This image was lost some time after publication.

Movable Type, the aging blog software made by San Francisco startup Six Apart, may be set to release a new 4.0 version on Tuesday. Barely. Which is bad, considering that Movable Type is still the company's cash cow. Byrne Reese, its product manager, accidentally forwarded an internal planning memo to an outside developers' mailing list. Reese says the memo, meant to help management decide on Monday whether or not the product is ready, is "intentionally pessimistic." But it has a laundry list of unresolved bugs. And Reese admits that some of the bugs will "dramatically slow" adoption of the new version. Despite that, he thinks the company can "manage" a release. And by manage, he means spin like a top: "We will need to be on our toes in regards to communication and PR to stay ahead of the curve with people who will say that we rushed the release." Here's a radical notion, for a blog-software company that claims to be about transparency: Why not just admit you did, in fact, rush the release? The full memo, after the jump:

The full memo:

A plan is coming together and I wanted to brief everyone on what it is and make sure that everyone is clear on what they need to do.


* Byrne has created a new release in FogBugz R16. This is a placeholder
right now for all bugs that will be fixed in 4.01.

* Byrne has moved all active issues in R15 to R16, leaving all resolved
issues in R16 to be verified and tested by QA.


* Maarten needs to check in any remaining EMEA translations.

* Byrne will create new builds of MT from the release-15 and place them on
koro64 for QA and Engineering.

* SAKK QA and Chris Hall is to smoke test these builds ASAP so that Byrne
can push them to ASAP.

* Once pushed, QA should continue to hammer the build to find any other
remaining issues.

MONDAY (US time)

* Some reports from the public suggest that there could be issues in
upgrading from 3.x that have lots of plugins installed. Once RC3 is pushed,
QA should devote a resource to investigating potential issues associated
with upgrading installs with lots of plugins.

* Product and Management team need to meet to review status and make a
go/no-go decision.


That's the plan. I will notify mt-dev as soon as builds are ready.

On 8/4/07 11:35 PM, "Byrne Reese" wrote:

> (sorry in advance for the long email)
> We have a big decision in front of us in the next day or two, and I wanted to
> outline from my perspective where I think we are with the product.
> Let me first provide an update:
> * RC2b was released Friday night - it contained a fix for a security
> vulnerability we found in mt-check.cgi
> There is a release branch called R15 (which would otherwise be RC3 if we
> decided to release RC3 as opposed to MT4). It contains the following issues
> (some are already fixed):
> P1 - 56173 - A design collapses on Confirm Publishing Configuration screen
> under IE6
> P1 (fixed) - 56174 - XHTMLification is not working in IE6
> P1 (fixed) - 56188 - Dynamic pages have wrong permalinks to entries
> P1 (fixed) - 56186 - Itemset actions broken due to unescaped label in
> javascript
> P2 - 55941 - IE6: Clicking 'arrow' icon broke the layout of the insert image
> screen
> P2 - 55896 - Cityscape banner images are cut off ...?
> P2 (fixed) - 56079 - Blog Templates: Standardize metadata on detail vs list
> pages
> P2 (fixed) - 56129 - Clicking Entry Name on "Edit Comment" screen yields error
> P2 (fixed) - 56136 - Delete blogs causes error with sqlite
> P2 (fixed) - 56087 - Blog Templates: Move Moderation message to comment
> confirmation page.
> P2 (fixed) - 56183 - Dynamic Author Monthly archive does not have his/her
> monthly archive list in dynamic publishing
> P2 (fixed) - 56189 - Can't write an entry on IIS5/6 with perlis.dll
> P3 - 56089 - L10N: Manage assets
> P3 - 56072 - L10N: System overview - List authors
> P3 - 56042 - L10N: Entry date
> P3 - 56080 - Blog Templates: Simplify Search Widget
> P3 - 56081 - Blog Templates: Standardize display of tags and categories
> P3 - 56082 - Blog Templates: Banner text overlapping illustrated headers
> graphics
> P3 - 56083 - Blog Templates: Content Nav Styling
> P3 - 56084 - Blog Templates: Style Search Form on search results listing
> P3 - 56085 - Blog Templates: Comment Confirmation: Link to "original entry"
> should refresh original entry
> P3 - 56086 - Blog Templates: Polish Styles on Preview Comment screen
> P3 - 56092 - Blog Templates: Add edit link if author is logged in via checking
> javascript cookie
> P3 - 56091 - Blog Templates: Style comments by authors on the blog
> P3 - 56090 - Blog Templates: Style author comments
> P3 - 56088 - Blog Templates: Add widgets to Comment Confirmation page
> P3 - 56217 - Rich text editor is adding editor-content.html to anchor links
> P3 (fixed) - 55690 - not including png as type
> P3 (fixed) 56171 - Filter message is wrong in list pings when it's filtered to
> show category trackbacks / entry trackbacks
> P3 (fixed) - 56028 - Store relative paths for asset file and URL in
> WXRImporter and AtomServer
> P3 (fixed) - 55947 - Warnings when DebugMode 1
> 56071 - L10N: TrackBack settings
> P3 (fixed) - 55683 - Republishing popup localized incompletely
> P4 - 55980 - Username isn't left-aligned in role users listing screen table
> P4 - 56004 - [IE6] strange navigation display on Grant Permissions:Select
> Blogs page
> P4 - 55203 - File upload options does not disappear when unchecking the option
> P4 - 55776 - Save, Delete and Preview button goes up a little when clicked in
> IE6/7
> P4 - 55438 - A part of Japanese comment is garbled under Expanded view mode
> P4 (fixed) - 55721 - MT::Component does not return an error when MT::Template
> cannot be created.
> P4 (fixed) - 55937 - IE6: Insert image screen is broken by hitting tab key
> time and again
> P4 (fixed) - 54746 - Wizard sometimes proceeds without mt-static successfully
> set up
> ——————
> Now in reviewing email on ProNet, mt-dev, and a few public case submissions
> here are some areas of risk and exposure that I think we also have:
> 1) Real world upgrade scenarios - our upgrade testing tests the upgrade
> process from one vanilla install of MT to another. None of these tests
> actually test what the upgrade process is like if you have a lot of plugins
> installed, as most of our users do. Normally this is not a big deal, but
> anecdotal evidence suggests that many plugins users have (especially those
> that have been folded into the core) will somehow conflict with the MT4
> counterparts resulting a poor upgrade experience, such as errors and the like.
> We are only starting to see these reports because until now because very few
> people have been willing to upgrade their production blogs from MT3 to MT4.
> Most have been using a clean MT4 install with no plugins.
> 2) Plugin API bugs - I have seen many developers having difficulty with the
> API, finding changes in behavior that result in unanticipated results. And
> also finding bugs. Some developers have also challenged, and legitimately so,
> our claims on backwards compatibility. Developers have been slow to build on
> top of MT4 because of a lack of documentation, so we have little data to truly
> predict how many bugs may exist in the API.
> 3) Few will Upgrade - Precious few 3rd party plugins have been updated to be
> MT4 compliant/compatible. This will dramatically slow the up take of MT4 in
> the market place by existing MT3 customers. The resolution here is simply more
> time and more documentation to help.
> ——-
> All in all, I think the product is in a place where we actually could manage a
> release. The operative word being "manage." We will need to be on our toes in
> regards to communication and PR to stay ahead of the curve with people who
> will say that we rushed the release.
> I think we will have some explaining to do with developers as they will be
> very mad that we released without giving them time or the resources to upgrade
> their plugins. Speaking from experience, it is frustrating having to answer
> support requests from users of my plugins that are frustrated that I have not
> updated them to the latest version of MT. Developers answered support requests
> from their users will no doubt be very willing to blame Six Apart for not
> giving them the documentation necessary to update their software.
> But there are no absolute show stopper issues IMHO. Yes, there are a number of
> high priority/low severity issues, but I think we can manage them. Especially
> in light of the fact that a 4.01 will be mandatory with the release of
> Enterprise Pack.
> ——-
> In summary, I will reserve judgement on whether I feel we should release or
> not until I can hear the perspectives of my colleagues. I also want to let
> everyone know that this email is intentionally pessimistic. I think it
> important we go into our decision making process with our eyes open to all the
> information, pleasant and unpleasant.
> I will end on an optimistic note however. The reason I feel that we can afford
> to launch on Tuesday is because despite any issues before us, MT4 is a stellar
> product that is light years ahead of MT3. Users, the community and customers
> will appreciate that. And despite any feelings they may have that are
> negative, everyone I have come in contact with is just floored by MT4. So I
> think we have some good will in the bank to draw upon, if we must. :)
> Byrne Reese
> Product Manager, MT