stuff in my head and lots of
this is my opinion so please read it in that context :-)
Suffice it to say that a moving to a framework such as Eclipse can
provide full API (or webservices) enabled application and all of the
ability to provide keyboard short cuts and other such goodness - this is
clearly a matter of design and implementation and should probably be
high up on the priority so we don't alienate current users. Here's my
current thinking...
1. Use Eclipse as the overall framework as a Rich Client Desktop tool
a. Design an intuitive tree view project management UI with a
contextualized editor panel.
+ The editor panel would adapt to the content being edited
b. Exploit the Eclipse "perspectives" paradigm to provide for
translator views, project manager views etc, etc.
c. Modularise core TM2 functionality
2. Identify the core components in openTM2 that can be "converted" into
plugins
a. Content Parsing/Generation
b. Leveraging
c. MT integration
d. Content Editors (source/target, WYSIWYG, XML editor etc)
e. Content Validation (language rule based checking?)
f. Workgroup/Collaboration (content synchronization etc)
g. Terminology Management
h. CMS integration
i. TMX/TBX import/export
.... Clearly this list is endless ;-)
One clear advantage to modularizing and moving core functionality to a
plugin-based metaphor is that we suddenly open the product up to a
community who can engage with us at a much more granular level. If I'm
a GUI guy, I can engage there, if I'm a search guy I can weigh in on the
leveraging plugin, if I'm a CMS guy, I can focus on integration plugins
and if I'm a content wizard I can start work on the parser plugins.
Right now, jumping into TM2 code is an onerous task due to its
monolithic origins.
Now, this is all fine, but I think that before we get ahead of
ourselves, two SUPER IMPORTANT areas need to be addressed as a
priority. These are the data model and the API/Webservices Layers.
The Data Model is basically the back end database, what's stored, what
meta-data is available to describe the content, whether it's an
extensible model and how it's accessed. A solid data model is a
foundation for a powerful toolset. I don't yet understand what meta
data is available in the current TM files within openTM2, but this is an
opportunity to review and revise it whilst providing backward
compatability. Core concepts such as multi-user synchronization,
content distribution, reporting etc should be addressed in this space.
Also, having a good data model means we could conceivably provide a
cloud based database with any time of interface, which leads on to....
I would also propose (and supporting Tex at the same time) that some
initial detailed design work also occur on the API/WebServices layer as
a priority to set a solid architectural foundation for all future
development. Having this layer clearly defined would allow
organizations to exploit the opensource framework to plugin their own
components (okay, I'm being very selfish here and have numerous ideas
for this!!)
Lastly, the Cloud, it's the buzzword de jour in the localization space,
sadly it's deeply misundertood (long conversation!), a sound data model
and exensible API (perhaps a webservices-based on) provides for security
and accessibility in the cloud - in fact, as we look at the complexities
of scale-up/scale-out services typified in SaaS models, these are
critical factors. I'd love to expand on this a lot more in the future,
but it's clear that for this opensource initiative to gain any traction
we need to pull openTM2 into the modern desktop application paradigm. I
just wanted to pop this in here as a placeholder so we continue to think
big!
Cheers
Steve
Chase Tingley
Stephen Holmes
Agreed - huge effort/risk. Is it the right thing? I think so. But to be honest, there are many other things that can occur in parallel and this conversation is currently only talking GUI. Also, C/C++ components can be wrapped and used in the framework without porting to Java per se - not optimal, might make an interesting prototype?
On a port taking years, hmm, I don't know - that really depends on the community that we build up around it. That is, community members who deliver code rather than strategy! I think at this early stage of the product life cycle that it's good to have the 37,000 feet discussions and then circle back on how to take us forward tactically in manageable chunks. In essence, we need a destination before we start a journey.
Early days still! :-)
Steve.
Elsebeth
item on the list. Adding sufficient function to make it a viable tool
should be.
I have actually used the full function IBM version quite a lot in the
past, and I agree with the evaluation provided in this thread by an
experienced translator ("Samarkand"):
http://www.proz.com/forum/cat_tools_technical_help/174563-experiences_of_opentm2.html
The UI may not be incredibly polished, but it allows the translators
to do what they need to do, and do it fast.
On Jul 20, 10:19 am, Stephen Holmes <stiofan.macthom...@gmail.com>
wrote:
> > >>> On 19 juil. 10, at 15:40, GerhardF wrote:
>
> > >>>> Now I'm wondering about the communities opinion on how a new
> > GUI could
> > >>>> look like:
> > >>>> - looking like a "standard WINDOWS application"?
> > >>>> - looking like a JAVA application?
> > >>>> - looking like an ECLIPSE application?
> > >>>> - should it be a GUI that can be changed via "faces" to fit
> > into a
> > >>>> user's personal
>
>
> read more »
Tex Texin
not what was released for open source.
I didn't find Samarkand's remarks that compelling either because at least
today tools do provide multiple dictionary and TM support, and allow font
customization etc. Maybe TM2 was advanced a few years ago.
I would prefer to see an enumeration of the productivity functions/features
that each user profile likes/dislikes to make sure they are
protected/revised going forward, rather than the high level debate over
which tool is better.
tex
--
Stephen Holmes
Steve
--
Stephen Holmes
(Sent from iPhone)
Cell : +353 86 7 944 955
Main: +353 1 254 1353
Skype: stephen.holmes
Elsebeth Flarup
I agree that the focus should be on function.