FLEx on Mac status?

[Patricia A]

This question hasn't been asked for about a year now, so I thought I'd resurface it - any news on FLEx for Mac?  Now that FLEx for Linux is out, is the Mac project something actively under development? Or is this a project that has been set aside for the time being?

Thanks!

Patricia

[Alexandre Arkhipov]

So, isn't there any news on FLEx for Mac?

Thanks,
Alexandre

--
You are subscribed to the publicly accessible group "FLEx list".
Only members can post but anyone can view messages on the website.
---
You received this message because you are subscribed to the Google Groups "FLEx list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to flex-list+...@googlegroups.com.
To post to this group, send email to flex...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/flex-list/27648b71-7399-481f-93bd-0ddc93d8b736%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Griz]

Hi Alexandre,

Since I'm not on the FLEx team, I can't give you an official reply. But, I can tell you that there have been thoughts of doing so. Putting FLEx on Linux has put the FLEx team in more of a position to do so. The effort to make FLEx cross-platform for Windows and Linux has demanded a huge commitment of personnel resources, since the goal has been to keep the versions of FLEx concurrent with each other.  There is a "wait and see" posture. We are waiting to see how much of a user base develops for Linux. If a user base does not develop for Linux, we will be wary of committing resources there in the future. Unfortunately, at this point, there probably won't be a Mac version forthcoming, unless there was a big demand for it.  Again, we would have to commit personnel resources to it.  We are also strained by the new demand to develop programs for Android. Of course, since it is Open Source, there is nothing preventing someone from developing a Mac version themselves. If you would consider donating to the development of FLEx, that would be certainly appreciated.

With warm regards, Ken

To view this discussion on the web visit https://groups.google.com/d/msgid/flex-list/543C8836.3040102%40mail.ru.

[Gwendolyn Hyslop]

Hi Ken,

With regard to your last point, are you suggesting a donation to the development of FLEx might help get a mac version developed? Such a development would be such a boon for my work that I would certainly consider a donation if it would help things moving. How much do you think it would cost? 

Best,

Gwen

To view this discussion on the web visit https://groups.google.com/d/msgid/flex-list/543C8F11.8010706%40gmail.com.

[David Rowbory]

It seems to me that the most useful future avenue for the development of Flex would actually be developing a cross-platform / portable API layer on which various GUIs can be build appropriate for each platform. Even if the API starts off very limited and simple, then a simple Fieldworks-compatible Mac app, iOS app, Android app etc would be much more easily achievable by community developers than a complete port. The low-hanging fruit to my mind would be Send/Receive and some interrogation of the lexicon. 

[Griz]

Hi Gwen,

That's an interesting question. In the case of one plugin for FLEx, the project was outsourced, and one SIL person was responsible for managing its development. I don't know how much the whole project would cost. Thousands, if not millions of dollars. Probably, a feasibility study would be in order to estimate the cost. Of course, software like that has to be maintained, so there is an ongoing cost, and commitment as well.

FLEx is free, but not cheap.

Kind regards,

Ken

To view this discussion on the web visit https://groups.google.com/d/msgid/flex-list/76B9934A-D7EB-4E84-8842-5D4454B26852%40gmail.com.

[Randy Regnier]

On 10/14/2014 4:05 PM, David Rowbory wrote:
> It seems to me that the most useful future avenue for the development
> of Flex would actually be developing a cross-platform / portable API
> layer on which various GUIs can be build appropriate for each
> platform. Even if the API starts off very limited and simple, then a
> simple Fieldworks-compatible Mac app, iOS app, Android app etc would
> be much more easily achievable by community developers than a complete
> port. The low-hanging fruit to my mind would be Send/Receive and some
> interrogation of the lexicon.

What you describe is possible, and perhaps desirable, but it isn't trivial.

The architecture (by Alistair Cockburn) to do it is called "Hexagonal
Architecture <http://alistair.cockburn.us/Hexagonal+architecture>" (aka:
'Ports & Adapters'). This architecture was a big influence on how we
managed to dump SqlServer (Windows only) as the data storage system in
FLEx and replace it with something that was more cross-platform
friendly. In the early testing stages, I had worked up 6-8 different
persistence systems that could be easily swapped out. Performance
testing then reduced the list of options to the two that are currently
being shipped. I had always hoped to do the same overhaul to the UI
using that same architecture, but I never got one of those "Round
Tuits". :-) There are parts of the FLEx code that are OS-specific that
are not UI related, so those could also need some attention on a Mac.

Randy Regnier

[Griz]

Hi Randy, David,

Okay, maybe we can downgrade my estimate from millions to thousands in
the case of porting the GUI, and parts of the code. Still, it would
require a team of developers to be assigned to the project, and it would
affect both Windows and Linux versions. New features would probably be
put on hold until that the portable API layer was completed. Maybe then,
as you suggest, community developers could take over from there.

Thanks for weighing in with your valuable insights.

K

[Kari Valkama]

Hi K or Griz, 

Greetings from Finland. 

If we talk about the cost of development, we need to have a context, i.e. we should not talk about Mac development in vacuum. The context here is development of software in general, independent of the platform. Therefore we should compare the cost of software development in general. All code needs developers, the Linux version and the Windows version, too. Whatever we do costs money and developers are needed to do all the software. (The cost can go down with good planning. For example, if one writes a Windows version with porting in mind from the beginning, then the cost of porting usually goes down significantly.) 

The fact that your estimation went down from millions to thousands, causes me pause. It may be that I have misunderstood you, but I get the impression that your argument is that, since the Mac version costs money and needs developers, perhaps SIL should not do it. 

My email is a response to that idea. 

Here is my response: 

I think we can all agree that the Windows version of FLEx cost the most, since it was done first. The Linux version cost less, since it benefited from the work already done for the Windows version. And the Mac version will cost the least, since it will benefit from the work already done on Windows and Linux. 

Additionally, I believe that:  

If we have the will, we will have the way. 

I think there is a need for the Mac version of FLEx. It is my understanding that there are at least as many Mac users as there are Linux users, who would like to use SIL software natively. 

My 2 cents worth. 

Yours, 

Kari 

-- 
You are subscribed to the publicly accessible group "FLEx list".
Only members can post but anyone can view messages on the website.
--- 
You received this message because you are subscribed to the Google Groups "FLEx list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to flex-list+...@googlegroups.com.
To post to this group, send email to flex...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/flex-list/543E0464.4070104%40gmail.com.

[Alexandre Arkhipov]

Thanks for all the comments in this thread!
Now I'd like to ask a more short-termed pragmatic question: those of you who had experience with using Flex (7 or 8) under some VM on a Mac (Wine, Parallels, whatever), what are your impressions of performance? Is it slower than running Flex on a native Windows system or not? What are estimated system requirements for a good performance?
(In other words, if I'm buying a new laptop and wish to use Flex on it, shall I consider Mac or better stick to a PC+Windows, and what Mac model would be suitable if the previous answer is yes?)

Thanks in advance,
Alexandre


Wed, 15 Oct 2014 10:26:14 +0300 от Kari Valkama <kariv...@mac.com>:

To view this discussion on the web visit https://groups.google.com/d/msgid/flex-list/F991CBD0-38DD-4CE5-A827-542EB13407BF%40mac.com.

For more options, visit https://groups.google.com/d/optout.

С уважением,
Alexandre Arkhipov
sar...@mail.ru

[Griz]

Hi Kari,

I like your reasoning. Yes, a Mac version would benefit from the Windows, and Linux development which has already been done. I'm not saying a Mac version shouldn't be done, I'm just saying it needs to be studied, and as you say, good planning is less costly than poor planning.

It would be interesting to do a poll of the FLEx list, and FLEx users in general, what platform they actually use.  If there are a significant number of Mac users out there (I'm not sure what the tipping point is), perhaps the Software Board could be persuaded to devote resources to its development.

Best Regards, Griz K ;-)

To view this discussion on the web visit https://groups.google.com/d/msgid/flex-list/F991CBD0-38DD-4CE5-A827-542EB13407BF%40mac.com.

[Kari Valkama]

Hi Alexandre, 

I have mostly used Paratext with my Mac, so I cannot answer based on experience with FLEx. 

Generally speaking, I would think that a Mac using a VM is not far behind of the Windows laptops. I remember reading comparisons between a Windows machine and a Mac using Windows. 

It is always good to have enough memory. The new Macs come with 8-16 GB (13”-15"). To make your computer future proof, 16 GB is better. (When I buy my next laptop, next year, I hope to get 16GB. The 15” Macbook Pros come with that, so I think I am set.) 

Secondly, the new Macs come with SSDs, which are also important for speed. 

Other than that the screen size is important. 15 inches is better than 13”. It is possible to get 16” external monitors which get their power from the USB3 port. I planned to buy such a monitor to our coordinator, but his Mac does not have a USB3, it is USB2. 

Hope this helps, 

Kari 

To view this discussion on the web visit https://groups.google.com/d/msgid/flex-list/1413385106.187393953%40f368.i.mail.ru.

[Jon C]

On 10/15/2014 9:58 AM, Alexandre Arkhipov wrote:
> those of you who had experience with using Flex (7 or 8) under some VM
> on a Mac (Wine, Parallels, whatever),

I think any true VM host should work fine. I've used a Windows XP guest
VM inside VirtualBox and it works well. So does Linux Mint Mate 16
inside VirtualBox. Parallels, VMWare, etc, should too. I can almost
guarantee that Wine (and likewise, Crossover) will not work any time
soon, and possibly never.

> what are your impressions of performance? Is it slower than running
> Flex on a native Windows system or not?

Yes, but I didn't find it all that painful after FW7 improved the
performance. You do need extra RAM so that both the host and guest have
enough, especially with newer guest OSes such as Win7. (My laptops have
had 3 or 4 GB each, but I had just had Win XP in the guest.)

Has anyone compared the speed and footprint of FLEx on a Linux VM as
compared to a WinXP or Win7 VM? I'm guessing the most efficient would be
a lightweight Ubuntu derivative such as Lubuntu or Linux Mint Mate, or
maybe Xubuntu. (I suppose a non-Ubuntu derivative of Debian might work
too, but you'd want to confirm that with the developers.)

> What are estimated system requirements for a good performance?

Even running natively, there are some things that are rather slow in
FLEx, like loading Dictionary view, and some sorts/filters. Different
people may have different tolerances there, but for ordinary editing
work, I think most "decent" laptops or desktops three years old or newer
could be made to do the job, given at least 3 or 4 GB of RAM. More is
better, of course, and a lot depends on how lean and clean you keep your
system, running apps/services, and anti-malware.

See also: http://fieldworks.sil.org/flex/faq/#run

Jon

[Jon C]

On 10/15/2014 9:58 AM, Alexandre Arkhipov wrote:

> if I'm buying a new laptop and wish to use Flex on it, shall I
> consider Mac or better stick to a PC+Windows

If you're planning to do much linguistic work on it, I'd recommend
Windows. It's nicer not to have to launch a VM every time.

BTW, one reason why it's so nice that FLEx now runs on Linux is that
Linux is a free OS, so a combination such as VirtualBox + Linux guest is
totally free. But one downside I just remembered that I hadn't mentioned
is this: "Guest Additions" for VirtualBox isn't as easy to get working
well with a Linux guest--and that can make it difficult for the
host/guest to see each others' shared folders.

Using a Windows OS as the guest requires buying an extra Windows license
(not cheap, especially on top of buying an expensive Mac).

Jon

[Gwendolyn Hyslop]

Hi all,

With regard to the poll idea, I think that's great. By my count there is an overwhelming number of Mac users who wish FLEx was easier for them to use. It's such a pain having to deal with parallels and windows or some other such configuration every time you set up a new computer. There would be a much bigger base of users if FLEx went straight to Mac. So yes! I think we should run a poll if that would help. 

To the question of buying a PC or a mac: I know I am a bit biased, but I still say mac even with the need to do a windows workaround. My current mac has lasted me 5 years and is still going strong. It rarely crashes despite the fact I carry it around like a trashy novel in my bag and almost never turn it off. It's quick, efficient, and virus free! My last Toshiba was in a much worse state after only 3 years. Sure, Macs are more upfront, but over time I don't think you lose because they don't have to be replaced as often. 

<stepping off soap box now> 

Gwen

To view this discussion on the web visit https://groups.google.com/d/msgid/flex-list/543E981A.4060108%40gmail.com.

[Kevin Penner]

I agree, Gwen! My wife and I have been using FLEx in Windows (XP, Win7 and Win8) virtual machines (Parallels and VMware Fusion) for years. The older versions of FLEx were pretty slow, but I think they were slow running natively on a Windows PC, too. FLEx 7 and 8 run much quicker and work fine in a VM, but the VM hogs a lot of system resources—still worth the hassle, though, IMHO. I have another buddy who’s doing the same thing, but we’d all love to run FLEx natively on our Macs!

Kevin

To view this discussion on the web visit https://groups.google.com/d/msgid/flex-list/FB533178-0DEE-4209-B7B7-4E57E726A7FF%40gmail.com.

[Hugh Paterson]

So, if this poll were to be put together, who would be asked? Windows users, who already have a solution, or linux user who already have a solution? But if the claim is that there would be more FLEx users if there were an OS X version then how does the Dev team reach them because they allegedly are not currently users because there is no solution for them... So I feel there are two separate embedded research questions here: 

1. How many current flex users deploy FLEx in a virtual environment?

1a. of these how many would rather deploy FLEx in OSX?

2. How many non-FLEx users would become FLEx users if an OS X native application existed?

By simply querying the FLEx Mailing List and asking for respondents, we are not actually asking FLEx users (because not all FLEx users are on the mailing list), nor are we reaching out of the FLEx "world of contact" to ask those people who are NOT currently FLEx users because there is no OS X version. If we really want we want to ask people who are also currently NOT Flex users we need an additional method of reaching that audience. 

This is the same research audience access challenge I had when I approached the archiving of FLEx databases question. Which can be read about here: http://hugh.thejourneyler.org/2013/lexical-database-archiving-questionnaire/

Full disclosure:

I currently deploy FLEx in Linux on OS X via Virtual-box. I wish there was a Greater range of resources on OSX from LSDev.

To view this discussion on the web visit https://groups.google.com/d/msgid/flex-list/FB533178-0DEE-4209-B7B7-4E57E726A7FF%40gmail.com.

[Eric Jackson]

Hi, all. As I understand it, part of the motivation for porting FLEx to Linux was not because there was a large group of people who were already using Linux, and who wanted to use FLEx natively. I admit, I was certainly an existing Linux user who was eagerly waiting for a native version of FLEx to use -- so such people did exist, I just think they would not have been considered a large group. There is probably a larger group of people who are currently users of OSX who are eagerly waiting for a native version of FLEx to use.

The motivation for having a version of FLEx that runs natively on Linux is that it then allows people to use FLEx who would not previously have had access to it on a licensed OS -- that is, many members of indigenous language communities who do not have the financial resources to purchase a computer with an actual licensed OS (whether that is Windows or OSX). Or, in some cases, community members may have had access to computers running Windows, but the versions they had access to were bootlegged or rife with viruses and security holes. (Wait, does that mean that there are versions of Windows that are not rife with security holes? But I digress... :-)

Projects like SIL's BALSA, whose goal (if I have it right) is to produce a lightweight appliance for the purpose of language documentation, language analysis, and language development by community members themselves, also benefit greatly from having a native FLEx on Linux, because now FLEx can be distributed and re-distributed as part of a complete open software system--OS and all--with a license which legally allows sharing. If you were a linguist who chose to purchase a Mac with OSX, you probably had ways to legally run FLEx. However, if it was not a readily sustainable model to have many language community members buying computers with legitimate copies of Windows in order to run FLEx, it certainly wasn't a sustainable model to have them buy Macs, either.

So, I don't mean to be disrespectful to Mac users, but I think that a proper consideration of the motivation for previously porting FLEx to Linux was not based on a large existing user base, but was based on the desire to make FLEx legally and freely available to a new user base, namely, language community members without great financial resources and potentially without much technical computer training. But I also don't disagree with Kari, Ken, and others: If there exists a large FLEx user base on OSX, though (as I suspect there is, given the number of linguists that I have personally encountered who use Macs), then I think it also makes sense to consider meeting the needs of those users -- balanced against the human and financial resources that SIL has available for software development.

--Eric

-- 
Eric M. Jackson, Ph.D.          芮建生博士
SIL Intl, East Asia Group       世界少数民族语文研究院东亚部
http://sil.academia.edu/EricJackson/

[Dennis Walters]

Alexandre and others

I did this recently for my wife on her MacBook Pro 15" 4GB RAM 2.53 GHz
mid-2009, now running Mac OS X Mavericks:
Installed VirtualBox
Set up a virtual machine for Ubuntu Linux Precise (12.04)
Installed Ubuntu Precise (stock) in that virtual machine
Added SIL repositories and keys, etc.
Installed FieldWorks and needed accessories, such as FlexBridge and kmfl

This installation works. Performance is a bit slow. After a couple of
weeks, I ordered more RAM and upgraded the MacBook Pro to 8GB RAM (about
US$70). Gave the virtual machine 4GB RAM. I was surprised I did not see
a big performance jump for FieldWorks in the virtual machine.

In short, this installation runs somewhat slower than my native FW
install on my main linux machine (ThinkPad T400 2.53 GHz 8GB RAM). The
disadvantages on the Mac are:
-Extra layer of complexity and drag inherent in using VirtualBox or
other virtualizing package
-Extra time it takes to start VirtualBox then FW when you want to use it
-Slightly degraded FW performance from running inside a virtual machine

*It's possible that performance would be better under a "lighter"
distribution of Linux in the virtual machine.* That question needs to be
addressed.

I wasn't perfectly clear which distro would be optimal for performance.
I wanted highest reliability, and fewest complications for the install
so I went with a stock distro.

For your new machine, I think your choice will depend on what things
you spend most of your time doing. If using FW is your main thing or
near to it, you might consider something other than a Mac.

My wife could not consider any such thing, as many Mac users would also
say. The many Mac benefits are worth a lot to her. FW is one of many
things she does/uses. She puts up with the virtual install so that she
can have access to live language data sometimes. It's not the main thing
she uses all day.

Dennis

> <//e.mail.ru/compose/?mailto=mailto%3agrizzl...@gmail.com>>

>> <//e.mail.ru/compose/?mailto=mailto%3aflex%2dlist%2bunsu...@googlegroups.com>.

>> To post to this group, send email to flex...@googlegroups.com

>> <//e.mail.ru/compose/?mailto=mailto%3aflex%2dl...@googlegroups.com>.

>> To view this discussion on the web
>> visithttps://groups.google.com/d/msgid/flex-list/543E0464.4070104%40gmail.com.

>> For more options, visithttps://groups.google.com/d/optout.

>
> --
> You are subscribed to the publicly accessible group "FLEx list".
> Only members can post but anyone can view messages on the website.
> ---
> You received this message because you are subscribed to the Google
> Groups "FLEx list" group.
> To unsubscribe from this group and stop receiving emails from it,
> send an email to flex-list+...@googlegroups.com

> <//e.mail.ru/compose/?mailto=mailto%3aflex%2dlist%2bunsu...@googlegroups.com>.

> To post to this group, send email to flex...@googlegroups.com

> <//e.mail.ru/compose/?mailto=mailto%3aflex%2dl...@googlegroups.com>.

> To view this discussion on the web visit

> https://groups.google.com/d/msgid/flex-list/F991CBD0-38DD-4CE5-A827-542EB13407BF%40mac.com
> <https://groups.google.com/d/msgid/flex-list/F991CBD0-38DD-4CE5-A827-542EB13407BF%40mac.com?utm_medium=email&utm_source=footer>.

> For more options, visit https://groups.google.com/d/optout.
>
>
>
> С уважением,
> Alexandre Arkhipov
> sar...@mail.ru
>

> --
> You are subscribed to the publicly accessible group "FLEx list".
> Only members can post but anyone can view messages on the website.
> ---
> You received this message because you are subscribed to the Google
> Groups "FLEx list" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to flex-list+...@googlegroups.com

> <mailto:flex-list+...@googlegroups.com>.

> To post to this group, send email to flex...@googlegroups.com

> <mailto:flex...@googlegroups.com>.

> To view this discussion on the web visit

> https://groups.google.com/d/msgid/flex-list/1413385106.187393953%40f368.i.mail.ru
> <https://groups.google.com/d/msgid/flex-list/1413385106.187393953%40f368.i.mail.ru?utm_medium=email&utm_source=footer>.

[Oumar Bah]

Hi

My two cents: why not a web based solution? By this way, there would not be need for a Flex version for each operating system.

Oumar

>>    send an email to flex-list+unsub...@googlegroups.com

>>    <//e.mail.ru/compose/?mailto=mailto%3aflex%2dlist%2bunsu...@googlegroups.com>.
>>    To post to this group, send email to flex...@googlegroups.com
>>    <//e.mail.ru/compose/?mailto=mailto%3aflex%2dl...@googlegroups.com>.
>>    To view this discussion on the web
>>    visithttps://groups.google.com/d/msgid/flex-list/543E0464.4070104%40gmail.com.
>>    For more options, visithttps://groups.google.com/d/optout.
>
>    --
>    You are subscribed to the publicly accessible group "FLEx list".
>    Only members can post but anyone can view messages on the website.
>    ---
>    You received this message because you are subscribed to the Google
>    Groups "FLEx list" group.
>    To unsubscribe from this group and stop receiving emails from it,

>    send an email to flex-list+unsub...@googlegroups.com

>    <//e.mail.ru/compose/?mailto=mailto%3aflex%2dlist%2bunsu...@googlegroups.com>.
>    To post to this group, send email to flex...@googlegroups.com
>    <//e.mail.ru/compose/?mailto=mailto%3aflex%2dl...@googlegroups.com>.
>    To view this discussion on the web visit
>    https://groups.google.com/d/msgid/flex-list/F991CBD0-38DD-4CE5-A827-542EB13407BF%40mac.com
>    <https://groups.google.com/d/msgid/flex-list/F991CBD0-38DD-4CE5-A827-542EB13407BF%40mac.com?utm_medium=email&utm_source=footer>.
>    For more options, visit https://groups.google.com/d/optout.
>
>
>
> С уважением,
> Alexandre Arkhipov
> sar...@mail.ru
>
> --
> You are subscribed to the publicly accessible group "FLEx list".
> Only members can post but anyone can view messages on the website.
> ---
> You received this message because you are subscribed to the Google
> Groups "FLEx list" group.
> To unsubscribe from this group and stop receiving emails from it, send

> an email to flex-list+unsub...@googlegroups.com
> <mailto:flex-list+unsub...@googlegroups.com>.

> To post to this group, send email to flex...@googlegroups.com
> <mailto:flex...@googlegroups.com>.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/flex-list/1413385106.187393953%40f368.i.mail.ru
> <https://groups.google.com/d/msgid/flex-list/1413385106.187393953%40f368.i.mail.ru?utm_medium=email&utm_source=footer>.
> For more options, visit https://groups.google.com/d/optout.

--
You are subscribed to the publicly accessible group "FLEx list".
Only members can post but anyone can view messages on the website.
---
You received this message because you are subscribed to the Google Groups "FLEx list" group.

To unsubscribe from this group and stop receiving emails from it, send an email to flex-list+unsub...@googlegroups.com.
To post to this group, send email to flex...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/flex-list/543F83AE.8030003%40sil.org.

[Kari Valkama]

Hi all, 

Interesting discussion concerning porting FLEx to Macs (and perhaps later to iPads). 

I deleted the older posts in order to save bandwidth. I hope that is OK.  

1. Cost of hardware in terms of purchase price for local people

Some argue that Macs are too expensive for local people and thus one does not need to port FLEx to Macs. 

a) My first counter argument is that developing for Macs does not take away the possibility to use FLEx on Windows or Linux. Developing for Macs increases choice. Therefore this argument is moot. (Yet you hear this argument guide a lot. I guess it is psychologically easier to say no on behalf of other people.) 

b) My second counter argument is that FLEx is not a program designed for local people, it is designed for linguists. Therefore the argument is moot on this account also. It is WeSay that is designed for local people. So, in those rare cases, where local people are independently working on dictionaries, i.e. they are not affiliated with SIL nor any other organization, they would use WeSay, which is available on Windows and Linux. Mac version of FLEx would not affect them. 

c) Thirdly, in those SIL projects where dictionary is being produced for publication, the funds for the hardware do not come from the local people, but from outside, e.g. the SIL field workers' ministry expenses, and SEED Co. Therefore in those cases the cost of hardware rarely prohibits buying Macs (or iPads in the future). And if it does, there is always the possibility to use the Windows or Linux version of FLEx. Porting FLEx to Macs will not bankrupt any SIL member. 

d) Fourthly, initial hardware cost is not the relevant figure, the total cost of ownership is. Macs cost more, but they last longer and give less hassle. I am writing this with a 15” Early 2008 model MacBook Pro. I plan to buy a new computer next year, i.e. 7 years after this model was introduced. Ordinarily one would have purchased two Windows machines for the local people in that time. Hence, it is OK, if the Macs cost twice as much. 

e) Fiftly, productivity. The less hassle part mentioned in #4 above, is important, since that means better productivity (also happier users). SEED Co has calculated that an average SIL translation project costs 1-5 million USD. In that light the initial purchase price of hardware is insignificant and even more irrelevant. If use of Macs increases productivity (and hence speeds up a project) by only 1%, that would save 10.000-50.000 USD from an average SIL translation project. 

2. Cost of development in money and human resources

Some argue that one should not develop for Macs, since it costs money and human resources. 

a) I argued earlier that developing FLEx for Macs is cheaper than developing it for Linux or Windows, since the Mac development team can build on top of what the Linux porting team did. Linux and Unix are similar, since Mac OS X is based on Unix . If one wants to throw numbers around, I'll put my neck out and suggest that developing a Mac version of FLEx costs 10% of the development of the original Windows version. Feel free to correct me. 

b) Based on the calculation of SEED Co, if the port of FLEx would cost 100.000 dollars in LSDev member support and work expenses (just to throw up a large enough number), that would only be 0.2 - 1% of the cost of an average SIL translation project. If there are 1000 projects, then that would be at the highest one thousandth of a percent. I think we as SIL can afford to port FLEx to Macs, if we can afford to translate the Bible. 

c) It has often been argued that SIL does not have enough programmers for the Mac. SIL has 5.000-6.000 members. SIL has not recruited Mac-developers, since there is a decision that LSDev will not do Mac development. If SIL were to officially decide and then communicate to the relevant general public that LSDev is going to build software for Macs also, SIL would get Mac developers. As long as Mac developers are not recruited, there will always be a lack of Mac developers at SIL. Hence, this argument is moot. 

3. Number of users. 

One reason Linux was produced was the fact that the OS could be made to support exotic scripts at the system or OS level. (Windows cannot do that.) The legality issue was also a factor. But that does not mean that the number of users or potential users is not important. Software development is a service, so the numbers matter. One should not write software if there are no users. 

a) Mac users (and iPad users). 

I have seen more and more SIL people using Macs. I have seen more and more non-SIL people using Macs. In some universities in the US, about 50% of students use Macs. Market share statistics back up these observations: usage of Macs has been increasing the last 10+ years. (And the market share of iPads is huge, about a third of tablets are iPads.) The world is moving towards more tablet usage (compared to computer usage), thus porting FLEx to Mac will open the door to that world, too. 

b) FLEx users. 

At the moment there are SIL members and their co-workers who use FLEx in a virtual machine. Then there are SIL members and their co-workers who prefer Macs but since there is no Mac version, they use FLEx with a Windows machine instead. Had there been a Mac version, they would be using FLEx with a Mac at the moment. Then there are members in the pipeline, who are thinking about what computer to use on the field. (Many of them study in US colleges, where Mac usage is higher than elsewhere.) Many of them would like to continue using Macs. Choice is good. (Who knows, it might even help in recruiting non-software development personnel.) 

In addition to SIL members, there is a linguistic community out there who needs dictionary apps. A surprising number of new linguists are choosing Toolbox. I attended a summer school of linguistics in Sweden last summer, and Peter Austin from SOAS taught how to use Toolbox, not how to use FLEx. He also gave reasons for that. One negative feature of FLEx was that it does not run on Macs. Toolbox runs on Macs using Wine, no Windows license needed. I have heard that there are up to 2000 Toolbox users at the moment, the great majority outside of SIL. Some of them might switch to FLEx, if there were a Mac version of FLEx. 

Summary: if there is a will, there is a way. 

Yours,

Kari 

[Alexandre Arkhipov]

Thank you Dennis and everyone who replied! Your help was great.
I also hope this discussion will be beneficial for the future of FLEx on
OS X ;-)

Best,
Alexandre

[Jon C]

Eric made some good points in his summary.

Oumar, regarding your question, web-based editing is indeed a project too. Check out languageforge.org if you want to know more. I'm not sure how many features it will have--my guess is it's similar to WeSay but somewhere between WeSay and FLEx ("Supports most FLEx fields").

http://languageforge.org/learn_language_forge

My understanding is that currently it can import from LIFT but will soon be able to do send/receive just like WeSay does.

Jon

To unsubscribe from this group and stop receiving emails from it, send an email to flex-list+...@googlegroups.com.

To post to this group, send email to flex...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/flex-list/1413463007.49191.YahooMailNeo%40web172706.mail.ir2.yahoo.com.

[Oumar Bah]

A very interesting project Jon. I visited Language forge some time ago but I got the impression it works only for WeSay users. Maybe funding this project is a better option than embarking on a new expensive Mac project. 

Oumar

To view this discussion on the web visit https://groups.google.com/d/msgid/flex-list/54403A1A.5090207%40gmail.com.