LAS 1.4 Debate

There is a new version of LAS in the works – LAS 1.4. There appears to be some concern about the fact that it is not backward compatible. See the recent blog comment.

I am doing some research into this issue. In the meantime if you have any information that you can provide please let us know.

I had heard that there was some debate around this release at ASPRS. Stay tuned.

This entry was posted in Orgs, Standards. Bookmark the permalink.

6 Responses to LAS 1.4 Debate

  1. The current proposal for a LAS 1.4 specification breaks away in compatibility from the LAS 1.0 – LAS 1.3 family. In this open letter I explain why this is a terrible idea and show how easy it is to fix the current proposal. So far the ASPRS committee has ignored these concerns that are shared among a number of reputed LAS developers. What can you do? Contact the ASPRS LAS committee, ask for a copy of the proposed specification, and join an open discussion.

    The proposed LAS 1.4 specification is not backwards compatible and also breaks the forward compatibility of the LAS 1.X family that, for example, had allowed a LAS 1.1 reader to read a LAS 1.3 file that contains only points of type 0 or 1. Furthermore, it – for the first time – requires LAS programmers to implement at least two different LAS specifications if they want their software to read and write LAS content other than LAS 1.4.

    Fixing this broken LAS 1.4 specification requires nothing more than to rearrange a few count and offset fields in the LAS header as detailed in this open letter.

  2. Lewis Graham says:

    Mr. Isenburg’s input will be considered during the public review cycle of the LAS 1.4 specificaion (which has not yet occurred). His issue is basically the argument of “should PowerPoint 2003 be able to successfully read/edit a PowerPoint 2007 file?” A requirement for forward compatibility is a bit counterproductive. The new format revision generally introduces some new features and there is usually a strong desire to stop producing data in the prior format. In the case of LAS 1.4, this will be 256 classes.

    In several forums, Mr. Isenburg has expressed concern regarding the closed nature of the LAS Working Group of the ASPRS. For everyone’s benefit, below is the operating protocol of the LWG:

    The LAS Working Group (LWG) is responsible for maintaining and updating the LAS point cloud data specification. The LAS Data Specification was conceived as a common format for vendors of kinematic laser scanners to transfer point cloud data to processing software. The primary focus of LAS today remains the vendors of kinematic laser scanning hardware. The entire process of revising the LAS specification is purposely designed to operate very quickly so as to accommodate new hardware features at their time of release. For this reason, it is a closed process prior to the public review period.

    The LWG is guided by a Chairperson. The Chairperson is appointed by the Division Director, ASPRS LIDAR Division.

    Members of the LWG are appointed by the Chairperson of the LWG. In general, a member must have a direct interest in the LAS format, meaning they are a commercial hardware and/or commercial (including Open Source) production software vendor. An LWG member must be a voting member of the ASPRS. The number of members is generally kept at 12 to 16 persons to make expeditious management of revisions possible.

    It is usually a member of the LWG that drives a new revision of the specification (but of course outside requests are always collected up and circulated to the group when a new revision is in progress).. For example, Leica drove the waveform addition that resulted in LAS 1.3 because they were adding this capability to their hardware and had to have file output support.

    One or mare LWG members authors a new revision draft. When the draft is to a reviewable state, it is circulated amongst the core group via email (with all members copied) with subsequent drafts authored until everyone within the LWG are in agreement. The group has not, to this point, had an issue sufficiently contentious or divisive to actually require a vote but this would be the final arbitration in such an event.

    When the LWG has converged on a consensus draft, the draft gets sent up to the ASPRS Board of Directors. The Board of Directors then sends the draft out to the full ASPRS membership for a review period of 30 days. Comments are sent directly to the Chairperson, LWG. While technically persons making comments are required to be voting members of the ASPRS, this rule is not currently enforced.

    Editorial comments that add clarity to the specification are generally accepted. Editorial comments that simply reflect differences of opinion of wording but make no material difference in the specification are generally rejected. Comments that are functional in nature will follow one of two paths: If the comment breaks the spirit of the revision (for example, the LWG is doing the change to support waveform but the comment deals with adding some totally new capability), the request is denied with a note back to the commenter that it will be added to consideration for the next revision cycle. If the functional comment is related to the current revision, it is circulated to the LWG for consideration (for example, someone wants to change the definition of mirror angle). The draft is revised accordingly. After the comment period, the final draft along with the adjudicated comments are sent to the ASPRS BOD for ratification. At this point, the new revision supersedes the prior revision.

  3. Michael P. Gerlek says:

    Lewis:

    * I’m glad you have put together a formal process document; while it does seem to leave an inordinate amount of control in the hands of the chairperson, it is a good first step.

    * Can you address Martin’s point that 1.4 does not provide the same level of forwards/backwards compatibility that previous releases have had (e.g. a LAS 1.1 reader could support a limited set of putative 1.3 inputs)? I’m wondering what the justification is for departing from this convention you’ve had until now.

    * Can you please list what vendors are currently represented on the committee?

    -mpg

  4. My argument is not “should PowerPoint 2003 be able to successfully read/edit a PowerPoint 2007 file?”.

    My argument is “should PowerPoint 2007 not be able to write files that PowerPoint 2003 can read, whenever the content of the presentation only uses PowerPoint 2003 functionality … especially when all it takes to do so is a slight rearrangement of 8 integer values in the file header?”.

  5. Michael P. Gerlek says:

    In the interest of full disclosure, Lewis should have noted that the “Division Director, ASPRS LIDAR Division” responsible for appointing the LWG chairperson is, in fact, Lewis Graham himself.

    -mpg

  6. Michael,

    I think this is just a temporary situation. The ASPRS LAS committee existed long before the newly announced ASPRS LIDAR Division, so it is more of a promotion up than a self-appointment down, if that makes sense 🙂

    Howard

Leave a Reply

Your email address will not be published. Required fields are marked *

*

This site uses Akismet to reduce spam. Learn how your comment data is processed.