Ad
Ad
Ad
Pages: « 1 [2] 3 »   Bottom of Page
Print
Author Topic: LR 1.0  (Read 16241 times)
Schewe
Sr. Member
****
Online Online

Posts: 5497


WWW
« Reply #20 on: August 17, 2006, 12:35:57 PM »
ReplyReply

Quote
Boku's question to Opgr about whether plugin developers are getting the API to make alternative plugins is also interesting and unanswered.
[a href=\"index.php?act=findpost&pid=73659\"][{POST_SNAPBACK}][/a]

Then you haven't been reading the Lightroom forums...Adobe has stated there -WILL- be an SDK, which they will work on -AFTER- version 1.0 ships. They will not spend time working on an SDK until such time as the core engine has been finalized and it would be foolish for 3rd party developers to be working on development for something that will likely change up till the final shipping version. This is a 1.0 product (meaning brand spanking new) and as such, nobody outside the actual engineering team will get access to the inner workings until it's done.
Logged
Fred Ragland
Full Member
***
Offline Offline

Posts: 155


WWW
« Reply #21 on: August 17, 2006, 01:20:01 PM »
ReplyReply

Quote
Have you seen this article or these discussions?
[a href=\"index.php?act=findpost&pid=73662\"][{POST_SNAPBACK}][/a]

Thank you for the references John.

Peter Krogh comments in the article that Lightroom "will undergo significant changes over the coming months" with regard to tracking and managing image files.  In the forum discussions, he suggests that "If you find that you like the way Lightroom converts RAW files, and that you like the way you can browse images with it, then you should keep any eye out for its capabilities as an Asset Manager."

...good things come for those who wait.
Logged
jani
Sr. Member
****
Offline Offline

Posts: 1604



WWW
« Reply #22 on: August 17, 2006, 01:37:49 PM »
ReplyReply

Quote
Then you haven't been reading the Lightroom forums...Adobe has stated there -WILL- be an SDK, which they will work on -AFTER- version 1.0 ships. They will not spend time working on an SDK until such time as the core engine has been finalized and it would be foolish for 3rd party developers to be working on development for something that will likely change up till the final shipping version.
I disagree. It would be good for plugin developers to work with the SDK during the beta process, so that the 1.0 won't be a plugin developer's beta process.
Logged

Jan
61Dynamic
Sr. Member
****
Offline Offline

Posts: 1442


WWW
« Reply #23 on: August 17, 2006, 02:11:36 PM »
ReplyReply

Quote
I disagree. It would be good for plugin developers to work with the SDK during the beta process, so that the 1.0 won't be a plugin developer's beta process.
[a href=\"index.php?act=findpost&pid=73673\"][{POST_SNAPBACK}][/a]
Unfortunately that just won't work until the means in which modules (not plugins - they want to distinguish between the two) are implemented is set. There are no guidelines for it yet. Any work a third-party developer may do for a module could turn out to be a foolish effort once everything is set in stone by 1.0 release.
Logged
David Mantripp
Sr. Member
****
Offline Offline

Posts: 696


WWW
« Reply #24 on: August 17, 2006, 02:30:57 PM »
ReplyReply

Quote
Unfortunately that just won't work until the means in which modules (not plugins - they want to distinguish between the two) are implemented is set. There are no guidelines for it yet. Any work a third-party developer may do for a module could turn out to be a foolish effort once everything is set in stone by 1.0 release.
[a href=\"index.php?act=findpost&pid=73679\"][{POST_SNAPBACK}][/a]

Especially if they're going to implement selective edits....   Since Photoshop plug-ins are designed to modify pixels, and Lightroom modifies metadata, I would imagine that except at a very high level of abstraction, plug-ins are going to be programmatically very different.

Maybe they're using PixelGenius to at least road test some 3rd party module API concepts. One could imagine that this might be the case... One could even hope so.
Logged

--
David Mantripp
http://www.snowhenge.net
jani
Sr. Member
****
Offline Offline

Posts: 1604



WWW
« Reply #25 on: August 17, 2006, 03:40:27 PM »
ReplyReply

Quote
Unfortunately that just won't work until the means in which modules (not plugins - they want to distinguish between the two) are implemented is set. There are no guidelines for it yet. Any work a third-party developer may do for a module could turn out to be a foolish effort once everything is set in stone by 1.0 release.
Once again, I disagree.

There are two issues here, and you are both confusing them:

1) The work third-party developers put into developing modules.
2) The feedback third-party developers can give to the SDK and API, so that the SDK and API can be stable from the release version.

It's the second that concerns me, and for these developers, Lightroom 1.0 will feature an alpha or beta API or SDK, unless third party developers have been involved already.
Logged

Jan
Schewe
Sr. Member
****
Online Online

Posts: 5497


WWW
« Reply #26 on: August 17, 2006, 03:47:30 PM »
ReplyReply

Quote
I disagree. It would be good for plugin developers to work with the SDK during the beta process, so that the 1.0 won't be a plugin developer's beta process.
[a href=\"index.php?act=findpost&pid=73673\"][{POST_SNAPBACK}][/a]

Are you a software developer?

I am and I can tell you that trying to engineer for a moving target is both difficult (at best) and a real waste of time and engineering resources...things at the core engine level are changing all the time...there is no guarantee that ANY engineering done now could even be used on the final 1.0 product.

No, working on the full SDK -AFTER- the final version is locked down is the only thing that makes sense...which is why the Lightroom engineers are taking this position.

Frankly, Lightroom is really at the pre-alpha stage. The only reason Adobe calls it "beta" is they figured most people wouldn't understand a "pre-alpha" concept. This is software engineering with all the guts exposed. And until they stuff the guts in and sew it back up, there's no reason for others sticking their hands in.
Logged
Schewe
Sr. Member
****
Online Online

Posts: 5497


WWW
« Reply #27 on: August 17, 2006, 03:52:37 PM »
ReplyReply

In the meantime, if 3rd parties are interested in getting prepared, they would do well to learn Lua...

Lua.org

You see, even the programing language is pretty radical...
Logged
john beardsworth
Sr. Member
****
Offline Offline

Posts: 2829



WWW
« Reply #28 on: August 17, 2006, 03:56:53 PM »
ReplyReply

Or if you want to dig around a little, why not take a look at the format of the templates and the SQLite database is open to prying eyes too - on Windows you can connect via ODBC if you want a peek. Yet when so much about the program is clearly missing or in flux, I wouldn't waste too much time yet.

John
« Last Edit: August 17, 2006, 04:10:10 PM by johnbeardy » Logged

David Mantripp
Sr. Member
****
Offline Offline

Posts: 696


WWW
« Reply #29 on: August 17, 2006, 04:19:00 PM »
ReplyReply

Quote
there is no guarantee that ANY engineering done now could even be used on the final 1.0 product.

Jeff, surely it isn't quite that black & white.  The whole purpose of abstraction is to decouple the interface from the implementation. An API can, at least in theory, remain quite stable whilst the implementation behind changes radically. A good example of this is the QuickTime API - or at least, it used to be.

And of course, no risk, no gain - a developer might be prepared to take a calculated risk developing on an unstable API in order to potentially be first to market...

I would have thought that by now there must be at least some basic ground rules set in Lightroom, so at least some kind of beta API could be built. I'd also venture that if the product is to be released in 2006, then given the requirements of a test cycle, of documentation, physical product design and all the rest of the stuff that a company like Adobe does, then surely there must be something very close to an internal Beta of Lightroom 1.0 (which of course might have very little in common with the Lightroom Public Beta).

I have to say I tend to agree that IF 1.0 was to have an API, then it must be in Beta now. But under strict NDA.

Nothing like wild speculation as an alternative to doing something useful with one's time, is there ?
Logged

--
David Mantripp
http://www.snowhenge.net
jani
Sr. Member
****
Offline Offline

Posts: 1604



WWW
« Reply #30 on: August 17, 2006, 04:52:29 PM »
ReplyReply

Quote
Are you a software developer?
Among other things, yes, although it isn't currently my main professional activity.

Quote
I am and I can tell you that trying to engineer for a moving target is both difficult (at best) and a real waste of time and engineering resources...things at the core engine level are changing all the time...there is no guarantee that ANY engineering done now could even be used on the final 1.0 product.
That's close to tautological, if one assumes that this part of the product would be in flux all the time up to a 1.0 release.

However, leaving that part of the product in flux for so long is bad software engineering, as you well know.

I've had to work as a third party developer against software like that (and when I write "against", I mean "against"), and it's something that I'd rather not do, but hey, it paid good money to fix the problems creatively.

Quote
No, working on the full SDK -AFTER- the final version is locked down is the only thing that makes sense...which is why the Lightroom engineers are taking this position.
It only makes sense if you assume that SDK and API are perfect in their first released incarnation.

If not, you'll end up with patch releases and API changes even less consistent than those of Microsoft Office, or you'll stick to an API that most likely has glaringly obvious faults and problems but which you'll keep around for ages because you want backwards compatibility (like the legacy Windows 16-bit API).

This is to the detriment both for third party developers as well as end users.

Quote
Frankly, Lightroom is really at the pre-alpha stage. The only reason Adobe calls it "beta" is they figured most people wouldn't understand a "pre-alpha" concept. This is software engineering with all the guts exposed. And until they stuff the guts in and sew it back up, there's no reason for others sticking their hands in.
Well, it's not quite pre-alpha, because it appears to be mostly stable both feature-wise and in terms of data corruption and runtime.

Also, very little of the guts are exposed, as opposed to e.g. the GIMP, where you can feel free to dig around in the guts all you want (ewwww).

As David (drm) mentions in his response, yes, there could be a beta API available by now, and it could be used by third party developers to provide presumably valuable feedback on the API.

The SDK could wait a bit longer, of course, since that in itself depends on the API.

But who knows, maybe there will be a market for a third-party API wrapper ...
« Last Edit: August 17, 2006, 04:53:19 PM by jani » Logged

Jan
Schewe
Sr. Member
****
Online Online

Posts: 5497


WWW
« Reply #31 on: August 17, 2006, 05:29:49 PM »
ReplyReply

Quote
I would have thought that by now there must be at least some basic ground rules set in Lightroom, so at least some kind of beta API could be built.
[a href=\"index.php?act=findpost&pid=73695\"][{POST_SNAPBACK}][/a]

Uh. . .I couldn't possibly comment on that silly statement...

::hint, hint-nudge, nudge::

:~)
Logged
John Camp
Sr. Member
****
Offline Offline

Posts: 1259


« Reply #32 on: August 17, 2006, 07:06:54 PM »
ReplyReply

Quote
I would have thought that by now there must be at least some basic ground rules set in Lightroom, so at least some kind of beta API could be built. I'd also venture that if the product is to be released in 2006, then given the requirements of a test cycle, of documentation, physical product design and all the rest of the stuff that a company like Adobe does, then surely there must be something very close to an internal Beta of Lightroom 1.0 (which of course might have very little in common with the Lightroom Public Beta).
[a href=\"index.php?act=findpost&pid=73695\"][{POST_SNAPBACK}][/a]

To quote Schewe back up the thread, "what part of 'due to ship end of 2006' you don't understand?"

JC
Logged
boku
Sr. Member
****
Offline Offline

Posts: 1493



WWW
« Reply #33 on: August 17, 2006, 07:32:19 PM »
ReplyReply

OK, from all the churn my little question to Opgr created, I learned some things. (I am a software architect.)

1) This is not Beta, but pre-Alpha.
2) [big red flag]This software was not designed, but developed. Designed software starts by defining the APIs, stubbing them out, the gradually realizing the underlying code.[/big red flag]
3) We are talking about modules, not plug-ins (I already remember hearing that). Modules are meant to replace, not augment, functionality in Lightroom. Personally, I feel the market will not readily accept that architectural decision.
4) Everyone here is pissed off.
5) This thread started off that way with "an attitude" about Adobe.
6) I will welcome Lightroom 2.0, but probably buy 1.0 because I need to stay on the edge for my training enterprise.
7) Jeff Schewe seems to be taking any negative comment about Lightroom to heart. I guess thats OK - at least he's passionate. No harm in that.
« Last Edit: August 17, 2006, 07:33:47 PM by boku » Logged

Bob Kulon

Oh, one more thing...
Play it Straight and Play it True, my Brother.
Josh-H
Sr. Member
****
Offline Offline

Posts: 1910



WWW
« Reply #34 on: August 17, 2006, 08:47:54 PM »
ReplyReply

Quote
Jeff Schewe seems to be taking any negative comment about Lightroom to heart. I guess thats OK - at least he's passionate. No harm in that

Agreed - he clearly has quite some investment of time in LR - which is very conmforting to me as it means he wants to make sure its right. I have watched the LR tutorials on the web, on LL Journal, listended to the pod casts etc.. and its obvious Jeff [sorry for talking about you Jeff!] is passionate about LR. As for taking it to heart - we all take critisism to heart when we are passionatte about something.

I only HOPE that the LR engineers are listening to some of the requests that are being made - both here and on the LR forums.

My biggest concern at the moment [its noted in another post] and I would be interested in Jeff's thoughts - is that the RAW conversions from LR are not as good as those from Canon's DPP 2.1. They are good - but just not as deeply saturated or as sharp as those from DPP2.1.

Any chance of using the Canon algorithim for processing Canon RAW files in LR?
Logged

61Dynamic
Sr. Member
****
Offline Offline

Posts: 1442


WWW
« Reply #35 on: August 17, 2006, 10:28:12 PM »
ReplyReply

Quote
3) We are talking about modules, not plug-ins (I already remember hearing that). Modules are meant to replace, not augment, functionality in Lightroom. Personally, I feel the market will not readily accept that architectural decision.
[a href=\"index.php?act=findpost&pid=73718\"][{POST_SNAPBACK}][/a]
I'm quite certain modules will not be replacing functionality. They - in all effect - are plugins as the rest of the world knows them but Adobe isn't using that term in order to differentiate LR's extensibility from that of Photoshops.

Each of the sections in LR right now (Develop, Web, Print, Etc.) are modules themselves. Based off what I've read/heard thus far, third-parties will develop their own modules that will appear in that same list at the top-right.

It's just that exactly how those will be executed hasn't been set in stone (publicly, if at all). As has been mentioned in the podcasts, Adobe wants to be very careful how they execute any feature in LR because once they set a way of doing things for 1.0, it'll have to remain that way for the life-span of the software.
Logged
Schewe
Sr. Member
****
Online Online

Posts: 5497


WWW
« Reply #36 on: August 17, 2006, 10:35:29 PM »
ReplyReply

Quote
...is that the RAW conversions from LR are not as good as those from Canon's DPP 2.1. They are good - but just not as deeply saturated or as sharp as those from DPP2.1.

Any chance of using the Canon algorithim for processing Canon RAW files in LR?
[a href=\"index.php?act=findpost&pid=73728\"][{POST_SNAPBACK}][/a]

Lightroom is using the Camera Raw pipeline, if "color or saturation" is "better" in DPP, I would argue you don't know how to set Camera Raw/Lightroom settings to adjust for the color & tone you want. A common problem I might add.

As far as "sharpness" that depends on the DPP settings and the Camera Raw settings. Camera Raw has a tendancy to product less "sharp" (but very sharpenable) and slightly noisier (unless you view @ 200% and fine-tune the luminance noise setting) at defaults. Canon also has a tendancy to really step on the low end of the tone curve to hide noise while Camera Raw produces flatter results that show the noise.

The demosaicing and sharpening/noise reduction is something Thomas wants to work on for the next major rev and with Micheal Jonsson from pixmantic as well as former Photoshop engineer Zalman Stern on board the Camera Raw team, there should be talent and people to make major improvements in Camera Raw/Lightroom raw processing.

As to the second part-using Canon's algorithm, no, Canon's stuff really doesn't preform all that well.  The only advantage that Canon has is using the proprietary metadata to do capture fixes that 3rd parties can't use unless they want to only use the Canon SDK and that would mean Camera Raw couldn't really use ANY of it's color/tone settings as currently designed, which just ain't gonna happen.
Logged
Schewe
Sr. Member
****
Online Online

Posts: 5497


WWW
« Reply #37 on: August 17, 2006, 10:37:21 PM »
ReplyReply

Quote
It's just that exactly how those will be executed hasn't been set in stone (publicly, if at all). As has been mentioned in the podcasts, Adobe wants to be very careful how they execute any feature in LR because once they set a way of doing things for 1.0, it'll have to remain that way for the life-span of the software.
[a href=\"index.php?act=findpost&pid=73735\"][{POST_SNAPBACK}][/a]

Which is why they will not allow 3rd parties in earlier than after 1.0 is finalized. Only then will the work of producing the SDK begin.
Logged
David Mantripp
Sr. Member
****
Offline Offline

Posts: 696


WWW
« Reply #38 on: August 18, 2006, 02:18:47 AM »
ReplyReply

Quote
To quote Schewe back up the thread, "what part of 'due to ship end of 2006' you don't understand?"

JC
[a href=\"index.php?act=findpost&pid=73715\"][{POST_SNAPBACK}][/a]

I'm sorry John, but I don't see the point of your, er, contribution. My whole post was based on the premise that I accept the announced due date.  


You wouldn't be "trying to suck up to teacher", would you ?  Heavens forbid  :-)
Logged

--
David Mantripp
http://www.snowhenge.net
David Mantripp
Sr. Member
****
Offline Offline

Posts: 696


WWW
« Reply #39 on: August 18, 2006, 02:30:19 AM »
ReplyReply

Those of us who know, or imagine we know, something about software engineering really should have a look at the Lua site (link posted by Schewe above).  And also Mark Hamburg's presentation slides given at the Lua conference held at Adobe (don't bother unless you're interested in software engineering - there's no secret info on Lightroom)

Whilst I don't pretend to have spent 10% of the time I'd need to understand 10% of this, I do get a few hints that possibly this language brings with it some powerful data sharing mechanisms, as well as a different methodology, which may have some impact on traditional software development cycles - such as, reading between the lines, greatly reduced testing overhead.  So, perhaps those of us pontificating on software matters (we're safe enough here...) might be on the wrong track when second guessing Adobe. Especially me.

Anyway, enough of that.  This is a photography forum.
Logged

--
David Mantripp
http://www.snowhenge.net
Pages: « 1 [2] 3 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad