diff --git a/api/api.go b/api/api.go
index 4e64653da0..76343cf7f3 100644
--- a/api/api.go
+++ b/api/api.go
@@ -4,7 +4,6 @@ import (
 	"context"
 
 	"github.com/brimdata/super/compiler/srcfiles"
-	"github.com/brimdata/super/lakeparse"
 	"github.com/brimdata/super/order"
 	"github.com/brimdata/super/pkg/field"
 	"github.com/brimdata/super/pkg/nano"
@@ -98,8 +97,7 @@ type EventBranch struct {
 }
 
 type QueryRequest struct {
-	Query string              `json:"query"`
-	Head  lakeparse.Commitish `json:"head"`
+	Query string `json:"query"`
 }
 
 type QueryChannelSet struct {
diff --git a/api/client/connection.go b/api/client/connection.go
index 0a0db7bfec..30285d16a5 100644
--- a/api/client/connection.go
+++ b/api/client/connection.go
@@ -19,7 +19,6 @@ import (
 	"github.com/brimdata/super/compiler/srcfiles"
 	"github.com/brimdata/super/lake"
 	"github.com/brimdata/super/lake/branches"
-	"github.com/brimdata/super/lakeparse"
 	"github.com/brimdata/super/runtime/exec"
 	"github.com/brimdata/super/zio/zngio"
 	"github.com/brimdata/super/zson"
@@ -296,15 +295,12 @@ func (c *Connection) Revert(ctx context.Context, poolID ksuid.KSUID, branchName
 //
 // As for Connection.Do, if the returned error is nil, the user is expected to
 // call Response.Body.Close.
-func (c *Connection) Query(ctx context.Context, head *lakeparse.Commitish, src string, filenames ...string) (*Response, error) {
+func (c *Connection) Query(ctx context.Context, src string, filenames ...string) (*Response, error) {
 	files, err := srcfiles.Concat(filenames, src)
 	if err != nil {
 		return nil, err
 	}
 	body := api.QueryRequest{Query: string(files.Text)}
-	if head != nil {
-		body.Head = *head
-	}
 	req := c.NewRequest(ctx, http.MethodPost, "/query?ctrl=T", body)
 	res, err := c.Do(req)
 	if ae := (*api.Error)(nil); errors.As(err, &ae) && len(ae.CompilationErrors) > 0 {
diff --git a/cmd/super/compile/shared.go b/cmd/super/compile/shared.go
index 75113d7745..aefec27259 100644
--- a/cmd/super/compile/shared.go
+++ b/cmd/super/compile/shared.go
@@ -76,12 +76,12 @@ func (s *Shared) Run(ctx context.Context, args []string, lakeFlags *lakeflags.Fl
 		}
 		return s.writeValue(ctx, ast.Parsed())
 	}
-	runtime, err := compiler.NewJob(runtime.DefaultContext(), ast, data.NewSource(nil, lk), nil)
+	runtime, err := compiler.NewJob(runtime.DefaultContext(), ast, data.NewSource(nil, lk), false)
 	if err != nil {
 		return err
 	}
 	if desc {
-		description, err := describe.AnalyzeDAG(ctx, runtime.Entry(), data.NewSource(nil, lk), nil)
+		description, err := describe.AnalyzeDAG(ctx, runtime.Entry(), data.NewSource(nil, lk))
 		if err != nil {
 			return err
 		}
diff --git a/cmd/super/db/branch/command.go b/cmd/super/db/branch/command.go
index 6d3c70f9df..2766c4a87a 100644
--- a/cmd/super/db/branch/command.go
+++ b/cmd/super/db/branch/command.go
@@ -131,7 +131,7 @@ func (c *Command) list(ctx context.Context, lake api.Interface) error {
 	if err != nil {
 		return err
 	}
-	q, err := lake.Query(ctx, nil, query)
+	q, err := lake.Query(ctx, query)
 	if err != nil {
 		w.Close()
 		return err
diff --git a/cmd/super/db/log/command.go b/cmd/super/db/log/command.go
index 5536483756..de2247884b 100644
--- a/cmd/super/db/log/command.go
+++ b/cmd/super/db/log/command.go
@@ -72,7 +72,7 @@ func (c *Command) Run(args []string) error {
 		return err
 	}
 	defer w.Close()
-	q, err := lake.Query(ctx, nil, query)
+	q, err := lake.Query(ctx, query)
 	if err != nil {
 		return err
 	}
diff --git a/cmd/super/db/ls/command.go b/cmd/super/db/ls/command.go
index bd41be1272..82d1723137 100644
--- a/cmd/super/db/ls/command.go
+++ b/cmd/super/db/ls/command.go
@@ -88,7 +88,7 @@ func (c *Command) Run(args []string) error {
 	if err != nil {
 		return err
 	}
-	q, err := lake.Query(ctx, nil, query)
+	q, err := lake.Query(ctx, query)
 	if err != nil {
 		w.Close()
 		return err
diff --git a/cmd/super/db/query/command.go b/cmd/super/db/query/command.go
index 2e694e7a44..79343b881e 100644
--- a/cmd/super/db/query/command.go
+++ b/cmd/super/db/query/command.go
@@ -5,7 +5,6 @@ import (
 	"os"
 
 	"github.com/brimdata/super/cli/outputflags"
-	"github.com/brimdata/super/cli/poolflags"
 	"github.com/brimdata/super/cli/queryflags"
 	"github.com/brimdata/super/cli/runtimeflags"
 	"github.com/brimdata/super/cmd/super/db"
@@ -33,7 +32,6 @@ func init() {
 type Command struct {
 	*db.Command
 	outputFlags  outputflags.Flags
-	poolFlags    poolflags.Flags
 	queryFlags   queryflags.Flags
 	runtimeFlags runtimeflags.Flags
 }
@@ -41,7 +39,6 @@ type Command struct {
 func New(parent charm.Command, f *flag.FlagSet) (charm.Command, error) {
 	c := &Command{Command: parent.(*db.Command)}
 	c.outputFlags.SetFlags(f)
-	c.poolFlags.SetFlags(f)
 	c.queryFlags.SetFlags(f)
 	c.runtimeFlags.SetFlags(f)
 	return c, nil
@@ -68,8 +65,7 @@ func (c *Command) Run(args []string) error {
 	if err != nil {
 		return err
 	}
-	head, _ := c.poolFlags.HEAD()
-	query, err := lake.Query(ctx, head, src, c.queryFlags.Includes...)
+	query, err := lake.Query(ctx, src, c.queryFlags.Includes...)
 	if err != nil {
 		w.Close()
 		return err
diff --git a/cmd/super/internal/lakemanage/scan.go b/cmd/super/internal/lakemanage/scan.go
index e4ccb87bed..14cfd1f745 100644
--- a/cmd/super/internal/lakemanage/scan.go
+++ b/cmd/super/internal/lakemanage/scan.go
@@ -74,7 +74,7 @@ type objectIterator struct {
 
 func newObjectIterator(ctx context.Context, lake api.Interface, head *lakeparse.Commitish) (*objectIterator, error) {
 	query := fmt.Sprintf(iteratorQuery, head.Pool, head.Branch, head.Pool, head.Branch)
-	q, err := lake.Query(ctx, nil, query)
+	q, err := lake.Query(ctx, query)
 	if err != nil {
 		return nil, err
 	}
diff --git a/compiler/ast/ast.go b/compiler/ast/ast.go
index 34a4f2fbe0..ad2bbab0d0 100644
--- a/compiler/ast/ast.go
+++ b/compiler/ast/ast.go
@@ -574,8 +574,10 @@ type (
 		Loc     `json:"loc"`
 	}
 	Delete struct {
-		Kind string       `json:"kind" unpack:""`
-		Loc  `json:"loc"` // dummy field, not needed except to implement Node
+		Kind   string       `json:"kind" unpack:""`
+		Pool   string       `json:"pool"`
+		Branch string       `json:"branch"`
+		Loc    `json:"loc"` // dummy field, not needed except to implement Node
 	}
 )
 
diff --git a/compiler/dag/op.go b/compiler/dag/op.go
index 7d31bdb685..623d6ac302 100644
--- a/compiler/dag/op.go
+++ b/compiler/dag/op.go
@@ -246,6 +246,9 @@ type (
 		Kind string `json:"kind" unpack:""`
 		Meta string `json:"meta"`
 	}
+	NullScan struct {
+		Kind string `json:"kind" unpack:""`
+	}
 )
 
 var LakeMetas = map[string]struct{}{
@@ -273,6 +276,7 @@ func (*DeleteScan) OpNode()     {}
 func (*LakeMetaScan) OpNode()   {}
 func (*PoolMetaScan) OpNode()   {}
 func (*CommitMetaScan) OpNode() {}
+func (*NullScan) OpNode()       {}
 
 func (*Lister) OpNode()  {}
 func (*Slicer) OpNode()  {}
diff --git a/compiler/dag/unpack.go b/compiler/dag/unpack.go
index 76d105da6f..572e49fcb1 100644
--- a/compiler/dag/unpack.go
+++ b/compiler/dag/unpack.go
@@ -42,6 +42,7 @@ var unpacker = unpack.New(
 	MapExpr{},
 	Merge{},
 	Mirror{},
+	NullScan{},
 	Output{},
 	Over{},
 	OverExpr{},
diff --git a/compiler/describe/analyze.go b/compiler/describe/analyze.go
index 61e65dec08..8845e11997 100644
--- a/compiler/describe/analyze.go
+++ b/compiler/describe/analyze.go
@@ -10,7 +10,6 @@ import (
 	"github.com/brimdata/super/compiler/parser"
 	"github.com/brimdata/super/compiler/semantic"
 	"github.com/brimdata/super/lake"
-	"github.com/brimdata/super/lakeparse"
 	"github.com/brimdata/super/order"
 	"github.com/brimdata/super/pkg/field"
 	"github.com/segmentio/ksuid"
@@ -31,21 +30,23 @@ type (
 		Meta string `json:"meta"`
 	}
 	Pool struct {
-		Kind     string      `json:"kind"`
-		Name     string      `json:"name"`
-		ID       ksuid.KSUID `json:"id"`
-		Inferred bool        `json:"inferred"`
+		Kind string      `json:"kind"`
+		Name string      `json:"name"`
+		ID   ksuid.KSUID `json:"id"`
 	}
 	Path struct {
-		Kind     string `json:"kind"`
-		URI      string `json:"uri"`
-		Inferred bool   `json:"inferred"`
+		Kind string `json:"kind"`
+		URI  string `json:"uri"`
+	}
+	Null struct {
+		Kind string `json:"kind"`
 	}
 )
 
 func (*LakeMeta) Source() {}
 func (*Pool) Source()     {}
 func (*Path) Source()     {}
+func (*Null) Source()     {}
 
 type Channel struct {
 	Name            string         `json:"name"`
@@ -53,26 +54,22 @@ type Channel struct {
 	Sort            order.SortKeys `json:"sort"`
 }
 
-func Analyze(ctx context.Context, query string, src *data.Source, head *lakeparse.Commitish) (*Info, error) {
+func Analyze(ctx context.Context, query string, src *data.Source) (*Info, error) {
 	ast, err := parser.ParseQuery(query)
 	if err != nil {
 		return nil, err
 	}
-	entry, err := semantic.Analyze(ctx, ast, src, head)
+	entry, err := semantic.Analyze(ctx, ast, src, false)
 	if err != nil {
 		return nil, err
 	}
-	return AnalyzeDAG(ctx, entry, src, head)
+	return AnalyzeDAG(ctx, entry, src)
 }
 
-func AnalyzeDAG(ctx context.Context, entry dag.Seq, src *data.Source, head *lakeparse.Commitish) (*Info, error) {
-	srcInferred := !semantic.HasSource(entry)
-	if err := semantic.AddDefaultSource(ctx, &entry, src, head); err != nil {
-		return nil, err
-	}
+func AnalyzeDAG(ctx context.Context, entry dag.Seq, src *data.Source) (*Info, error) {
 	var err error
 	var info Info
-	if info.Sources, err = describeSources(ctx, src.Lake(), entry[0], srcInferred); err != nil {
+	if info.Sources, err = describeSources(ctx, src.Lake(), entry[0]); err != nil {
 		return nil, err
 	}
 	sortKeys, err := optimizer.New(ctx, src).SortKeys(entry)
@@ -102,14 +99,14 @@ func AnalyzeDAG(ctx context.Context, entry dag.Seq, src *data.Source, head *lake
 	return &info, nil
 }
 
-func describeSources(ctx context.Context, lk *lake.Root, o dag.Op, inferred bool) ([]Source, error) {
+func describeSources(ctx context.Context, lk *lake.Root, o dag.Op) ([]Source, error) {
 	switch o := o.(type) {
 	case *dag.Scope:
-		return describeSources(ctx, lk, o.Body[0], inferred)
+		return describeSources(ctx, lk, o.Body[0])
 	case *dag.Fork:
 		var s []Source
 		for _, p := range o.Paths {
-			out, err := describeSources(ctx, lk, p[0], inferred)
+			out, err := describeSources(ctx, lk, p[0])
 			if err != nil {
 				return nil, err
 			}
@@ -117,19 +114,21 @@ func describeSources(ctx context.Context, lk *lake.Root, o dag.Op, inferred bool
 		}
 		return s, nil
 	case *dag.DefaultScan:
-		return []Source{&Path{Kind: "Path", URI: "stdio://stdin", Inferred: inferred}}, nil
+		return []Source{&Path{Kind: "Path", URI: "stdio://stdin"}}, nil
+	case *dag.NullScan:
+		return []Source{&Null{Kind: "Null"}}, nil
 	case *dag.FileScan:
-		return []Source{&Path{Kind: "Path", URI: o.Path, Inferred: inferred}}, nil
+		return []Source{&Path{Kind: "Path", URI: o.Path}}, nil
 	case *dag.HTTPScan:
-		return []Source{&Path{Kind: "Path", URI: o.URL, Inferred: inferred}}, nil
+		return []Source{&Path{Kind: "Path", URI: o.URL}}, nil
 	case *dag.PoolScan:
-		return sourceOfPool(ctx, lk, o.ID, inferred)
+		return sourceOfPool(ctx, lk, o.ID)
 	case *dag.Lister:
-		return sourceOfPool(ctx, lk, o.Pool, inferred)
+		return sourceOfPool(ctx, lk, o.Pool)
 	case *dag.SeqScan:
-		return sourceOfPool(ctx, lk, o.Pool, inferred)
+		return sourceOfPool(ctx, lk, o.Pool)
 	case *dag.CommitMetaScan:
-		return sourceOfPool(ctx, lk, o.Pool, inferred)
+		return sourceOfPool(ctx, lk, o.Pool)
 	case *dag.LakeMetaScan:
 		return []Source{&LakeMeta{Kind: "LakeMeta", Meta: o.Meta}}, nil
 	default:
@@ -137,16 +136,15 @@ func describeSources(ctx context.Context, lk *lake.Root, o dag.Op, inferred bool
 	}
 }
 
-func sourceOfPool(ctx context.Context, lk *lake.Root, id ksuid.KSUID, inferred bool) ([]Source, error) {
+func sourceOfPool(ctx context.Context, lk *lake.Root, id ksuid.KSUID) ([]Source, error) {
 	p, err := lk.OpenPool(ctx, id)
 	if err != nil {
 		return nil, err
 	}
 	return []Source{&Pool{
-		Kind:     "Pool",
-		ID:       id,
-		Name:     p.Name,
-		Inferred: inferred,
+		Kind: "Pool",
+		ID:   id,
+		Name: p.Name,
 	}}, nil
 }
 
diff --git a/compiler/file.go b/compiler/file.go
index 3c80c08bab..8182e7ddc6 100644
--- a/compiler/file.go
+++ b/compiler/file.go
@@ -23,7 +23,7 @@ func NewFileSystemCompiler(engine storage.Engine) runtime.Compiler {
 }
 
 func (f *fsCompiler) NewQuery(rctx *runtime.Context, ast *parser.AST, readers []zio.Reader) (runtime.Query, error) {
-	job, err := NewJob(rctx, ast, f.src, nil)
+	job, err := NewJob(rctx, ast, f.src, len(readers) > 0)
 	if err != nil {
 		return nil, err
 	}
@@ -45,7 +45,7 @@ func (f *fsCompiler) NewQuery(rctx *runtime.Context, ast *parser.AST, readers []
 	return optimizeAndBuild(job, readers)
 }
 
-func (*fsCompiler) NewLakeQuery(_ *runtime.Context, ast *parser.AST, parallelism int, head *lakeparse.Commitish) (runtime.Query, error) {
+func (*fsCompiler) NewLakeQuery(_ *runtime.Context, ast *parser.AST, parallelism int) (runtime.Query, error) {
 	panic("NewLakeQuery called on compiler.fsCompiler")
 }
 
diff --git a/compiler/job.go b/compiler/job.go
index 15b74f291b..3e7b06c41f 100644
--- a/compiler/job.go
+++ b/compiler/job.go
@@ -2,6 +2,7 @@ package compiler
 
 import (
 	"errors"
+	"fmt"
 
 	"github.com/brimdata/super/compiler/dag"
 	"github.com/brimdata/super/compiler/data"
@@ -28,7 +29,7 @@ type Job struct {
 	entry     dag.Seq
 }
 
-func NewJob(rctx *runtime.Context, ast *parser.AST, src *data.Source, head *lakeparse.Commitish) (*Job, error) {
+func NewJob(rctx *runtime.Context, ast *parser.AST, src *data.Source, extInput bool) (*Job, error) {
 	// An AST always begins with a Sequential op with at least one
 	// operator.  If the first op is a From or a Parallel whose branches
 	// are Sequentials with a leading From, then we presume there is
@@ -39,10 +40,7 @@ func NewJob(rctx *runtime.Context, ast *parser.AST, src *data.Source, head *lake
 	// caller via runtime.readers.  In most cases, the AST is left
 	// with an ast.From at the entry point, and hence a dag.From for the
 	// DAG's entry point.
-	if len(ast.Parsed()) == 0 {
-		return nil, errors.New("internal error: AST seq cannot be empty")
-	}
-	entry, err := semantic.AnalyzeAddSource(rctx.Context, ast, src, head)
+	entry, err := semantic.Analyze(rctx.Context, ast, src, extInput)
 	if err != nil {
 		return nil, err
 	}
@@ -131,10 +129,14 @@ func VectorCompile(rctx *runtime.Context, query string, object *vcache.Object) (
 		return nil, err
 	}
 	src := &data.Source{}
-	entry, err := semantic.Analyze(rctx.Context, ast, src, nil)
+	entry, err := semantic.Analyze(rctx.Context, ast, src, true)
 	if err != nil {
 		return nil, err
 	}
+	if _, ok := entry[0].(*dag.DefaultScan); !ok {
+		panic("DAG assumptions violated")
+	}
+	entry = entry[1:]
 	puller := vam.NewVectorProjection(rctx.Zctx, object, nil) //XXX project all
 	builder := kernel.NewBuilder(rctx, src)
 	outputs, err := builder.BuildWithPuller(entry, puller)
@@ -153,7 +155,7 @@ func VectorFilterCompile(rctx *runtime.Context, query string, src *data.Source,
 	// Eventually the semantic analyzer + kernel will resolve the pool but
 	// for now just do this manually.
 	if !src.IsLake() {
-		return nil, errors.New("non-lake searches not supported")
+		return nil, errors.New("non-lake vectorized search not supported")
 	}
 	poolID, err := src.PoolID(rctx.Context, head.Pool)
 	if err != nil {
@@ -163,18 +165,23 @@ func VectorFilterCompile(rctx *runtime.Context, query string, src *data.Source,
 	if err != nil {
 		return nil, err
 	}
-	ast, err := parser.ParseQuery(query)
+	spec, err := head.FromSpec("")
+	if err != nil {
+		return nil, err
+	}
+	ast, err := parser.ParseQuery(fmt.Sprintf("%s | %s", spec, query))
 	if err != nil {
 		return nil, err
 	}
-	entry, err := semantic.Analyze(rctx.Context, ast, src, nil)
+	entry, err := semantic.Analyze(rctx.Context, ast, src, false)
 	if err != nil {
 		return nil, err
 	}
-	if len(entry) != 2 {
+	// from -> filter -> output
+	if len(entry) != 3 {
 		return nil, errors.New("filter query must have a single op")
 	}
-	f, ok := entry[0].(*dag.Filter)
+	f, ok := entry[1].(*dag.Filter)
 	if !ok {
 		return nil, errors.New("filter query must be a single filter op")
 	}
diff --git a/compiler/kernel/op.go b/compiler/kernel/op.go
index 876556d838..f39793445e 100644
--- a/compiler/kernel/op.go
+++ b/compiler/kernel/op.go
@@ -287,6 +287,9 @@ func (b *Builder) compileLeaf(o dag.Op, parent zbuf.Puller) (zbuf.Puller, error)
 			scanners = append(scanners, scanner)
 		}
 		return zbuf.MultiScanner(scanners...), nil
+	case *dag.NullScan:
+		//XXX we need something that implements the done protocol and restarst
+		return zbuf.NewPuller(zbuf.NewArray([]super.Value{super.Null})), nil
 	case *dag.Lister:
 		if parent != nil {
 			return nil, errors.New("internal error: data source cannot have a parent operator")
@@ -744,7 +747,7 @@ func isEntry(seq dag.Seq) bool {
 		return false
 	}
 	switch op := seq[0].(type) {
-	case *dag.Lister, *dag.DefaultScan, *dag.FileScan, *dag.HTTPScan, *dag.PoolScan, *dag.LakeMetaScan, *dag.PoolMetaScan, *dag.CommitMetaScan:
+	case *dag.Lister, *dag.DefaultScan, *dag.FileScan, *dag.HTTPScan, *dag.PoolScan, *dag.LakeMetaScan, *dag.PoolMetaScan, *dag.CommitMetaScan, *dag.NullScan:
 		return true
 	case *dag.Scope:
 		return isEntry(op.Body)
diff --git a/compiler/lake.go b/compiler/lake.go
index 1287f88e0c..3f4694e8e3 100644
--- a/compiler/lake.go
+++ b/compiler/lake.go
@@ -31,8 +31,8 @@ func NewLakeCompiler(r *lake.Root) runtime.Compiler {
 	return &lakeCompiler{src: data.NewSource(storage.NewRemoteEngine(), r)}
 }
 
-func (l *lakeCompiler) NewLakeQuery(rctx *runtime.Context, ast *parser.AST, parallelism int, head *lakeparse.Commitish) (runtime.Query, error) {
-	job, err := NewJob(rctx, ast, l.src, head)
+func (l *lakeCompiler) NewLakeQuery(rctx *runtime.Context, ast *parser.AST, parallelism int) (runtime.Query, error) {
+	job, err := NewJob(rctx, ast, l.src, false)
 	if err != nil {
 		return nil, err
 	}
@@ -57,7 +57,7 @@ func (l *lakeCompiler) NewLakeQuery(rctx *runtime.Context, ast *parser.AST, para
 }
 
 func (l *lakeCompiler) NewLakeDeleteQuery(rctx *runtime.Context, ast *parser.AST, head *lakeparse.Commitish) (runtime.DeleteQuery, error) {
-	job, err := newDeleteJob(rctx, ast, l.src, head)
+	job, err := newDeleteJob(rctx, ast, head.Pool, head.Branch, l.src)
 	if err != nil {
 		return nil, err
 	}
@@ -76,15 +76,15 @@ func (InvalidDeleteWhereQuery) Error() string {
 	return "invalid delete where query: must be a single filter operation"
 }
 
-func newDeleteJob(rctx *runtime.Context, parsed *parser.AST, src *data.Source, head *lakeparse.Commitish) (*Job, error) {
-	if err := parsed.ConvertToDeleteWhere(); err != nil {
+func newDeleteJob(rctx *runtime.Context, parsed *parser.AST, pool, branch string, src *data.Source) (*Job, error) {
+	if err := parsed.ConvertToDeleteWhere(pool, branch); err != nil {
 		return nil, err
 	}
 	seq := parsed.Parsed()
 	if len(seq) != 2 {
 		return nil, &InvalidDeleteWhereQuery{}
 	}
-	entry, err := semantic.Analyze(rctx.Context, parsed, src, head)
+	entry, err := semantic.Analyze(rctx.Context, parsed, src, false)
 	if err != nil {
 		return nil, err
 	}
diff --git a/compiler/parser/api.go b/compiler/parser/api.go
index 72c1dcf0e1..5a47cd2ccc 100644
--- a/compiler/parser/api.go
+++ b/compiler/parser/api.go
@@ -24,11 +24,15 @@ func (a *AST) Files() *srcfiles.List {
 	return a.files
 }
 
-func (a *AST) ConvertToDeleteWhere() error {
+func (a *AST) ConvertToDeleteWhere(pool, branch string) error {
 	if len(a.seq) == 0 {
 		return errors.New("internal error: AST seq cannot be empty")
 	}
-	a.seq.Prepend(&ast.Delete{Kind: "Delete"})
+	a.seq.Prepend(&ast.Delete{
+		Kind:   "Delete",
+		Pool:   pool,
+		Branch: branch,
+	})
 	return nil
 }
 
@@ -39,6 +43,9 @@ func ParseQuery(query string, filenames ...string) (*AST, error) {
 	if err != nil {
 		return nil, err
 	}
+	if files.Text == "" {
+		return &AST{files: files}, nil
+	}
 	p, err := Parse("", []byte(files.Text), Recover(false))
 	if err != nil {
 		if err := convertParseErrs(err, files); err != nil {
diff --git a/compiler/reader.go b/compiler/reader.go
index c601bc6240..a308ae8529 100644
--- a/compiler/reader.go
+++ b/compiler/reader.go
@@ -21,7 +21,7 @@ func (i *anyCompiler) NewQuery(rctx *runtime.Context, ast *parser.AST, readers [
 	if len(readers) != 1 {
 		return nil, fmt.Errorf("NewQuery: Zed program expected %d readers", len(readers))
 	}
-	job, err := NewJob(rctx, ast, data.NewSource(nil, nil), nil)
+	job, err := NewJob(rctx, ast, data.NewSource(nil, nil), true)
 	if err != nil {
 		return nil, err
 	}
@@ -30,7 +30,7 @@ func (i *anyCompiler) NewQuery(rctx *runtime.Context, ast *parser.AST, readers [
 
 // XXX currently used only by group-by test, need to deprecate
 func CompileWithSortKey(rctx *runtime.Context, ast *parser.AST, r zio.Reader, sortKey order.SortKey) (*exec.Query, error) {
-	job, err := NewJob(rctx, ast, data.NewSource(nil, nil), nil)
+	job, err := NewJob(rctx, ast, data.NewSource(nil, nil), true)
 	if err != nil {
 		return nil, err
 	}
@@ -42,7 +42,7 @@ func CompileWithSortKey(rctx *runtime.Context, ast *parser.AST, r zio.Reader, so
 	return optimizeAndBuild(job, []zio.Reader{r})
 }
 
-func (*anyCompiler) NewLakeQuery(rctx *runtime.Context, ast *parser.AST, parallelism int, head *lakeparse.Commitish) (runtime.Query, error) {
+func (*anyCompiler) NewLakeQuery(rctx *runtime.Context, ast *parser.AST, parallelism int) (runtime.Query, error) {
 	panic("NewLakeQuery called on compiler.anyCompiler")
 }
 
diff --git a/compiler/semantic/analyzer.go b/compiler/semantic/analyzer.go
index 9a2752b239..7726822ce0 100644
--- a/compiler/semantic/analyzer.go
+++ b/compiler/semantic/analyzer.go
@@ -10,42 +10,36 @@ import (
 	"github.com/brimdata/super/compiler/data"
 	"github.com/brimdata/super/compiler/parser"
 	"github.com/brimdata/super/compiler/srcfiles"
-	"github.com/brimdata/super/lakeparse"
 )
 
 // Analyze performs a semantic analysis of the AST, translating it from AST
 // to DAG form, resolving syntax ambiguities, and performing constant propagation.
 // After semantic analysis, the DAG is ready for either optimization or compilation.
-func Analyze(ctx context.Context, ast *parser.AST, source *data.Source, head *lakeparse.Commitish) (dag.Seq, error) {
+func Analyze(ctx context.Context, ast *parser.AST, source *data.Source, extInput bool) (dag.Seq, error) {
 	files := ast.Files()
-	a := newAnalyzer(ctx, files, source, head)
-	s := a.semSeq(ast.Parsed())
-	s = a.checkOutputs(true, s)
-	return s, files.Error()
-}
-
-// AnalyzeAddSource is the same as Analyze but it adds a default source if the
-// DAG does not have one.
-func AnalyzeAddSource(ctx context.Context, ast *parser.AST, source *data.Source, head *lakeparse.Commitish) (dag.Seq, error) {
-	files := ast.Files()
-	a := newAnalyzer(ctx, files, source, head)
-	s := a.semSeq(ast.Parsed())
-	s = a.checkOutputs(true, s)
-	if err := files.Error(); err != nil {
-		return nil, err
-	}
-	if !HasSource(s) {
-		if err := AddDefaultSource(ctx, &s, source, head); err != nil {
-			return nil, err
+	a := newAnalyzer(ctx, files, source)
+	seq := a.semSeq(ast.Parsed())
+	if !HasSource(seq) {
+		if a.source.IsLake() {
+			if len(seq) == 0 {
+				return nil, errors.New("query text is missing")
+			}
+			seq.Prepend(&dag.NullScan{Kind: "NullScan"})
+		} else if extInput {
+			seq.Prepend(&dag.DefaultScan{Kind: "DefaultScan"})
+		} else {
+			// This is a local query and there's no external input
+			// (i.e., no command-line file args)
+			seq.Prepend(&dag.NullScan{Kind: "NullScan"})
 		}
 	}
-	return s, nil
+	seq = a.checkOutputs(true, seq)
+	return seq, files.Error()
 }
 
 type analyzer struct {
 	ctx     context.Context
 	files   *srcfiles.List
-	head    *lakeparse.Commitish
 	opStack []*ast.OpDecl
 	outputs map[*dag.Output]ast.Node
 	source  *data.Source
@@ -53,11 +47,10 @@ type analyzer struct {
 	zctx    *super.Context
 }
 
-func newAnalyzer(ctx context.Context, files *srcfiles.List, source *data.Source, head *lakeparse.Commitish) *analyzer {
+func newAnalyzer(ctx context.Context, files *srcfiles.List, source *data.Source) *analyzer {
 	return &analyzer{
 		ctx:     ctx,
 		files:   files,
-		head:    head,
 		outputs: make(map[*dag.Output]ast.Node),
 		source:  source,
 		scope:   NewScope(nil),
@@ -66,8 +59,11 @@ func newAnalyzer(ctx context.Context, files *srcfiles.List, source *data.Source,
 }
 
 func HasSource(seq dag.Seq) bool {
+	if len(seq) == 0 {
+		return false
+	}
 	switch op := seq[0].(type) {
-	case *dag.FileScan, *dag.HTTPScan, *dag.PoolScan, *dag.LakeMetaScan, *dag.PoolMetaScan, *dag.CommitMetaScan, *dag.DeleteScan:
+	case *dag.FileScan, *dag.HTTPScan, *dag.PoolScan, *dag.LakeMetaScan, *dag.PoolMetaScan, *dag.CommitMetaScan, *dag.DeleteScan, *dag.NullScan:
 		return true
 	case *dag.Fork:
 		return HasSource(op.Paths[0])
@@ -77,34 +73,6 @@ func HasSource(seq dag.Seq) bool {
 	return false
 }
 
-func AddDefaultSource(ctx context.Context, seq *dag.Seq, source *data.Source, head *lakeparse.Commitish) error {
-	if HasSource(*seq) {
-		return nil
-	}
-	// No from so add a source.
-	if head == nil {
-		seq.Prepend(&dag.DefaultScan{Kind: "DefaultScan"})
-		return nil
-	}
-	// Verify pool exists for HEAD
-	if _, err := source.PoolID(ctx, head.Pool); err != nil {
-		return err
-	}
-	fromHead := &ast.From{
-		Kind: "From",
-		Elems: []*ast.FromElem{{
-			Kind: "FromElem",
-			Entity: &ast.Name{
-				Kind: "Name",
-				Text: "HEAD",
-			},
-		}},
-	}
-	headSeq := newAnalyzer(ctx, &srcfiles.List{}, source, head).semFrom(fromHead, nil)
-	seq.Prepend(headSeq[0])
-	return nil
-}
-
 func (a *analyzer) enterScope() {
 	a.scope = NewScope(a.scope)
 }
diff --git a/compiler/semantic/op.go b/compiler/semantic/op.go
index 9bdd53f44a..77d8a87ce8 100644
--- a/compiler/semantic/op.go
+++ b/compiler/semantic/op.go
@@ -339,14 +339,6 @@ func (a *analyzer) semPool(nameLoc ast.Node, poolName string, args *ast.PoolArgs
 		meta = nullableName(args.Meta)
 		tap = args.Tap
 	}
-	if poolName == "HEAD" {
-		if a.head == nil || a.head.Pool == "" {
-			a.error(nameLoc, errors.New("cannot resolve unknown HEAD"))
-			return badOp()
-		}
-		poolName = a.head.Pool
-		commit = a.head.Branch
-	}
 	poolID, err := a.source.PoolID(a.ctx, poolName)
 	if err != nil {
 		a.error(nameLoc, err)
@@ -422,20 +414,16 @@ func (a *analyzer) semDelete(op *ast.Delete) dag.Op {
 		a.error(op, errors.New("deletion requires data lake"))
 		return badOp()
 	}
-	// delete-where only supports deleting at head
-	pool := a.head.Pool
-	commit := a.head.Branch
-
-	poolID, err := a.source.PoolID(a.ctx, pool)
+	poolID, err := a.source.PoolID(a.ctx, op.Pool)
 	if err != nil {
 		a.error(op, err)
 		return badOp()
 	}
 	var commitID ksuid.KSUID
-	if commit != "" {
+	if op.Branch != "" {
 		var err error
-		if commitID, err = lakeparse.ParseID(commit); err != nil {
-			commitID, err = a.source.CommitObject(a.ctx, poolID, commit)
+		if commitID, err = lakeparse.ParseID(op.Branch); err != nil {
+			commitID, err = a.source.CommitObject(a.ctx, poolID, op.Branch)
 			if err != nil {
 				a.error(op, err)
 				return badOp()
diff --git a/compiler/srcfiles/list.go b/compiler/srcfiles/list.go
index 069ffb4bf6..0f95f5e09e 100644
--- a/compiler/srcfiles/list.go
+++ b/compiler/srcfiles/list.go
@@ -42,9 +42,6 @@ func Concat(filenames []string, query string) (*List, error) {
 		b.Write(bb)
 		b.WriteByte('\n')
 	}
-	if b.Len() == 0 && query == "" {
-		query = "*"
-	}
 	// Empty string is the unnamed query text while the included files all
 	// have names.
 	files = append(files, newFile("", b.Len(), []byte(query)))
diff --git a/compiler/ztests/head.yaml b/compiler/ztests/head.yaml
index 4bb728d6cd..38f8f1d536 100644
--- a/compiler/ztests/head.yaml
+++ b/compiler/ztests/head.yaml
@@ -9,11 +9,11 @@ script: |
 outputs:
   - name: stdout
     data: |
-      reader
+      null
       | head 1
       | output main
       ===
-      reader
+      null
       | (
         const x = 1
         
diff --git a/compiler/ztests/merge-filters.yaml b/compiler/ztests/merge-filters.yaml
index c9a62b4f49..609bb3c533 100644
--- a/compiler/ztests/merge-filters.yaml
+++ b/compiler/ztests/merge-filters.yaml
@@ -1,5 +1,5 @@
 script: |
-  super compile -C -O 'where a | where b'
+  super compile -C -O 'from foo | where a | where b'
   echo ===
   super compile -C -O 'from ( file a => where b | where c file d => where e | where f ) | where g'
   echo ===
@@ -10,7 +10,7 @@ script: |
 outputs:
   - name: stdout
     data: |
-      reader filter (a and b)
+      file foo filter (a and b)
       | output main
       ===
       fork (
@@ -21,13 +21,13 @@ outputs:
       )
       | output main
       ===
-      reader
+      null
       | over a => (
         where b and c
       )
       | output main
       ===
-      reader
+      null
       | fork (
         =>
           where a and b
diff --git a/compiler/ztests/remove-passops.yaml b/compiler/ztests/remove-passops.yaml
index 2dc99db9bf..a1ffab814e 100644
--- a/compiler/ztests/remove-passops.yaml
+++ b/compiler/ztests/remove-passops.yaml
@@ -1,7 +1,7 @@
-script: super compile -O -C 'x>1 | pass | pass | x>2 | pass'
+script: super compile -O -C 'from foo | x>1 | pass | pass | x>2 | pass'
 
 outputs:
   - name: stdout
     data: |
-      reader filter (x>1 and x>2)
+      file foo filter (x>1 and x>2)
       | output main
diff --git a/compiler/ztests/tail.yaml b/compiler/ztests/tail.yaml
index aefff3b78a..2d129e46ed 100644
--- a/compiler/ztests/tail.yaml
+++ b/compiler/ztests/tail.yaml
@@ -9,11 +9,11 @@ script: |
 outputs:
   - name: stdout
     data: |
-      reader
+      null
       | tail 1
       | output main
       ===
-      reader
+      null
       | (
         const x = 1
         
diff --git a/compiler/ztests/udf-implied-where.yaml b/compiler/ztests/udf-implied-where.yaml
index 6febdd3383..079ef66ff4 100644
--- a/compiler/ztests/udf-implied-where.yaml
+++ b/compiler/ztests/udf-implied-where.yaml
@@ -4,7 +4,7 @@ script: |
 outputs:
   - name: stdout
     data: |
-      reader
+      null
       | (
         func h(e): (
           has(e)
diff --git a/docs/libraries/go.md b/docs/libraries/go.md
index 712d45f723..fed92aeb93 100644
--- a/docs/libraries/go.md
+++ b/docs/libraries/go.md
@@ -134,7 +134,7 @@ func main() {
 	if err != nil {
 		log.Fatalln(err)
 	}
-	q, err := lake.Query(ctx, nil, "from Demo")
+	q, err := lake.Query(ctx, "from Demo")
 	if err != nil {
 		log.Fatalln(err)
 	}
diff --git a/fuzz/fuzz.go b/fuzz/fuzz.go
index 383cca5e11..7e29d90cd8 100644
--- a/fuzz/fuzz.go
+++ b/fuzz/fuzz.go
@@ -109,7 +109,7 @@ func RunQuery(t testing.TB, zctx *super.Context, readers []zio.Reader, querySour
 	// Infer demand
 	// TODO This is a hack and should be replaced by a cleaner interface in CompileQuery.
 	source := data.NewSource(engine, nil)
-	dag, err := semantic.AnalyzeAddSource(ctx, ast, source, nil)
+	dag, err := semantic.Analyze(ctx, ast, source, true)
 	if err != nil {
 		t.Skipf("%v", err)
 	}
diff --git a/lake/api/api.go b/lake/api/api.go
index 8b1e260b3e..4c8ed8509b 100644
--- a/lake/api/api.go
+++ b/lake/api/api.go
@@ -11,7 +11,6 @@ import (
 	"github.com/brimdata/super/api/client"
 	"github.com/brimdata/super/lake"
 	"github.com/brimdata/super/lake/pools"
-	"github.com/brimdata/super/lakeparse"
 	"github.com/brimdata/super/order"
 	"github.com/brimdata/super/zbuf"
 	"github.com/brimdata/super/zio"
@@ -22,7 +21,7 @@ import (
 
 type Interface interface {
 	Root() *lake.Root
-	Query(ctx context.Context, head *lakeparse.Commitish, src string, srcfiles ...string) (zbuf.Scanner, error)
+	Query(ctx context.Context, src string, srcfiles ...string) (zbuf.Scanner, error)
 	PoolID(ctx context.Context, poolName string) (ksuid.KSUID, error)
 	CommitObject(ctx context.Context, poolID ksuid.KSUID, branchName string) (ksuid.KSUID, error)
 	CreatePool(context.Context, string, order.SortKeys, int, int64) (ksuid.KSUID, error)
@@ -55,7 +54,7 @@ func IsLakeService(u string) bool {
 func LookupPoolByName(ctx context.Context, api Interface, name string) (*pools.Config, error) {
 	b := newBuffer(pools.Config{})
 	zed := fmt.Sprintf("from :pools | name == '%s'", name)
-	q, err := api.Query(ctx, nil, zed)
+	q, err := api.Query(ctx, zed)
 	if err != nil {
 		return nil, err
 	}
@@ -79,7 +78,7 @@ func LookupPoolByName(ctx context.Context, api Interface, name string) (*pools.C
 
 func GetPools(ctx context.Context, api Interface) ([]*pools.Config, error) {
 	b := newBuffer(pools.Config{})
-	q, err := api.Query(ctx, nil, "from :pools")
+	q, err := api.Query(ctx, "from :pools")
 	if err != nil {
 		return nil, err
 	}
@@ -97,7 +96,7 @@ func GetPools(ctx context.Context, api Interface) ([]*pools.Config, error) {
 func LookupPoolByID(ctx context.Context, api Interface, id ksuid.KSUID) (*pools.Config, error) {
 	b := newBuffer(pools.Config{})
 	zed := fmt.Sprintf("from :pools | id == hex('%s')", idToHex(id))
-	q, err := api.Query(ctx, nil, zed)
+	q, err := api.Query(ctx, zed)
 	if err != nil {
 		return nil, err
 	}
@@ -122,7 +121,7 @@ func LookupPoolByID(ctx context.Context, api Interface, id ksuid.KSUID) (*pools.
 func LookupBranchByName(ctx context.Context, api Interface, poolName, branchName string) (*lake.BranchMeta, error) {
 	b := newBuffer(lake.BranchMeta{})
 	zed := fmt.Sprintf("from :branches | pool.name == '%s' branch.name == '%s'", poolName, branchName)
-	q, err := api.Query(ctx, nil, zed)
+	q, err := api.Query(ctx, zed)
 	if err != nil {
 		return nil, err
 	}
@@ -147,7 +146,7 @@ func LookupBranchByName(ctx context.Context, api Interface, poolName, branchName
 func LookupBranchByID(ctx context.Context, api Interface, id ksuid.KSUID) (*lake.BranchMeta, error) {
 	b := newBuffer(lake.BranchMeta{})
 	zed := fmt.Sprintf("from :branches | branch.id == 'hex(%s)'", idToHex(id))
-	q, err := api.Query(ctx, nil, zed)
+	q, err := api.Query(ctx, zed)
 	if err != nil {
 		return nil, err
 	}
diff --git a/lake/api/local.go b/lake/api/local.go
index 7cb0091a13..0432c16882 100644
--- a/lake/api/local.go
+++ b/lake/api/local.go
@@ -105,12 +105,12 @@ func (l *local) Compact(ctx context.Context, poolID ksuid.KSUID, branchName stri
 	return exec.Compact(ctx, l.root, pool, branchName, objects, writeVectors, commit.Author, commit.Body, commit.Meta)
 }
 
-func (l *local) Query(ctx context.Context, head *lakeparse.Commitish, src string, srcfiles ...string) (zbuf.Scanner, error) {
+func (l *local) Query(ctx context.Context, src string, srcfiles ...string) (zbuf.Scanner, error) {
 	ast, err := parser.ParseQuery(src, srcfiles...)
 	if err != nil {
 		return nil, err
 	}
-	q, err := runtime.CompileLakeQuery(ctx, super.NewContext(), l.compiler, ast, head)
+	q, err := runtime.CompileLakeQuery(ctx, super.NewContext(), l.compiler, ast)
 	if err != nil {
 		return nil, err
 	}
diff --git a/lake/api/remote.go b/lake/api/remote.go
index d2e5a61856..9bca3a1895 100644
--- a/lake/api/remote.go
+++ b/lake/api/remote.go
@@ -119,8 +119,8 @@ func (r *remote) Revert(ctx context.Context, poolID ksuid.KSUID, branchName stri
 	return res.Commit, err
 }
 
-func (r *remote) Query(ctx context.Context, head *lakeparse.Commitish, src string, srcfiles ...string) (zbuf.Scanner, error) {
-	res, err := r.conn.Query(ctx, head, src, srcfiles...)
+func (r *remote) Query(ctx context.Context, src string, srcfiles ...string) (zbuf.Scanner, error) {
+	res, err := r.conn.Query(ctx, src, srcfiles...)
 	if err != nil {
 		return nil, err
 	}
diff --git a/lake/ztests/checkout-dash-p.yaml b/lake/ztests/checkout-dash-p.yaml
index 4acddd4844..013f4b9fc3 100644
--- a/lake/ztests/checkout-dash-p.yaml
+++ b/lake/ztests/checkout-dash-p.yaml
@@ -7,8 +7,7 @@ script: |
   super db branch -q child
   super db use -q @child
   super db load -q b.jsup
-  super db use -q POOL
-  super db query -z "*"
+  super db query -z "from POOL"
 
 inputs:
   - name: a.jsup
diff --git a/lake/ztests/debug.yaml b/lake/ztests/debug.yaml
index 628d5c5594..cc0463e40f 100644
--- a/lake/ztests/debug.yaml
+++ b/lake/ztests/debug.yaml
@@ -3,7 +3,7 @@ script: |
   super db init -q
   super db create -use -q test
   echo '{x: "foo"}' | super db load -q -
-  super db query -z 'debug f"debug: {x}"'
+  super db query -z 'from test | debug f"debug: {x}"'
 
 outputs:
   - name: stdout
diff --git a/lake/ztests/delete-where-missing.yaml b/lake/ztests/delete-where-missing.yaml
index b0d2802af3..d4b89e3d1b 100644
--- a/lake/ztests/delete-where-missing.yaml
+++ b/lake/ztests/delete-where-missing.yaml
@@ -4,9 +4,9 @@ script: |
   super db create -use -q test
   super db load -q in.jsup
   ! super db delete -q -where 'uid==C3UeSqaSOFRReHD68'
-  super db query -z 'count:=count()'
+  super db query -z 'from test | count:=count()'
   super db delete -q -where 'uid=="C3UeSqaSOFRReHD68"'
-  super db query -z 'count:=count()'
+  super db query -z 'from test | count:=count()'
 
 inputs:
   - name: in.jsup
diff --git a/lake/ztests/delete-where.yaml b/lake/ztests/delete-where.yaml
index b7ad6fc86a..4cb29157d1 100644
--- a/lake/ztests/delete-where.yaml
+++ b/lake/ztests/delete-where.yaml
@@ -6,9 +6,9 @@ script: |
     super db create -use -q -S 1KB -orderby ts:$order test
     seq 1000 | super -c '{ts:this-1,s:f"val{this-1}"}' - | super db load -q -
     super db delete -q -where 'ts > 400 and ts <= 500'
-    super db query -z 'count:=count()'
+    super db query -z 'from test | count:=count()'
     super db delete -q -where 's == "val1" or s == "val999"'
-    super db query -z 'count:=count()'
+    super db query -z 'from test | count:=count()'
     super db drop -f -q test
   done 
 
diff --git a/lake/ztests/delete.yaml b/lake/ztests/delete.yaml
index 722ea78a33..f0161fe04d 100644
--- a/lake/ztests/delete.yaml
+++ b/lake/ztests/delete.yaml
@@ -5,10 +5,10 @@ script: |
   super db load -q 1.jsup
   id=$(super db query -f text "from test@main:objects | cut id:=ksuid(id) | tail 1")
   super db load -q 2.jsup
-  super db query -z "*"
+  super db query -z "from test"
   echo === | tee /dev/stderr
   super db delete -q $id
-  super db query -z "*"
+  super db query -z "from test"
   echo === | tee /dev/stderr
   ! super db delete -q 27aaaaaaaaaaaaaaaaaaaaaaaaa
 
diff --git a/lake/ztests/query-error.yaml b/lake/ztests/query-error.yaml
index 9c7e14ed3d..ba6856b4c7 100644
--- a/lake/ztests/query-error.yaml
+++ b/lake/ztests/query-error.yaml
@@ -9,8 +9,8 @@ script: |
 outputs:
   - name: stderr
     data: |
-      query must include a 'from' operator
-      cannot resolve unknown HEAD at line 1, column 6:
+      query text is missing
+      HEAD: pool not found at line 1, column 6:
       from HEAD
            ~~~~
       unknown lake metadata type "unknownmeta" in from operator at line 1, column 6:
diff --git a/lake/ztests/revert-revert.yaml b/lake/ztests/revert-revert.yaml
index c7f3f865ee..4a4663e50f 100644
--- a/lake/ztests/revert-revert.yaml
+++ b/lake/ztests/revert-revert.yaml
@@ -7,10 +7,10 @@ script: |
   super db query -z "from test | sort this"
   r=$(super db revert $a | awk '{print $5}')
   echo ===
-  super db query -z "sort this"
+  super db query -z "from test | sort this"
   super db revert -q $r
   echo ===
-  super db query -z "sort this"
+  super db query -z "from test | sort this"
 
 inputs:
   - name: a.jsup
diff --git a/lake/ztests/revert.yaml b/lake/ztests/revert.yaml
index bdf20e7f4a..0eda7088ae 100644
--- a/lake/ztests/revert.yaml
+++ b/lake/ztests/revert.yaml
@@ -4,13 +4,13 @@ script: |
   super db create -use -q test
   a=$(super db load a.jsup | head -1 | awk '{print $1}')
   b=$(super db load b.jsup | head -1 | awk '{print $1}')
-  super db query -z "sort this"
+  super db query -z "from test | sort this"
   super db revert -q $a
   echo ===
-  super db query -z "sort this"
+  super db query -z "from test | sort this"
   super db revert -q $b
   echo ===
-  super db query -z "sort this"
+  super db query -z "from test | sort this"
 
 inputs:
   - name: a.jsup
diff --git a/lake/ztests/seek-index-boundaries.yaml b/lake/ztests/seek-index-boundaries.yaml
index 92ba896fb5..ac21413213 100644
--- a/lake/ztests/seek-index-boundaries.yaml
+++ b/lake/ztests/seek-index-boundaries.yaml
@@ -2,14 +2,13 @@ script: |
   export SUPER_DB_LAKE=test
   super db init -q
   for o in asc desc; do
-    echo // $o | tee /dev/stderr
     super db create -use -q -seekstride 1B -orderby ts:$o $o
     seq 20 | super -c 'yield {ts:this}' - | super db load -q -
-    source query.sh 'ts == 5'
-    source query.sh 'ts < 2'
-    source query.sh 'ts <= 2'
-    source query.sh 'ts > 19'
-    source query.sh 'ts >= 19'
+    source query.sh "from $o | ts == 5"
+    source query.sh "from $o | ts < 2"
+    source query.sh "from $o | ts <= 2"
+    source query.sh "from $o | ts > 19"
+    source query.sh "from $o | ts >= 19"
   done
 
 inputs:
@@ -20,53 +19,49 @@ inputs:
 outputs:
   - name: stdout
     data: |
-      // asc
-      // ts == 5
+      // from asc | ts == 5
       {ts:5}
-      // ts < 2
+      // from asc | ts < 2
       {ts:1}
-      // ts <= 2
+      // from asc | ts <= 2
       {ts:1}
       {ts:2}
-      // ts > 19
+      // from asc | ts > 19
       {ts:20}
-      // ts >= 19
+      // from asc | ts >= 19
       {ts:19}
       {ts:20}
-      // desc
-      // ts == 5
+      // from desc | ts == 5
       {ts:5}
-      // ts < 2
+      // from desc | ts < 2
       {ts:1}
-      // ts <= 2
+      // from desc | ts <= 2
       {ts:2}
       {ts:1}
-      // ts > 19
+      // from desc | ts > 19
       {ts:20}
-      // ts >= 19
+      // from desc | ts >= 19
       {ts:20}
       {ts:19}
   - name: stderr
     data: |
-      // asc
-      // ts == 5
+      // from asc | ts == 5
       {bytes_read:2,bytes_matched:2,records_read:1,records_matched:1}
-      // ts < 2
+      // from asc | ts < 2
       {bytes_read:2,bytes_matched:2,records_read:1,records_matched:1}
-      // ts <= 2
+      // from asc | ts <= 2
       {bytes_read:4,bytes_matched:4,records_read:2,records_matched:2}
-      // ts > 19
+      // from asc | ts > 19
       {bytes_read:2,bytes_matched:2,records_read:1,records_matched:1}
-      // ts >= 19
+      // from asc | ts >= 19
       {bytes_read:4,bytes_matched:4,records_read:2,records_matched:2}
-      // desc
-      // ts == 5
+      // from desc | ts == 5
       {bytes_read:2,bytes_matched:2,records_read:1,records_matched:1}
-      // ts < 2
+      // from desc | ts < 2
       {bytes_read:2,bytes_matched:2,records_read:1,records_matched:1}
-      // ts <= 2
+      // from desc | ts <= 2
       {bytes_read:4,bytes_matched:4,records_read:2,records_matched:2}
-      // ts > 19
+      // from desc | ts > 19
       {bytes_read:2,bytes_matched:2,records_read:1,records_matched:1}
-      // ts >= 19
+      // from desc | ts >= 19
       {bytes_read:4,bytes_matched:4,records_read:2,records_matched:2}
diff --git a/lake/ztests/seek-index-null.yaml b/lake/ztests/seek-index-null.yaml
index 9b8fafaa57..74190af99a 100644
--- a/lake/ztests/seek-index-null.yaml
+++ b/lake/ztests/seek-index-null.yaml
@@ -5,7 +5,7 @@ script: |
     echo // $o | tee /dev/stderr
     super db create -q -seekstride 2KB -orderby ts:$o $o
     super db load -q -use $o babble.jsup null.jsup
-    super db query -z -s -use $o "ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z"
+    super db query -z -s "from $o | ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z"
   done
 
 inputs:
diff --git a/lake/ztests/seek-index-ts.yaml b/lake/ztests/seek-index-ts.yaml
index 4ed5706bf9..f4b884de33 100644
--- a/lake/ztests/seek-index-ts.yaml
+++ b/lake/ztests/seek-index-ts.yaml
@@ -2,12 +2,11 @@ script: |
   export SUPER_DB_LAKE=test
   super db init -q
   for o in asc desc; do
-    echo // $o | tee /dev/stderr
     super db create -use -seekstride 2KB -orderby ts:$o -q $o
     super db load -q babble.jsup
-    source query.sh "ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z"
-    source query.sh "ts == 2020-04-21T23:59:26.06326664Z"
-    source query.sh "ts == 2020-04-21T23:59:26.06326664Z or foo == 'bar'"
+    source query.sh "from $o | ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z"
+    source query.sh "from $o | ts == 2020-04-21T23:59:26.06326664Z"
+    source query.sh "from $o | ts == 2020-04-21T23:59:26.06326664Z or foo == 'bar'"
 
   done
 
@@ -22,37 +21,33 @@ inputs:
 outputs:
   - name: stdout
     data: |
-      // asc
-      // ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z
+      // from asc | ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z
       {ts:2020-04-21T23:59:26.06326664Z,s:"potbellied-Dedanim",v:230}
       {ts:2020-04-21T23:59:29.06985813Z,s:"areek-ashless",v:266}
       {ts:2020-04-21T23:59:38.0687693Z,s:"topcoating-rhexis",v:415}
-      // ts == 2020-04-21T23:59:26.06326664Z
+      // from asc | ts == 2020-04-21T23:59:26.06326664Z
       {ts:2020-04-21T23:59:26.06326664Z,s:"potbellied-Dedanim",v:230}
-      // ts == 2020-04-21T23:59:26.06326664Z or foo == 'bar'
+      // from asc | ts == 2020-04-21T23:59:26.06326664Z or foo == 'bar'
       {ts:2020-04-21T23:59:26.06326664Z,s:"potbellied-Dedanim",v:230}
-      // desc
-      // ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z
+      // from desc | ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z
       {ts:2020-04-21T23:59:38.0687693Z,s:"topcoating-rhexis",v:415}
       {ts:2020-04-21T23:59:29.06985813Z,s:"areek-ashless",v:266}
       {ts:2020-04-21T23:59:26.06326664Z,s:"potbellied-Dedanim",v:230}
-      // ts == 2020-04-21T23:59:26.06326664Z
+      // from desc | ts == 2020-04-21T23:59:26.06326664Z
       {ts:2020-04-21T23:59:26.06326664Z,s:"potbellied-Dedanim",v:230}
-      // ts == 2020-04-21T23:59:26.06326664Z or foo == 'bar'
+      // from desc | ts == 2020-04-21T23:59:26.06326664Z or foo == 'bar'
       {ts:2020-04-21T23:59:26.06326664Z,s:"potbellied-Dedanim",v:230}
   - name: stderr
     data: |
-      // asc
-      // ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z
+      // from asc | ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z
       {bytes_read:16403,bytes_matched:87,records_read:500,records_matched:3}
-      // ts == 2020-04-21T23:59:26.06326664Z
+      // from asc | ts == 2020-04-21T23:59:26.06326664Z
       {bytes_read:8141,bytes_matched:31,records_read:250,records_matched:1}
-      // ts == 2020-04-21T23:59:26.06326664Z or foo == 'bar'
+      // from asc | ts == 2020-04-21T23:59:26.06326664Z or foo == 'bar'
       {bytes_read:32889,bytes_matched:31,records_read:1000,records_matched:1}
-      // desc
-      // ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z
+      // from desc | ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z
       {bytes_read:16403,bytes_matched:87,records_read:500,records_matched:3}
-      // ts == 2020-04-21T23:59:26.06326664Z
+      // from desc | ts == 2020-04-21T23:59:26.06326664Z
       {bytes_read:8141,bytes_matched:31,records_read:250,records_matched:1}
-      // ts == 2020-04-21T23:59:26.06326664Z or foo == 'bar'
+      // from desc | ts == 2020-04-21T23:59:26.06326664Z or foo == 'bar'
       {bytes_read:32889,bytes_matched:31,records_read:1000,records_matched:1}
diff --git a/lake/ztests/vacuum.yaml b/lake/ztests/vacuum.yaml
index 0a1dfd3fd4..7d49ddff1f 100644
--- a/lake/ztests/vacuum.yaml
+++ b/lake/ztests/vacuum.yaml
@@ -8,7 +8,7 @@ script: |
   super db vacuum -dryrun
   super db vacuum -f
   super db revert -q $r
-  ! super db query -z '*'
+  ! super db query -z 'from test'
 
 outputs:
   - name: stdout
diff --git a/runtime/compiler.go b/runtime/compiler.go
index 2d58046b01..bb5353663b 100644
--- a/runtime/compiler.go
+++ b/runtime/compiler.go
@@ -14,7 +14,7 @@ import (
 
 type Compiler interface {
 	NewQuery(*Context, *parser.AST, []zio.Reader) (Query, error)
-	NewLakeQuery(*Context, *parser.AST, int, *lakeparse.Commitish) (Query, error)
+	NewLakeQuery(*Context, *parser.AST, int) (Query, error)
 	NewLakeDeleteQuery(*Context, *parser.AST, *lakeparse.Commitish) (DeleteQuery, error)
 }
 
@@ -44,9 +44,9 @@ func CompileQuery(ctx context.Context, zctx *super.Context, c Compiler, ast *par
 	return q, nil
 }
 
-func CompileLakeQuery(ctx context.Context, zctx *super.Context, c Compiler, ast *parser.AST, head *lakeparse.Commitish) (Query, error) {
+func CompileLakeQuery(ctx context.Context, zctx *super.Context, c Compiler, ast *parser.AST) (Query, error) {
 	rctx := NewContext(ctx, zctx)
-	q, err := c.NewLakeQuery(rctx, ast, 0, head)
+	q, err := c.NewLakeQuery(rctx, ast, 0)
 	if err != nil {
 		rctx.Cancel()
 		return nil, err
diff --git a/runtime/sam/expr/filter_test.go b/runtime/sam/expr/filter_test.go
index 0e597634ca..ccdacc5103 100644
--- a/runtime/sam/expr/filter_test.go
+++ b/runtime/sam/expr/filter_test.go
@@ -7,6 +7,7 @@ import (
 
 	"github.com/brimdata/super"
 	"github.com/brimdata/super/compiler"
+	"github.com/brimdata/super/compiler/data"
 	"github.com/brimdata/super/compiler/parser"
 	"github.com/brimdata/super/runtime"
 	"github.com/brimdata/super/runtime/sam/expr"
@@ -54,7 +55,7 @@ func runCasesHelper(t *testing.T, record string, cases []testcase, expectBufferF
 			t.Helper()
 			ast, err := parser.ParseQuery(c.filter)
 			require.NoError(t, err, "filter: %q", c.filter)
-			job, err := compiler.NewJob(runtime.DefaultContext(), ast, nil, nil)
+			job, err := compiler.NewJob(runtime.DefaultContext(), ast, &data.Source{}, true)
 			require.NoError(t, err, "filter: %q", c.filter)
 			err = job.Optimize()
 			require.NoError(t, err, "filter: %q", c.filter)
diff --git a/runtime/sam/op/meta/ztests/nulls-max.yaml b/runtime/sam/op/meta/ztests/nulls-max.yaml
index 50387e11c8..9a3806c29f 100644
--- a/runtime/sam/op/meta/ztests/nulls-max.yaml
+++ b/runtime/sam/op/meta/ztests/nulls-max.yaml
@@ -11,7 +11,7 @@ script: |
     echo '{ts:1}' | super db load -q -
     super db query -z "from $o:objects | drop id, size"
     echo "// ==="
-    super db query -z 'head 1'
+    super db query -z "from $o | head 1"
   done
 
 outputs:
diff --git a/service/auth_test.go b/service/auth_test.go
index 0001ced203..55f57a4333 100644
--- a/service/auth_test.go
+++ b/service/auth_test.go
@@ -37,7 +37,7 @@ func TestAuthIdentity(t *testing.T) {
 	core, conn := newCoreWithConfig(t, service.Config{
 		Auth: authConfig,
 	})
-	_, err := conn.Query(context.Background(), nil, "from [pools]")
+	_, err := conn.Query(context.Background(), "from [pools]")
 	require.Error(t, err)
 	require.Equal(t, 1.0, promCounterValue(core.Registry(), "request_errors_unauthorized_total"))
 
@@ -59,7 +59,7 @@ func TestAuthIdentity(t *testing.T) {
 		UserID:   "test_user_id",
 	}, res)
 
-	_, err = conn.Query(context.Background(), nil, "from :pools")
+	_, err = conn.Query(context.Background(), "from :pools")
 	require.NoError(t, err)
 }
 
diff --git a/service/client_test.go b/service/client_test.go
index 61333ee991..372d4aabb4 100644
--- a/service/client_test.go
+++ b/service/client_test.go
@@ -48,7 +48,7 @@ func (c *testClient) TestBranchGet(id ksuid.KSUID) (config lake.BranchMeta) {
 }
 
 func (c *testClient) TestPoolList() []pools.Config {
-	r, err := c.Query(context.Background(), nil, "from :pools")
+	r, err := c.Query(context.Background(), "from :pools")
 	require.NoError(c, err)
 	defer r.Body.Close()
 	var confs []pools.Config
@@ -80,7 +80,7 @@ func (c *testClient) TestBranchPost(poolID ksuid.KSUID, payload api.BranchPostRe
 }
 
 func (c *testClient) TestQuery(query string) string {
-	r, err := c.Connection.Query(context.Background(), nil, query)
+	r, err := c.Connection.Query(context.Background(), query)
 	require.NoError(c, err)
 	defer r.Body.Close()
 	zr := zngio.NewReader(super.NewContext(), r.Body)
diff --git a/service/handlers.go b/service/handlers.go
index 08da636bb6..a27ede26d1 100644
--- a/service/handlers.go
+++ b/service/handlers.go
@@ -62,7 +62,7 @@ func handleQuery(c *Core, w *ResponseWriter, r *Request) {
 		w.Error(srverr.ErrInvalid(err))
 		return
 	}
-	flowgraph, err := runtime.CompileLakeQuery(r.Context(), super.NewContext(), c.compiler, ast, &req.Head)
+	flowgraph, err := runtime.CompileLakeQuery(r.Context(), super.NewContext(), c.compiler, ast)
 	if err != nil {
 		w.Error(srverr.ErrInvalid(err))
 		return
@@ -182,7 +182,7 @@ func handleQueryDescribe(c *Core, w *ResponseWriter, r *Request) {
 		return
 	}
 	src := data.NewSource(storage.NewRemoteEngine(), c.root)
-	info, err := describe.Analyze(r.Context(), req.Query, src, &req.Head)
+	info, err := describe.Analyze(r.Context(), req.Query, src)
 	if err != nil {
 		w.Error(srverr.ErrInvalid(err))
 		return
diff --git a/service/ztests/curl-delete-where.yaml b/service/ztests/curl-delete-where.yaml
index 3380e8cf8c..5cbd08e7e7 100644
--- a/service/ztests/curl-delete-where.yaml
+++ b/service/ztests/curl-delete-where.yaml
@@ -7,10 +7,10 @@ script: |
   curl -s -d '{where:"x <= 4"}' $SUPER_DB_LAKE/pool/test/branch/main/delete |
     sed -E 's/0x[0-9a-f]{40}/xxx/'
   echo ===
-  super db query -z '*'
+  super db query -z 'from test'
   echo ===
   curl -w 'code %{response_code}\n' -d '{where:"x <= 4"}' $SUPER_DB_LAKE/pool/test/branch/main/delete
-  super db query -z '*'
+  super db query -z 'from test'
 
 inputs:
   - name: service.sh
diff --git a/service/ztests/curl-query-error.yaml b/service/ztests/curl-query-error.yaml
index 08ccf79f31..29c61d6a0c 100644
--- a/service/ztests/curl-query-error.yaml
+++ b/service/ztests/curl-query-error.yaml
@@ -13,11 +13,11 @@ inputs:
 outputs:
   - name: stdout
     data: |
-      {"type":"Error","kind":"invalid operation","error":"no pool name given"}
+      {"type":"Error","kind":"invalid operation","error":"query text is missing"}
       code 400
-      {"type":"Error","kind":"invalid operation","error":"no pool name given"}
+      {"type":"Error","kind":"invalid operation","error":"query text is missing"}
       code 400
-      {"type":"Error","kind":"invalid operation","error":"cannot resolve unknown HEAD at line 1, column 6:\nfrom HEAD\n     ~~~~","compilation_errors":[{"Msg":"cannot resolve unknown HEAD","Pos":5,"End":8}]}
+      {"type":"Error","kind":"invalid operation","error":"HEAD: pool not found at line 1, column 6:\nfrom HEAD\n     ~~~~","compilation_errors":[{"Msg":"HEAD: pool not found","Pos":5,"End":8}]}
       code 400
       {"type":"Error","kind":"invalid operation","error":"unknown lake metadata type \"unknownmeta\" in from operator at line 1, column 6:\nfrom :unknownmeta\n     ~~~~~~~~~~~~","compilation_errors":[{"Msg":"unknown lake metadata type \"unknownmeta\" in from operator","Pos":5,"End":16}]}
       code 400
diff --git a/service/ztests/debug.yaml b/service/ztests/debug.yaml
index 938747c206..3019f05c0b 100644
--- a/service/ztests/debug.yaml
+++ b/service/ztests/debug.yaml
@@ -2,7 +2,7 @@ script: |
   source service.sh
   super db create -use -q test
   echo '{x: "foo"}' | super db load -q -
-  super db query -z 'debug f"debug: {x}"'
+  super db query -z 'from test | debug f"debug: {x}"'
 
 inputs:
   - name: service.sh
diff --git a/service/ztests/delete-where.yaml b/service/ztests/delete-where.yaml
index 87f7bbb212..0facc1ca83 100644
--- a/service/ztests/delete-where.yaml
+++ b/service/ztests/delete-where.yaml
@@ -5,11 +5,11 @@ script: |
     super db create -use -q -S 1KB -orderby ts:$order test
     seq 1000 | super -c '{ts:this-1,s:f"val{this-1}"}' - | super db load -q -
     super db delete -q -where 'ts > 400 and ts <= 500'
-    super db query -z 'count()'
+    super db query -z 'from test | count()'
     super db delete -q -where 's == "val1" or s == "val999"'
-    super db query -z 'count()'
+    super db query -z 'from test | count()'
     ! super db delete -q -where 's == "val1" or s == "val999"'
-    super db query -z 'count()'
+    super db query -z 'from test | count()'
     super db drop -f -q test
   done 
 
diff --git a/service/ztests/delete.yaml b/service/ztests/delete.yaml
index f254c6144d..c8a79c1e60 100644
--- a/service/ztests/delete.yaml
+++ b/service/ztests/delete.yaml
@@ -4,10 +4,10 @@ script: |
   super db load -q 1.jsup
   id=$(super db query -f text "from test@main:objects | cut id:=ksuid(id) | tail 1")
   super db load -q 2.jsup
-  super db query -z "*"
+  super db query -z "from test"
   echo === | tee /dev/stderr
   super db delete -q $id
-  super db query -z "*"
+  super db query -z "from test"
   echo === | tee /dev/stderr
   ! super db delete -q 27aaaaaaaaaaaaaaaaaaaaaaaaa
 
diff --git a/service/ztests/query-describe.yaml b/service/ztests/query-describe.yaml
index 48afda3425..6a64c9369a 100644
--- a/service/ztests/query-describe.yaml
+++ b/service/ztests/query-describe.yaml
@@ -20,10 +20,10 @@ inputs:
       ) | put foo := "bar"
   - name: agg.spq
     data: |
-      count() by key1:=v1, key2
+      from test1 | count() by key1:=v1, key2
   - name: agg-no-keys.spq
     data: |
-      sum(this)
+      from test1 | sum(this)
   - name: two-channels.spq
     data: |
       fork (
@@ -32,7 +32,7 @@ inputs:
       )
   - name: agg-sort.spq
     data: |
-      sum(this) by foo | sort x
+      from test1 | sum(this) by foo | sort x
   - name: scope.spq
     data: |
       type port = uint16
@@ -50,14 +50,12 @@ outputs:
               {
                   "kind": "Pool",
                   "name": "test1",
-                  "id": "XXX",
-                  "inferred": false
+                  "id": "XXX"
               },
               {
                   "kind": "Pool",
                   "name": "test2",
-                  "id": "XXX",
-                  "inferred": false
+                  "id": "XXX"
               }
           ],
           "channels": [
@@ -80,8 +78,7 @@ outputs:
           "sources": {
               "kind": "Pool",
               "name": "test1",
-              "id": "XXX",
-              "inferred": true
+              "id": "XXX"
           },
           "channels": [
               {
@@ -103,8 +100,7 @@ outputs:
           "sources": {
               "kind": "Pool",
               "name": "test1",
-              "id": "XXX",
-              "inferred": true
+              "id": "XXX"
           },
           "channels": [
               {
@@ -120,14 +116,12 @@ outputs:
               {
                   "kind": "Pool",
                   "name": "test1",
-                  "id": "XXX",
-                  "inferred": false
+                  "id": "XXX"
               },
               {
                   "kind": "Pool",
                   "name": "test2",
-                  "id": "XXX",
-                  "inferred": false
+                  "id": "XXX"
               }
           ],
           "channels": [
@@ -159,8 +153,7 @@ outputs:
           "sources": {
               "kind": "Pool",
               "name": "test1",
-              "id": "XXX",
-              "inferred": true
+              "id": "XXX"
           },
           "channels": [
               {
@@ -186,8 +179,7 @@ outputs:
           "sources": {
               "kind": "Pool",
               "name": "test1",
-              "id": "XXX",
-              "inferred": false
+              "id": "XXX"
           },
           "channels": [
               {
@@ -214,8 +206,7 @@ outputs:
           "sources": {
               "kind": "Pool",
               "name": "test1",
-              "id": "XXX",
-              "inferred": false
+              "id": "XXX"
           },
           "channels": [
               {
diff --git a/service/ztests/query-error.yaml b/service/ztests/query-error.yaml
index 567e5e61e6..eae67f0e90 100644
--- a/service/ztests/query-error.yaml
+++ b/service/ztests/query-error.yaml
@@ -12,8 +12,8 @@ inputs:
 outputs:
   - name: stderr
     data: |
-      status code 400: no pool name given
-      cannot resolve unknown HEAD at line 1, column 6:
+      status code 400: query text is missing
+      HEAD: pool not found at line 1, column 6:
       from HEAD
            ~~~~
       unknown lake metadata type "unknownmeta" in from operator at line 1, column 6:
diff --git a/service/ztests/seek-index-null.yaml b/service/ztests/seek-index-null.yaml
index f4230a4a72..1d7cf04d02 100644
--- a/service/ztests/seek-index-null.yaml
+++ b/service/ztests/seek-index-null.yaml
@@ -4,7 +4,7 @@ script: |
     echo // $o | tee /dev/stderr
     super db create -q -seekstride=2KB -orderby ts:$o $o
     super db load -q -use $o babble.jsup null.jsup
-    super db query -z -s -use $o "ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z"
+    super db query -z -s "from $o | ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z"
   done
 
 inputs:
diff --git a/service/ztests/seek-index.yaml b/service/ztests/seek-index.yaml
index 557a1bbdd5..322d16354c 100644
--- a/service/ztests/seek-index.yaml
+++ b/service/ztests/seek-index.yaml
@@ -4,9 +4,9 @@ script: |
     echo // $o | tee /dev/stderr
     super db create -use -seekstride 2KB -orderby ts:$o -q $o
     super db load -q babble.jsup
-    source query.sh "ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z"
-    source query.sh "ts == 2020-04-21T23:59:26.06326664Z"
-    source query.sh "ts == 2020-04-21T23:59:26.06326664Z or foo == 'bar'"
+    source query.sh "from $o | ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z"
+    source query.sh "from $o | ts == 2020-04-21T23:59:26.06326664Z"
+    source query.sh "from $o | ts == 2020-04-21T23:59:26.06326664Z or foo == 'bar'"
   done
 
 inputs:
@@ -22,36 +22,36 @@ outputs:
   - name: stdout
     data: |
       // asc
-      // ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z
+      // from asc | ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z
       {ts:2020-04-21T23:59:26.06326664Z,s:"potbellied-Dedanim",v:230}
       {ts:2020-04-21T23:59:29.06985813Z,s:"areek-ashless",v:266}
       {ts:2020-04-21T23:59:38.0687693Z,s:"topcoating-rhexis",v:415}
-      // ts == 2020-04-21T23:59:26.06326664Z
+      // from asc | ts == 2020-04-21T23:59:26.06326664Z
       {ts:2020-04-21T23:59:26.06326664Z,s:"potbellied-Dedanim",v:230}
-      // ts == 2020-04-21T23:59:26.06326664Z or foo == 'bar'
+      // from asc | ts == 2020-04-21T23:59:26.06326664Z or foo == 'bar'
       {ts:2020-04-21T23:59:26.06326664Z,s:"potbellied-Dedanim",v:230}
       // desc
-      // ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z
+      // from desc | ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z
       {ts:2020-04-21T23:59:38.0687693Z,s:"topcoating-rhexis",v:415}
       {ts:2020-04-21T23:59:29.06985813Z,s:"areek-ashless",v:266}
       {ts:2020-04-21T23:59:26.06326664Z,s:"potbellied-Dedanim",v:230}
-      // ts == 2020-04-21T23:59:26.06326664Z
+      // from desc | ts == 2020-04-21T23:59:26.06326664Z
       {ts:2020-04-21T23:59:26.06326664Z,s:"potbellied-Dedanim",v:230}
-      // ts == 2020-04-21T23:59:26.06326664Z or foo == 'bar'
+      // from desc | ts == 2020-04-21T23:59:26.06326664Z or foo == 'bar'
       {ts:2020-04-21T23:59:26.06326664Z,s:"potbellied-Dedanim",v:230}
   - name: stderr
     data: |
       // asc
-      // ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z
+      // from asc | ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z
       {bytes_read:16403,bytes_matched:87,records_read:500,records_matched:3}
-      // ts == 2020-04-21T23:59:26.06326664Z
+      // from asc | ts == 2020-04-21T23:59:26.06326664Z
       {bytes_read:8141,bytes_matched:31,records_read:250,records_matched:1}
-      // ts == 2020-04-21T23:59:26.06326664Z or foo == 'bar'
+      // from asc | ts == 2020-04-21T23:59:26.06326664Z or foo == 'bar'
       {bytes_read:32889,bytes_matched:31,records_read:1000,records_matched:1}
       // desc
-      // ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z
+      // from desc | ts >= 2020-04-21T23:59:26.063Z and ts <= 2020-04-21T23:59:38.069Z
       {bytes_read:16403,bytes_matched:87,records_read:500,records_matched:3}
-      // ts == 2020-04-21T23:59:26.06326664Z
+      // from desc | ts == 2020-04-21T23:59:26.06326664Z
       {bytes_read:8141,bytes_matched:31,records_read:250,records_matched:1}
-      // ts == 2020-04-21T23:59:26.06326664Z or foo == 'bar'
+      // from desc | ts == 2020-04-21T23:59:26.06326664Z or foo == 'bar'
       {bytes_read:32889,bytes_matched:31,records_read:1000,records_matched:1}
diff --git a/service/ztests/vacuum.yaml b/service/ztests/vacuum.yaml
index 6bfd4c13c5..0cd1837a62 100644
--- a/service/ztests/vacuum.yaml
+++ b/service/ztests/vacuum.yaml
@@ -7,7 +7,7 @@ script: |
   super db vacuum -dryrun
   super db vacuum -f
   super db revert -q $r
-  ! super db query -z '*'
+  ! super db query -z 'from test'
 
 inputs:
   - name: service.sh
diff --git a/zfmt/ast.go b/zfmt/ast.go
index 5230fee868..42224ef6b2 100644
--- a/zfmt/ast.go
+++ b/zfmt/ast.go
@@ -605,7 +605,7 @@ func (c *canon) op(p ast.Op) {
 			c.expr(p.Expr, "")
 		}
 	default:
-		c.open("unknown proc: %T", p)
+		c.open("unknown operator: %T", p)
 		c.close()
 	}
 }
diff --git a/zfmt/dag.go b/zfmt/dag.go
index 5fbe6c4fef..a306207c02 100644
--- a/zfmt/dag.go
+++ b/zfmt/dag.go
@@ -505,6 +505,9 @@ func (c *canonDAG) op(p dag.Op) {
 			c.expr(p.Filter, "")
 			c.write(")")
 		}
+	case *dag.NullScan:
+		c.next()
+		c.write("null")
 	case *dag.FileScan:
 		c.next()
 		c.write("file %s", p.Path)
@@ -548,7 +551,7 @@ func (c *canonDAG) op(p dag.Op) {
 		c.write("output %s", p.Name)
 	default:
 		c.next()
-		c.open("unknown proc: %T", p)
+		c.open("unknown operator: %T", p)
 		c.close()
 	}
 }
diff --git a/zfmt/ztests/debug.yaml b/zfmt/ztests/debug.yaml
index cb866860f2..a53db1698d 100644
--- a/zfmt/ztests/debug.yaml
+++ b/zfmt/ztests/debug.yaml
@@ -7,7 +7,7 @@ outputs:
     data: |
       debug f"debug: {this}"
       | head 1
-      reader
+      null
       | mirror (
         =>
           yield "debug: "+cast(this, <string>)
diff --git a/zfmt/ztests/join.yaml b/zfmt/ztests/join.yaml
index 71f44d5474..771b74ccd2 100644
--- a/zfmt/ztests/join.yaml
+++ b/zfmt/ztests/join.yaml
@@ -10,7 +10,7 @@ outputs:
         from "test.jsup"
       ) on x=x p:=a
       ===
-      reader
+      null
       | fork (
         =>
           pass
diff --git a/zfmt/ztests/output.yaml b/zfmt/ztests/output.yaml
index 58e77b924d..d97c29dc9b 100644
--- a/zfmt/ztests/output.yaml
+++ b/zfmt/ztests/output.yaml
@@ -6,7 +6,7 @@ script: |
 outputs:
   - name: stdout
     data: |
-      reader
+      null
       | fork (
         =>
           output foo
@@ -15,7 +15,7 @@ outputs:
           | output main
       )
       // ===
-      reader
+      null
       | switch x (
           case "foo" =>
             output foo
diff --git a/zfmt/ztests/over.yaml b/zfmt/ztests/over.yaml
index 3814ceb5e2..00b9db16e3 100644
--- a/zfmt/ztests/over.yaml
+++ b/zfmt/ztests/over.yaml
@@ -42,7 +42,7 @@ outputs:
         search g
       )
       ===
-      reader
+      null
       | yield (
         over a
         | where search("b")
diff --git a/zfmt/ztests/precedence-dag.yaml b/zfmt/ztests/precedence-dag.yaml
index a6cd7d1a88..807d7e6e0d 100644
--- a/zfmt/ztests/precedence-dag.yaml
+++ b/zfmt/ztests/precedence-dag.yaml
@@ -16,12 +16,12 @@ inputs:
 outputs:
   -  name: stdout
      data: |
-        reader
+        null
         | yield (10+2)/8
         | output main
-        reader
+        null
         | where (y==2 or x==4) and z==5
         | output main
-        reader
+        null
         | where !(ts<=2)
         | output main
diff --git a/zfmt/ztests/switch.yaml b/zfmt/ztests/switch.yaml
index dfb95cb525..00f2510b29 100644
--- a/zfmt/ztests/switch.yaml
+++ b/zfmt/ztests/switch.yaml
@@ -17,7 +17,7 @@ outputs:
           tail
       )
       ===
-      reader
+      null
       | switch (
           case search("a") =>
             head 1
@@ -34,7 +34,7 @@ outputs:
           tail
       )
       ===
-      reader
+      null
       | switch (
           case search("a") =>
             head 1
diff --git a/zfmt/ztests/type-value.yaml b/zfmt/ztests/type-value.yaml
index 5b9f54e4e9..86921bd3c7 100644
--- a/zfmt/ztests/type-value.yaml
+++ b/zfmt/ztests/type-value.yaml
@@ -6,6 +6,6 @@ outputs:
   - name: stdout
     data: |
       search is(<foo>) and bar
-      reader
+      null
       | where is(<(uint16,ip)>) and search(80)
       | output main
diff --git a/zfmt/ztests/yield-shortcut.yaml b/zfmt/ztests/yield-shortcut.yaml
index a1e9447279..d6c9e21a85 100644
--- a/zfmt/ztests/yield-shortcut.yaml
+++ b/zfmt/ztests/yield-shortcut.yaml
@@ -20,18 +20,18 @@ outputs:
       yield |{"foo":1,"bar":2}|
       yield <(int64,string)>(1)
       ===
-      reader
+      null
       | yield {x:1,...y}
       | output main
-      reader
+      null
       | yield [1,2,3]
       | output main
-      reader
+      null
       | yield |["foo","bar"]|
       | output main
-      reader
+      null
       | yield |{"foo":1,"bar":2}|
       | output main
-      reader
+      null
       | yield cast(1, <(int64,string)>)
       | output main