X-Chrome-Variations logic refactoring

chrome/common/metrics/variations/variation_ids.h

 // Copyright (c) 2012 The Chromium Authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 #ifndef CHROME_COMMON_METRICS_VARIATIONS_VARIATION_IDS_H_
 #define CHROME_COMMON_METRICS_VARIATIONS_VARIATION_IDS_H_
 
 namespace chrome_variations {
 
 // A list of Chrome Variation IDs. These IDs are associated with FieldTrials
@@ skipping 117 matching lines @@
   // reuse experiment IDs.
   kSuggestTrialStarted2012Q4IDMin = 3310029,
   kSuggestTrialStarted2012Q4IDMax = 3310048,
 
   // Name: Instant channel field trial.
   // Range: 3310050 - 3310059
   kChannelInstantIDBeta   = 3310050,
   kChannelInstantIDDev    = 3310051,
   kChannelInstantIDStable = 3310052,
 
+  // Name: OmniboxSearchSuggestStarted2013Q1

[Mark P, 2012/12/14 16:56:41]

SuggestStarted
->
SuggestTrialStarted

Joel pointed out to me I forgot the word Trial in the previous time I claimed
variation IDs.  Let's add the word Trial here to make this comment correctly
match the actual field trial name.

[Bart N., 2012/12/14 17:06:07]

Done. In other places it seems fine, so I'm changing anything else.

+  // Range: 3310060 - 3310079
+  // Suggest (Autocomplete) field trial, 20 IDs.  This differs from
+  // the earlier omnibox suggest field trial in this file because

[Mark P, 2012/12/14 16:56:41]

trial
->
trials

[Bart N., 2012/12/14 17:06:07]

Done.

+  // we created a new trial (with a new name) in order to shuffle IDs.
+  // We assign new experiment IDs because it's a good practice not to
+  // reuse experiment IDs.
+  kSuggestTrialStarted2013Q1IDMin = 3310060,

[Mark P, 2012/12/14 16:56:41]

Judging by the format of this file, it seems the preference is to not skip
blocks of IDs.  I also understand that not having round numbers makes it a
little harder to mentally translate field trial group numbers to experiment IDs.
 I'll let Steve decide which feature is more important and therefore whether you
should skip numbers here as you do.

[Bart N., 2012/12/14 17:06:07]

I'm confused about your comment.

I think we are just lucky - I haven't skipped any block, I simply followed the
comment below which asked to use kNextID (which was ...060) and increment it
accordingly. If there a was block skipped, it must have been decided in the
previous change.

[Mark P, 2012/12/14 17:09:55]

Interesting.  I didn't notice the kNextId had already skipped some stuff. 
(Maybe they're reserving nearby ID space?)  In that case, this sounds fine to
me.

[Bart N., 2012/12/14 18:21:42]

Acknowledged.

+  kSuggestTrialStarted2013Q1IDMax = 3310079,
+
   // NEXT ID: When adding new IDs, please add them above this section, starting
   // with the value of kNextID, and updating kNextID to (end of your reserved
   // range) + 1.
-  kNextID = 3310060,
+  kNextID = 3310080,
 
 
   // USABLE IDs END HERE.
   //
   // The largest possible Chrome variation ID in the reserved range. When
   // defining your variation IDs, DO NOT exceed this value.
   kMaximumID = 3399999,
 };
 
 }  // namespace chrome_variations
 
 #endif  // CHROME_COMMON_METRICS_VARIATIONS_VARIATION_IDS_H_